United States Industrial Environmental Research EPA-600/2-79-082
Environmental Protection Laboratory April 1979
Agency Research Triangle Park NC 27711
Research and Development
vxEPA Coke Quench Tower
Emission Testing Program
-------
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.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-79-082
April 1979
Coke Quench Tower
Emission Testing Program
by
A.M. Laube and B.A. Drummond
York Research Corporation
One Research Drive
Stamford, Connecticut 06906
Contract No. 68-02-2819
W.A. 1
Program Element No. 1AB604
EPA Project Officer: Robert V. Hendricks
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ACKNOWLEDGEMENT
The following personnel performed the field
tests under the hostile conditions of the coke
quench process:
Barbara Drummond
John Gale
John Jeffery
Richard Keith
Roderick Lamothe
Paul Wade
Michael Ziskin
Arthur D. Little, Inc., Boston, Massachusetts
performed the major portion of the analytical
work including all of the organic emissions
test samples and the organic analysis of water
samples.
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES
LIST OF TABLES
1 . 0 SUMMARY !
1.1 Organic Emissions ^
1.2 Organics in 'Quench Water 5
1.3 The Source of Quench Tower Organic Emissions 5
1.4 Benzene Soluble Residue, Benzene and
Total Hydrocarbons 6
1.5 Biological Test 6
1.6 Particulate Emissions 6
2 . 0 CONCLUSIONS 9
3 . 0 RECOMMENDATIONS 10
4.0 INTRODUCTION 11
5.0 PROCESS DESCRIPTION 13
5.1 The Coking Process 13
5.2 The Quenching Process 19
6.0 TEST PROCEDURES 28
6.1 Test Program 28
6.2 Sampling Problems 36
6.3 Sampling Equipment Design 38
6.4 Sampling Technique 42
6.5 Test Sequence 45
6.6 Coke Quench Tower Emission Tests 48
6.7 Sample Recovery 51
6.8 Precision of Sampling Methods and Estimated
Probable Error for Analytical Procedures 54
7.0 ANALYTICAL METHODS 58
7.1 Organic Emission Tests and Water Samples 58
7.1.1 Particulate 58
7.1.2 Sample Extraction 61
7.1.3 T.C.O., GRAV, and IR 61
7.1.4 Liquid Chroma tography (LC) 62
7.1.5 Low Resolution Mass Spectroscopy (LRMS) 63
7.1.6 Gas Chroma tography/Mass Spectrometry
(GC/MS) 63
-------
TABLE OF CONTENTS
(continued)
Page
7.2 Benzene Soluble Residue Tests 66
7.3 Benzene and Total Hydrocarbons 67
7.4 Total Organic Carbon in Quench Water 67
7.5 Electrical Conductivity of Quench Water 67
7.6 Biological Test 67
7.6.1 Ames Bacterial Assay 68
7.6.2 Clonal Cytotoxicity Assay 70
8.0 ORGANIC EMISSIONS 72
8.1 Organic Emission Test Results 72
8.2 Sample Contamination 83
8.3 Benzene Soluble Residue Test Results 86
8.4 Benzene and Total Hydrocarbons (THC) 88
8.5 Biological Test Results 88
9.0 PARTICULATE EMISSIONS 94
9.1 Particulate Emission Results 94
9.2 Particle Size 98
9.3 Comparison of Quench Tower Particulate Emission 102
9.4 Baffles 106
10.0 QUENCH WATER ORGANICS AND WATER FLOW 112
10.1 Organics in Quench Water 112
10.2 Quench Water Flow 116
10.3 Mass Balance Around the Quench Tower 119
11.0 SOURCE OF ORGANIC EMISSIONS 124
11.1 Statistical Analyses ' 124
11.2 Discussion 128
REFERENCES 136
APPENDICES
A. Summary of Coke Quench Emission Sampling Data A-l
B. Gaseous Emissions - 1976 - Lorain Study B-!
C. Calculations -
-------
LIST OF FIGURES
Number Page
5-1 Coke Oven Plant Schematic 14
5-2 Diagram of Lorain Coke Plant
Contaminated Quench Water System 18
5-3 Service and Process Water, Coke
Plant, Lorain Works 20
5-4 The Push of Incandescent Coke
From the Oven to the Quench Car 21
5-5 Quench Car with Incandescent
Coke Entering the Quench Tower 22
5-6 "Car In" 22
5-7 "Water On" 22
5-8 Start of Quench Water Flow 23
5-9 Fugitive Emissions a Few
Seconds Before the Up-stack
Draft is Well Established 23
5-10 A View Showing the first
Eruption of Steam 24
5-11 Another View Showing the
First Eruption of Steam 24
5-12 Quench Tower Plume 25
5-13 The Stack Flow Well Established,
About 10 Seconds After "Water On" 25
6-1 Quench Tower Schematic 29
6-2 Diagram of the Quench Tower Area 30
6-3 to 6-6 Photographs Illustrating Coke Greenness
Ratings (ratings of 0,2,3, and 4) 33
6-7 Organic Matter and Particulate
Sampling Train Using Modified
EPA Method Five 40
6-8 Velocity Profile 44
-i-
-------
LIST OF FIGURES CONT.
Number page
6-9 Sample Point Location 46
6-10 Fugitive Emissions at Start of Quench 49
6-11 Fugitive Emissions a Few Seconds
After Beginning of Quench 49
7-1 Analytical Procedures for Samples
from Tests 5,7, and 14 59
7-2 Analytical Procedures for Samples
from Test 2B, 3,4,5,6,7,8,9,10, 11,
12,14, 15,16, and 17 60
8-1 EC50 Determination - Coke Quench
Tower Emission Sample 93
9-1 D50 Cut Point in Micrometers vs. Cubic Feet
per Minute for Cyclone Used in Sample Train 101
9-2 Baffled Section of Quench Tower 107
9-3 Details of Baffles 108
9-4 Open Area Between Baffles 109
9-5 Baffled Section Showing Missing Boards 110
10-1 Water Balance 117
11-1 Graph of Total PAH versus Coke Greenness 134
-11-
-------
LIST OF TABLES
Number Page
1-1 Summary of Coke Quench Tower 3
Organic Emissions
1-2 Summary of Coke Quench Tower
Particulate Emissions 8
5-1 Average Oven Temperatures (G Battery)
During Test Period 16
6-1 Index of Photographic Slide Number
and Observers Greenness Ratings
and Comments on a Series of Pushes 32
6-2 Coke Quench Tower Emissions Tests 35
6-3 Summary of Measurements - Coke
Quench Tower Emission Tests 37
6-4 Potential Problems and Solutions in
Testing Quench Tower Emissions 39
6-5 Isokinetics 47
6-6 Fugitive Emissions • 50
6-7 Probable Analytical Errors 57
7-1 Instrument Conditions for T.C.O.
Analysis 61
7-2 Operating Conditions for GC/MS
Analyses 65
7-3 Ames Mutagenicity Assay Test Conditions 69
8-1 Organic Emissions Summary 73
8-2 Polycyclic Aromatic Hydrocarbons (yg/m ) 75
8-3 Polycyclic Aromatic Hydrocarbons
(grams/metric ton of coal) 76
8-4 Polar Fraction Cng/m ) 77
8-5 Polar Fraction (grams/metric ton of coal) 78
8-6 Organic Analysis of Train Components
j79
-111-
-------
LIST OF TABLES CONT.
Number
Page
8-7 Organic Analysis of Train Components
(grams/metric ton of coal) 80
8-8 Benzo (a) pyrene in Coke Quench Tower
Emissions 81
8-9 Organic Species Found in Emissions at
Levels Potentially Harmful to Health 82
8-10 Presence of Silicone Grease and
Phthalates as Indicated by IR 84
8-11 Presence of Silicone Grease and
Phthalates as Indicated by IR
(sampling train components) 85
8-12 Benzene Soluble Residue 87
8-13 Benzene and THC Emissions 88
8-14 Results of Mutagenicity and Toxicity Assays
of a Coke Quench Tower Emission Sample 91
8-15 Results of the CHO Clonal Cytotoxicity
Assay of a Coke Quench Tower Emission
Sample:Colony Counts and Percent Relative
Survival 92
9-1 Coke Quench Tower Particulate Emissions 95
9-2 Benzene Soluble Residue Tests - Total
Particulate 96
9-3 Particulate Emission Summary 97
9-4 Cut Size of "In Stack" Cyclone at Test
Conditions 99
9-5 Quench Tower Particulate Emissions, U.S.
Steel Corporation Lorain Works, Quench
Tower No. 1 - November, 1976 (1) 103
9-6 Quench Tower Particulate Emissions, Dom-
inion Foundries and Steel, Ltd., August 1977 105
9-7 Baffle Open Area HI
10-1 Quench Water Analysis (mg/1 x 10~3) 113
10-2 Quench Water Analysis (grams/metric ton
of coal)
-iv-
-------
LIST OF TABLES CONT.
Number
10-3 Results of T.O.C. Analysis of Water
Samples 115
10-4 Water Flow (gals/quench) 116
10-5 Lip-Stack Water Flow as Extrapolated
from Cyclone, Condenser and Impinger
Catch 120
10-6 Mass Balance of Organic Compounds 122
11-1 Analysis of Variance - F Values (22) 126
11-2 Simple Correlations (using water quality,
coke greenness, individual organic com-
pounds, and total PAH values for all tests) 129
11-3 Simple Correlations (using coke green-
ness, individual organics, and total PAH
values for clean water tests only) 131
-v-
-------
1.0 SUMMARY
York Research Corporation was contracted by the Environmental
Protection Agency to conduct a study of coke quench tower
organic emissions. Testing was conducted at quench tower No. 1,
U.S. Steel's Lorain Works, Lorain, Ohio in November, 1977.
The primary objectives of the study were to:
• Characterize and quantify the organic matter in
coke quench tower emissions
• Identify the possible origins of organic emissions
by evaluating the effects of various process condi-
tions (quench water and coke) on organic emissions.
Stack samples were collected by modified EPA Method 5 sampling
methods and subjected to extensive organic chemical analysis
for identification and quantification of individual organic
compounds. Sufficient samples were taken under controlled
process conditions to provide a statistically confident basis
for emission factor determination. The process conditions under
consideration were clean quench water, contaminated quench water
(flushing liquor and blowdown from other plant processes), green
coke (not fully distilled) and nongreen coke.
Supplemental objectives of the test program were the following:
• Determine the total organic carbon content of
the quench water in order to study the mass balance
around the quench tower.
• Determine the amount of benzene soluble material
in coke quench tower emissions.
• Measure benzene and total hydrocarbons in the stack
aerosol (2 samples).
• Obtain a stack sample for bioassay.
1.1 Organic Emissions
Fifteen tests were performed for the determination of quench
tower organic emissions: six with clean water and nongreen
-1-
-------
coke, five with clean water and green coke, and four with
contaminated water and nongreen coke. The test results are
summarized in Table 1-1, showing the large quantities of Poly-
cyclic Aromatic Hydrocarbons (PAH) and certain polar compounds
(heterocyclic oxygen and nitrogen compounds) found in the quench
tower plume. Together, these organic classes represent fifty-
three different organic species detected in emissions samples.
Those species found at the highest concentrations include:
1) Naphthalene
2) Methyl Naphthalenes
3) Acenaphthylene/Biphenylene
4) Dimethyl Naphthalenes
5) Fluorene
6) Dibenzofuran/Methyl Biphenyl
7) Anthracene/Phenanthrene
8) Methyl Anthracenes
9) Phenol
10) Cresol
11) Methyl Cresol
12) Quinoline
It may be observed in Table 1-1 that the concentration of organics
is much greater in the contaminated water tests than in the clean
water tests. Also, the clean water - green coke tests show
higher organic concentrations than the clean water - nongreen
coke tests.
A specific analysis for benzo (a) pyrene (BaP) in the organic
emission samples was performed and revealed substantial amounts
of BaP to be present (see Table 1-1). Although BaP levels are
consistently higher in the contaminated water tests than in the
clean water tests, the data were scattered in regards to coke
greenness. BaP was not detected in every test but when it was
its concentration exceeded the Minimum Acute Toxicity Effluent
(16)
values.
-2-
-------
TABLE 1-1
SUMMARY OF COKE QUENCH TOWER ORGANIC EMISSIONS
PAH
All Tests
Green Coke Tests
Clean Water
ug/m'
767C
g/metric ton of coal
0.758
(170 to 1,900)" (0.156 to 1.90)
1160
(660-1900)
1.17
(0.679 to 1.90)
Contaminated Water
ug/m
31,000
(13,700 to 47,300)
g/metric ton of coal
32.1
(15.5 to 46.4)
Nongreen Coke Tests
442
(170 to 745)
0.419
(0.156 to 0.646)
31,000
(13,700 to 47,300)
32.1
15.5 to 46.4)
I
U)
I
Polar Compounds Total
All Tests
Green Coke Tests
Nongreen Coke Tests
771
(60 to 6090)
1540
(118 to 6090)
134
(60 to 253)
1.10
(0.0575 to 6.89)
1.70
(0.119 to 6.89)
0.606
(0.0575 to 3.13)
540,000 581
(243,000 to 922,000) (185 to 1009)
540,000 581
(243,000 to 922,000) (185 to 1009)
Denzo (a) Pyrene
All tests
Green Coke Tests
Nongreen Coke Tests
e 19
(ND to 66)
13
(NO to 66)
24
(ND to 66)
(ND
(ND
(ND
0.019
to 0.072)
0.012
to 0.060)
0.024
to 0.072)
76
(36 to 99)
—
76
(36 to 99)
0.081
(0.040 to 0.12)
0.081
(0.040 to 0.12)
a Average value.
b (ttange)
c Pht
d Results from an analysis specific for benzo (a) pyrene which is separate from the PAH analysis above.
e ND = Not Detected.
-------
Each of three organic emission tests were analyzed according
to three separate sections of the sampling train. These results
indicated that most of the PAH found for each test was in the
organic adsorber unit, which means this material is either
vaporous or associated with particles smaller than 0.3 micro-
meters .
A comparison of the organic emissions data with various rating
systems showed that the following compounds considered to be
toxic, hazardous, or carcinogenic (.17,18) are present in the
quench tower plume:
1) Benzo (a) pyrene
2) 3-Methyl cholanthrene
3) 7, 12-Dimethyl benz (.a) anthracene
4) Dibenz (a,h) anthracene
5) Dibenzo (a,h) pyrene
6) Dibenzo (a,i) pyrene
7) Benz (a) anthracene (s)
8) Pyridine
9) Indeno (1,2,3-cd) pyrene
10) Phenanthrene
11) Phenol
12) Cresol
13) Quinoline
Silicone grease and phthalates were detected in the organic
emissions samples in large amounts. These substances are con-
sidered to be sample contaminants but their origin and method of
entry into the test samples is not well defined. Silicone grease
was used to seal some glass connections in the back half of the
sampling train, however, its appearance in a front half sample
(cyclone) and in some water samples and blanks indicates that
another source may have been present. Phthalates were found in
all of the organic emissions tests and in consistently greater
quantities in contaminated water tests than in clean water tests.
Phthalates have also been found in other tests of quench tower
-4-
-------
emissions and coke oven door leaks. ' ' Due to these findings
it has been speculated that phthalates could originate from coke
oven processes.
1.2 Organics in Quench Water
Quench water samples were found to contain many of the same organic
species present in quench tower emissions. It was observed that as
the molecular weights of the species increases, the ratio of the
quantity of a species in the stack emission to the quantity intro-
duced in the makeup liquor becomes less and less. The lower the
molecular weight the higher the boiling point, therefore, the lower
molecular weight species are readily stripped from the quench water
by the evaporation and distillation process of the quench. In a
similar manner, the contaminated makeup liquor is stripped of lower
molecular weight species and the higher molecular weight species
tend to remain in the quench water. A few of the higher molecular
weight compounds do not even appear in the stack emissions, however,
almost all of the other species are present in the stack emissions
in equivalent or larger quantities than found in the makeup water.
Samples of clean and contaminated quench water were taken and
analyzed for Total Organic Carbon (T.O.C.). A mass balance around
the coke quench tower reveals that of those individual organic com-
pounds which were measured there were greater quantities found in
the stack emissions than came in via the makeup water. Coke is
suspected as an additional source of these organic emissions.
1.3 The Source of Quench Tower Organic Emissions
A statistical analysis of the organic emissions data showed
that there is a significant realtionship between the three
different test conditions of: clean water and green coke,
clean water and nongreen coke, and contaminated water and non-
green coke; and the concentration of total PAH and most individual
organic compounds in coke quench tower emissions. The concentra-
tion of organics in the quench tower plume increases when contam-
inated water is used to quench. The organic emissions also in-
crease with the quenching of greener coke. The effect of
-5-
-------
quench water quality is much greater than the effect of coke
greenness.
1.4 Benzene Soluble Residue, Benzene, and Total Hydrocarbons
Benzene soluble residue in coke quench tower emissions was
determined for three tests. A substantial amount of benzene
solubles (at least 234 grams per metric ton of coal) was found,
with the greater quantity collected in the adsorber.
Two grab samples of quench tower emissions were taken and
analyzed for benzene and Total Hydrocarbons (THC). With clean
quench water in use, 0.005 ppm and 0.040 ppm benzene and 8.54
ppm and 17.34 ppm THC were detected.
1.5 Biological Test
A high volume sample of coke quench tower emissions was taken
with clean quench water in use. This sample was subjected to
several bioassays by Litton Bionetics. The Ames bacterial assay
was run with and without metabolic activation using rat liver
extract and showed the quench tower emission sample was not
mutagenic under these conditions. The results of toxicity
tests employing the same bacterial strains were also negative.
The clonal cytotoxicity assay was performed on the emission
sample in order to determine the sample's potential cytoxicity
through its effect on the colony forming ability of cultured
Chinese hamster cells (CHO). A sample concentration between 74
and 100 ul/nil reduced the number of colonies by 50% (EC50 value) .
Compared with a standard range for low toxicity of 60 to 600
Ul/ml the quench tower emission sample was determined to be of
low toxicity.
1.6 Particulate Emissions
The organic emission tests and benzene soluble residue tests were
-6-
-------
also analyzed for particulate. These test results are summarized
in Table 1-2. There was a striking reduction in the quantity
of particulate emissions when the switch was made from contam-
inated to clean quench water. As shown in Table 1-2, the front
half emission was reduced from 1.1 kg per metric ton of coal
during the contaminated water tests to 0.68 kg per metric ton
of coal during the clean water tests.
Although carloads of green coke are smoky and emit much parti-
culate, there seems to be no correlation of coke greenness to
the concentration of particulate in the stack aerosol. One
possible explanation for this dramatic reduction in quench tower
emissions from green cars is the scrubbing mechanism taking place
when the top layers of coke are cooled and then wetted by the
continued spraying of the quench water. This reduction in plume
particulate minimizes the effect of coke greenness on stack
emissions.
The weight of larger particles collected by the cyclone is about
45% of the total weight emitted when quenching with clean water
and about 75% of the contaminated water quench emission. The
average size of the particulate in the plume aerosol was less
than 4 micrometers in previous tests at Lorain^ but this would
be shifted towards a larger size in the present study due to the
greater cyclone collection.
The particulate results for the benzene soluble residue tests
are shown in Table 1-2 to be only half that found for the front
half in the organic emission tests.
-7-
-------
TABLE 1-2
SUMMARY OF COKE QUENCH TOWER PARTICULATE EMISSIONS
•(in kg/metric ton of coal)
Cyclone
Probe/Nozzle
Filter
Total Front
Half
Organic Emission Tests
CLEAN QUENCH WATER
Nongreen
Coke
0.32
0.15
0.32
Green
Coke
0.29
0.039
0.29
Nongreen
Coke
0.89
0.10
0.16
0.79
0.62
CONTAMINATED QUENCH WATER
Green
Coke
1.1
Cyclone
Probe/Nozzle
Filter
Total Front
Half
Total Clean Tests
0.30
. 0.075
0.30
0.68
Total Contaminated Tests
0.89
0.10
0.16
1.1
Benzene Soluble Residue Tests
Clean Water-Green Coke
Cyclone
Probe/Nozzle
Filter
Subtotal-Front Half
0.26
0.042
0.0044
0.31
Contaminated Water
Nongreen Coke
0.42
0.033
0.038
0.49
Total (including front
half,condenser,
adsorber)
0.52
0.99
-8-
-------
2.0 CONCLUSIONS
The organic emissions from the coke quench tower tested were
found to be substantial. Polycyclic aromatic hydrocarbon
emissions ranged from 0.15 to 46 grams per metric ton of coal
and polar compound totals ranged from 0.057 to 1000 grams per
metric ton of coal. Most of these organics are either vaporous
or associated with particles smaller than 0.3 micrometers.
Bioassays of one coke quench tower emission sample were negative
for mutagenicity and toxicity, and showed the sample to be of
low cytotoxicity. However, out of fifty-three different organic
species detected in quench tower emissions, thirteen have been
designated as either toxic, hazardous, or carcinogenic.(17'18)
Among these potentially harmful species is benzo (a) pyrene,
which although not detected in all tests, was found to exceed
Minimum Acute Toxicity Effluent values for the concentrations
measured.
The process conditions of quench water quality and coke greenness
have a definite effect on organic emissions from the quench
tower. The use of contaminated quench water rather than clean
water increases the average PAH concentration 40 times while the
average concentration of polar materials increases 500 times.
The quenching of green coke rather than nongreen coke increases
both PAH and polar compound emissions by a factor of 3. The
quality of the quench water has a decidedly greater effect on
the quantity of organic emissions than does coke greenness.
These conclusions are supported by statistical analysis of the
data.
Particulate emissions from the coke quench tower are also sub-
stantial, ranging from 290 to 1220 grams per metric ton of coal.
Again, quench water quality has a great effect on the quantity
of particulate emitted, with an average of 680 grams per metric
ton of coal for clean water increasing to an average of 1200
grams per metric ton of coal for contaminated water. The quality
of the coke being quenched did not appear to have any effect on
the amount of particulate emissions.
-9-
-------
3.0 RECOMMENDATIONS
Large quantities of silicone grease and phthalates were found
in the coke quench tower emission samples and the presence of
these substances interfered with the determination of total
organics. In addition, there is much uncertainty as to the
origin of these compounds. For these reasons it is recommended
that all sources of phthalates and silicone grease in the coking
and quenching processes be determined prior to another test
program for quench tower organic emissions. It is also recom-
mended that all use of silicone grease on sampling equipment be
avoided in similar tests and that other methods (i.e. teflon
sleeves) be used to obtain tight seals between train components.
-10-
-------
4.0 INTRODUCTION
It has been a common practice in the U.S. steel industry to use
contaminated water in the quenching of incandescent coke, that
is, water contaminated by prior use in either quenching or other
processes within the plant. Typically, 10-20 ton loads of hot
coke (upwards of 2000°F) are quenched by 6,000 to 12,000 gallons
of water. Each quench takes 2-3 minutes and produces huge
billowing clouds of steam, water droplets, and air contaminants.
In order to draft these emissions out of the work area quenching
takes place under towers which are open at the bottom to admit
the coke car. Baffles are often fitted inside the towers in
order to reduce the amount of large diameter particles emitted,
and at the same time reduce water losses and thus the amount of
makeup water required.
It had been theorized that significant amounts of coke quench
water contaminants are transformed by the quenching process
into air pollutants which pass through baffles and enter the
(4)
atmosphere. Studies performed by York Research Corporation
(YRC) quantified particulate emissions under varied conditions
of quench water quality and also found certain gaseous emissions
and organic material to be present in the stack aerosol. How-
ever, major consideration had not been given to these organic
emissions or to the effect of the varied process condition of
coke greenness. Thus, York Research Corporation was contracted
by the U.S. Environmental Protection Agency to further evaluate
the coke quench operation particularly concerning organic emissions
and identification of the source of these pollutants.
The site selected for the coke quench tower emissions testing
program was U.S. Steel Company's Lorain Works, Lorain, Ohio, a
fully integrated steel plant producing finished and semi-finished
steel products. Since the previous quench tower work performed
by YRC had been done at this site a broad data base for particu-
late and gaseous emissions, quench water flow rates and water
-11-
-------
contaminants already existed for use in support of a new test
series. A wide range of coke greenness had been observed at
this plant. In addition, a sampling methodology devised to
handle the difficult test conditions presented by the quench
had been successfully used in YRC's first test program. The
coke quench tower emission study was conducted in November of
1977.
-12-
-------
5.0 PROCESS DESCRIPTION
The integrated steel mill manufactures coke to be used as a
basic raw material for the blast furnace. There are two
generally accepted methods for manufacturing coke in this
country. These are known as the beehive process and the by-
product or chemical recovery process. This latter process
produces about 99 percent of all metallurgical coke and is
the process used at Lorain.
In the by-product coke manufacturing process, bituminous coal
is heated in an oxygen-deficient atmosphere (coke oven) and
volatile components of the coal are driven off. At the com-
pletion of this process the residue remaining in the oven is
coke, the volatile components having been recovered and pro-
cessed in the by-product plant to produce gas, tar, light oils,
and other materials. The coke is removed from the oven and
cooled by water sprays, after which it is stored for eventual
use in the blast furnace.
Much of the following process description was taken from the
1976 study at Lorain.(1)
5.1 The Coking Process
The coking process is accomplished in narrow, rectangular,
silica brick ovens which are arranged side-by-side in groups
called batteries. At Lorain, G battery provided the coke
quenched at tower No. 1, the test site.
The regular coal handling system is located between I and J
batteries (see Figure 5-1). During the test period, H battery
was being reconstructed and G battery could not be supplied
with coal in the usual manner. Instead, coal was trucked to
a nearby area and was supplied to G battery from ground level
by a conveyor feeding the charge car.
-13-
-------
FIGURE 5-1
COKE OVEN PLANT SCHEMATIC
COKE OVEN BATTERIES
G
H
I
COKE OVEN BATTERIES
J
K
L
Quench
Tower #1
Quench Tower #2 #3
Quench
Tower .#5
Quench Tower #4
-------
In order to charge the ovens, coal is loaded into a hopper
car. This car holds approximately 13.1 tons of coal in three
hoppers. At Lorain, there were scales for weighing the coal,
but they were not used. After loading, the hopper car moves
along the top of a battery where the coal is dumped through
three ports in the oven top. This coal fills the oven to a
point level with the top of the leveling bar door (oven density:
46-50 pounds per cubic foot). From 13.1 tons of coal per charge
each oven yields:
9.6 tons of coke oven product
- 1.0 tons of breeze
Total = 8.6 tons furnace coke
The coke ovens at Lorain are 40 feet 1-3/4 inches long, 10 feet
1-7/8 inches high, 21 inches wide at the car end and 17 inches
wide at the pusher end, averaging 19 inches in width. The
working volume below the leveling bar is 560 cubic feet.
There are 413 coke ovens in use in seven batteries, each
battery consisting of 59 ovens. G battery was manufactured
by Koppers and built in 1956. It was shut down temporarily
in 1976 and its present condition is "rebuilt". The ovens are
normally operated for three shifts per day, seven days per week,
and convert upwards of 7,000 tons of coal per day to coke. Normal
coking time is 17.3 hours. However, this may vary with the
condition of the oven.
Typical oven temperatures were obtained from production records
and some daily averages during the test period are listed in
Table 5-1. The average coke oven temperature for G battery
during the test period was 2338 F.
-15-
-------
TABLE 5-1
AVERAGE OVEN TEMPERATURES (G BATTERY) DURING TEST PERIOD
Average Oven Temperature
Date (G Battery) °F
11/1/77 2337
11/2/77 2338
11/5/77 2352
11/6/77 2389
11/7/77 2360
11/8/77 2352
11/9/77 2346
11/10/77 2283
11/11/77 2234
11/15/77 2338
11/17/77 2286
11/19/77 2390
Average for test period 2338°F
The coke ovens are separated from each other by a space between
the walls of the adjacent ovens. Gas is burned in these spaces
to provide the heat necessary for the coking process. This gas
is about 40% of that produced from the coking process. The
balance of the gas available from coking is used elsewhere in
the steel plant. The combustion products from these inner
wall furnaces are drafted out of the area by tall stacks. The
ovens are operated under a very slight positive pressure so
that any leakage occurs from the oven to the atmosphere or into
the adjacent furnaces.
Three factors affecting the composition of the coke oven gases
are:
• Coking temperature (this mostly determines the
hydrogen to hydrocarbons ratio. The higher the
-16-
-------
temperature, the higher the hydrogen content
of the gas and the lower the hydrocarbon content).
• Coal compositions.
• The amounts of air and combustion products drawn
into the coke oven. Oxygen reacts with the gas
and coke to form more C02, CO, and N. (In the case
of Lorain the oven was under positive pressure and
the effect was minimal.
During the total coking period, gases and other volatile
material from the coking process are bled off the top of each
oven through ascension pipes into the collection main which
runs the length of the battery. Where the gas main leaves
the battery a back pressure valve maintains a positive pressure
on the oven side against the negative pressure developed by
the exhaust fans. The hot gases from the ovens are cooled to
below 212°F with sprays of flushing liquor (blowdown from plant
processes) which are then decanted to remove the crude tar.
After the decanter the flushing liquor is recycled. Excess
flushing liquor is stored and eventually used at the quench
tower. (See Figure 5-2 for a diagram of the contaminated
quench water system.) The partially cooled gases are further
reduced in temperature (to 110°F) by primary coolers which
utilize service water (river water) for cooling the gas with-
out direct contact between the cooling water and the chemicals
in the gas stream. The gases pass through exhaust turbines
and electrostatic precipitators to remove entrained tar. They
then pass to an ammonia scrubber, a final cooler, a light oil
scrubber, and are finally burned either in the coke plant or
elsewhere in the steel mill.
This project did not include any investigation of the operations
of the chemical by-products processes. However, since the make-
up water for the quench tower sump included flushing liquors
and other excess wash waters from the by-products process, it
is appropriate to include a brief description of the chemical
recovery operations.
-17-
-------
FIGURE 5-2
DIAGRAM OF LORAIN COKE PLANT CONTAMINATED QUENCH WATER SYSTEM
FLUSHING
LIQUOR
SYSTEM
TAR DECANTER
PRIMARY COOLER
TAR FLUSH
(RECYCLE STREAM)
FLUSHING
LIQUOR
SYSTEM
FINAL COOLER
NAPHTHALENE
SKIMMER
RIVER WATER
54,000 GPD
99,000 GPD RIVER WATER
393,000 GPD
TO SULFATE PLANT
BAROMETRIC CONDENSER
495, OC
600,0
BLOWDOWN FRC
SULFATE PLA1,
.g GPD
)0 GPD
105,000
-4GPD *
00 GPD
» ti
T
^ RIVER WAI
PRIMARY LIGHT-
OIL DECANTER
BLOWDOWN FROM WASH
OIL DECANTER
SECONDARY LIGHT-
OIL DECANTER
TO QUENCH TOWER SOS. 1,2,3,4
(QUENCH TOWER NO. 5 ON D-BATTERY USES
ONLY RIVER WATER FOR QUENCHING)
-18-
-------
Ammonia is recovered as an ammonium sulfate. Crude light
oil and crude coal tar are also recovered. No tar distilla-
tion, phenol recovery or light oil purification is employed.
All of the excess flushing liquors and blowdown from the
ammonium sulfate plant are used to quench the coke produced
in batteries G, H, I, J, K and L. D battery is normally
quenched with river water. Figures 5-2 and 5-3 illustrate
the various operations in the chemical by-product plant.
5.2 The Quenching Process
At the end of the coking period the doors are removed from
each end of an oven and the pushing machine pushes the in-
candescent coke into the quenching car. The quench car is
then moved by a small electric engine to the closest quench
tower. (See Figures 5-4 and 5-5). At this point the car
operator pulls a switch which activates a valve in the 16 inch
pipe leading from the head tank to the spray nozzles. After
a safety delay of about 15 seconds, 7-9 thousand gallons of
water flow through a header and 10 nozzles onto the incandescent
coke. Figures 5-6, 5-7, and 5-3 show photographs taken a few
seconds apart through the first 20-30 seconds of the quench.
The hot coke is quickly cooled to about 250 F by the evapora-
tion of about 20 percent of this water. Violent jets of super-
heated steam result from this process; and the steam, being
less dense than the surrounding air, flows up the tower inducing
a flow of cooler air through the car ports. Figures 5-9, 5-10,
and 5-11 show the first eruption of steam, while 5-12 and 5-13
show the tower (stack) flow well established. During the rapid
cooling, some of the coke fractures, and small particles are
broken off and carried away by the gas and water stream. Some
of the particles are ejected from the car as if by an explosion.
The quench towers are built to contain the violence and turbu-
lence of this cooling process, and the geometry of the tower
directs most of this hot mixture of steam, air, water droplets,
-19-
-------
WEAK LIQUOR
SERVICE
WATER
FLUSHING
LIQUOR
BATTERIES
D,G,H,I,J,K,L
SERVICE
WATER
EVAPORATOR
OUTFALL
NO. 002
AMMONIA UACID
RECYCLED
COOLING
WATER ,
L.O. VAPOR \
CONDENSATE
OUTFALL NO. 002
T
CRUDE L.O.
FOR SHIPMENT
SERVICE
WATER
DE-BENZOLIZED
SERVICE I
WATERJ .CONDENSED
OUTFALL NO. 002 <:
SERVICE AND PROCESS
WATER COKE PLANT
LORAIN WORKS
-20-
-------
FIGURE 5-4
THE PUSH OF INCANDESCENT COKE
FROM THE OVEN TO THE QUENCH CAR
FIGURE 5-5
QUENCH CAR WITH INCANDESCENT
COKE ENTERING THE QUENCH TOWER
-21-
-------
FIGURE 5-6
"CAR IN"
FIGURE 5-7
"WATER ON"
-22-
-------
FIGURE 5-8
START OF QUENCH WATER FLOW
FIGURE 5-9
FUGITIVE EMISSIONS A FEW SECONDS
BEFORE THE UP-STACK DRAFT IS WELL ESTABLISHED
-23-
-------
FIGURE 5-10
A VIEW SHOWING THE FIRST ERUPTION OF STEAM
FIGURE 5-11
ANOTHER VIEW SHOWING THE FIRST ERUPTION OF STEAM
-24-
-------
FIGURE 5-12
QUENCH TOWER PLUME
FIGURE 5-13
THE STACK FLOW WELL ESTABLISHED
ABOUT 10 SECONDS AFTER "WATER ON"
-25-
-------
and coke particles upwards so that it can be partially removed
from the immediate working area, and so that as much of the
cooling water as possible may be recovered.
A small amount of water is mechanically entrained as droplets
and carried upward by the jets of steam and air leaving the
car. Most of the large droplets are eliminated by a single
row of baffles. In 1976 it was concluded that the baffles in
the quench tower at Lorain had no significant effect on water
droplets below about 40 micrometers, nor on particulate matter
below about 30 micrometers. The average size of the aerosol
which passed through the baffles and exited the stack was about
3 micrometers. In 1977, 10% of the tower cross section was
unbaffled open area. Apparently, more of the wooden baffles
had burned out because only 6% was unbaffled in 1976.
About four percent of the total water flow continues up the
stack as droplets, unhindered by baffles. The balance of the
water drains from the car and returns to a storage sump. The
sump is designed to collect the quench water for re-use, while
allowing the settling of coke particles. This settling action
results in a layer of coke sediment in the bottom of the sump.
This sediment is cleaned out and hauled away whenever a truck
load or two accumulates.
In addition to the physical processes of evaporation, and the
cooling and fracturing of the coke, chemical reactions are
also ongoing. For example, the coke and water react in the
reducing atmosphere to form hydrogen and carbon monoxide.
Some of the other components of the quench water (such as
ammonia) dissociate to form hydrogen. Additionally, certain
components of the quench water, including ammonia, phenol,
and cyanide, evaporate from the sump. These substances are
also found in the stack gases.
The cooled coke is dumped on an inclined area, called the
wharf, which feeds a conveyor belt where it is transferred
-26-
-------
for storage and eventual use in the steel making process.
During normal operation, quench towers No. 1 and 2 are used
for G, H, and I batteries; towers No. 3 and 4 are used for
J, K, and L batteries; and tower No. 5 is used for D battery.
The hot coke is taken to the nearest tower, with a quench
occurring every 10-12 minutes (16-18 minutes in the case of
D battery).
All testing was performed on quench tower No. 1. This tower
was used only for quenching the pushes from G battery because
H battery was being rebuilt and construction blocked the
quench car tracks at the G & H boundary.
The operation of the coke ovens at Lorain appears to be typical
of the industry so far as operating temperature, coking time,
and oven size are concerned. Quench tower emissions would also
be expected to be typical, except for the factors of inefficient
baffles (already described) and coke greenness. YRC has no
broad data base to use for comparing the number of green pushes
and is not aware of any such study having been previously made.
-27-
-------
6.0 TEST PROCEDURES
6.1 Test Program
The test site was the number one quench tower at Lorain, des-
cribed in Figure 6-1. The plan of the area around the quench
tower is shown in Figure 6-2.
In November 1977, eighteen tests were performed to sample coke
quench tower emissions. Emission tests were conducted from a
test platform surrounding the tower, 95 feet above the ground
(See Figure 6-1.) The operation of quench tower No. 1 and the
coke ovens using it presented much the same sampling situation
as experienced at that tower in the 1976 tests. Therefore,
planned sampling train modifications enabled isokinetic sampling.
The quench tower was placed on clean water and the sump dredged.
Conductivity measurements were taken as follows:
October 27 1400 micromhos/cm
October 28 1100 micromhos/cm
October 30 980 micromhos/cm
These measurements are close to those of 855 micromhos/cm and
1450 micromhos/cm for clean sumps in 1976. Although conductivity
has no direct relationship to organics in the water, it indicates
here whether flushing liquor (containing much dissolved solids
and organics) or river water was present. Thus, clean quench
water was in use for the tests beginning November 3, 1977.
Later, on November 10, 1977, the makeup water was switched to
flushing liquor. After five days conductivity measurements
showed the sump water to be at the same level of contamination
as in the 1976 contaminated water tests, and testing was resumed.
The other process condition, coke greenness, was determined for
each quench tested. When the oven doors are opened and the coke
is pushed into the waiting car there is a glow from the incandes-
cent coke and varying amounts of flame. If the volatile matter
has been completely removed there will be very little smoke
-28-
-------
© J I
©
(D
S
'-'
.
©
INCANDESCENT CCKE
QUENCHED COKE
EXHAUST GASES
CONTAMI.NATEC WATES (FLUSHING LIQUOR MAKE-UP)
SERVICE WA7E3 (CLEAN l^AKE-UP)
HEAD TAMX - OVERFLOW
MEAC TANK- STA.NO .°!PE
NOZZLE HEADER(INLET WATER)
NOZZLE HEACES DRAIN
EMISSIONS TESTING STATION
SUMP
RETURN DRAIN OITCH
2A.-FLE3
Quench Tower
FIGURE 6-1
-29-
-------
Figure 6-2
o
I
TEST PLATFORM-
STACK-
SUMP
COKE BATTERIES
COKE CAR TRACKS
PUMP
HOUSE
OBSERVER_
STATION"
TEMPORARY.
WHARF
TRAILER
-------
(nongreen coke). Occasionally, however, some of the volatile
matter may not be distilled from the coal during the coking
process. As a result quantities of gray or black smoke may
arise from the coke as it is pushed from the oven. As the car
is moved towards the quench tower changing amounts of these
dark colored volatile materials arise from certain sections of
the car (green coke). On such occasions, after the car is in
the quench tower there is a discoloration of the heated air
exiting from the stack before the steam plume starts. Each
quench car was rated by two observers stationed on the pumphouse
roof (see Figure 6-2) just before the car entered the quench
tower.
The density of the smoke and the area of the coke's surface
emitting smoke were both visually evaluated in accordance with
an arbitrary scale from 0 to 5. A rating of 5 designated a
very smoky push usually with flames and a rating of 0 designated
no smoke and no flames. A reading of 0, 1, or 2 indicated non-
green coke and one of 3, 4, or 5 indicated green coke. The
pushes that became part of one test were either all green or
all nongreen. If there was disagreement between the observers
as far as green or nongreen ratings, the push was not tested.
A series of coke oven pushes were photographed to provide doc-
umentation of the coke greenness rating system. Table 6-1 pre-
sents the data recorded for these photographs and is an example
of the data recorded on coke greenness for each emission test.
In addition, photographs of coke greenness in Figures 6-3 to
6-6 give examples of greenness ratings of 0, 2, 3, and 4. This
rating procedure was also used in the 1976 study at Lorain and
resulted in very consistent ratings for the several observers.
A summary of the quench tower emission tests and corresponding
process conditions of water quality and coke quality is presented
in Table 6-2.
-31-
-------
TABLE 6-1
INDEX OF PHOTOGRAPHIC SLIDE NUMBER AND OBSERVERS
GREENNESS RATING AND COMMENTS ON A SERIES OF PUSHES
i
u>
Photo-
graphic
slide
number
1-2-3
4-5
6-7
8-9-10
11-12
13-14-15
16-17
18-19-20
21-22-23
NA
NA
NA
24-25-26
27-28
29-30-
31-32
NA
NA
NA
33-34
35-36-
37-38
39-40-41
42-43
44-45-46
47-48
49-50
51-52
53-54
Push
No.
1 NT*
2 NT
3 NT
4 NT
5 NT
6 NT
1 V
2 NT
3 T
4 T
5 NT
6 T
7 NT
1 T
2 T
3 NT
4 T
5 T
1 NT
2 NT
3 NT
4 NT
5 NT
6 NT
7 NT
Oven
No.
24
34
44
54
64
6
26
36
46
56
8
18
28
38
48
58
1
11
21
31
41
45
51
61
3
Greenness
(Note)
4
3
1
1
2
3
4
4
0
4
2
2
1
3
1
1
3
2
0
4
3
2
2
2
3
4
4
0
4
f
2
2
1
3
1
0
3
2
0
Coke Level
High
High
High
High
High
Very High
High
High
High
High
High
High
High
High
High
High
High
High
High
Flames
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
NO
NO
No
Yes
Yes
Yes
Yes
Yes
Smoke Color
dark gray
dark gray
gray
gray
light gray
dark gray
dark gray
dark gray
light gray
black
light gray
gray
light gray
gray
light gray
light gray
gray
gray
light gray
Note: Two observers rated the greenness of the car
All pushes were from G battery.
-------
PHOTOGRAPHS ILLUSTRATING COKE GREENNESS RATINGS
FIGURE 6-3
OVEN 3 PUSH 7 GREENNESS RATING OF 0
OVEN 1 PUSH 4
FIGURE 6-4
GREENNESS RATING OF 2
-33-
-------
PHOTOGRAPHS ILLUSTRATING COKE GREENNESS RATINGS
OVEN 34 PUSH 2
FIGURE 6-5
GREENNESS RATING OF 3
OVEN 28-3 PUSH 7
FIGURE 6-6
GREENNESS RATING OF 4
-34-
-------
TABLE 6-2
COKE QUENCH TOWER EMISSIONS TESTS
Clean Quench Water
Date
11/3/77
11/4/77
11/5/77
11/5/77
11/6/77
11/7/77
11/7/77
11/8/77
11/8/77
11/8/77
11/8/77
11/9/77
11/9/77
11/9/77
Test Type and No.
1 Benzene soluble
2 Benzene soluble
2B Organic emissions
3 Organic emissions
4 Organic emissions
5 Organic emissions
6 Organic emissions
7 Organic emissions
8 Organic emissions
9 Organic emissions
10 Organic emissions
11 Organic emissions
12 Organic emissions
Biological test
Average
Greenness
Rating
3.0
1.8
3.4
3.3
1.6
4.0
1.6
1.2
1.8
1.6
1.5
3.2
3.7
Coke
Greenness
G
NG
G
G
NG
G
NG
NG
NG
NG
NG
G
G
Contaminated Quench Water
Average
Greenness
Date
11/15/77
11/16/77
11/19/77
11/19/77
11/19/77
11/19/77
Test Type and No. Rating
13
14
15
16
17
18
Benzene
Organic
Organic
Organic
Organic
Organic
soluble
emissions
emissions
emissions
emissions
emissions
1.
1.
0.
0.
0.
5
5
5
3
5
Coke
Greenness
NG
NG
NG
NG
NG
-
NG = Nongreen (0-2 greeness rating); G
(3-5 greenness rating)
= Green
-35-
-------
The various parameters measured concurrent with the emission
tests are identified in Table 6-3, along with measurement and
analytical methods, and sampling location. Samples of quench
water were taken on November 9th, 16th, and 19th from the clean
water makeup and inlet (positions 5 and 7 on Figure 6-1) and
from the contaminated water makeup and inlet (positions 4 and
7). Water flow rates to the quench process were not measured
but were assumed to be the same as in the 1976 tests. An exam-
ination of the pumping and piping system revealed no significant
change and supported such an assumption.
6.2 Sampling Problems
Previous test techniques and methods used in attempts to charac-
terize air pollution emissions from quench towers were often
plagued with difficulties. These included the exclusion of
certain size ranges of particulate matter in particulate measure-
ments and limitations in obtaining measurements of important
process parameters. Early attempts at particle measurement using
greased plates and petri dishes succeeded in measuring only
large diameter, heavy droplet and/or solid particles. Further
attempts by State and local agencies to measure particulate
emissions utilizing standard EPA sampling equipment were limited
in their scope and also encountered a number of sampling diffi-
culties . (4 ' 5)
Major problems occurred because, in operation, the quench tower
generated short violent rushes of steam which hampered efforts
to accurately read and adjust the sampling instrumentation to
maintain isokinetic conditions. In this case, the EP~A guidelines
were not usable because they recommended measuring velocities
and making adjustments every 3 to 5 minutes. This time interval
was more than the total duration of a quench. The short duration
of the quenches made capturing the prescribed volume of gases
difficult. The square shape of most quench towers, and the use
of internal partitions compounded these problems by causing
uneven flows across the tower cross-section. In addition to
-36-
-------
TABLE 6-3 SUMMARY OF MEASUREMENTS-COKE QUENCH TOWER EMISSION TESTS
Measured
Parameter
Organic matter loading
95 clean water quenches
19 contaminated water
quenches
Benzene soluble matter
(26 quenches)
Mutagenicity and
cytotoxicity
u> Gaseous Hydrocarbons
71 (Benzene and THC)
Particulate
Size Distribution:
Particulate
Water Quality
Total solids
Total organic carbon
Greenness of coke
Oven coking time
Oven Heat temperature
No. of
Tests
15
2
n
29
140
140
140
Measurement
Method
EPA 5 Modified with
porous polymer
resin (XAD-2)
EPA 5 Modified
Hi Vol sample using
modified EPA 5 train
Grab sample
Component of modified
EPA 5 train
20 ym cyclone
Composite samples
taken from lines or
from sump
Visible Observation
(photo and movies)
Plant data
Plant data
Analytical
Method
IR, LRMS
Combined gas
chromotography/
mass spectrometry
EPA 5 Modified
Sampling
Location
(Figure 6-1)
(10)
(10)
Ames Mutagenicity Assay (10)
Clonal Cytotoxicity Assay
Gas Chromatograph
Weight collected
Weight collected
EPA Methods for
Chemical Analysis
of water and
wastes
(10)
(10)
Flushing Liquor (4)
Service water (5)
Inlet (.1)
Sump
Pump house
roof (8)
-------
the above problems, droplets in the exhaust stream plugged
filters, and made determinations of the molecular weight of
the gases nearly impossible.
Potential problems that had existed during earlier testing
attempts were overcome in 1976 by an extensive planning period
in which previous tests were scrutinized, test equipment was
researched, and engineering judgments were made as to the most
effective methods available for conducting the tests. In 1977,
the use of special equipment such as the organic adsorber unit
required further modification to the EPA Method 5 train.
Table 6-4 lists the major problems, their possible effects on
test accuracy, and the solutions that were utilized to overcome
these problems.
6.3 Sampling Equipment Design
Organic Characterization Sampling Train (OCST)
To determine organic emission rates, tests were conducted using
a high volume sampling train with an organic adsorbent unit
(Figure 6-7). This was equipment frequently used for EPA Method
5 but modified as described below.
A button hook sample nozzle of stainless steel (SS) was connected
by (SS) swagelock fittings to a special cyclone. This cyclone
was constructed of (SS) and was designed to have a 50% cut size
of approximately 20 micrometers at the flow rates used in the
tests. The 50% cut size (or D--) designates the size at which
the collection efficiency is 50 percent - where half the part-
icles encountered are captured and half escape. The cup of the
cyclone was fitted with a Viton gasket. The cyclone was connected
to the probe with (SS) swagelock fittings.
The 10 foot (SS) probe was heated to 250 F. The probe was con-
nected with (SS) swagelock fittings to a filter holder. The
filter holder was made of (SS) and coated inside with Teflon.
-38-
-------
TABLE 6-4 POTENTIAL PROBLEMS AND SOLUTIONS IN TESTING QUENCH TOWER EMISSIONS
Potential Problems
Anticipated small differential
in TOS between clean and con-
taminated quench water.
Tower Configuration
Short duration of quench re-
sulting in velocity profiles
changing faster than can be
measured or recorded by usual
methods.
Mater droplets in stack sample
Droplets and grit in stack
gases
Fugitive emissions
Sampling locations
Effect on Accuracy
A wide TDS range desired to
show effects on clean and
contaminated quench water
on emissions.
Square, squat, sectionalized
towers likely to produce
non-uniform flow.
Velocities move up and down
so rapidly as to make ac-
curacy difficult to achieve.
Inability to measure actual
moisture content, molecular
weight of stack gas, and
filter plugging.
Plugging of pitot tubes
Emissions exiting from en-
trances of tower causing
non-representative sampling.
Possible uneven velocity
patterns throughout stack.
Solution
At Lorain, the ratio of contam-
inated water TDS to clean water
TDS was on the order of 10:1.
Tall circular tower with no
internal buttressing was
selected.
Use of Hastings-Raydist velocity
meter with continuous recording.
Cyclone (50% cut size about 20
microns) was fitted to the front
of the probe.
Used Hastings Raydist velocity
meter with continuous purge.
Entrance emissions monitored and
testing halted when excessive.
Performed traverse of 12 points
velocity and selected position
Bl for coke quench tower emission
tests (18 tests).
Number of quenches
Skewed flow
Greenness of car varies
from push to push.
EPA Method 5 would require
sampling 20 quenches to get
a one hour test, however, this
would have resulted in plug-1
ging and other errors.
The vertical component of
velocity differs from the
linear velocity along a
skewed flow.
Unless greenness can be char-
acterized for each car its
effect on organic loading in
the quench effluent might
negate any attempt to deter-
mine origin(s) of organics.
The 1976 tests showed that four
(4) quenches would provide suffi-
cient weight of particulate
(over 100 mg) and volumes of dry
stack gas (over 15 cubic ft.)
using a high volume sampler for
EPA Method 5. However, additional
quenches were tested in 1977 be-
cause of the lower sample rates
required when the train included
an organic adsorbent unit.
Vane measurements were made to
determine probe positioning.
A rating system (developed in 1976)
was expanded and movies and photos
were taken to document a typical
series of observations.
-39-
-------
Figure 6-7
ORGANIC MATTER AND PARTICULATE SAMPLING TRAIN
USING MODIFIED EPA METHOD FIVE
STACK WALL
-FILTER
inn
HEATED PROBE
o
I
•CYCLONE FOR PARTICULATE
AND WATER DROPLET SIZING
HASTINGS
VELOCITY
METER &
RECORDER
Resin Cartridge
Condenser
Impinger train
IMPINGER TRAIN
VACUUM GUAGF:
COARSE ADJ. VALVE
ORIFICE P DRY GAS VACUUM
MAGNEHELIC METER PUMP
-------
Viton and Teflon gaskets were used to seal the two halves of
the filter holder. Spectrograde Type AE glass fiber filters
with an efficiency of 99.99% for particles larger than 0.3
micrometers were used. A (SS) braided line connected the
filter holder to a (SS) female impinger connector. This con-
nector was greased with silicone and placed into a glass im-
pinger. The glass impinger was in an ice bath where it was
cooled to approximately 68°F. Water and some organic material
were recovered in this condenser. The adsorbent cartridge
located after the condenser employed a porous polymer resin
(XAD-2) contained in a tube sealed at top and bottom. The unit
contained approximately 32 grams of XAD-2. A thermocouple was
placed in the inlet (SS) fitting to the adsorber to monitor
the temperature. The outlet fitting (glass) from the adsorber
was connected to a standard Lexan impinger unit. Two of these
impingers contained 250 ml of water, the third was empty and
the fourth contained 350 grams silica gel. Viton gaskets and
small amounts of silicone grease were used to form a vacuum
proof seal on the impingers. The use of silicone grease in
this location would not interfere with samples collected since
the djnpingers are downstream of all samples taken. However,
the silicone grease used on the condenser might introduce this
substance to the samples.
Velocity pressure ( A p) was measured with a Hastings-Raydist
meter which operates as follows. A continuous strip chart re-
corder was connected to a Hastings meter to record the velocity
head pressure ( A p) , and to aid in verifying any uneven flow
patterns. Purge gas (air) was injected into a pneumatic bridge
arrangement formed by the velocity transducer, manifold and
pitot tube. At zero velocity, the bridge was balanced so that
no flow occurred through the velocity transducer and purge gas
exhausted equally through both openings of the pitot tube.
As flow across the tip occurred, a differential pressure was
developed, unbalancing the bridge and causing a small amount of
-41-
-------
purge gas to flow through the transducer. The transducer mea-
sured the flow which was related to the main gas flow at the
tip of the pitot tube. Purge gas still exhausted through both
openings, but at slightly unequal rates.
The purge gas continually exhausted into the stack, thereby
preventing water and large particles from plugging the pitot
lines. This instrument allowed accurate measurement of flow
even though high particulate concentrations were present.
Benzene Soluble Sampling Train (BSST)
The train configuration, all train components, and sampling
methods were the same as those described for the OCST, except
for the sample recovery phase.
Biological Sampling Train (BST)
The primary function of this train was to collect a large sample
of gaseous emissions. This "hi-vol" sample was collected spec-
ifically for mutagenicity and cytotoxicity tests. This sampling
train was also identical to the OCST, except for the sample re-
covery phase.
6.4 Sampling Technique
In November 1977 several tests were performed prior to beginning
the full field test. Velocity profiles were obtained for 12
points. These preliminary profiles were analyzed for each one
second time interval for magnitude of A P. The A p's of each
interval were then added together and averaged by dividing the
time of each quench into the total. Visual comparison of over-
laid profiles showed that roughly the same pattern, time and
velocity heads were found for each quench, suggesting that vel-
ocity heads for a particular point in the stack could be pre-
dicted, and these 1977 profiles were similar to those obtained
in the 1976 tests.
-42-
-------
Analysis of the above preliminary velocity work indicated that
isokinetic sampling conditions could be maintained throughout
the quench process by running a velocity test during the first
quench in a series. This test could be quickly analyzed and
graphed as velocity versus elapsed time. Data from the velocity
profile were then utilized to determine:
• Velocity Pressure (A p)
• Velocity (feet per minute)
• Required nozzle size
• Predicted sampling rates
If the preliminary test indicated the presence of an erratic
flow pattern, the test was delayed until a predictable flow
was established. If stack flows were predictable, testing
could be continued and the sampling rates would constantly be
adjusted based upon the graph. Thus, the sampling flow rate
would increase and decrease coincidentally with the tower's
exhaust flow rate. This type of graphical analysis is illus-
trated in Figure 6-8. It will be noted that three sections of
the profile can be defined, namely:
• Ramp Up - From the time the coke car entered
the tower until the water hit the incandescent
coke.
• Plateau - The time during which the plume
velocity is either flat or gently rising to
a peak and then holding.
• Ramp Down - The period during which the velocity
of the plume starts to fall rapidly.
In order to set sampling rates from these Ap's, velocity equa-
tions were computed utilizing data obtained from the particulate
sampling in October, 1976, consisting of:
• Moisture data obtained in accordance with
EPA Method 4.
• Orsat samples obtained in accordance with
EPA Method 3 with the use of two condenser units
to trap entrained water droplets.
-43-
-------
FIGURE 6-8
VELOCITY PROFILF
Sampling
Rate ( ]
Stack
Velocity
10 SECONDS/CM
The Velocity Profile (%) represents the scale of the Hastings flow meter.
Full scale on this meter is 50%; the % readings correspond to velocity head
pressures (Ap) measured in inches of water. The conversion to Ap is dependent
upon the pre-calibration range determined for the Hastings meter.
-------
These data were verified during the first 1977 tests. In
addition to the velocity traverses, initial measurements were
performed to determine the maximum angle of flow in the tower.
Data from these findings along with the 1976 data indicated
that sampling should be conducted near 110 degrees measured
from the horizontal.
The requirements for the organic matter tests suggested that
both the clean and contaminated water tests including the green
coke versus the non-green coke tests be conducted at one point.
Based on this preliminary test information and after reviewing
the 1976 data it was decided to perform the organic matter tests
at point Bl (Figure 6-9) . In order to assure as large a sample
as possible each test sampled four to six quenches. In order
to reflect the total quench emissions, sampling began when the
quench car entered the tower and ended when it left.
6.5 Test Sequence
Since velocity results were reproducible within a time frame of
approximately one hour, profiles were taken preceding each test
run to determine sampling rates for the test. Sampling rates
for an average test profile are shown in Figure 6-8. The problem
was to maintain the correct sampling rate (A H) over the entire
quench period (from ear-in to car-out), allowing for the rise
and fall of both the velocity head (A p) and the stack tempera-
ture. For each quench test the sampling rate (AH) was gradually
increased from the signal ear-in to approximately 0.50 inches of
water over a thirty second period. This maximum sampling rate
was maintained until the signal water-off was given and a ramp-
down procedure was initiated to a pre-determined rate (approxi-
mately 0.25 inches of water) which was established by averaging
the ramp-down portion of the velocity profile. Sampling was
terminated when the car was completely out of the tower.
-45-
-------
FIGURE 6-9
SAMPLE POINT LOCATION
B
D
Pt.No.
1
2
3
4
5
6
7
8
9
C
Inches from stack wall
71.60
55.60
44.40
35.40
27.50
20.50
14.10
8.25
2.65
94.25
-46-
-------
If a major velocity variation occurred during sampling, (10-20%
change from velocity profile) a new sampling rate was determined
and implemented within 5-10 seconds.
The area under the profile curve was not used to determine sample
rates required during field tests. Instead, 6-second blocks were
averaged to establish the AP readings and, in turn, the necessary
isokinetic sampling rates ( AH). The averaging method was used
to reduce calculation time since a field computer was not avail-
able.
The techniques applied to measurement and plotting of velocity
profiles assured that 13 of the 17 tests were within the required
range for standard EPA Method 5 (- 10 percent of 100 percent
isokinetic). The tests listed below were not within this range:
TABLE 6-5
Test Number Below 100% Above 100%
9 138
11 86
14 70
17 89
Individual stack flows for the tests varied by 17% more than
the average and 11% less than the average. These variations
and the variation from quench to quench may have been due to
differences in:
• Moisture content of the stack gas
• Probe position
• Wind speed and direction
• Total tonnage of coke quenched
• Coke temperature
Test results are included in Appendix A. Once a successful
velocity profile had been run, 4 to 6 subsequent quenches were
sampled to make a test. The normal test sequence went as
follows:
-47-
-------
Incandescent coke was pushed from an oven into a quench car
which was driven down the tracks to quench tower No. 1. The
two observers on the pump house roof recorded their ratings of
coke greenness and if the ratings disagreed between green and
nongreen,. no sampling would be performed on that quench.
When the coke car entered the tower the observers notified the
sampling crew and the test began. When quench, water began to
spray from the nozzles the signal "water on" was given.
Fugitive emissions began to escape from the bottom of the tower.
During a period of approximately 25 seconds immediately after
the quench water started flowing there was more steam generated
than could be drafted up the tower. See Figures 6-10 and 6-11.
Frequency and duration of fugitive emissions from the eastern
opening of the quench tower were noted by one of the observers.
The quantity of steam involved in the fugitive emissions was
difficult to judge, but if the total fugitive emissions from
both portals was deemed excessive, the test was cancelled by
the project director. The fugitive emissions averaged 17% of
the total sample time as shown in Table 6-6.
The proper adjustments were made in the sampling rate from the
previous signal of car in to water off and car out. The total
sampling time was recorded by instrumentation installed in the
field trailer.
6.6 Coke Quench Tower Emission Tests
During several of the test runs with clean quench water, holes
were found in the center of the filters. It was concluded in
the 1976 test report that particulate matter caught on the
filter would be smaller than 5 pm, ranging downward toward sub-
micron sizes. During the clean water tests, there was a rela-
tively small quantity of particulate matter retained on the in-
tact portion of the filter, while during the contaminated water
-48-
-------
FIGURE 6-10
FUGITIVE EMISSIONS AT START OF QUENCH
FIGURE 6-11
FUGITIVE EMISSIONS A FEW SECONDS AFTER
BEGINNING OF QUENCH
-49-
-------
TABLE 6-6
FUGITIVE EMISSIONS
i
ui
o
t
Test
Number
1
1
1
1
1
2
2
2
2
2
2B
2B
2B
2B
2B
3
3
3
3
3
4
4
4
4
4
5
5
5
5
5
Fugitive
Duration
(Sec.)
15
13
7
10
19
29
40
25
42
37
15
16
19
13
15
15
13
16
15
16
22
36
18
17
20
22
21
25
38
29
Sampling
Duration
(Min.)
2.4
2.3
2.8
2.4
2.4
2.7
3.1
2.7
3.0
2.8
2.6
2.3
2.4
2.4
2.3
2.4
2.3
2.4
2.3
2.2
2.5
2.3
2.4
2.4
2.4
2.5
2.6
2.4
2.4
3.0
Test
Number
6
6
6
6
6
7
7
7
7
7
8
8
8
8
9
9
9
9
9
10
10
10
10
11
11
11
11
12
12
12
Fugitive
Duration
(Sec.)
17
21
17
15
33
18
30
38
17
16
22
19
17
20
NA
12
20
27
27
25
20
15
21
18
NA
7
25
7
11
27
Sampling
Duration
(Min.)
2.3
2.4
2.4
2.4
2.4
2.4
2.4
2.4
2.5
2.5
2.3
2.3
2.4
2.4
2.0*
2.5
2.7
2.4
2.5
2.3
2.4
2.3
2.5
2.4
2.6*
2.7
2.4
Test
Number
13
13
13
13
13
13
14
14
14
14
14
15
15
15
15
16
16
16
16
17
17
17
17
17
17
2.7 Fugitive
3 . 3 Duration
2 2
*• • *• •* M«>4. •! r*r*
Fugitive
Duration
(Sec.)
27
15
25
26
20
19
61
40
55
125
55
58
51
44
39
51
44
39
20
23
38
25
22
26
27
2175
Duration T
x 100 = 17%
T 11 *3 s*i y»l 4 tf\ 4- s*\
Sampling
Duration
(Min.)
2.4
2.7
2.4
2.4
2.4
2.4
2.4
2.6
3.7
2.4
2.5
2.4
2.4
2.5
2.4
2.4
2.4
2.5
2.4
2.6
2.4
3.0
2.6
2.5
2.7
208.0
Sampling
•t-^,1
-------
tests, a larger quantity was caught. This larger quantity of
particulate matter may have been initially deposited on the
filter directly in front of its inlet, thus protecting this area
from the continuing abrasive action of particles smaller than
0.3 ym as they passed through the filter. Test work conducted
by YRC at Research Triangle Park ^ ' for EPA involving wind
tunnel tests with low grain loadings and small diameter particles
produced much the same effect. Redesign of the inlet to the
filter should be considered so as to reduce the velocity of the
incoming particles and to disperse them over a larger area of
the filter.
6.7 Sample Recovery
Sample recovery was performed in two stages. First, each sampling
train was partially disassembled and all water catches stoppered
at the sampling location. Then, further disassembly and clean-up
were performed in the laboratory trailer. All sample recovery
procedures were the same for each, of the three test types:
organics, benzene solubles, and biological tests, except that
different solvents were used to wash the train components during
clean-up. The washes used were:
Test Solvent
Organics Methylene chloride
Benzene soluble Distilled water then benzene
Biological Distilled water
At the completion of each test the probe and cyclone assembly
was removed from the stack and its exterior wiped clean. The
cyclone cup was removed, its contents poured into a clean,
labeled container and the cup replaced in the cyclone assembly.
The condenser (impinger) was removed and capped. The filter
holder, umbilical line, and adsorber-impingers assembly were
disconnected and each was capped. After inspection and recording
of any anomalies, the individual units were lowered from the
test platform and taken to the laboratory trailer for clean-up.
-51-
-------
The samples generated from each sampling train during labora-
tory procedures were:
1. Cyclone catch and wash
2. Probe wash
3. Filter
4. Condenser catch and wash
5. Adsorber unit
6. Solvent blanks
The volumes of the cyclone catch and the condenser catch were
recorded. Water caught in the back impingers and in the silica
gel was used to determine moisture values for each test. No
organic analyses were performed on train components positioned
after the adsorber unit.
In the laboratory trailer, clean-up and disassembly of the
sampling train were conducted in the following order:
1. The cyclone catch was measured to the nearest
milliliter and returned to its container. The
inside of the cyclone was cleaned by rinsing
with solvent and brushing between each rinse
with a precleaned nylon brush. This was con-
tinued until the rinse showed no visible part-
icles. The rinses were added to the cyclone
catch.
2. The nozzle was carefully removed and the in-
side surface cleaned in a manner similar to the
cyclone until the rinse showed no visible part-
icles. All rinses were added to the probe wash.
3. The probe liner was rinsed by squirting solvent
into the upper end, while tilting and rotating
it to assure that all inside surfaces were cleaned.
The water was drained from the lower end into the
sample container. A second rinse using the same
procedure was then performed with the aid of a
probe brush, which was pushed through the entire
-52-
-------
length of the liner using a twisting action.
The brushing and rinsing operation was repeated
until no particulate matter remained in the
probe liner upon visual inspection. Upon com-
pletion of the brushing and rinsing operation,
the brush was rinsed with solvent and the liner
was given a final rinse. These rinsings were
collected in a glass jar.
4. The inside of the front half of the filter holder
was cleaned by double brushing with a nylon
bristle brush and rinsing with solvent until all
visible particulate was removed. The brush and
the inside surface of the front half of the filter
holder were then given a final rinse. All rinses
were added to the probe wash.
5. The filter was removed from the holder and was in-
spected for tears, punctures and other deformations
before being placed in a clean glass jar.
6. The volume of the condenser catch (.first impinger
before adsorber) was measured to the nearest milli-
liter and the inside of this impinger brushed with
a nylon brush and rinsed with, solvent. The catch
and rinses were poured into a glass jar.
7. The inside of the back half of the filter holder
and the inside of the tubing (.filter holder to con-
denser and condenser to adsorber1 was cleaned by
brushing with a nylon brush and rinsing with solvent.
These washes were added to the condenser catch.
8. The adsorber unit was fully disconnected and capped.
9. The liquid in the first three impingers and the con-
densate from the umbilical cord were measured to the
nearest milliliter and recorded. The umbilical cord
was washed, this wash was not collected.
10. The silica gel from the last impinger was transferred
to a pretared container, weighed to the nearest 0.1 g
and recorded.
-53-
-------
A blank test was prepared for inclusion in the organic emission
test series by additional rinsing of a train with methylene
chloride. Upon completion of Test No. 17, the nozzle, cyclone,
probe, filter holder, and condenser were rinsed a second time.
A third wash was done and saved as the blank. An unused organic
adsorber and glass fiber filter were also labeled as part of
the blank test.
Each of the five samples generated from the laboratory clean-up
(except the adsorber unit) was put in a glass jar. The lids
were sealed tightly with teflon tape. All samples were stored
under refrigeration and out of sunlight both on-site and during
shipment to York Research in Stamford or Arthur D. Little in
Boston, for extraction and/or analysis.
6.8 Precision of Sampling Methods and Estimated Probable Error
for Analytical Procedures
In order to place a measure of reliability to the data obtained
from this program, it is necessary to examine the various pro-
cesses and measurements carried out in achieving the final num-
erical result. Once the reliability of each step is assessed,
it is then possible to compute the probable error associated
with the data point.
Probable error is defined as a plus or minus quantity within
which limits the actual accidental error is likely as not to
fall. It is an indication of precision and does not signify
either the actual error, or the error most likely to occur.
To evaluate quantities derived from field test data, many
equations are used in ways that combine one or more measured
quantities resulting in determination of such things as the
Total Air Emission, Volumetric Flow, and Volume of Sample
tested; to mention a few. With each such quantity there exists
an associated maximum possible error. When several quantities
are combined in some fashion, as in an equation, these errors
tend to accumulate and compound themselves. A common procedure
-54-
-------
for determining the maximum probable error for any combination
of quantities is discussed on Page 1, 6, 7 of the Civil Engineer's
Handbook, and was used in the following calculations for maximum
probable error.
This method is given as follows for calculating the probable
error (Es) for related quantities of differing reliabilities:
(eq.l)
.+ E
n
where
quantities.
through E are the probable errors for the independent
As an example of the procedure used in calculating the probable
error, the following example is given:
Air Emission (Ib/T.C.) =
V
m
std
scfm
where: mg is the sample weight
V
m
is the volume of the sample tested
std
scfm is the volumetric flow
C is conversion constants (C, , C~,...C )
12 n
The first three quantities contain other independent quantities
that also have errors associated with them; they too must first
be analyzed for maximum probable error in order to determine
a maximum probable error for the total Air Emission results.
As a part of the overall calculation, it was necessary to cal-
culate the volume of the sample tested. Maximum errors for each
component of the calculation were obtained from a paper by
Shigehara, et al.
(8)
This sample calculation is shown below in its entirety. It
should be noted that dimensions, like conversions, have no
bearing on maximum probable error determinations and have not
been included.
-55-
-------
V
mstd
(T
m
C3>
where: V = volume (meter)
m
P, =
P =
m
atmospheric pressure
pressure at meter
Individual Probable Errors
Vm; + .004%
P.; + .21%
P ; + .42%
m —
T; + .10%
(8)
Then using Eq. 1,
E for V
s m
= + 1.10%
std
In the same manner, the maximum probable error for scfm was
found to be +_ 3.2%. Laboratory analysis determined the weight
measurements to be accurate + 0.003% of the total weight. Table
6-7 shows the estimated error for each type of analysis. The
effect of this error on the Total Air Emission figure was neg-
ligible.
With maximum probable errors for mg, V , and scfm all known,
mstd
Eqn. (1) is then evaluated using the procedure previously mentioned
to yield a maximum probable error for the Total Air Emission
figures.
Maximum Probable Error for Total Air Emission is - 3.4%
-56-
-------
TABLE 6-7
PROBABLE ANALYTICAL ERRORS
Parameter
Particulate analysis of
organic emission tests
probe wash
cyclone
filter
condenser
Benzene soluble residue
tests (each train component)
benzene solubles
particulate weight
Total Organic Carbon
(in quench water)
Total Hydrocarbons
Benzene
Biological Test
Estimated Error
+ 0.05 mg
+ 0.05 mg
+ 0.05 mg
+ 0.05 mg
+ 0.05 mg
+ 0.05 mg
± 2%
+_ 1 ppm
NA*
Source
EPA Method 5
EPA Method 5
EPA Method 5
EPA Method 5
Estimate
Estimate
Function of
instrument
Function of
instrument
* NA = Not Available
-57-
-------
7.0 ANALYTICAL METHODS
7.1 Organic Emission Tests and Water Samples
Samples from the fifteen organic emission tests (plus one blank)
were analyzed for particulate and organic content. Four water
samples were also analyzed for organics.
Three tests (Nos. 5, 7, and 14) representing three different
test conditions (clean water - green coke, clean water - nongreen
coke, and contaminated water - nongreen coke) were analyzed by
sampling train component. The analytical scheme for these tests
is depicted in Figure 7-1. Particulate content was determined
separately for each of three samples per test: The cyclone
wash, probe wash, and filter. Then, soxhlet extractions were
performed on the dried cyclone catch, dried probe wash plus
filter, and the XAD-2 resin. The XAD-2 extract and methylene
chloride extract of the condensate were combined. Three extracts
(cyclone wash, probe wash plus filter, and XAD-2 resin plus
condensate) were subsequently analyzed for organic content.
The other twelve tests and the blank were each analyzed similarly
by train component for particulate. A composite sample of ex-
tracts from particulates, condensate, and XAD-2 resin was then
used for organic analysis. These procedures are outlined in
Figure 7-2.
EPA Level 1 analytical methods were followed for all of the
described analyses. ' The only exceptions are the Gas
Chromatography/Mass Spectrometry analyses (GC/MS. '
7.1.1 Particulate
Total particulate was determined by drying each sample
in a tared evaporating dish at 50 C and desiccating at
room temperature to constant weight. These procedures
were carried out under conditions of dim light and samples
were subsequently stored in the dark at 4 C.
-58-
-------
FIGURE 7-1
ANALYTICAL PROCEDURES FOR SAMPLES FROM TESTS 5, 7, AND 14
Cyclone
Wash
Probe
Wash
Filter
Condensate
XAD-2
Resin
Water
Samples
D
H
• - -
s:
n>
H-
•i -
X
^
Combine
Soxhlet
Extraction
•1 • •
^
Liquid
Extraction
•^
^
Combine
Extracts
Concentrate
Extract
M
n
* 8
*V O
0 0
ju 2
03 O
> 0
I
en
vo
I
-------
FIGURE 7-2
ANALYTICAL PROCEDURES FOR SAMPLES FROM TESTS 2B, 3, 4, 6, 8, 9, 10, 11, 12, 15, 16, AND 17
O
I
Cyclone
Wash
Probe
Wash
Filter
Condensate
XAD-2
Resin
D
3
s
(D
p-
iQ
cr
/
Combine
\
/
Soxhlet
Extraction
^^
Liquid
Extraction
x.
\
^
^
Combine
Extracts
V
\
S>
Concentrate
Extract
H
50
f
n
s|
1) O
o n
Qj J<*
H R
03 O
iS§
-------
7.1.2 Sample Extraction
Samples were prepared for organic analysis by extraction
with high purity methylene chloride (Burdick and Jackson,
distilled-in-glass). Extraction was performed over a
24-hour period with about 500 ml of methylene chloride
for XAD-2 resin samples and with about 300 ml methylene
chloride for particulate samples. Four quench water
samples (clean make-up and inlet, and contaminated make-up
and inlet) were extracted with methylene chloride in
separatory funnels fitted with Teflon stopcocks. The pH
of the aqueous sample was adjusted to 2.8 with hydrochloric
acid and then to 12.0 with sodium hydroxide. Two extrac-
tions were performed at each pH using a sample to methylene
chloride ratio of about 20/1. The quench water extracts
were then subjected to the same analyses described below.
7.1.3 T.C.O., GRAY, and IR
A Total Chromatographable Organics (T.C.O.) analysis was
performed to determine the amount of organic material with
boiling points from 100°C to 300°C in each sample. A gas
chromatograph with flame ionization detector was used at
the conditions shown in Table 7-1. The concentration of
T.C.O. was calculated from the ratios of the peak areas
of each sample to those of the known standards.
TABLE 7-1
INSTRUMENT CONDITIONS FOR T.C.O. ANALYSIS
Column: 10% OV-101 on 100/120 mesh Supelcoport
Injector temperature: 270 C
Detector temperature: 305 C
Temperature program: Room temperature for 5 minutes,then
programmed at 20°C/min up to 250°C
Gas flow rates: He at 30 ml/min
H2 at 30 ml/min
Air at 300 ml/min
-61-
-------
A gravimetric analysis was performed for organic material
with boiling points higher than 300°C. A one or five ml
aliquot of sample extract was dried at room temperature
in a desiccator to constant weight.
The Infrared (JR) spectra of all samples as potassium
bromide micro pellets were obtained on a Perkin Elmer
521 grating spectrometer.
7.1.4 Liquid Chromatography (.LC)
Subsequent analyses required another preparation step,
Liquid Chromatography (LC) . Samples were concentrated to
10 ml using Kuderna Danish apparatus then concentrated to
1 ml under a nitrogen stream and subjected to three con-
secutive solvent exchanges with cyclopentane. The cyclo-
pentane solutions were chromatographed on a silica gel
column. Seven fractions were collected by elution with
solvent mixtures of increasing polarity.
Two extracts per sample were prepared for analysis by
combining LC fractions 2,3, and 4 (.aromatic) and fractions
5,6, and 7 Cpolar). This was necessary due to the large
amounts of silicone grease and phthalates present in most
of the samples (.as shown by their IR spectra) . The origin
of these substances (.in particular, whether or not they
are sample contaminants) is not known. (See Section 8.2 for
a detailed discussion.) However, their large quantity in
most samples necessitated a reevaluation of future analytical
procedures. An IR analysis of the EPA Level 1 LC fractions
of two samples indicated that most of the silicone grease
was found in fractions 2, 3, and 4 and that the phthalates
were collected in fractions 5, 6, and 7. This separation
of the silicone grease and phthalates allowed the IR
spectra of other materials in the samples to be observed.
-62-
-------
Therefore, all samples were separated into aromatic and
polar fractions prior to further analysis.
Although the separation of silicone grease and phthalates
aids in avoiding some analytical problems silicone grease
may affect the LC procedure itself. During the concentra-
tion step prior to LC separation, the presence of silicone
grease often prevented reduction of the extract to or below
10 ml. This will in effect raise the limit of detection
for the GC/MS analyses. In addition, the large amounts of
silicone grease present may alter the LC separation efficiency,
If the LC separation becomes ineffective, the compounds for
analysis may be spread throughout the LC fractions collected
lowering the analytical levels observed for such samples in
the combined extracts.
7.1.5 Low Resolution Mass Spectroscopy (LRMS)
LRMS was conducted on a Dupont 21-110B spectrometer. Samples
were usually run at 15 ev and 70 ev ionization potentials
over a temperature range of 70 - 350 C.
7.1.6 Gas Chromatography/Mass Spectrometry (GC/MS)
The two extracts obtained for each sample were subject to
Gas Chromatography/Mass Spectrometry (GC/MS) analysis.
For this, a Finnigan Model 4000 GC/MS system with a Finnigan
Model 6110 data system was used. Each sample was separated
into its component parts by GC, then specific organic com-
pounds were identified and their concentrations determined
by mass spectrometry.
The aromatic fractions were analyzed for a standard group
of polycyclic aromatic hydrocarbons (PAH), ranging from
fluorene to higher molecular weight aromatics. Analysis.
-63-
-------
was also performed for other PAH compounds including lower
molecular weight aromatics which were found in test 15 (a
highly loaded sample). GC/MS conditions for these analyses
are listed in Table 7-2. PAH species were identified by
a calibration mixture which contained standards for most
of the species analyzed. For those species not present
in the calibration mixture, response factors were approxi-
mated from a linear regression analysis of the response
factors of the materials in the calibration mixture as a
function of molecular weight. This procedure is possible
since it has been shown that the mass spectral response/
weight for PAH's of quite different molecular weight has
a nearly linear correlation with molecular weight. The
lower molecular weight materials are subject to the highest
analytical error and their values are likely to represent
an over-estimation of their levels in the samples. There-
fore, the levels of compounds with molecular weights be-
low that of fluorene represent worst case levels.
The silicone grease present in most of the aromatic frac-
tions would not be expected to interfere with the PAH
analyses. Interferences in the GC/MS analysis are caused
by coeluting compounds which have ions in common with the
analytical ions of interest. Silicones tend to have few
ions in common with PAH species.
The polar fractions were analyzed for heterocyclic oxygen
and nitrogen compounds under the GC/MS conditions specified
in Table 7-2. A highly loaded sample (TestlS) was again
used as an indicator for compounds expected to be found in
this fraction. Not all of the compounds detected in the
survey sample were available for calibration standards.
Six of the eighteen compounds identified were available in
pure form (pyridine, aniline, phenol, cresol, quinoline,
and acridine). Quantitation of nine more compounds (which
are alkyl substituted derivatives of these six) was achieved
-64-
-------
TABLE 7-2
OPERATING CONDITIONS FOR GC/MS ANALYSES
PAH
Polar
BAP
GC CONDITIONS
Column
Temperature
program
Helium flow rate
Sample Size
Internal Std
MS CONDITIONS
Mass Ranges
Integration Times
Electron Multi-
plier Voltage
Electron Energy
Filament Emission
Scan Rate
Dexsil 400
Isothermal oper-
ation at 170° C
for 1 min . Linear
operation to 300°C
at 15°C/min, iso-
thermal operation
at 300°C for 30
min.
30 mL/min
2 - 4 uL
9-phenyl anthra-
cene
70-210,211-270,
271-350
2, 5, 13
1800 V
50 V
45 ma
1 sec/spectrum
Superpak 20M
Isothermal oper-
ation at 80°C
for 1 min. Lin-
ear operation to
to 250°C at
15 C/min , thermal
operation at
250°C for 30
min.
30 mL/min
1 - 2 pL
9-phenyl anthra-
cene
70-210,211-270,
271-350
2, 5, 13
1400 V
50V
45 ma
1 sec/spectrum
SP-301
(liquid crystal)
Isothermal
operation at
260°C.
30 mL/min
2 yL
9-10-diphenyl
anthracene
240-260
320-340
25, 25
1500 V
50 V
45 ma
1 sec/spectrum
-65-
-------
through using the unsubstituted compound as a model. The
similarity in the mass spectra of the model and the sub-
stituted analog, and the generally flat system response
within this range allow this method of quantitation. Cal-
ibration values were approximated for the few other compounds
(i.e., benzonitrile, toluidine, and indole) by again finding
models in the calibration mixture with similar mass spectra
and spectral sensitivities. These calibration procedures
lead to a higher relative error in the reported absolute
values, a factor of 2 or 3 for the substituted analogs and
a factor of 5 for the other few compounds. Their relative
values remain quite accurate, however.
A separate GC/MS analysis, specific for benzo (a) pyrene,
was also performed on each aromatic fraction. The operating
conditions listed in Table 7-2 allow benzo (a) pyrene to
be selectively separated from the other PAH species. The
sensitivity and accuracy of the GC/MS system are improved
with this analysis.
7.2 Benzene Soluble Residue Tests
All samples from the three benzene soluble residue tests were
analyzed for both benzene soluble residue and particulate.
The cyclone catch, probe wash, and condenser catch from each test
were analyzed in the same manner. Each sample contained water
and benzene fractions. The entire sample was extracted with
benzene. The water layer was evaporated to dryness at 103 C,
desiccated and weighed to constant weight for a particulate deter-
mination. Insoluble particulate was filtered from the benzene
layer and the benzene fraction was then evaporated at room temp-
erature, desiccated and taken to constant weight for determination
of benzene soluble residue. The particulate weight of each sample
was found by adding the particulate weight from the water fraction,
plus benzene-in-soluble particulate and benzene soluble residue
from the benzene fraction.
-66-
-------
The filter and adsorber from each train were Soxhlet extracted
with benzene according to Parma Standard Test Methods, Union
Carbide Corporation (PSM-1013). Particulate weight for the
filter was found by adding the benzene-insoluble particulate
to the benzene soluble residue. The benzene soluble weight for
the adsorber was also called "particulate" and added into the
total particulate weight for each train.
7.3 Benzene and Total Hydrocarbons
Two grab samples of coke quench tower emissions were analyzed
using a gas chromatograph with flame ionization detector for
determination of Total Hydrocarbons (T.H.C.) and benzene.
7.4 Total Organic Carbon in Quench Water
Total Organic Carbon (T.O.C.) in clean and contaminated quench
water (makeup to the sump and inlet) was automatically determined
by a Dohrmann Envirotech Total Organic Carbon Analyzer. The
results were read directly as mg/liter T.O.C.
7.5 Electrical Conductivity of Quench Water
These tests were conducted on the sump at the Lorain Plant.
A Hach Meter was used and measurements were taken in micromhos/cm,
7.6 Biological Test
A composite sample was prepared by extraction with methylene
chloride of the filter; the distilled water wash of the probe,
cyclone, condenser, and the porous polymer (XAD-2) adsorbent
trap. The extraction procedures followed the protocol outlined
in the "Technical Manual for Analysis of Organic Materials in
Process Streams" * ', and were approved by EPA.
The combined sample was prepared for the Ames bacterial assay and
the clonal cytotoxicity assay by solvent exchange with dimethyl
-67-
-------
sulfoxide (DMSO) and concentration to 3 ml. These analyses
(14)
were performed by Litton Bionetics, Inc.
7.6.1 Ames Bacterial Assay
Mutagenicity potential of the sample was determined by
a test of its genetic activity as indicated by its ability
to revert certain Salmonella strains from histidine dependence
to histidine independence, with and without mammalian meta-
bolic activation. The number of revertants observed on the
histidine-free medium reflects the degree of genetic activity.
The test conditions for the Ames Mutagenicity Assay are
described in Table 7-3.
g
Approximately 10 cells from an overnight culture of each
indicator strain were added to separate test tubes contain-
ing 2.0 ml molten agar supplemented with biotin and a trace
of histidine. For nonactivation tests, 0.01, 0.1, 1.0 and
10 ml/plate were added to the contents of the appropriate
tubes and poured over the surfaces of selective agar plates.
In activation tests, four dose levels of the test chemical
were added to the appropriate tubes with cells. Just prior
to pouring, an aliquot of reaction mixture (0.5 ml contain-
ing the 9,000 x g_ liver homogenate) was added to each of
the activation overlay tubes which were then mixed, and
the contents poured over the surface of a minimal agar
plate and allowed to solidify. The plates were incubated
for 48 hr at 37°C and scored for the number of colonies
growing on each plate. Positive and solvent controls using
both directly positive chemicals and those that require
metabolic activation were run with each assay.
A toxicity test was performed at each dose level with and
without metabolic activation. The methodology is similar
to the plate test method described above, except that approx-
p
imately 200 cells instead of 10 were used. The Vogel-Bonner
-68-
-------
TABLE 7-3
AMES MUTAGENICITY ASSAY
TEST CONDITIONS
1) Indicator Micro-organisms:
Salmonella strains :
Salmonella Typhimurium
TA-1535
TA-1537
TA-98
TA-100
2) Activation System:
a) Reaction mixture
Component
TPN (sodium salt)
Glucose -6-phosphate
Sodium phosphate (dibasic)
MgCl2
KC1
Final Concentration/ml
4 pi/moles
5 yl/moles
100 yl/moles
8 yl/moles
33 pi/moles
b) S9 homogenate
A 9,000 x g supernatant was prepared from Sprague-
Dawley adult male rate liver induced by Aroclor 1254
five days prior to kill according to the procedure
of Ames et al. (1975). G9 samples ware coded by lot
number and assayed mg protein/ml and relative P443/
P450 activity by methods described in LSI Technical
Data on Rat Liver S9 Product.
3) Positive Control Chemicals
Salmonella strain Chemical
Dose
TA-1535
TA-1537
TA-98
TA-100
N-Methyl,N-Nitro,N-Nitrosoguandidine 1 yg/plate
9-Aminoacridine 50 yg/plate
2-Nitro fluorene 10 yg/plate
-69-
-------
minimal glucose agar plates were replaced by the standard
method nutrient agar plates.
The numbers of colonies on each plate were counted and
recorded. These raw data were analyzed in a computer
program for revertants/plate (mutagenicity assays) and
population/plate (toxicity assays) for each indicator
strain.
7.6.2 Clonal Cytotoxicity Assay
The cytotoxicity of the coke quench tower emission sample
was determined by measuring the reduction in colony-forming
ability of cultured Chinese hamster cells (CHO). Following
24 hours exposure to the sample and a period of recovery
and growth, the number of colonies present in treated
cultures was compared to the number in untreated cultures.
The concentration of test material responsible for a 50%
reduction in colony number was estimated and referred to
as the EC50 value.
A Chinese hamster cell line, CHO-KI (ATCC No. CCL 61),
was used for this assay. The cell type was originally
derived from ovarian tissue and has spontaneously trans-
formed to a hypo-diploid line of rounded, fibroblastic
cells with unlimited growth potential. Monolayer cultures
have a fast doubling time of 10 to 12 hours and untreated
cells can normally be cloned with an efficiency of 80% or
greater. The CHO-KI cell line was maintained in Ham's
F12 culture medium, containing 3 x 10 M L-proline (the
cell line has an absolute requirement for proline), and
supplemented with 10% fetal bovine serum, 100 units/ml
penicillin, and 100 yg/ml streptomycin.
Cells from a monolayer stock culture were trypsinized,
counted by hemocytometer, and reseeded into a series of
-70-
-------
100 mm culture dishes at 500 cells/dish. Each dish
contained 10 ml of growth medium. The cultures were
incubated overnight at 37°c to allow attachment of the
cells and recovery of growth rate.
A volume of 2.2 ml of test material was combined with
7.8 ml of growth medium to yield a stock concentration
of 220 yl/ml. Other stocks were prepared by dilutions
with growth medium such that 1 ml additions to the 10 ml
cell cultures would yield the following final concentra-
tions: 20 pi/ml,10 Ml/ml, 5 yl/ml, 2 yl/ml and 0.5 Ml/ml.
Each concentration of test material was applied to three
culture dishes. After a 24-hr exposure period the
medium was aspirated and the cells washed twice with
Hank's balanced solution (prewarmed to 37°C). Fresh
medium (20 ml) was placed on each culture, and incubation
continued for an additional 6 days to allow colony develop-
ment.
Medium was drained from the cultures after the incubation
period and the surviving colonies were washed with phos-
phate-buffered saline (PBS), fixed in methanol, and
stained with Giemsa. Colonies were counted by eye; tiny
colonies of approximately 50 to 100 cells were excluded.
The controls consisted of one culture of untreated cells
and two cultures exposed to DMSO (solvent control) at a
final concentration of 20 yl/ml (20% by volume). DMSO
in the treated cultures was a maximum of 2% at the highest
concentration of test material. The solvent control dishes
provided the reference cloning efficiency for determining
the effect of the test material.
-71-
-------
8.0 ORGANIC EMISSIONS
8.1 Organic Emission Test Results
The results of the organic emission tests are presented in
Tables 8-2 to 8-8. Values for individual organic species are
given in ug/m and grams/metric ton of coal as determined by
GC/MS analysis. Results of the TCO, gravimetric, infrared and
LRMS analyses are not available due to interference from the
large amounts of silicone and phthalates in the samples.
Total organic emissions from the quench tower are therefore
unknown since all of the organics could not be quantified.
Polycyclic aromatic hydrocarbons were found in the emission
samples in substantial quantities as shown in Tables 8-2 and
8-3. Compounds which together contributed an average of 80%
(at least 64%) of the total PAH for each clean water test are:
1) Naphthalene
2) Methyl Naphthalenes
3) Acenaphthylene/Biphenylene
4) Dimethyl Naphthalenes
5) Fluorene
6) Dibenzofuran/Methyl Biphenyl
7) Anthracene/Phenanthrene
8) Methyl Anthracenes
The major contributors in contaminated water tests include all
of the above except dimethyl naphthalenes and methyl anthracenes,
These major compounds account for an average of 85% (at least
78%) of the total PAH in contaminated water tests. The primary
contributor among these is naphthalene. These organic species
are of lower molecular weight, ranging from m/e 128 to m/e 192.
-72-
-------
Polar compound results are reported in Tables 8-4 and 8-5 and
also show substantial concentrations to be present. In par-
ticular, phenol is the primary constituent in the clean water
tests and phenol, cresol, methyl cresol, and quinoline are
major contributors in the contaminated water tests.
The average concentrations of PAH and polar compunds in coke
quench tower emissions are shown according to process con-
dition in Table 8-1.
TABLE 8-1
ORGANIC EMISSIONS SUMMARY
Clean Water
Total PAH:
average 767
Grange L C17CI-19.QO}
Green coke 1160
(660-1900)
Nongreen coke 442
(170-745L
Polar Fraction:
average 771
(range) 060-6090)
Green coke 1540
O18-6090L
Nongreen coke 134
(60-253)
grams/taetric
ton of coal
Contaminated Water
grams/metric
ton of coal
yg/m3
0-758 31,000 32.1
CO.156-1.901 (13,700-47,300) CIS.5-46.4)
1.17
CO.679-1.901
0.419 31,000 32.1
CQ.156-0.646) (13,700-47,300) (15.5-46.4)
1.10 540,000 581
CO.0575-6.891 (243,000-922,000) (185-1009)
1.70
CO. 119-6.89).
0.606 540,000 581
CO.0575-3.13). (243,000-922,000) (185-1009)
-73-
-------
An increase in organic emissions is distinctly shown for both
PAH (^ 40 times) and polar materials (^ 500 times) when con-
taminated water is in use. This trend also appears with the
use of green coke where a threefold increase is found for both
PAH and polar compounds.
The test results for the three sampling trains analyzed by
component are summarized in Tables 8-6 and 8-7. Most of the
PAH material found for each test was in the organic adsorber
unit/ meaning that this material is smaller than 0.3 micro-
meters Cor it would have been caught on the filter). Tests by
Broddin et al have shown that "over 90% of polycyclic aromatic
hydrocarbons is found on the particles smaller than 3 ym
which means that most of these compounds are within the respir-
atory size range.^ It should be noted that organic species
of higher molecular weight (m/e > 216) were found in the
cyclone samples (which contain larger size particles) but not
in the filter samples. Most PAH compounds that have been
determined to be carcinogenic do have a molecular weight
greater than 216.
The GC/MS analysis specific for benzo(al pyrene (BaP) generated
the results found on Table 8-8. BaP in quench tower emissions
is four times greater when contaminated water is in use (0.081
g/metric ton of coall than clean water (0.019 g/metric ton of
coalI. However, the BaP data are very scattered in regards to
coke greenness.
A number of the organic compounds detected in the coke quench
tower emission tests exceeded the Minimum Acute Toxicity
Effluent CMATEf values for health for some tests ^. These
compounds also showed a high "Degree of Hazard" . The
compounds are identified in Table 8-9 along with the process
-74-
-------
TABLE 8-2
POLYCYCLIC AROMATIC HYDROCARBONS
jiy/m
CLEAN HATER TESTS
Tost 1
m/o Spates [Goneness —
116 irklune
128 NJiJhtheilcne
134 BLUizoth iophene
142 Hjthvl Naphthalenes
152 tt^aaphthylene/Biphenylene
114 Riphenyl
156 Dimethyl Nachdtalenes
166 Fluorenu
167 Carbazole
168 DibenzofuranMithyl biphenyl
178 Anthraceno/Ftenanthrene
184 Dibenzothiophenc
192 rtothvl anthracenes
202 Fliioranthene
202 Pyrenf!
204 Ci6»12 PAH
204 Ci6"12 PAH _
204 C16»12 PAH
208 Cl6»16 PAH —
«fcthyl Fluoranthenea
216 methyl Pyrenes
218 Dih^rclcnzofl.ioreno
234 "=.[* l-hr^nrnt-h i nphfint!
242 "A^fhyl -^hrytynps
252 Bcinzofluoranthenes
B^nzo (e) pyrene
Benzo (al pvreno
254 9-ffrcnvl anthracene (IS)
266 tt^tiiyl-tenzQ pyrenes i
"?fift "^-^pt-hy 1 (7hnl^nl'hn:>[V>
278 Diltnzo (a hi anthracene
300 f torn lent:
302 Dibanzo (a.i.t, ah.) ^n>nt.«
TOTAL
7
1.2
99
2.3
55 —
TI —
25
35
18
260
16
85
31
29
13
89
745
10
1.5
81.7
"9.86
1.34
4.75
37.0
2771
4.45
30.4
14.9
16.1
9
1.6
80.8
JO. 2
27.6
1B.1
23.2
164
TT76-
109
"49.5
a. 39
19.1
18.2
IK4,
26.3
227 620
4
1.6
£.41
42.5
170
6
1.6
74.5
'17.6
T797-
4.57
52.4
'49.6
8.47
8.43
TO-
29.0
29.8
323
a
1.8
1.J9
£2.1
16.5
5.19
54.3
517?
5.52
86.5
nm
17.9
28.7
565
2B
135
"78. 6~
28.8
114
38.6
.4
-140
'83.2
25.6
J9.B
11.4
5.75
^713~
11
-T55
25.6
11. S '
~171
\4
T
—f
"S
3
rrs
'•*
•E9~
U.I
745 660
3
52.1
34.5
86. 7~
67.2
"ITS
-143
40.7
8.79
14.2
1040
12
527
~2TE —
233
208
82.7
14.4
15
1900
5
4.2
135
2TH
478
183
20
— rr*
1443
CONTAMINATED
WATER TESTS
16
0.3
273
24100
9BO
1850
"3070
— 139
—737 —
IbT5~
22.6
5.27
7.38
58.6
44.6
.280
13.6
52.3
7 71
7.00
40500
15
0.5
31.3
26330
1030
1540 '
— 1260
4990
280
4T9~
15.8
40.5
9.61
21.9
145
2.59
17.8
88.9
2.41
.783
47300
17
0.5
165
424
6B.1
997
39.3
~757T
8.04
2.45
~57l4
33.5
.47
10.0
28.4
.3
5.09
13700
14
1.5
69
— IFJI
352
> 4005
219
«ID
— n —
120
2.3
— n —
118
22844
Ul
I
Nate: Polynuclear Aranatic Hydrocarbons
apocics was be-low tike detection 1<
five to twenty turus with a lower
UoLoction Limit in iio/m-* I <90 I <
l~ 1—:
were measured by Gas Chromatorjraphy/Hiss S|«ctroit:try. All blanks indicate that the concentration of the
»vul. The detection limit for each PAH sjecies varies with imlucular weight, reaching differences from
limit at naphthalene and a hiyher limit at coronene. Thia 'upiier limit' for each test io given below:
&72 | <92 | &104 | <110 | S76 | 177 | j61
^102
<84
1£
9B
i76
S?8 I <102
-------
TABLE 8-3
POLVCVCLIC AROMATIC HYDROCARBONS
(grams/metric ton of coal)
Tu'St It
m/e Species Greenecs
116 Indone
128 Naphthalene
134 Bcnzothiophene
142 Methyl Naphthalenes
152 Acenaphthylene/Biphenylene
154 Biphcnyl
156 Dimethyl Naphthalenes
166 Fluorcne
167 Carbazole
168 Dibenzofuran/methyl
bi phenyl
170 Anthracene/phenanthrene
184 Dibenzothiophene
192 Methyl anthracenes
202 Fluoranthene
202 Pyrcne
204 C ,1 PAH
204 CM PA..
204 C S, - PAH
208 Cie"i6 Pft"
216 McL-1y1 Fluoranthcnes
Methyl Pyrenes
218 Dihydrobunzof luorene
228 Chryscne/Benz (a)
anthracenes
234 Naphthobonzothiophene
242 Methyl-chrysencs
2S2 Benzof luoranthcnes
Benzo (e) pyrene
Benzo (a) pyrene
252 Perylene
254 9-|ihonyl anthracene (IS)
256 7, 12 Dimethyl Benz (a)
anthracene
266 Methyl -taenzo pyrenou
267 Dihenzo (c,g) carbazole
268 3-methyl cholanthrene
276 Inclono (1,2,3-cd) pyrene
276 Benzo (ghi) perylene
278 Dibenzo (a,h) anthracene
300 Coronene
302 Dibenzo (ai i ah) pyrenes
TOTAL
CLEAN WATER TESTS
7
1.2
0.0895
ip_.00191l
t.0219
0.00970
0.00590
0.0211
0.0308
0.0158
0.224
0.0135
0.0730
0.0263
0.0246
0.0112
0.0765
0.646
10
1.5
0.0750
0.0100
0.00126
0.00500
0.0350
0.0250
0.00500
0.0300
0.0150
0.0150
0.216
9
1.6
0.0770
0.0192
0.0263
0.0173
0.0222
0.15G
0.0111
0.104
0.0470
0.0303
0.00790
0.0182
0.0174
0.0109
0.0251
0.590
4
1.6
0 . 094 5
0.0168
0.00590
0.0390
0.156
6
1.6
0.0815
0.0193
0.0110
0.00432
0.00496
0.0570
0.0540
0.0925
0.00920
0.0379
0.0317
0.0326
0.436
8
l.B
0.181
0.00119
0.0530
0.0141
0.00500
0.0462
0.0443
0.00500
0.0735
0.0140
0.0155
0.0245
0.00800
0.485
2D
3.4
0.135
0.07U5
0.02U8
0.00870
0.114
0.0385
0.0575
0.140
0.0830
0.0256
0.0198
0.0114
0.00575
0.00312
0.750
11
3.2
0.112
0.0263
0 .0106
0.193
0.0118
0.113
0. 147
0.0159
0.028
0.0099:
0.0114
0.679
3
3.3
0.282
0.0590
0.0810
0.0288
0.0393
0.0474
0.0795
0.0975
0.0760
0.134
0.162
0.0462
0.0227
0.00995
0.0161
1.18
12
3.7
0.525
0.216
0.0895
0.0204
0.232
0.0715
0.283
0.206
0.0720
0.0720
0.0820
0.0144
0.0149
1.90
5
4.0
0.159
0.00400
0.0750
0.0290
0.0100
0.124
0.0635
0.0625
0.435
0.00950
0.169
0.0400
0.0345
0.0185
0.0105
0.0770
1.32
CONTAMINATED
WATER TESTS
16
0.3
0.263
23.2
0.940
1.78
3.62
0.293
0.303
2.69
0.266
1.06
2.95
0.134
0.228
0.520
0.454
0.00990
0.0217
0.00505
0.00710
0.0565
0.0327
0.0430
0.000269
0.0130
0.0500
0.00740
0.00670
39. 0
15
0.5
0.0307
25.8
1.01
1.51
3.77
0.326
0.264
3.47
0.539
1.23
4.91
0.274
0.412
1.13
1.18
0.0155
0.0396
0.00940
0.0215
0.143
0.0795
0.0795
0.00254
0.0174
0.0870
0.00236
' 0.000770
46. 4
17
0.5
0.186
9.60
0.349
0.477
1.25
0.0770
0.0765
0.875
0.258
0.294
1.12
0.0443
0.0845
0.278
0.309
0.00382
0.00915
0.00267
0.00575
0.0374
0.0149
0.0302
0.00496
0.0115
0.0321
0.0138
0.00575
15.5
14
1.5
0.0830
8.30
1.14
1.98
2.28
0.412
0.424
2.65
0.565
1.54
4.85
0.246
0.496
1.14
0.890
0.0555
0.0141
0.0134
0.145
0.0496
0.107
0.0027
0.0167
0.143
0.0910
27.6
NOTC: l'oly,,uclc-ac Aromatic Hydrocarbons were measured by Gas Chromatoyraphy/Mass Spectrometry. All blanks indicate that the concentration of the
species was below l.he do Lection level.
-------
TABLE 8-4
POLAR FRACTION
e
254 9-Phenvlanthracene (IS)
37.6
1.66
0.952
77.3
1.65
0.47
>l'iSOO
10
1.5
181
72.1
> 2fl60
9
1.6
60.1
>is7n
4
1.6
179
>4470
6
1.6
90.4
>1970
8
1.8
75.2
23.6
>1640
2B
l.B
118
>lfl400
11
3.2
779
30.0
61.1
272
3
3.3
2830
2800
329
132
> 16200
12
3.7
78. <
98. (
19.=
>5380
5
4.0
2.7(
4.3;
0.56(
0.14'
128
0.11J
o.nt
>9.49C
CONmMINATfcD WATER
16
0.3
4810
2120
1610
88800
2730
750
975
103000
341
885
13400
9430
8030
920
^lQ3QflO_
15
0.5
1470C
530C
35K
50300C
180C
1240C
55 15200C
232
17
0.5
2610
2510
679
59600
1490
1520
176
75100
1330
376
11800
6430
614
14
1.5
25900
22700
4640
272000
3970
14200
2060
404000
3200
2230
34800
36300
490
1570
3800
>72800
390
2.41
* Phthalates are considered to be a contaminant.
Nate: All blanks indicate that the concentration of the species was below the detection level. The detection limit for each PAH species varies with
molecular weight, reaching differences from five to twenty tines with a lower limit at naphthalene and a higher limit at coronene. This "upper
limit" for each test is given below:
Detection Limit in ug/m3 I <90 I <116 I <102 I
-------
TABLE H-5
POLAR FRACTION
(grams/metric ton of coal)
Test II
m/e Species Greenest
79 Pyridine
93 Methyl Pyridine
93 Aniline
94 Phenol
103 Bonzonitrile
107 Dimethyl/ethyJ. pyridine
107 Toluidine
108 cresol
117 Indole
121 Triraethyl pyridine
122 Nuthyl cresol
^j 129 Quinoline
CO 129 Isoquinoline
1 136 Triraethyl phenol
143 Methyl quinoline
149 Phthalates*
157 Dimethyl/ethyl quinoline
179 Acridine
254 9-phenylanthracene (IS)
CLEAN WATER
7
1.2
0.0324
0.00143
I). 00080!
0.0670
0.00141
0.00039;
>13.3
10
1.5
2.24
0.890
>3.53
9
1.6
0.0575
>1.50
4
1.6
0.165
>4.12
6
1.6
0 . 0990
>4.35
8
1.8
0.0640
0.0202
>1.40
2B
3 .4
0.119
>18.4
11
3.2
0.799
0 .0308
0.0625
0.278
>14. ]
3
3.3
3.20
3.17
0.372
0.150
>18.4
12
3.7
0.0780
0.0975
0.0194
>5.35
5
4.0
0.00253
0.00397
0.000489
0.000123
0.117
0.0000980
0.0000980
>8.70
CONTAMINATED
WATER
16
0.3
4.61
2.04
1.54
85.0
2.62
0.720
0.935
98.5
0.328
0.850
12.9
9.045
7.70
0.885
>105
15
0.5
14.4
5.20
3.44
491
1.76
12.2
0.540
342
0.630
2.89
25.0
0.945
1.54
>149
0.228
17
0.5
2.94
2.83
0.765
67.0
1.67
1.70
0.200
84.5
1.50
0.425
13.3
7.25
O.G90
>125
14
1.5
31.2
27.5
5.60
330
4.79
17.1
2.49
490
3.87
2.70
42.1
43.8
0.590
1.90
4.59
>70.0
0.472
0.00291
All blanks indicate that the concentration of the species was below the detection level.
* Phthalates are considered to be a contaminant.
-------
TASLE 8-6
0? TRAIN COMPONENTS
Saecies
Rjiynuclear Aromatic Hydrocarbons
Indene
iimhtnalene
Ber-zotmoonane
"tethvl KaorKhalaMS
rteenaohtiivlene/Biahenvlene
Bisnenyl
DiiiBKV?. Nanhthalenes
^USSSS
"•"fr'Tnlf
CipeazQfciran/frBttivl bieaeayl
taihraeene/Bhenar.threne
Methvl anthracenes
Fluoranthene
Pvrene
CISH,, PAH
C1KH12 EM 1
.ClfiHK HM
Methyl Pluarmrhgng i Methyl Pvnme
Djjwdrobenzoflunnme
>iaph^hobmgnt~iigpripnB
;*»thvl-chrvsenes
Benzo£luoiantnene,benzo(s) pyrene
Bangalat pyrme
jtarylano
tMhensrl anriraM"*. (T§\ |
7.1.' gunethvl Rpn?(Bl anriiraoeng
Q^gr^o (g.ffl <-ar^.~n1fl
.i-mec-ivl eholanemeng
Indeno (1.2.3-odl rare™.
Benzc (cr-1 o=-^leno
Diber-zo (ajil ac«r»<=na
nihpn?r. fai _t^ah; pjmanoe
Total
Polar Cemcunds
MTlfl-nP
Msthvl Pvridine
Phenol
Dimgehvl/eghvl mnriding
"!-"~<'h"1 "J^"^^~
?tethyl, cresol 1
TsnqiannT ua L
T>imm->Vj.1 phmnl
Bh*hala»B* 1
frinM-Iyi /pthyl qiunn1ir» I
v^rlrtH^ r
coke cireenssE = i.C
'filter en=
cyclaie probe wash
extract 1 extract
A. ,_t
*•*
46
~T?
[16
22
5.?
110
1.51
2.64
12E
>"BO
1
0.60
0.96
TT!
150
-3.0
1.44
10
186
0.244
0.142 I
»»*>
XK>2
ej^ract
iSo
32
10
S3
67
280
7-3
174
K
16
15 ,
94
1147
1.C1
i.ee j
g-iis
>28«n 1
1S_
Test N=. 7 clear, water
cote creeness = 1.2
cyclone
extract
1"
6-1
o-ig
142
11.7
64.5 |
>7BQ!i J
fil^r and
probs wash
ej.tr act
=T 1
5-8
U
10
9-6
130
O.Sfi
1 1-66
1 1-65
1 0-*7
><900 !
XAD-2
extract
2.30 1
4.1 1
| 1« 1
1 gs
473
2=.2
Li.S .
>2750 •
Test No. i; contaminated water
cake creeness = 1.5
i filter are-
cyclone probe wash
1 1-"
«0
98
1 465
1 ^8
36
95
E-S 1
10
S7
2s ,
26
1-9P |
K-t
1
>ifi9oo ;
39
1 3^°
. 24
. " I
509
65 '
<=s
51-0
>moo t
3-ai
XAD-2
63
1630
344
2110
3150
151
35Q
S21
62S
Si_
12
11
J2__
I 2.26 .
2.6
;T
21736
25300
2270C
4640
272°00
404POQ
34800
1=70
379Q
24 1
* Phthalates are considered to be a contaminant.
*• These results ware obtained from a special analysis with high sensitivity for benzolalpyrene.
PAH
and
Detection Limit in
£.74
<104
<7£
<6S
iSO
-79-
-------
TABLE 8-7
ORGANIC ANALYSIS OF TRAIN COMPONENTS
(grama/metric ton of coal)
Tc
Polynuclear Aromatic Hydrocarbons
Methvl Nauhthalenes
Biphenvl
Dimethvl Naohthalenes
FluQjffinc
Carbazole
nihpnzo fur an/me thvl binhenyl
Methvl anthracenes
Fluor an thene
Pyrene
C_rH.-PAH
C«H"PMI
C"H::PAH
C"H"PAH
Methyl Fluoranthene s,
Hechyl Pvrene
Dihydrobenzofluorene
Chrysene/Benz (a) Anthracenes
Napnthobenzothiophene
Methyl-chryscnes
Benzofluoranthene.benzo(e)pyrene
Benzo(a)pyrene
Perylene
p-phenyl anthracene (IS)
7,12 Dimethyl Benz (a) anthracene
Methyl-benzo pyrenes
Dibenzo (c,g) carbazole
Indcno 3.72
0.025
1.00350
) 000500
,
0 00100
0.0135
0 00150
O 00300
). 00150
). 00950
XAD-2
extract
0.00350
0.0750
0.0290
0.00900
0.123
0.0480
O.061O
O.O06SO
0.159
0.0145
0.0145
~~]
3.00150
0.173
0.000221
0.000196
0.000123
2.32
0.0150
0 . 00600
0.0770
1.05
0.000910
0.000098
>2.64
0.035
filter and
cyclone probe wash
extract extract
0.00200
0.00250
0.0010C.'
0.00500
0.000150
o . ooioo:
0.112
0.0101
0.0555
>6.70
0.00250
0.00500
0.0110
0.00850
0 . 00800
0.0715
0.112
0.00055!
0.00143
0.00016'
0.00141
0.00039:
>4.20
XAD-2
extract
0.0220
0.00950
0.00350
0.0210
0.0235
0.0155
0.0730
0.00700
0.0110
0.0110
0.00350
0.406
0.0217
0.00062
0.0111
>2.37
Test No. 14 concam. water
coke greeness'1.5
filter and
cyclone probe wash
extract extract
0.00172
0.00275
0.0013S
0.0485
0.118
O. 00975
0.0334
(1.0435
0.103
0.115
0.0313
0.0103
0.0590
0.0123
0.0810
0.0910
1.33
0.00240
0.0170
0.0115
>2.055
0.04 70
O. 0077 5
0.0290
0.0404
0.0170
0.615
0.0610
0.550
0.000439
0.-0615
13.5
0.00291
0.066 0.023
XAD-2
extract
0.0830
8.30
1.98
2.28
0.412
0.424
2.56
0.447
1.52
0.214
0.424
0.990
0.760
0.0555
0.0141
0.0134
0.114
0.0395
0.0477
0.00274
0.00439
0.062
25.7
31.2
27.5
5.60
329
4.81
17.1
2.49
490
3.80
2.70
42.1
43.8
0.575
1.90
>54.0
0.029
• Phthalates are considered to be a contaminant.
••These results were obtained from a special analysis with high sensitivity for benzo(a)pyrene.
-80-
-------
TABLE 8-8
BENZO(A) PYRENE IN COKE QUENCH TOWER EMISSIONS
Test Number
7
10
9
4
6
8
11
3
2B
12
5
Coke
Greenness Rating
CLEAN WATER
1.2
1.5
1.6
1.6
1.6
1.8
3.2
3.3
3.4
3.7
4.0
3
yg/m
TESTS
*ND
ND
38
42
66
ND
ND
ND
ND
ND
66
g/metric
Ton of Coal
0.036
0.039
0.072
0.060
CONTAMINATED WATER TESTS
16
15
17
14
0.3
0.5
0.5
1.5
99
72
36
98
0.095
0.071
0.040
0.12
TESTS ANALYZED BY SAMPLING TRAIN COMPONENT
Component
Cyclone
Probe/Filter
Adsorber
Total
Test 5
, g/metric
ii ton of coal
Test 7
g/metric
Ug/m ton of coal
Test 14
g/metric
pg/m" ton of coal
28
ND
38
66
0.025
0.035
0.060
ND
ND
ND
ND
ND
ND
ND
ND
im «•*• , •.,
55
19
24
98
0.066
0.023
0.029
0.12
* ND - Not Detected. The detection limits for the benzo(a)pyrene
analysis for each test are as follows:
Test No.
Detection Limit
in pg/nu3: Total
Cyclone
Probe/Filter
Adsorber
2B 3 4 5 6 7 8 9 10 11 12 14 15 16 17
8
7 i 8
10 |10
7
6
10.-
9
8
7
9
8
10
12 ' 10 1 9
!
!
Q
5
5
6
8
11
8
-81-
-------
TABLE 8-9
ORGANIC SPECIES FOUND IN EMISSIONS
AT LEVELS POTENTIALLY HARMFUL TO HEALTH
Process Conditions
Clean Water Clean Water Contaminated Water
Species Nongreen Coke Green Coke Nongreen Coke
3-Methyl Cholanthrene x x x
Benz Cal anthracene* X X
Phenanthrene x
Benzo CaJ pyrene X
Phenol x
Cresol x
Quinoline x
Note: X designates the presence of the species in emission tests
under the noted process conditions.
*Assuming the total weight found at m/e 228 is attributable
to benzCaL anthracene.
Carcinogenic Hazard
Species Potential Potential
Benzo (a), pyrene +++ XXX
3-Methyl cholanthrene ++++ XXX
7,12-Dimethyl benz (a)
anthracene ++++ XXX
Dibenz (a,hi anthracene +++ XXX
Dibenzo Ca,h} pyrene +-M-
Dibenzo Ca,il pyrene -H-+ XX
Benz CaJ anthracene + XX
Pyridine x
Indeno (1,2,3-cdlpyrene +•
Rating system:, -M-, +++, or ++++ strongly carcinogenic
+ carcinogenic
XXX most hazardous
XX very hazardous
X hazardous
-82-
-------
conditions under which they where found. It should be noted
that other compounds were also present that fall into the
categories labelled "carcinogens" by a Public Health Survey"
and "hazardous" by Multimedia Environmental Goals (MEGs) *16*.
These compounds are also identified in Table 8-9. Application
of these four rating systems to the organic emission results
has shown thirteen compounds present in quench tower emissions
to be potentially harmful to some degree.
8.2 Sample Contamination
Silicone grease and phthalates were detected in coke quench
tower samples by infrared analysis. The presence of these
substances in each test is indicated in Tables 8-10 and 8-11.
Although silicone grease and phthalates are considered sample
contaminants for the purpose of this report, their origin and
method of entry into the test samples is not well defined.
Silicone grease is reported in Table 8-10 as present in service
water C#4I , inlet water CH)_, blank methylene chloride (trace)
and blank water (tracej_ samples. It would have been difficult
for the field test crew to contaminate the samples of water
with silicone grease. It might be speculated that silicone
grease could enter the water supply through its common use on
pump and valve packing glands.
The presence of silicone grease in the cyclone catch of Test 5
Csee Table 8-11T may also indicate some other source of con-
tamination since silicone grease was used only on the con-
denser and impingers. No silicone grease was detected in the
probe/filter samples for Tests 5, 7 and 14 which supports the
statement that no silicone grease was used in the front half
of the train.
-83-
-------
TABLE 8-10
PRESENCE OF SILICONE GREASE AND PHTHALATES
AS INDICATED BY IR
Test No. Silicone Grease
2B
3
4
5
6
7
8
9
10
11
12
14
15
16
17
18
Service water (#4)
Inlet water (#4)
Liquor water (#9)
Inlet water (#9)
Blank methylene
chloride
Blank water
yes
yes
-
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
-
yes
yes
yes
yes
-
-
trace
trace
Phthalates
yes
yes
yes
yes
yes
yes
-84-
-------
TABLE 8-11
PRESENCE OF SILICONS GREASE AND PHTHALATES
AS INDICATED BY IR
Sample Silicone Grease Phthalate
Test #5
Cyclone yes
Probe & Filter
XAD-2 Module yes
Test #7
Cyclone - yes
Probe & Filter
XAD-2 Module yes
Test #14
Cyclone - yes
Probe & Filter - yes
XAD-2 Module yes yes
Test #18 yes yes
-85-
-------
Phthalate was reported as a contaminant in three of the eleven
clean water tests and in two of the four contaminated water
tests (see Table 8-10). But phthalate was detected by GC/MS
in some quantity in every one of the organic emission tests
(from 1.4 to 149 grams per metric ton of coal). Phthalate was
also found in the 1976 organic emission tests ^. At that
time the most likely candidates were thought to be:
• Butyl benzyl phthalate
• Dimethyl cyclohexyl phthalate
• Mixed alcohol phthalates
Phthalatic acid would be one of the oxidation products of any
benzene derivative having only two side-chains in the 0-
position, such as the oxidation of naphthalene. Therefore
some phthalates would be expected in coke quench tower emis-
sions and phthalate has been measured in coke oven door
leakageC24).
Although phthalate was present in all of the tests, in the
contaminated water tests phthalates are present at levels of
at least one order of magnitude greater than in the clean
water tests (see Tables 8-4 and 8-5). This substantiates the
theory that a significant portion of the phthalates may have
come from coke oven processes.
8.3 Benzene Soluble Residue Test Results
The quantity of benzene soluble residue found in coke quench
tower emissions appears to be substantial as shown in Table
8-12 with at least 234 grams per metric ton of coal. Most of
the benzene solubles Cat least 50%). were collected in the
adsorber. Vaporous material and particles smaller than 0.3
micrometers are caught in this component.
-86-
-------
TABLE 8-12
BENZENE SOLUBLE RESIDUE
G/Metric
Sample mg/m3 Ton of Coal
Test No. 1 - clean,3a
probe wash 38.2 42.3
cyclone 13.2 14.6
filter 3.95 4.35
condenser 1.26 1.39
adsorber 155 172
Total 212 234
Test No. 2 - clean,1.8
probe wash 10.9 14.9
cyclone 5.19 7.10
filter no sample
condenser0 101 138
adsorber 129 177
Total 246 337
Test No. 13
contaminated,1.5
probe wash 18.0 17.6
cyclone 8.74 8.55
filter 38.7 38.0
condenser 81.6 80
adsorber 160 156
Total 307 300
Quench water quality, Coke greenness
Filter burned during test
This benzene soluble sample appeared contaminated.
-87-
-------
8.4 Benzene and Total Hydrocarbons (THC)
Benzene and THC concentrations detected in grab samples are
given in Table 8-13.
TABLE 8-13
BENZENE AND THC EMISSIONS
grans/metric
Date ppm ton of coal
Benzene: 11/7/77 0.005 0.01
11/8/77 0.040 0.13
THC: 11/7/77 8.54 30
(as Hexane) 11/8/77 17.34 60
Note: Clean quench water was in use on the test dates.
The concentrations of THC are comparable to the value of 12
ppm THC found in the 1976 Lorain tests^ .
8.5 Biological Test Results
One of the bioassays to which the coke quench tower biological
sample was subjected is the Ames bacterial test for mutagenicity.
Strains of Salmonella typhimurium unable to synthesize the
amino acid histidine (due to a mutation, his") are used in the
test. These bacteria cannot grow unless histidine is supplied
to them. However, the event of a back mutation to his"1" would
allow the bacteria to grow without the external histidine
supply'. The potential mutagenic properties of a material such
as an extract of coke quench tower emissions may be tested in
this way when the material is applied to his" bacteria. In
addition, it is sometimes not the original form of such materials
-88-
-------
that is mutagenic, but one of its metabolites. Thus, the
quench tower sample was also subjected to enzymes of a rat-
liver extract (the site of major metabolic processes) in order
that the sample metabolites also be tested.
The results of the Ames test are presented in Table 8-14.
It is apparent that the number of revertant colonies is not
significantly different from the solvent controls nor close to
results for the positive controls. The test results for
nonactivation and metabolic activation systems are negative ,
Toxicity tests on the same strains of Salmonella were also
performed and the results given in Table 8-14. The population
counts may be compared to those of the solvent controls and
positive controls to show that the test material was not toxic
to any of the indicated organisms employed. It was noted in
the original data that for nonactivation assay results for
strain TA-1537 the solvent control colony counts were low, as
were those for the highest dose of emission sample. This may
be due to a technical error in diluting stock cells .
Another bioassay, the clonal cytotoxicity assay, was performed
on the coke quench, tower biological sample. This test measures
the effects of varied doses of the sample on the colony forming
ability of cultured Chinese hamster cells (CHO). The sample
concentration that reduces the number of colonies by 50% (EC50
value!" is determined and compared with defined toxicity levels.
The results of the clonal cytotoxicity assay are presented in
Table 8-15. The original sample had been concentrated 8-fold
prior to application to the test colonies, thus the "applied
concentrations" are multiplied by 8 to obtain the original
concentration values. The cloning efficiency of 72.0% for the
untreated control culture indicates that the cells were in a
-89-
-------
healthy state and good cloning conditions were provided. The
lower cloning efficiency of 59.4% for the solvent control may
reflect artificially low colony numbers for this control since
the number of colonies obtained with the lower concentrations
of test material exceeded both the solvent and untreated con-
trols(14).
Comparable numbers of colonies appear for the triplicate
dishes at each sample concentration level, except for the
original concentration of 4.0 yl/ml. The percent relative
survival for each sample dose was derived from a comparison of
the average number of colonies per test dish to the average
number of colonies in the solvent control. These survival
values were plotted in Figure 8-1 to estimate the sample's EC50
value. A smooth dose-response curve was obtained which passed
through the 50% survival level at 100 yl/ml. However, the
real position of the EC50 is likely to be less than 100 yl/ml
since the survival curve itself rises far above the 100% sur-
vival level (reflecting the low colony count for the solvent
control). Assuming that 153% survival at 4 yl/ml represents
the true 100% survival level, the true EC50 would occur at 76%
survival in Figure 8-1, corresponding to approximately 74
yl/ml. Thus, the EC50 probably lies between 74 and 100 yl/ml
The EC50. range for low toxicity is defined as 60 to 600 yl/ml
in Litton Bionetics1 response to Technical Directive No. 301.
According to this definition, the coke quench tower emission
sample is of low
-90-
-------
TABLE 8-14
RESULTS OF MUTAGENICITY AND TOXICITY ASSAYS
OF A COKE QUENCH TOWER EMISSION SAMPLE
I
VO
Salmonella strains:
Solvent Controls .
Positive Controls
Mg/plate of sample:
0.01000
0.10000
1.00000
10.00000
Solvent Controls .
Positive Controls
Mg/plate of sample:
0.01000
0.10000
1.00000
10.00000
(Results
TA-1535
1
t
MUTAGENICITY
given in revertants per
TA-1537
1
2
TA-98
1
2
plate)
TA-100
1
2
TOXICITY
(Results given in population
countB -10^ dilution - per plate)
TA-1535
TA-1537
TA-98
TA-100
Non Activation Tests
8
708
6
B
10
11
4
45
3
9
8
7
7
696
7
8
7
8
9
46
7
8
3
9
21
168
1 j
16
16
16
19
293
16
25
25
12
11
167
20
18
16
14
20
309
15
23
15
17
48
0
45
50
54
64
Rat
35
1,072
41
31
31
31
54
296
67
53
49
59
Liver
33
923 1,
31
33
19
33
146
0
145
176
168
174
165
2,666
158
159
133
163
Activation
155
889
215
185
185
204
195
1,844
171
179
178
188
42
21
19
31
40
27
Tests
59
27
43
44
39
52
6
0
16
14
22
6
66
6
36
53
46
50
196
190
55
121
167
188
0
233
210
210
231
229
56
64
70
68
68
54
56
44
86
83
67
73
aSolvent used was dimethyl sulfoxide, 50 pi/plate
Positive controls for each Salmonella strain were:
TA-1535, TA-100 N-Methyl, N-Nitro, N-Nitrosoguanidine 1 UG/Plate
TA-1537 9 - Aminoacridine 50 UG/Plate
TA-98 2 - Nitrofluorene 10 UG/Plate
Note: These data are the results of tests performed by Litton Bionetios.
-------
TABLE 8-15
RESULTS OF THE CHO CLONAL CYTOTOXICITY ASSAY OF A COKE QUENCH TOWER EMISSION SAMPLE:
COLONY COUNTS AND PERCENT RELATIVE SURVIVAL
K)
I
SamPle
Applied
Concentration
pi/ml
Original
Concentration'
lil/ml
Colony Numbers
DishDishDish
#1 «2 #3
Avg
Relative
Survival
%
Cloning
Efficiency
%
Untreated Control
Solvent Control
Test
Test
Test
Test
Test
Material
Material
Material
Material
Material
0
20
0.
2.
5.
10.
20.
5
0
0
0
0
—
—
4.
16.
40.
80.
160.
0
0
0
0
0
360
284
573
399
355
199
12
„_ _
310
454
465
345
211
19
_ _
—
339
400
308
210
20
360 121.2 72.0
297 100.0 59.4
455 153.2
421 141.8
336 113.1
207 69.7
17 5.7
The original sample was concentrated 8-fold during the solvent exchange to DMSO.
Solvent used was dimethyl sulfoxide.
Note: These data are the results of tests performed by Litton Bionetics. (14)
-------
FIGURE 8-1
ECSO DETERMINATION
COKE QUENCH TOWER EMISSION SAMPLE
cc
20
2 4 6 10 20 40 60 100
TEST CONCENTRATION,
1000
Note: This data analysis supplied by Litton Bioneticsfl4)
-93-
-------
9.0 PARTICULATE EMISSIONS
9.1 Particulate Emission Results
Samples from the quench tower emission tests for organic matter
and for benzene soluble residue were also analyzed for particu-
late; these results are reported in Table 9-1 and Table 9-2
respectively.
The breakdown of particulate results by sampling train component
(Table 9-1) reveals similar concentrations for the cyclone and
filter in clean water tests. However, the cyclone concentrations
show a substantial increase in contaminated water tests, whereas
filter concentrations show a decrease. These data indicate a
fairly even particle size distribution for clean water tests and
an increased concentration of larger particles for contaminated
water tests. Particle size is discussed further in Section 9.2.
The data in Table 9-1 also show that total particulate emissions
are greater for the contaminated water tests and that there is
no substantial difference in test results for nongreen coke tests
versus green coke tests.
Although no correlation of greenness to the concentration of
particulate in the stack aerosol is revealed in Table 9-1, visual
observations indicate that 1) there is a relationship between
coke greenness and the visible particulate above the quench car
(see Section 6.1); and 2) there is a relationship between coke
greenness and a dark haze observed exiting the quench tower
before the water is turned on to a car of green coke. The
dramatic reduction in tower particulate emissions during the
actual quenching of the green coke might be due to the scrubbing
mechanism taking place when the top layers of coke are cooled
and then wetted by the continued spraying of the quench water.
This reduction in plume particulate could minimize the effect
of coke greenness on quench tower emissions.
-94-
-------
Table 9-1
COKE QUENCH TOWER PARTICULATE EMISSIONS
Test No.
Coke
Cyclone
Nozzle/Probe FiJ
.ter
Total
g/metric g/metric g/metric g/metric
Greenness ton ton ton ton
(avg.) mg/rn-^ coal ma/m^ coal mg/m^ coal mg/m^ coal
Clean
Water Tests
7
10
9
4
6
8
11
3
2B
12
5
Contaminated
Water Tests
16
15
17
14
1.2
1.5
1.6
1.6
1
1
3
3
3
3
4
0
0
0
1
.6
.8
.2
.3
.4
.7
.0
.3
.5
.5
.5
445
278
220
277
456
306
194
275
263
415
412
962
891
633
863
382
261
209
255
500
261
198
311
262
412
377
924
874
710
1040
279 240
67 63.3
98 93.5
No Sample
59
64
63
28
32
62
35
61
123
89
89
64.5
55.0
64.1
31.8
32
61.8
31.8
58.8
120
100
108
405 348
Negative
Negative
457 420
Negative
345
27.0
355
397
295
27.7
400
397
Negative
360
167
202
178
59.0
330
160
197
200
71.0
1130 970
N/A
N/A
N/A
N/A
716
283
659
692
611
290
743
691
N/A
807
1190
1216
901
1011
739
1140
1190
1010
1220
"NA" = Not Available
"Negative" indicates a negative sample weight. It has been deter-
mined that this result is most likely due to holes in the filter
which formed during sampling, allowing particulate to escape to the
back half of the train. See Section 6.6 for a detailed discussion.
-95-
-------
Table 9-2
BENZENE SOLUBLE RESIDUE TESTS
TOTAL PARTICULATE
Benzene insoluble particulate determined for each test is added
to the benzene soluble residue to yield total particulate.
mg/m-
g/metric
ton of coal
SAMPLE
Test No. 1 - clean,3*
cyclone
nozzle/probe
filter
condenser
adsorber
Total
Test No. 2 - clean, 1.8
cyclone
nozzle/probe
**filter
***condenser
adsorber
Total
Test No. 13 - contaminated, 1.5
cyclone
nozzle/probe
filter
condenser
adsorber
Total
* Quench water quality, coke greenness
** Filter burned during test
*** This benzene soluble sample appeared contaminated
236
38.2
3.95
34.6
155
468
219
22.8
no sample
146
129
517
1.5
434
33.7
38.7
349
160
015
262
42.
4.
38.
172
519
301
31.
201
177
710
424
33.
38.
342
157
994
4
35
2
3
1
0
-96-
-------
-J
I
TAUI.b' 9-3 '
I'AKTICULATi: EMISSION SUMMARY
Organic Emission Tests
CONTAMINATED QUENCH WATER
Sampling
Train average
Component: (range)
Cyclone (20iim,
50% cut size)
Probe/Nozzle
Filter
Total Front Half
Cyclone (20pm,
50% cut size)
Probe/Nozzle
Filter
Total Front Half
Cyclone (20pm
50% cut size)
Probe/Nozzle
Filter
Subtotal - Front Half
Nongreen Coke
(Tests 7,8)
kg/metric
mg/m3 ton coal
380 0.32
(310-450) (0.26-0.38)
170 0.15
(64-280) (0.055-0.24)
370 0.32
(350-410) (0.30-0.35)
920 0.79
(720-1100) (0.61-0.97)
Total (Tests 2B,
mg/mj
320
(190-450)
84
(28-280)
320
(27-400)
720
(280-1100)
CLEAN WATER - GREEN
mg/m-1
240
38
4.0
280
Total (including front 470
half condenser,
adsorber)
Green Coke
(Tests 20,3,5,11)
kg/metri c
mg/m ton coal
290 0.29
(190-410) (0.20-0.41)
40 0.039
(32-63) (0.032-0.061)
280 0.29
(27-400) (0.027-0.40)
610 0.62
(280-810) (0.29-0.74)
3,5,7,8,11)
kg/metric ton coal
0.30
(0.20-0.50)
0.075
(0.032-0.24)
0.30
(0.027-0.42)
0.68
(0.29-0.97)
Benzene Soluble
COKE (Test l)b
kg/metric ton coal
0.26
0.042
0.0044
0.31
0.52
Nongreen Coke
(Tests 14-16)
kg/metric
mg/m^ ton coal
840 0.89
(630-960) (0.71-1.04)
91 0.10
(61-120) (0.059-0.12)
150 0.16
(59-200) (0.071-0.20)
1080 1.1
(900-1200) (1.0-1.2)
Total (Tests
mg/mj
840
(630-960)
91
(61-120)
150
(59-200)
1080
(900-1200)
Residue Tests
Green Coke
(No Tests)
kg/metric
mg/m3 ton coal
_
-
— "~
— —
14-17)
kg/metric ton coal
0.89
(0.71-1.0)
0.10
(0.059-0.12)
0.16
(0.071-0.20)
1.2
(1.0-1.2)
CONTAMINATED WATER - NONGREEN COKE (Test 13)
mg/mj
430
34
39
500
1020
kg/metric ton coal
0.42
0.033
0.038
0.49
0.99
aSampling train components for some tests (4,6,9,10,12) have negative results or no sample (see Table 9-1). Results for these
tests were not incorporated into this summary. No major difference occurs in the average values when this procedure is used.
bTest No. 2 was not included in these summary results since there war, no filter sample and the condenser sample appeared
contaminated.
-------
Particulate data from the benzene soluble residue tests (Table
9-2) were obtained by adding the benzene insoluble particulate
results to the benzene soluble residue (reported in Table 8-10) .
In addition to the front half (cyclone, nozzle/probe, and filter)
the condenser and adsorber were also analyzed for each test.
A summary of all the particulate data is presented in Table 9-3.
Considering the organic tests, the clean water-nongreen coke
tests average 0.79 kg/metric ton of coal and the clean water -
green coke tests average 0.62 kg/metric ton of coal. The clean
water tests average 0.68 kg/metric ton of coal and the contamin-
ated water tests, 1.1 kg/metric ton of coal. Again, there is
a substantial increase in particulate concentration for contam-
inated water tests, but no notable difference in results for
green versus nongreen coke tests. The benzene soluble residue
tests show only half the concentration of front half particulate
found in the organic tests.
9.2 Particle Size
Some important aspects of the size distribution of particulate
in coke quench tower emissions are revealed in Table 9-3. The
sampling train provides a breakdown of particle size. The
cyclone was designed to capture aerosols generally large enough
to be easily visible - the 50% cut size was about 20 micrometers.
The D-,. cut size of the "in-stack" cyclones at test conditions
was calculated for each test as shown in Table 9-4, Column 11.
The D,-n cut size of any given cyclone is determined by a number
of variables. These variables include gas density, gas viscosity,
gas inlet velocity, particle density, and particle size and
shape. The value of the gas density is so small relative to
particle density that, in most cases, variation in gas density
will have a negligible effect on cyclone separation efficiency
or calibration. For a given set of conditions in which particle
density, gas density, and gas viscosity are assumed to be fixed,
-98-
-------
TABLE 9-4
CUT SIZE OF "IN STACK" CYCLONE AT TEST CONDITIONS
TEST
NUMBER
(1)
1
2
2B
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
TEST
TIME/MIN
CAR IN
CAR OUT
(2)
12.3
14.3
12.1
11.6
12.1
12.9
11.9
12.4
9.4
12.1
9.6
10.4
3.2
14 .8
13.5
9 .7
9.7
15.8
GAS
DSCF PER TEST STACK LBS DRY GAS
PERIOD TEMPERATURE DURING TEST
(METER DATA) °F PERIOD
(3) (4) (5)
20.78 175 1.56
27.65 175 2.07
9.7 175 0.73
10.41 175 0.78
7.95 175 0.60
10.17 175 0.76
10.69 175 0.80
9.67 175 0.72
9.96 175 0.75
10.30 175 0.77
9.11 175 0.68
10.43 175 0.78
7.15 175 0.54
12.28 175 0.92
12.47 175 0.93
8.74 175 0.66
9.56 175 0.72
13.77 175 1.03
WATER
VOLUME OF
VAPOR
CONDENSED
(ml)
(6)
102.5
237.6
75.6
54.3
39.8
74.4
73.1
52.1
107.6
116.9
59.9
41.7
67.7
82.2
56.7
64.2
76.7
90.8
VOL. OF MIXTURE DRY GAS AND IMPINGER
HATER AT STACK TEMP.
LBS H2O/LBS DRY FLOW OF
GAS IN SAMPLE ACF PER MIXTURE
AFTER CYCLONE POUND ACF ACFM
(7) (3) (9) (10)
0.14 19.6 30.6 2.5
0.25 22.4 46.4 3.2
0.23 21.8 15.9 1.3
0.15 19.8 15.4 1.3
0.15 19.8 11.9 1.0
0.22 21.6 16.4 1.3
0.20 21.1 16.9 1.4
0.16 20.3 14.6 1.2
0.32 23.7 17.3 1.9
0.33 23.8 10.3 1.5
0.19 20.6 14.0 1.4
0.12 18.7 14.6 1.4
0.20 22.5 12.1 1.5
CYCLONE
D50
CUT SIZE
MICRONS
(11)
11
10
24
24
30
24
22
26
16
21
22
22
21
Clean Test Average 18
0.20 21.1 19.4 1.3
0.13 19.5 13.1 1.3
0.21 21.3 14.1 1.5
0.23 21.8 15.7 1.6
0.19 20.6 21.2 2.2
Contaminated Test Average
24
24
21
20
15
21
-------
the Dj-f, cut size of a given cyclone is determined largely by
(19)
the gas inlet velocity and the particle size and shape.
Calibration data for this series of cyclones has been utilized
to construct Figure 9-1. This graph relates sample flow rate
in actual cubic feet per minute to the D_Q cut size in micro-
meters. ACFM is, in this case, used as a convenient indication
of inlet gas velocity. It is an average value based on the
measured amount of dry gas recorded by the dry gas meter (Table
9-4, Column 3), the elapsed time of the test (Table 9-4, Column
2), and the volume of water vapor. These values are corrected
to actual stack conditions at the temperature of the test
(Table 9-4, Column 4). Following the determination in ACFM
of the gas flow for each test (Table 9-4, Column 10), the D5Q
cut size in micrometers is read from Figure 9-1. Since the
inlet velocity was an average, the D_Q cut size is also an
average.
In the following example, the D__ cut size is determined for
Test 4:
Example
7.95 dscf x 0.07495 Ibs/cf = 0.60 Ibs dry gas during test period.
The volume of water vapor condensed and collected in the impinger
and the condenser (Table 9-4, Column 6) is converted to pounds
of water vapor per pound of dry gas (Table 4, Column 7) as
follows:
39.8 ml H2O =0.15 Ibs water/lb dry gas
454 ml/lb H20 x 0.60 Ibs dry gas
This value (the absolute humidity) is used with a reference
table (2°) to determine the ACF/lb of the mixture of dry gas and
water vapor. The water droplets collected in the cyclone is
not included in these calculations.
-100-
-------
Figure 9-L
Cut point in micrometers vs cubic feet per minute
0
0
— 1
o
in
o
V
o
r*l
O
(N
O
-^
in
0
C
\
\x
\
.01*- 0.5 1 23
^
x
\
I
s
>
\
N
\
v\
x\
i 10
• SASS 10 Density 1.0 g/cubic centimeter
<3 SASS 10 Density 2.0g/cubic centimeter
4r Actual cubic feet per minute
-101-
-------
ACF per pound (Table 9-4, Column 8) is converted to ACF (Table
9-4, Column 9):
19.8 ACF/lb x 0.60 = 11.8 ACF
ACF of mixture is converted to ACF per minute (Table 9-4, Column
10) by dividing the elapsed time of the test (Table 9-4, Column
2):
11.8 ACF = 0.97 ACFM (Table 9-4, Column 10).
12.1
This actual cubic feet per minute value is entered on the ordinate
of Figure 9-1. Assuming a particle density of 2.0 gm/cm , the
Dj-n cut size for conditions of test 4 is read as 32 micrometers.
The DSO cut sizes for each test are tabulated in Table 9-4. The
equivalent mass diameters for the clean and dirty tests are as
follows:
Equivalent mass diameter - Dirty 21 micrometers
Equivalent mass diameter - clean 18 micrometers
The weight of dissolved solids later recovered from the cyclone
water was not used in calculating these equivalent mass diameters.
If it were used, the average would be about 0.5 micrometers
smaller.
9.3 Comparison of Quench Tower Particulate Emission
The fraction of particulate caught in each part of the sampling
train during the 1977 tests (Table 9-3) may be compared to the
results shown in Table 9-5 of previous tests at Lorain on the
same quench tower in 1976. The 10 ym cyclone used in 1977
would collect basically the same fraction as the D_0, 20 um
cyclone used in 1977, since by definition, 10 micrometer part-
icles would be collected with 50% efficiency, greater than 10
micrometer particles with greater than 50% efficiency, and less
than 10 micrometer particles with less than 50% efficiency.
Only the 10 micrometer to 20 micrometer particles would not
have been as efficiently collected in the 1977 tests. Since
the cyclone collection was greater in 1977, any difference in
-102-
-------
TABLE 9-5
QUENCH TOWER PARTICULATE EMISSIONS
U.S. STEEL CORPORATION LORAIN WORKS
QUENCH TOWER NO. 1 - NOVEMBER 1976^
Sampling
Train
Component
Clean Make-up Water
(13 Test Runs)
Kg/metric ton of coal
Average (range)
Contaminated
Make-up Water
(12 Test Runs)
Kg/metric ton of coal
Average (range)
Cyclone (10 micrometers, 0.14 (0.048-0.027)
50% cut size)
Probe
0.53 (0.15-0.90)
Filter (99.99% efficiency 0.0075 (0-0.037)
for particles 0.3 micro-
meters and larger)
Front half of sampling 0.73 (0.34-1.0)
train catch
Back half of sampling
train catch
0.080 (0.025-0.16)
0.59 (0.34-0.34)
0.58 (0.24-1.2)
0.23 (0.15-0.37)
1.4 (0.37-2.2)
0.40 (0.18-0.83)
Total Quench Tower
Particulate Emissions
0.31 (0.38-1.2)
1.8 (1.1-3.0)
-103-
-------
particle collection due to this fraction would be relatively
minimal. The fact that the total amount of particulate col-
lected in the cyclone was greater than in 1976 could have been
the result of the increase in the open area in the center of
the baffles (see Table 9-7) .
The total front half of the sampling train catch in the 1977
tests is very similar to the 1976 results. Again there was
a dramatic decrease in the emissions when clean water was used
rather than contaminated water. However, the distribution of
size fractions as collected in different parts of the train
is somewhat different from that in the 1976 tests. The weight
of larger aerosols collected by the initial cyclone in 1977..
was about 45% of the total weight emitted when quenching with
clean water and about 75% of the contaminated water quench
emission. The average size of the particulate in the plume
aerosol was less than 4 micrometers in the 1976 tests but this
would be shifted towards a larger size in the 1977 Lorain tests
due to the greater cyclone collection.
The results of tests at a Canadian plant (Table 9-6) indicate
that the particulate per ton of coal in the Canadian tests was
about 20% of that found during the 1977 Lorain tests. The
large size fraction caught in the cyclone during the 1977
Lorain test (larger than 20 micrometers) was about 12 times as
much as in Canada. This would be expected considering the
open area in the Lorain baffles. The amount caught in the probe
nozzle and condenser at Lorain was about the same as that caught
in the probe and nozzle during the Canadian tests. When compared
to the Canadian tests which used clean water for makeup, the
amount of particulate caught on the filter was 12 times greater
during the organic matter tests at Lorain using clean water as
makeup.
-104-
-------
TABLE 9-6
QUENCH. TOWER PARTICULATE EMISSIONS
DOMINION FOUNDRIES AND STEEL, LTD.
AUGUST 1977(21)
Sampling Train
Component
Cyclone (10 micrometers,
50% cut size)
Probe
Filter (99.99% efficiency
for particles 0.3 micro-
meters and larger)
Front half of sampling
train catch
Back half of sampling
train catch
''Clean" Make-up Water
(9 Test Runs)
Kg/metric ton of coal
0.025 (0.0050 - 0.075)
0.090 (0.060 - 0:13)
0.025 (0.015 - 0.060)
0.140 (0.095 - 0.20)
0.10 (0.055 - 0.19)
Tower Quench Tower
Particulate Emissions
0.24 (.0.16 - 0.31)
-105-
-------
9.4 Baffles
Quench tower No. 1 at Lorain was equipped with baffles (Figure
9-2) as described in Figures 9-3 and 9-4 which were based on
U.S. Steel Drawing No. LA-100891 . These figures were some-
what modified however, as described below.
As can be seen in Figure 9-4 the baffles were constructed in
such a way as to allow a vertical area of one inch between the
bottom of one baffle and the top of another. This verticle
open area allowed a substantial portion of the stack flow to
pass straight through the baffles with little opportunity for
even larger size particles to impinge on the baffle surfaces.
Several baffles had been removed and this provided additional
open area. The U.S. Steel drawing calls for 11 slanted boards
in each baffle section (See Figure 9-3) but only 10 full boards
in 5 of the sections and 6 full boards in one section were
actually in place (Figure 9-5) at the time of the test. The
effect of these open areas would be to reduce the resistance
to gas flow and increase the draft up the tower. There were
come charred remains which indicated some of,these baffles had
been burned out. The effect of removing baffles was to increase
the vertical open area as can be seen in Table 9-7 which compares
the open area of the baffled section in 1976 and 1977.
-106-
-------
FIGURE 9-2
BAFFLED SECTION OF QUENCH TOWER
i
i-1
o
i
^mm\\\\\\\\
BAFFLED
SECTION
\V\\\\^\
o
r~
i
i
o
P
BAFFf.RS
OOOOOO
COKE QUENCH
CAR
H
SPRAY
NOZZLES
/ /
d
/ i
o
1
1
,
BAFFLES
v»
V
-------
.Figure 9-3
DETAILS OF BAFFLES
IV
I
Ik"
o
oo
I
(10 Baffles)
7'-6"
Six sections installed
each section 6'-3V1 x 9'-0" overall
6'-OV x 9'-3" inside
-------
FIGURE 9-4
OPEN AREA BETWEEN BAFFLES
I
I-1
o
I
V
9"
Typical
.6-3/4"
8"
-------
Figure 9-5
BAFFLZD SECTION SHOWING
MISSING BOARDS
N
Two rows of baffles out at time of 1976 tests
Additional 3 baffles i
1977 tests
Six sections desined for 11 baffles each
-110-
-------
TABLE 9-7
BAFFLE OPEN AREA
Square feet
Total open area of the tower before
any baffles were installed from
Dwg. LA-100891 (U.S.S. Corp.)
19' - 2" x 19' - 2"
Total vertical open area of the
tower after the baffles were installed.
Based on U.S. Steel Corp. Dwg.LA-100891-
11 rows of baffles in each section
Total vertical open area with 10 rows
of baffles in 5 sections and 6 rows in
one section as actually installed at time
of test program
Open area in center of tower due to 1976
Omission of baffles 1977
367
70
106
24
36
Open area
as a % of
total area
100
19
29
6
10
The efficiency of these baffles and more detailed discussion is
to be found in the report of the 1976 test series
-111-
-------
10.0 QUENCH WATER ORGANICS AND WATER FLOW
10.1 Organics in Quench Water
The amounts of organic species detected by GC/MS in the quench
water (both the inlet of the spray nozzles and makeup to the
sump) are given in Tables 10-1 and 10-2.
The flushing liquor used as makeup during the contaminated water
tests contained almost all of the species eventually found in the
stack emission. However, as molecular weight increases the ratio
of the quantity of species in the stack emission to the quantity
introduced in the makeup liquor becomes less and less. The
lower the molecular weight the higher the boiling point; therefore,
the lower molecular weight species are readily stripped from the
quench water by the evaporation and distillation process of the
quench. The same relationship is also apparent when comparing
the quantity of each organic species introduced to the quench
system by the makeup liquor and the quantity of each species
available from the inlet water which was being used repeatedly
to quench the coke. As would be expected, the inlet water is
stripped of lower molecular weight species and the higher molecular
weight species tend to remain in the quench water. A few of
these higher molecular weight compounds do not even appear in
the stack emissions. There is no sharp cut off due to the high
temperature of the uncooled portions of incandescent coke and
the stripping action of the steam which tends to lower the boil-
ing point of the higher molecular weight compounds.
In the case of the clean quench water samples only a small
quantity of organics was found. The organic matter in the stack
emissions from clean tests could come from a number of sources
including the sludge in the sump (where organics from previously
used flushing liquor could collect) or the more likely possibility
that organics from the coke itself are released during the quench-
ing process and become part of the quench tower emission.
-112-
-------
TAULE 10-1
QUENCH WATER 'ANALYSIS
(in
POLYCYC1.1C AROMATIC HYUKOCARUONS
Clean Wjter
Service Inlet
Species Water to Nozzle
Indcne
Naphthalene 3.09 5.86
Ben zoth iophene
Methyl Naphthalenes
Acenaphthylene/Uiphenylene 0.275 1.59
Biphenyl 1.04 1.47
Dimethyl Naphthalenes
Fl uorcne
Carbazole
UibenzoCurcin/methyl bi phony 1
Anthracene/phenanthrenc 65.6 11.1
Dibenzoth iophene
Methyl anthracenes 39.8 4.09
Fluoi-anthene 1.92 9.48
Pyrene 3.03 13.9
C16H12PAH
C16II12PAII
C16II12PAH
C16II16PAH
Methyl Fluoranthene & Methyl
Pyrene
Dihydrobenzofluorene
Chrysene/Benz (a)anthriicencs 6.27
Naphthobenzoth iophene
Methyl-chrysenes
Benzof 1 uoranthene , benzo (e) pyrene &
Benzo (a)pyrene
Perylene
7,12 Dimethyl Bc-nz (a) anthracene
Methyl-benzo pyrenes
Ui benzo (c.g)carbazolc
3-methyl cholanthrene
Indeno (1,2, 3-cd) pyrene
Benzo I ghi) per ylene
Di benzo (a , h) anthracene
Coronene
Dibenzoldi & ah) pyrenes
Totul 135 5J.8
Benzo(a) pyrene 27
Contijnii nat
Flushing
L iquor
8,240
1,270
1.610
3,080
242
233
1.670
1,320
994
4,230
194
310
1,270
801
324
569
85.7
400
483
0.589
79.0
17.1
39.1
126
27.9
27.600
300
oil Wate
r
Inlet
to Nozzle
4
>2,240
162
216
668
33
35
394
1,610
193
1,380
44
111
692
675
4
16
4
10
150
295
37
54
259
364
74
79
244
38
70
10,200
440
.82
.1
.8
.9
.49
.6
.29
.3
.4
.6
.8
.8
.9
.0
POLAR COMPOUNDS
Clean Mater Contaminated Water
Polar Compounds
Pyridine
Methyl Pyridine
Aniline
Phenol
Benzonitrile
Dimethyl/ethyl pyridine
Toluidine
Cresol
Indole
Trimethyl pyridine
Methyl cresol
Quinoline
Isoqulnoline
Trimethyl phenol
Methyl quinoline
'Phthalates
Dimethyl/ethyl quinoline
Acridine
Service Inlet Flushing
Water to Nozzle Liquor
1.14 7,880
1.14 4,700
3,000
2,770
3,080
0.762 8,200
2,060
1,410
13,500
1,560
23,400
6,900
1,980
>1,660 >2,380
93.
745
Inlet
to Nozzle
134
1,510
888
2,250
866
1,420
175
122,000
22,900
338
10,600
7,090
1,040
>1,620
3 87.5
531
569
85.7
400
483
0.589
79.0
17 . 1
39.1
126
27.9
295
37.4
54.6
259
364
74.8
79.8
244
38.9
70.0
• Phthalates are considered to be a contaminant.
** These values for benzo (a) pyrene were determined by a special analysis
with high sensitivity for BaP.
All blanks indicate that the concentration of the species wa,s be Ion
the detection level. The detection limit for each PAH species varies
with molecular weight, reaching differences from five to twenty
times with a lower limit at naphthalene and a higher limit at
coroneiic. This "upper limit* for each test is as follows:
Clean Water Contaminated Water
Service Inlet Flushing Inlet
Water to Nozzle Liquor to Nozzle
Detect ion
Limit in <48 $66 <110 <.56
pg/i
-------
TABLE 10-2
QUENCH WAT12H ANALYSIS
(in grams/metric ton of coal)
POLYCYCLIC AROMATIC IIYUHUCAKBOHS
POLAK COMPOUNDS
Clean Hater Contaminated
Service Inlet Flushing
Species
Indene
Naphthalene
Benzoth iophene
Methyl Naphthalenes
Acenaphthyltne/Biphenylene
Bi pheny 1
Dimethyl Naphthalenes
Fl no rone
Carbazole
Dibenzofuran/inethyl biphenyl
Anthracenu/phenanthrene
Dibenzoth iophene
Methyl anthracenes
Fluoranthene
Pyrene
C16H12PAII
Cl6H12l'AH
CI6II12PAII
C,6II16PAH
Water to Nozzle Liquor
0.00187 0.00354 4.
0 .
0.
0.000166 0.000960 1.
0.000630 0.000890 0.
0.
1.
0.
0.
0.0515 0.00670 2.
0.
0.0241 0.00247 0.
0.00116 0.00575 0.
0.00183 0.00840 0.
98
765
975
86
146
141
01
800
600
56
117
187
765
485
Wa to r
Clean Water Contaminated
Inlet to
Nozzle
0
>1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.00291
.36
.0980
.131
.404
.0200
.0216
.238
.970
.117
.835
.0271
.0670
.418
.408
.00271
.0100
.00259
.00620
Species
Pyridine
Methyl Pyridine
Aniline
Phenol
Benzoni trile
Dimethyl/ethyl pyridine
Toluidine
Cresol
Indole
Trimethyl pyridine
Methyl Cresol
Quinol ine
Isoquinil ine
Trimethyl phenol
Methyl quinoline
Phthalates*
Dimethyl/ethyl quinoline
Acridine
Service Inlet Flushing
Water to Nozzle Liquor
0.000685 4
0.000685 2
1
1
1
0.000462 4
1
0
8
0
14
4
1
>i.oo ?1
0
0
.77
.84
.81
.67
.86
.96
.24
.850
.15
•945
.1
.16
.19
.44
.0565
.450
Water
Inlet to
Nozzle
0
0
0
1
0
0
0
73
13
0
6
4
0
>0
0
0
.0810
.910
.540
. 36
.523
.860
.106
.5
.9
.204
.40
.28
.630
.980
.053
.321
Methyl Fluoranthene and
Methyl Pyrene
Di hydrobenzofluorene
Chrysene/Denzo(a)anthracenes
Naphthobenzothiophene
Methyl-chrysenea
Benzofluoranthene, Benzo(e)pyrene
and Benzo(a)pyrene
Perylene
7,12 Dimethyl Benz(a)anthracenes
Methyl-benzo Pyrenes
Dibenzo(c,g)carbazole
3-methyl cholanthrene
Indeno (l,2,3)-cd pyrene
Benzo(ghi)perylene
Dibenzo(ah)anthracene
Coronene
Dibenzo (ai & ahlpyrenes
Total PAH
Benzo(a)pyrene**
0. 196
0.0905
0.00379
0
0
0
0
0
0
. 344
.0520
.242
.292
.000356
.0477
0
0
0
0
0
0
.178
.0226
.0330
.157
.220
.0452
0.01035
0.0237 0.0482
0.0760 0.148
0.0169
0.0235
0.00423
0.0810 0.0325 16.7 6.10
0.0161 0.181 0.267
• Phthlates are considered to be a contaminant.
*• These values for benzo(a)pyrene were determined by a special analysis
with high sensitivity for BaP.
All blanks indicate that the concentration of the species was below
the detection level.
7192 liters (1900 gallons) of make-up water were added to the sump per
quench. Therefore, 7192 liters Mere used as the basis for converting
these values from my/liter to g/metric ton of coal.
-------
TABLE 10-3
RESULTS OF T.O.C. ANALYSIS OF WATER SAMPLES
Clean Water
DATE
11/9/77
11/9/77
11/9/77
11/9/77
11/19/77
11/19/77
11/19/77
11/19/77
MAKEUP
SAMPLE
IDENTIFICATION
13819
13820
13821
13822
20222
20223
20224
20225
PPM
26
26
28
23
28
29
28
30
DATE
11/9/77
11/9/77
11/9/77
11/9/77
11/9/77
11/9/77
11/9/77
INLET
SAMPLE
IDENTIFICATION
13814
13815
13816
13824 .
13825
13826
13827
PPM
30
18
25
28
27
25
27
1977 Average = 27
Range = (23-30)
1976 Lorain Test
Average 3.8
1977 Average = 26
Range =(18-30)
1977 Lorain Test
Average 9.6
Contaminated Water
11/16/77
11/16/77
11/16/77
11/16/77
11/19/77
11/19/77
11/19/77
1977
20202
20203
20204
20205
20217
20218
20219
Average =
2,200.0
2,422.0
2,317.0
2,130.0
2,282.0
2,088.0
2,252.0
2,242
11/16/77
11/16/77
11/16/77
11/16/77
11/19/77
11/19/77
11/19/77
1977 ;
20207
20208
20209
20210
20212
20213
20214
Average =
1,510.0
1,547.0
1,545.0
1,507.0
1,198.0
1,195.0
1,197.0
1,386
Range = (2,088-2,422)
1976 Lorain Test
Average = 1,195
Range = (1,195-1,547)
1976 Lorain Test
Average = 1,098
-115-
-------
The total organic carbon found in quench water samples averaged
27 ppm in clean makeup water and 2,242 ppm in contaminated make-
up water. The average for clean inlet water was 26 ppm and for
contaminated inlet water it was 1,386 ppm. The results of the
TOC analysis are given in Table 10-3. As noted in the Table,
these test results are comparable to those for the 1976 tests
at Lorain
(1)
10.2 Quench Water Flow
Water flow quantities at quench tower No. 1, U.S. Steel Co. Lorain
Works were identified during EPA tests in 1976 and are presented
in Table 10-4.
TABLE 10-4 WATER FLOW
Description
Stack Flow
Inlet to Nozzle
(Sprayed on Coke)
Losses
- retained by coke
- car drainage
- evaporation
Return to sump
Make up to sump
Moisture in Induced Air
(gals/quench)
Identifier
Ql
Q2
Q3
Q4
Q5
Q6
Min.
900
8100
Max.
2500
10500
Avg.
1800
9200
NA
NA
NA
7
NA
NA
NA
50
150
7300
1900
20
See Figure 10-1
The flow identifiers (Ql to Q6) are used to describe the water
flow into the quench tower process in Table 10-4 and Figure 10-1.
During the 1976 tests, certain water flows such as Q, and Q2 were
measured, while others such as Q,., Qc and Q, were calculated. Q0
4 b b J
was an estimated value. The reliability of the quantities shown
in Table 10-4 was reinforced by thermodynamic considerations.
-116-
-------
FIGURE 10-1
WATER BALANCE
Ql
\_
4 ,
I 4
OD
J 1
04 SUMP
Q5
oo
i 1
^ '"Tot.
Ql =
Q2 -
3
Stack Flow Q4 = Return to Sump
Inlet to Nozzles Q5 = Makeup to Sump
Losses Qg = Moisture in
Induced Air
Q3
-117-
-------
No actual water flows were measured during the 1977 series. After
reviewing the engineering aspects of the quench water system
(there was no significant change in the equipment during the
period November 1976 to November 1977) and in view of the
thermodynamics involved it was concluded that the water flows
during the 1977 tests could be assumed to be the same as the
water flows determined in 1976.
As shown in the previous report^ even a major change in the
amount of water (Q_) sprayed on the coke will not affect the
amount of water used in cooling (evaporated). Furthermore,
with any given arrangement of nozzles and storage tanks the
single most influential factor determining the amount of water
evaporated per quench will be the amount and temperature of
incandescent coke in each car. Since these parameters are the
same during the 1977 series as they were in the 1976 series no
significant change in the amount of water used (evaporated) in
cooling the coke would be expected.
As in the 1976 tests the stack flow was measured as one of the
necessary parameters used in calculating isokinetics. In this
series of tests the EPA Method 5 sampling train was modified by
the addition of a condenser and an adsorber between the filter
and the impingers (see Figure 6-7) . The amount of water vapor
recovered in the condenser was added to the amount removed by
the cooled impingers.
The amount of water droplets collected in the in-stack cyclone
was about 60% of that collected in the 1976 series. The average
velocity of the stack flow was 23.6 fps in the 1977 series of
tests which is 76% of the 30.9 fps average during the 1976 tests.
This lower velocity would entrain fewer and smaller drops from
the baffles and the quenching area. The amount of water vapor
(the amount collected in the condenser, the impinger and the
silica gel) was 58% of that found in the 1976 test.
-118-
-------
The detailed up-stack water flow (at test point Bl) is presented
in Table 10-5. This information is included for comparison
purposes. However, for the mass balance of organic materials
the 1976 water flow data will be used. The 1977 test series
was run at one test position (Bl, See Figure 6-9) and would not
be as representative of the total stack flow as would the 1976
series. The 1976 tests were designed to provide an average of
all stack flow parameters across the cross section of the tower.
The following calculation shows how Table 10-5 was constructed:
Taking Test No. 1 (which consisted of five separate quenches)
as an example:
34 ml of water caught in the cyclone (water drops)
50 ml of water caught in the condenser
36 ml of water caught in the impinger
17 ml of water caught in the silica gel
137 ml of water as vapor
A total of 137 ml of water was collected during the five quenches.
The gas flow up the stack was apparently not uniform and the con-
centration of water droplets was probably not uniform across the
tower cross section, however, for purposes of comparing this data
to the 1976 data it is assumed that such uniformity exists.
Based upon such an assumption, the quantity of water evaporated
up the stack is given as:
Quantity of Water Collected (ml) x (Tower Diameter) = gal/test
2
3785 ml/gal. (Orifice Diameter)
For Test No. 1
137 x (188.5)2 = 4100 gal/test
3785 (0.56)2
4100 gal/test x 8.34 Ibs water = 6838 Ibs water/quench
5 quenches/test gallon
10.3 Mass Balance Around the Quench Tower
The basic functions of the quench tower are to quickly provide a
-119-
-------
Table 10-5
Up-Stack Water Flow as Extrapolated
from Cyclone, Condenser and Impinger Catch
r Cutcli (ml) par test
Quantity up-stank(pounds per
No . 0 1"
No. [JUL- Cc-sit
1 b
2 b
-" ;>
) b
•4 5
b b
1 r' 3
M '' ^
° 8 4
9 b
III 4
1 1 4
1. 2 3
1 J (,
14 b
L5 4
1 6 4
17 b
Cyc: 1 one
34
36
J5
28
12
30
28
11
27
12
17
JO
15
24
111
23
27
24
Condenser J
50
160
60
50
28
55
61
42
LOO
100
50
40
60
58
50
60
70
84
fll^ilKJ,
36
76
9
1
3
4
0
4
0
8
II
0
0
15
0
0
0
0
Sil Jc-a
e r Ge 1
17
2
7
3
10
15
12
6
8
9
10
2
8
9
7
4
7
7
'
Cyclone
1698
1796
1544
2883
1235
2426
2883
8'JO
2114
970
1719
783
2022
1618
1409
2325
2730
1617
32662
1815
Condei-
plus Impinger
plus sjlica qel
5140
11875
7824
5S59
4221
5985
7535
4206
8455
9463
6066
3288
9166
5527
4463
6470
778'j
6133
119142
661'l Average
Gallons
820
1639
1123
820
654
1009
1247
611
1267
1251
933
488
1341
857
704
1266
1261
929
18220
1012
t-u^*rv |-**^ <- y|u<
Pounds
6838
13672
9368
8433
5456
8411
10398
5096
10569
10433
7785
4071
11188
7145
5872
10561
10515
7750
153561
8531
-------
large quantity of water to cool the incandescent coke below
its ignition temperature, and to recover the excess water for
re-use.
It is concluded in the 1976 Lorain report that the contaminants
in the quench water primarily originated from two sources - the
coke itself, and the flushing liquor which has been previously
used to cool the gases produced by the coking process. Most
of the volatile material present in the coal is driven off by
the coking process. However, a certain amount of both organic
and inorganic materials are introduced to the quench process
by the coke. Some of these materials then enter the atmosphere
during a subsequent quenching operation, some are to be found
in the sludge which is removed from the sump and some are retained
or readsorbed by the quenched coke. When process flushing liquor
is used as a make up source for the quench water, additional con-
taminants are introduced because this flushing liquor was used
to cool the gases produced during the coking operations.
In 1976 the makeup to the sump (Q5) was 1900 gallons and the up
stack flow was 1800 gallons per quench. Using these data a
mass balance of the T.O.C. entering with makeup liquor and exit-
ing the quench system can be constructed. The total organic
carbon (T.O.C.) in water samples is summarized in Table 10-6.
Makeup used during the contaminated tests had 50 times more
T.O.C. than the relatively clean service water.
This mass balance (.Table 10-6) reveals that of the organic com-
pounds identified (PAH and polar) there was a larger quantity
found in the stack, emissions than was detected in the makeup
water. It is apparent that these organics were introduced by
the coke into the quench emissions in both the contaminated
and clean tests, since there is 2 to 10 times as much PAH., and
10 to 500 times as much polar material in the emissions as
would be expected from the amount introduced by the makeup water.
-121-
-------
TABLE 10-6
MASS BALANCE OF ORGANIC COMPOUNDS
(grams per metric ton of coal)
Clean
Contaminated
T.O.C. entering the quench
process in the makeup water:
Polar Compounds
PAH
T.O.C. entering the quench
process in the coke
20
0.002
0.1
NA
1,000
50
17
NA
Organic matter leaving the
quench process in the stack
gases:
Polar Compounds
PAH
T.O.C. left in the coke
T.O.C. destroyed or formed
in the quench
T.O.C. deposited in the sump
sludge
1
1
NA
NA
NA
600
30
NA
NA
NA
NA = Not Available
-122-
-------
The following mechanisms would account for the balance of the
T.O.C.
• T.O.C. would be oxidized by the uncooled areas of
incandescent coke.
• The cooled coke would have a large activated surface
which would adsorb T.O.C.
-123-
-------
11.0 SOURCE OF ORGANIC EMISSIONS
The two variables in the coke quenching process being investi-
gated in this test program are the quality of quench makeup
water ("clean" water from the Black River versus "contaminated"
plant flushing liquor) and the quality of the coke being quenched
(green coke versus nongreen coke) . The 1976 EPA tests at Lorain
and tests conducted at Dominion Foundries and Steel showed
that the quality of quench water (concentration of total dis-
solved solids) had a definite effect on the concentration of
air contaminants in the quench tower plume. But, the effect
of water quality on organic emissions specifically, and any
effect on these emissions produced by coke greenness had not
been studied. As discussed in Section 6, this test program
was designed to address the influence of these two parameters.
11.1 Statistical Analyses
Two types of statistical analyses were performed on the organic
emissions data: analysis of variance and simple correlations.
The following sets of data (in Ibs/ton of coal) were used in
the analyses:
1) Total PAH, all tests (15 tests)
2) Total PAH, clean water tests (11 tests)
3) Each organic compound, all tests
4) Each organic compound, clean water tests
For the analysis of variance for data from all tests (.1 and 3
above) , the following hypothesis was used:
The effect of differences between the population means
for the three groups is identical, that is y]_ = u2 = ^3 '
where
y, represents the population mean of clean
water - green coke
For example:
The sum of the PAH concentrations for
each of the 5 clean water - green coke
tests divided by 5
-124-
-------
= the sample mean (which represents the
population mean for clean water-green
coke)
y_ . represents the population mean of clean
water-nongreen coke
p., represents the population mean of contam-
inated water - nongreen coke
When data from the clean water tests only was analyzed, this
hypothesis was used:
The effect of differences between the two population
means is identical, that is y, = y», where
M, and \i~ are defined as above
In statistical terms, these hypotheses were rejected at the 10%
significance level, indicating that there are less than ten
chances in 100 that the effect of differences between the popu-
lation means will be the same. The results of the analysis of
variance are shown in Table 11-1. Significant relationships
are defined as those where the different process conditions do
not have the same effect on organic emissions, within
90% confidence limits. Such results are denoted with a star(s).
The more stars appearing for a particular value, the more signifi-
cant the relationship (see explanatory note at bottom of Table).
It is obvious that the concentration of almost every species
listed and total PAH are related to the process conditions of
water quality and coke greenness. And, when the factor of
water quality is held constant, the concentration of most com-
pounds and of total PAH are related to coke greenness. Overall,
the F values generated for all tests are higher and more signifi-
cant than those for the clean water tests. This points out
that in general, water quality has a greater effect on the con-
centration of these organic emissions than does coke greenness.
In order that the relationship of organic emissions to these
two process conditions could be defined and better studied,
simple correlations for these sets of data were also run.
-125-
-------
TABLE 11-1
ANALYSIS OF VARIANCE - F VALUES
(22)
Organic Species
Total PAH
Naphthalene
Benzothiophene
Methyl naphthalenes
Acenaphthalene/biphenylene
Biphenyl
Dimethyl naphthalenes
Fluorene
Dibenzofuran/methyl biphenyl
Anthracene/phenanthrene
Dibenzothiophene
Methyl anthracenes
Fluoranthene
Pyrene
C16 H12 PAH
Methyl fluoranthene/methyl pyrene
Dihydrobenzofluorene
Chrysene/benz(a)anthracenes
3-methyl cholanthrene
Benzo(a) pyrene
Pyridine
Methyl pyridine
Aniline
Phenol
Dimethyl/ethyl pyridine
Tolui dine
Cresol
Trimethyl pyridine
Quinoline
Methyl quinoline
Phthalates
All test
conditions
Clean
Water Tests
31.519***
19.662***
35.425***
24.820***
30.004***
20.830***
8.660***
12.359***
21.098***
19.445***
15.229***
7.435***
15.832***
16.126***
4.367**
8.743***
4.527**
6.443***
0.275
6.558**
6.157**
3.418*
10.144***
8.289***
5.731**
6.217**
9.726***
10.733***
9.160***
5.391**
51.524***
18.621**
5.361**
0.108
5.994**
9.372**
5.419**
11.258***
7.268**
4.197*
4.051*
0.427*
10.226**
1.370
5.044*
0.818
4.295*
0.002
0.446
0.015
0.514
0.668
1.235
0.045
0.112
0.818
0.818
1.360
1.227
3.887*
0.108
7.931**
***
**
*
Significant at 1% rejection level.
Significant at 5% rejection level.
Significant at 10% rejection level.
-126-
-------
The average concentration of each organic compound for all tests
was run versus
• Quench water quality (designated as clean or
contaminated
• Coke greenness (using numerical ratings from
0 to 5)
• The average concentration of every other organic
compound
The results of this analysis are presented in Table 11-2. Correla-
tion values from 0.8 to 1.0 and from -0.8 to -1.0 are taken to be
significant. The concentrations of some compounds for some tests
were below the detection limits of the analytical procedures em-
ployed. In these cases a value of zero (0) was given to the data
point; otherwise, the point would be excluded as though that
species had never been tested for or analyzed. It should also
be noted that the computer program for simple correlations could
not handle all of the data at once, so the data were split into
two parts. Because of this, there are no correlations listed
between compounds on the first page and compounds on the second
page of each table.
Table 11-2 reinforces the conclusion that water quality is a more
dominant factor than coke greenness. Also, the positive correla-
tion values for water indicate that for total PAH and for all
compounds except 3-methyl cholanthrene (No. 20), the concentration
of these organic emissions is higher when contaminated water is
in use than when clean water is in use. There appears to be no
correlation between coke greenness and organic emissions in Table
11-2. However, since water quality dominates over coke greenness,
a true assessment of the influence of coke greenness cannot be
made until the factor of water quality is excluded from the data.
The simple correlations analysis for evaluation of the effect of
coke quality on organic emissions included correlations between
the average concentration for clean water tests of each organic
compound versus
-127-
-------
• Coke greenness (using numerical ratings from
0 to 5)
• The average concentration of every other organic
compound
Table 11-3 includes the results of this analysis and shows that
the correlations between each compound and coke greenness are
much higher here than when quench water is also considered (Table
11-2). Total PAH concentrations are plotted against coke greenness
in Figure 11-1 and again a correlation appears between these two
variables. Among the green (or nongreen) tests themselves, how-
ever, the correlation is not good. Also, the computer correla-
tions do not fall into the "significant" range (0.8 to 1.0). The
visual rating system for coke greenness is very useable but
deviations of one unit in either direct:
may account for the lower correlations.
deviations of one unit in either direction may occur . This
11.2 Discussion
It has been shown that two sources of organic matter found in
coke quench tower emissions are contaminated quench makeup water
and green coke. The process that is common to both of these
parameters is the coking of the coal. As described in Section 5,
some of the gases, oils, and tars that are distilled from the
coal during coking are combined with blowdown from other plant
processes to produce flushing liquor (contaminated water) which
is used to quench coke. The process of destructive distillation
going on in the coke ovens could be the producer of PAH that is
distilled off to mix with flushing liquor or trapped in the coke
when this distillation has not been completed and the coke is
"green". The formation of PAH in combustion processes has been
/ i g \
studied , but PAH formation in an oxygen deficient atmosphere
such as a coke oven has not been widely discussed in the litera-
ture. However, previous studies of coke oven charging emissions
and coke oven door emissions have shown the presence of PAH.
compounds. It would seem reasonable, therefore, that the coking
process itself may be the true source of organics which are
later found in quench tower emissions.
-128-
-------
TABLE 11-2
SIMPLE CORRELATIONS
individual organics, and total PAH values for all tests)
MATt'N
mm-
NAPHTHAI.
HKN7.HTMI
•NF.THYI.NA
ACFNAPMT
RIPHK.KYI.
niNFTHYl.
FMlORKNt
ANTHRACF
niRENZDT
MFTHYLAN
FMIOKAhT
PHRFNF.
MF.THYL
nlHYMRMH
CHRVSeNF.
HeTHVI.l
PMKNIII,
CRESflL
QIIINOI.IN
PHTHAl.AT
HF.MZO
niHFNZflT
HFTHY1.AN
FLUDRANT
PHRENK
MFTHYI,
DIHYOROn
CHPYSF.NF.
HKTHYI.3
PHFNOI.
PRFSni,
ourNoi.iN
PHTHAl.AT
BFNZO
2
i
4
5
b
7
9
10
11
12
1 1
14
Ib
17
II
20
21
22
21
24
25
11
14
15
16
17
IH
20
21
22
24
25
( us i nc
WATKR
2
| .OOOO
-0.6570
O.H753
O.H963
0.9128
O.MR04
0.60HO
O.R203
O.»870
O.R736
0.8469
0.6658
O.R510
0.8S12
0.77S9
0.6557
0.7101
-0.7060
0.7616
0.7R64
0.7771
0.9421
0.7096
nlBFNZOT
11
1 .0000
O.R660
0.9904
0.9H71
O.RH41
O.R190
O.R 136
-0. 3337
0.963R
0.«»4H7
0.9739
0 . 7 H h 3
0.7301
I water quc
riiKK
1 .0000
-0.6100
-0.566H
-0.5101
-0.6121
-0.4941
-O.O060
-0.4746
-0.4994
-0.4919
-0.5091
-0.1151
-0.4699
-0.4749
-0.7210
-0.4272
-0.5059
0.0407
-0.4492
•0.4061
-0.1A94
-0.61 19
-0.4561
MKTHYI.AN
14
1.0000
0.8744
O.8461
O.R7SR
0.6915
0.7124
-0.2169
O.H246
O.H709
0.590S
0.6929
ility, cokt
NAPHTHAI.
4
1 .OOOO
O.H7I4
O.M62
0.97R9
(J.H701
0.551 1
0.6624
0.8183
0.8156
0.8268
0.6102
0.7886
0.8162
0.6716
0.7221
0.5912
-0.2032
0.7658
0.6314
0.5914
0.9158
0.6| 16
FLOOKANT
15
1 .0000
0.9R40
0.9157
O.M369
0.8242
-O.V714
O.9590
0.96M9
0.94H9
0.7791
0 . 7 11 7
: greenness
5
I .0000
0.9905
0.9511
0.9918
0.7 161
O.H656
0.9920
0.97H5
0.9610
0.9519
0.9292
0.8)41
0.7593
0.79H3
-0.2955
O.H6 )8
0.9008
O.H92R
O.H270
0.7855
PHRF.NK
16
1 .0000
0.9084
O.M4S2
0.7945
-0.2H64
0.9792
0.9273
O.R925
O.R 119
0.6S15
;, inaiviui
HKTHYI.nA
6
1 .0000
0.9J61
0.9895
0. 7762
0.9069
0.9H51
0.9565
0.9)110
0.8220
0.9218
O.RH67
0.82R1
0. M7R
0.7715
-0.2955
O.R076
O.H M7
O.M722
0.7748
0.7854
MF.THYL
17
1 .0000
0.771 1
0.7 369
-0.2293
0.8750
0.6685
0.5762
Id i- UJ. 14111.1-1
ACfNAPNT
J
1 .00<>0
0.91 79
(I.6S19
0.7759
0.9158
0.9167
0.9026
0.7146
O.H748
O.H92I
0.7561
0.7561
0.6H90
-0.2471
O.HIHI
0.7S12
0.7251
0.8944
0.7014
DIHYDHOH
18
1 .0000
0.5552
0.1219
O.ri491
0.7511
0.7) 18
0.6701
O.S607
--•-»( VtliU «•**!
B1PHKNYL
8
1 .OOOO
0.7617
H. MM 4
0.99H7
0.9820
0.9651
0.8529
0.9648
0.929M
O.SbHJ
0.7605
0.8169
-O.llbl
0.86H6
0.9264
0.9224
0.76bl
0.7771
CHHYSKNK
19
1 .OOOO
-0.4661
0.7710
0.81R7
O.R i45
0.6223
0.5609
DlMKTHYI.
9
1.0000
0.6979
0.7521
0.7418
0.7002
0.9097
0.7054
0.6692
0.7241
0.51H4
0.5197
-0.2069
0.6064
0.7001
0.7112
0.5174
0.5000
MF.THYL1
20
i.oono
-0. 3058
-0.1224
-0.1190
-0.1701
-0.0915
Fl.tlORFNK
10
1.0000
0.8544
0.7646
0.7109
O.blRl
0.7172
0.6160
0.6077
0.4559
0.6492
-0.2268
0.5098
0.7014
0.7197
0.6185
0.7629
PHKNOI.
21
1 .OOOO
0.9200
0.876H
0.7606
0.6078
niHKNZOK
11
l .0000
0.9851
0.9701
0.8522
0.9685
0.9117
0.8579
0.7615
0.8179
-0.1215
0.8771
0.9110
0.9282
0.7691
0.7881
CRESUI.
22
1.0000
0.9950
0.6661
0.7105
ANTHHArE
12
| .0000
0.9954
0.8795
0.9889
0.9741
0.8910
0.8021
0.8285
-0.1119
0.9409
0.9477
0.9291
0.7987
0.7541
OIIINOLIN
23
1 .0000
o!7212
PHTHAI.AT MKN/II
71 ?S
IU N/n
?4
I . i
I . iI'MIll
* Soe the Key to Tables 11-2 and il-3.
-------
TABLE 11-2 continued - SIMPLE CORRELATIONS
WATFR
COKE
Tl 6H1 2PA
PYR1DINF
VPTHYLPY
ftNILINF
HI "F.THYI,
Tni.nmiN
TRIMFTHY
METHYI.OU
2
)
4
b
b
7
a
9
10
1 1
HATFH
2
1 .0000
-0.6570
0.647H
n.7117
H. 6072
0.7927
O.h9fi9
0.71 34
O.flOOH
0.6R7h
THKF
1 .0000
-0.3017
-0. 3102
-O.2779
-0.4093
-O. M6f>
-o. jif^q
-0.4422
-0.333H
ClhH12PA
4
1
0
0
0
0
0
0
0
.0000
.9736
.9798
.9418
.909S
,9»H2
.HI HO
.9626
PYWlllME
b
1
0
II
0
0
0
0
.0000
. 9 h ) 5
.9H6R
.•^7 19
.9458
.9107
.9696
MKTHYI.PY
h
1 .0000
0.9I3H
0 . H 9 1 9
0.9S45
0.7696
0.9697
AM LINK
7
I
0
0
0
0
.0000
.9751
.9297
.9574
.9451
D1MKTHYI. TOLU1DIN TRIMETHV MKTHYI.OU
8 9 10 II
1.
0.
•>.
o.
(XI 00
HS17 t.OOOO
9b90 0.7916 J.OOOO
9151 0.9527 0.8336 1.0000
WATER
Ui
O
I
WATER
COKE
TOTAL PAH
COKE
2 3
2 1.0000
3 -0.6520 1.0000
4 0.9163 -0.6017
TOTAL PAH
1.0000
-------
TABLE 11-3
SIMPLE CORRELATIONS
COKF
NAPHTHA!.
HFNZDTHI
MRTHYLNA
ACFNAPHT
RIPHF.NYI,
OIMFTHYI.
nlRFNZriF
ANTHRACE
DIHFMZnT
METHYI.AN
FLUOPANT
PHRENE
METHYL
DTHYDKOR
CHRYSENF
METHYL3
PHENOL
CHFSni,
QII1NOLIN
PHTHALAT
HENZn
METHYLAN
FI.IH1RANT
PHRENE
MRTHYL
DIHYPHOB
CHRYSKNE
METHYL3
PHFNflL
CRESni,
QII|Nni,|N
PHTHALAT
HENZO
1
4
5
6
7
8
9
0
1
2
1
4
15
16
17
18
19
20
21
22
23
24
1 1
14
15
16
17
18
19
20
21
22
21
24
CflKF
7.
\ . 00(10
0.216S
0.7076
O.SH77
0. 7106
0.7191
0.5570
0.61O8
-0.2000
0 . 7 O 9 4
0. 1166
0.5144
0.5B48
-0.0191
-0. 1046
-0.0812
0.0177
0.2835
0.4205
0.5127
-0. 1026
MFTHYI.AN
13
1 .0000
0.2002
0.1171
0.4471
-0.0761
-0.0984
0.2092
-0.2731
-0.0571
0.1180
0.4515
-n.OROH
NAPMTHAI,
1
1 .00(10
-0. I 157.
0.912Q
0.9077
0.72H4
0.5762
0.7412
O.OS79
0. 1614
-0.11 12
0.5349
0.4441
0.6154
0.9254
0.0693
-0.1757
0.0911
0.0913
0.1049
0.3560
0.1094
-0.3633
FIJJOHANT
14
1 .OOOO
0.7525
0.6172
0.4867
-0.0834
0.5746
0.4411
0.7415
0.5855
0.2067
0.1487
HKhZriTHI
4
1 .0000
0.01 09
-0.087?
O.OH46
-(1.0056
0.4632
0. 3280
0.7127
0.6151
0.1568
-0. 1604
-0.0881
-0.0651
-0.4653
0.4598
-0.5272
-0.2664
-0. 1753
-0.2651
-0.0131
0.2953
PHRENE
15
1 .0000
0.6556
0.1221
0.0054
0. 1521
0.5955
0.8985
0.841?
0.4608
-0. 3546
MKTHYI.NA
1 . ilOOl)
O.HI 1 4
0.6160
0.797S
0.0759
0.5117
-0.2127
0.570J
0.7B74
0. 1929
0.077}
-0.0550
-0.1726
-0.0416
-0. 1405
0.0470
0.0770
0.0522
-0.2609
MFTHYI,
16
1 .0000
0.1612
-0.0412
0. 1560
O.?059
0.4088
0.1111
0.0945
-0.2162
6
1 .00110
0.9212
0.4095
0.7466
0.4112
0. 3266
-O.?890
0.4295
0.6*12
0.8)80
0.94/0
0.1614
-0. 1 101
0. 1045
0.3412
0.6122
0.5420
0. 3021
-0. 3030
DIHVDHilb
17
1 .0000
-0. 1210
0.8159
0.5012
0.2729
0. 1062
0.0554
0.0931
7
1 .0000
0. 1695
0.7512
0.6141
0. »197
-0.0716
0.2192
0.6H53
0.8806
0.8310
0.0410
0.1591
-0.0737
0.4822
0.7529
0.5511
0.3517
-0.1521
CHHYSENE
18
1 .0000
-0. 3897
-0.0891
-0.0050
-0.0999
0.2102
-0.2236
DIMETHYL
8
1 .0000
0.4523
-0.0677
0.5101
-0.1963
0.9013
-0.0024
0.2406
0.3652
-0.0448
-0.2146
0.0724
-0.2951
-0. 1 169
0.3578
0.5160
-0.3086
HEFHYL1
19
1 .0000
0.2330
0.1832
0.3129
-0.0529
0.2593
FLUUHENE
9
1.0000
0.4754
0.7211
0. 1280
0.5051
0.3394
0.5516
0.7164
-0.3189
0.1213
-0.3189
-0.0666
0.2406
0.1677
0.1847
-0.2271
PHENOL
20
1 .0000
0.7992
0.5940
0.1841
-0.3256
* j t.a — *-"* ' _y f
OIBENZOF
10
1.0000
0.3311
0.1777
0.0771
0.4780
0.5862
0.2226
-0.12BO
0.0637
-0.2257
0.3948
0.6291
0.3R77
0.4655
0.0334
CKESOL
21
1 .0000
0.8219
0.4020
-0.2374
ANTHHACE
11
1.0000
0.5435
0.7195
0.0661
0. 1699
0.3996
-0.3977
0.2473
-0.2727
-0.3622
-0.1509
-0.0635
0.1822
0.1515
OUINOLIN
22
1 .0000
0.6731
-0.3555
DIBENZOT
12
1 .0000
0.1239
-0. 1095
-0. 1128
-0.1543
-0.5301
0.6588
-0.2718
-0.2942
-0.2019
-0.3023
-0.0876
0.2205
PHTHALAT
23
1 .0000
-0.4054
HENZO
74
24 i.onnn
* See the Key to Tables 11-2 and 11-3.
-------
TABLE 11-3 continued - SIMPLE CORRELATIONS
U)
N)
I
COKK
C16H12PA
PYHiniNF
METHYI.PY
ANII.INF
niMETHYl,
TRIMfTHY
METHYI.OII
2
1
4
b
6
7
8
9
10
COKF
2
1 .OOOO
-0.2A49
•0. 1S19
O.217h
-O.OR4H
-0.1902
-0.1902
0.4870
COKE
2
C 1 hHl 2PA
i .onoo
-0. 1 (IR1
-0.
-o.
-n.
-o.
-0.
-0.
TOTAL
1005
1419
1000
1000
1000
1197
PAH
3
PYKIfMNF
4
1 .00(10
-0.1073
O.HH06
0.9970
0.9970
-0.0221
-0.1514
MfclTHYLPY »N1
5
1 .0000
-0. I41f< 1
-0.09')2 0
-0.0992 0
-0.0970 0
0.1701 -0
LINE (>|M£THYI, Tnl.UlDTN THlHETHV MLTHYLQU
6 7 H 9 10
.0000
.8412 1.0000
.8412 1.0000 1.0000
.4519 -0.1000 -0.1000 1.0000
.2011 -0.1397 -0.1197 -O.U97 1.0000
COKE 2 1.0000
TOTAL PAH 3 0.8374
1.0000
-------
KEY TO TABLES 11-2 AND 11-3
SIMPLE CORRELATIONS
The correlation value indicates the relationship between
two variables. As a positive correlation value approaches
+1, the greater the tendency that x and y will increase
together. As a negative correlation value approaches -1,
the greater the tendency that x and y will decrease
together. In other words, the closer the correlation value
comes to il, the greater the effect of one variable on
another. As any correlation value approaches 0, the lesser
the effect of one variable on another. Correlation values
between io.8 and il are taken to be significant.
* Table 11-2 - For example, the correlation value for
methylnaphthalene (6) and benzothiophene (5) is 0.9905,
indicating a strong tendency for concentrations of each
compound to increase together.
** Table 11-3 - For example, the correlation value for
fluorene (9) and coke (2) is 0.7393, indicating a tendency
for concentrations of fluorene to increase as coke greenness
increases.
-133-
-------
2.0-'
(0
8 i.s-r
c 1.6-f
O
-P
•3 1.44
i-i
4J
(0
^ 1.
c
•H
as 0.8--
(0 0.6 --
-P
0.4 -•
0.2 _.
FIGURE 11-1
GRAPH OF
TOTAL PAH VERSUS COKE GREENNESS
FOR THE CLEAN WATER TESTS
2 3 4
Coke Greenness Ratings
-134-
-------
Benzo (a) pyrene is often used as an indicator for the presence
(23)
of other carcinogenic compounds and subsequently as an indi-
cator for PAH. Though BaP may be representative of the presence
of PAH, the subject study showed that although BaP levels did
correlate well with water quality they did not correlate with
coke greenness or with other PAH compounds.
The presence of phthalates in the quench tower samples was
originally judged to be contamination. However, substantial
concentrations of phthalate have been reported in coke oven door
(24)
emissions . Also, the analysis of variance and simple correla-
tions do show phthalate concentration to increase substantially
when contaminated water and/or green coke is in use.
-135-
-------
REFERENCES
(1) Laube, A.H., Jeffery, J.; and Sommerer, D., Evaluation
of Quench Tower Emission Part IT - Draft Report, Contract
Mo. 68-01-4138, Task No. 3, U.S. Environmental Protection
Agency, Research Triangle Park, N.C., December, 1977.
(2) Fullerton, R.W., "Impingement Baffles to Reduce Emission
From Coke Quenching", Journal of the Air Pollution Control
Association, Vol. 17, December, 1967, pp. 807 - 809.
(3) Air Pollution Control by Coking Plants, United Nations
Report, Economic Commission For Europe, St/ECE/Coal, 26,
1968, pp. 3-27.
(4) Sidlow, A.F., Source Test Report - Kaiser Steel Plant,
Fontana, California, San Bernadino County Air Pollution
Control District, February 29, 1972.
(5) Memorandum From Robert A. Armbrust, Region IX, New York
State Department of Environmental Conservation, To Bernard
Bloom, U.S. Environmental Protection Agency, Office of
Enforcement, November 6, 1975.
(6) Comparative US/USSR Tests of a Hot-side Electrostatic
Precipitator, NTIS PB-261-918, U.S. Environmental
Protection Agency, Research Triangle Park, N.C.,
January, 1977.
(7) Civil Engineer's Handbook, New York, McGraw Hill, 1962.
(8) Shigehara, R.T., Todd, W.F., and Smith, W.S., "Significance
of Errors in Stack Sampling Measurements", Air Pollution
Control Association Annual Meeting, St. Louis, Missouri,
June 1970.
(9) Hamersma, J.W., S.L. Reynolds, and R.F. Maddalone,
IERL/RTP Procedures Manual: Level 1 Environmental
Assessment EPA-600-2-76-160a, June 1976.
(10) Revised Organic Analysis Procedures for Level 1 Environmental
Assessment, under Sub-case T.D. 10102, ADL Monthly Report
to EPA 68-02-2150, October 1977.
(11) Lao, R.C., R.S. Thomas and J.L. Monkman, "Application of
GC-MS to the Analysis of PAH in Environmental Samples" in
Carcinogenesis, Vol. 1, Polynuclear Aromatic Hydrocarbons:
Chemistry, Metabolism and Carcinogenesis, edited by R.I.
Freudenthal and P.W. Jones, Raven Press, New York (1976).
(12) Hase, A., P.H. Lin, and Ronald A. Hites, "Analysis of
Complex Polycyclic Aromatic Hydrocarbon Mixtures by
Computerized GC-MS", ibid.
-136-
-------
(13) Technical Manual for Analysis of Organic Materials in
Process Streams, EPA-600/2-76-072, U.S. Environmental
Protection Agency, Research Triangle Park, N.C., 1976.
(14) Ames Mutagenicity Assay and Clonal Cytotoxicity Assay
(CHO) of a Coke Quench Tower Emission Sample, Unpublished
Report under Contract No. 68-02-2681, Technical Directive
No. 102, U.S. Environmental Protection Agency, Research
Triangle Park, N.C. 27711.
(15) Broddin, G., Van Vaeck, L. , and Van Cauwenberghe, K.,
"On the Size Distribution of Polycyclic Aromatic Hydro-
carbon Containing Particles From a Coke Oven Emission
Source", Atmospheric Environment, Great Britain, Pergamon
Press, Vol. 11, 1977, p. 1061.
(16) Cleland, J.G. and Kingsbury, G.L., Multimedia Environmental
Goals for Environmental Assessment - Volume I, EPA-600/7-
77-136b, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., November, 1977.
(17) Schalit, L.M. and Wolfe, K.J., SAM/IA:A Rapid Screening
Method for Environmental Assessment of Fossil Energy
Process Effluents, EPA-600/7-78-015, U.S. Environmental
Protection Agency, Research Triangle Park, N.C.,
February, 1978.
(18) Particulate Polycyclic Organic Matter, National Academy
of Sciences, Washington, D.C., 1972, pp. 4-14.
(19) Stern, A., Air Pollution IV, 1977.
(20) Zimmerman and Lavine, Psychometric Charts and Tables.
(21) Jeffery, J., Test Report for Quench Tower Emission Tests,
Dominion Foundries and Steel> Ltd., 1977.
(22) Snedecor, George W. and Cochran, William G., Statistical
Methods, Ames, Iowa, The Iowa State University Press,
1967, pp. 560, 564.
(23) Bee, R.W., Erskine, G., Shaller, R.B., Spewak, R.W.,
Wallo III, A., and Wheaton, W.L., Coke Oven Charging
Emission Control Test Program - Volume I, EPA-650/2-
74-062, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., July, 1974.
(24) Barrett, R.E., Margard, W.L., Purdy, J.B., Strup, P.E.,
Sampling and Analysis of Coke-Oven Door Emissions, EPA-
600/2-77-213, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., October, 1977.
-137-
-------
APPENDIX A
SUMMARY OF COKE QUENCH EMISSION SAMPLING DATA
-------
SUMMARY OF COKE QUENCH EMISSION SAMPLING DATA
>
\
TEST NO.
DATE
TIME PERIOD
POSITION IN STACK
NUMiJER OUENCIIES SAMPLED
NET SAMPLING TIME (HIN.)
FLOW PARAMETERS
MET % ISOKIIIET1C
MEAN VERTICAL VELOCITY (fpm)
MEAN STACK TEMPERATURE (°F)
MEAN VOLUMETRIC FLOW (SCFM)
% MOISTURE BY VOLUME
MOLECULAR WEIGHT OF STACK GAS
MOLE FRACTION OF DRY GAS
GAS ANALYSIS (DRY PERCENT BASIS)
CARBON DIOXIDE
OXYGEN
CARBON MONOXIDE
NITROGEN
SAMPLING TRAIH PARAMETERS
NOZZLE DIAMETER (in.)
PITOT TUBE CORRECTION FACTOR
VOLUME OF DRY GAS SAMPLED (SCFD)
VOLUME OF WATER COLLECTED:
CYCLONE - DROPLETS (ral)
CONUErtSER - VAPOR (ml)
IMPIMGERS - VAPOR (ral)
SILICA GEL - VAPOR (ml)
SUUTOTAL - VAPOR (ml)
TOTAL (ml)
1
11/3/77
1600-1630
Bl
5
12.3
101
1445
175
189287
19.0
26. 8
0.81
0.0
20.0
0.0
80.0
0.56
1.00
20.78
34
50
36
17
103
137
2
11/4/77
1440-1444
Bl
5
14.3
109
1755
175
201538
28.9
25.7
0.71
0.0
20.0
0.0
80.0
0.56
1.00
27.65
36
160
76
2
233
274
2B
11/5/77
1205-1240
Ul
5
12.1
108
1469
175
173382
27.0
25.9
0.73
0.0
20.0
0.0
80.0
0.39
1.00
9.70
15
60
9
7
76
91
3
11/4/77
1357-1620
Bl
5
11.6
102
1583
175
205145
19.8
26.7
0.80
0.0
20.0
0.0
80.0
0.39
1.00
10.41
28
50
1
3
54
92
4
11/6/77
1333-1646
Bl
5
12.1
96
1224
175
159993
19.2
26.7
0.81
0.0
20.0
0.0
80.0
0.39
1.00
7.95
12
2B
2
10
40
52
5
11/7/77
1158-1250
Bl
5
12.9
97
1244
175
149295
25.7
26.02
0.74
0.0
20.0
0.0
80.0
0.44
1.00
10.17
30
55
4
15
74
104
6
11/6/77
1422-1430
Ul
5
11.9
109
1582
175
193104
24.5
26.26
0.75
0.0
20.0
0.0
80.0
0.39
1.00
10.69
28
61
0:
12
73
101
7
11/8/77
1012-1050
Dl
5
12.4
93
1133
175
145836
20.4
26.6
0.79
0.0
20.0
0.0
80.0
0.44
1.00
9.67
11
42
4
6
52
63
8
11/8/77
1219-1250
'Bl
4
9.4
98.1
1428
175
152725
33.9
25.1
0.66
3.0
20.0
0.0
80.0
0.50
1.00
9.96
27
100
0
8
108
135
-------
SUMMARY OF COKE QUENCH EMISSION SAMPLING DATA
TEST HO.
DATE
TIME PERIOD
POSITION IN STACK
NUMBER QUENCHES SAMPLED
NUT SAMPLING TIME (MIN.)
FLOW PARAMETERS
(JET % ISOKIIIETIC
MEAN VERTICAL VELOCITY (fpm)
MEAN STACK TEMPERATURE (°F)
MEAN VOLUMETRIC FLOW (SCFM)
% MOISTURE UY VOLUME
MOLECULAR WEIGHT OF STACK GAS
MOLE FRACTION OF DRY GAS
GAS ANALYSIS (DRY PERCENT BASIS)
CARBON DIOXIDE
OXYGEN
CARBON MONOXIDE
NITROGEN
SAMPLING TRAIN PARAMETERS
NOZZLE DIAMETER (in.)
P1TOT TUBE CORRECTION FACTOR
VOLUME OF DRY GAS SAMPLED (SCFD)
VOLUME OF WATER COLLECTED:
CYCLONE - DROPLETS (ml)
COI40ENSER - VAPOR (ml)
IMPINGERS - VAPOR (ml)
SILICA GEL - VAPOR (ml)
SUBTOTAL - VAPOR (ml)
TOTAL (ml)
9
11/8/77
1344-1415
Bl
5
12.1
138
1575
175
165575
35.0
25.0
0.65
0.0
20.0
0.0
80.0
0.44
l.OU
10.30
12
100
B
9
117
129
10
11/8/77
1619-1655
Bl
4
9.6
106
1337
175
164733
23.8
26.2
0.76
0.0
20.0
0.0
80.0
0.44
1.00
9.11
17
50
0
10
60
77
11
11/9/77
1029-1240
Bl
4
10.1
86
1253
175
170344
15.9
27.1
0.84
0.0
20.0
0.0
80.0
0.50
1.00
10.43
10
40
0
2
42
52
12
11/9/77
1233-1420
Al
3
8.2
105
1373
175
153179
31.0
25.5
0.69
0.0
20.0
0.0
80.0
0.44
1 .00
7.15
15
60
0
8
68
83
13
11/15/77
1206-1413
Bl
6
14.8
91
135
175
166719
24.1
26.2
0.76
0.0
20.0
0.0
80.0
0.44
1.00
12.28
24
58
15
9
82
106
14
11/17/77
1 159-1240
Bl
5
13.5
70
1415
175
188192
17.7
26.9
0.82
0.0
20.0
0.0
80.0
0.50
1.00
12.47
18
50
0
7
57
75
15
11/19/77
1444-1435
Bl
4
9.7
98
1414
175
169542
25.8
26.0
0.74
0.0
20.0
0.0
80.0
0.44
1.00
8.74
23
60
0
4
64
87
16
11/19/77
1640-1612
Al
4
9.7
109
1419
175
166213
27.5
25.8
0.72
0.0
20.0
0.0
80.0
0.44
1.00
9.56
27
70
0
7
77
104
17
11/19/77
1438-1439
Bl
6
15.8
89
1459
175
179708
23.6
26.2
0.76
0.0
20.0
0.0
80.0
0.44
1.00
13.77
24
84
0
7
91
115
-------
APPENDIX B
GASEOUS EMISSIONS - 1976 LORAIN STUDY
-------
TABLE B-l
GASEOUS EMISSIONS - 1976 LORAIN STUDY
(ppm by weight)
Oxygen
Carbon Dioxide
Carbon Monoxide
Total Hydrocarbon
Sulfur Dioxide
Sulfide
Clean Quench
Water
2.2
0.003
Contaminated
Quench Water
115,402
16,806
681
12 (by volume)
187
0.003
Note; The concentration of these gases in the quench tower emission
was determined by the following methods:
Oxygen
Carbon Dioxide
Carbon Monoxide
Total Hydrocarbon
Sulfur Dioxide
Sulfide
Sampling
Grab flask -
Grab flask -
Grab flask -
Grab flask -
EPA 6 Train
with moisture
trap
EPA 6 Train
with moisture
trap
Analysis
Orsat 40
Orsat 40
Gas Chromatography
Gas Chromatography
EPA 6
EPA 6
B-l
-------
APPENDIX C
CALCULATIONS
-------
CALCULATIONS
Nomenclature used:
An.£ = Area nozzle used at each point - Sq. Ft.
Fs = Pitot tube factor
MD = Mole fraction dry gas
PP = Barometric Pressure - In. Hg. abs.
Pm = Average orifice pressure drop - In. HgO
Pm£ = Orifice pressure drop at one point - In.
Ps = Stack Pressure - In. Hg. abs.
Xm = Average gas meter temperature - °F
Tm£ = Gas meter temperature at one point - °F
TS£ = Stack temperature at one point - °F
Ttj[ = Elapsed sampling time for one point - Min.
Vm = Volume of dry gas sampled at meter conditions - Ft.
Vm- = Dry gas meter- reading at the end of a point - Ft.3
Vy/1 = Total volume of water collected - ml.
^P^ = Velocity pressure at one point - In. F^O
N = Number of Quenches Tested
Other parameters are developed and defined in the calculations.
In order to calculate micrograms per cubic meter and grains per
metric ton of coal, the following calculations are required:
Starting with the sample weight of the sample collected
in some component of the stack gas sample train (modified
EPA Method 5),
1) yg = \ig sample x scfd
m3 V f. 0.02832 scmd
sera
2) grams _ grams sample
metric ton of coal ~ Vscfd
dscfm x min x quench
quench 11.9 metric tons of coal
where:
Vscfd = Volume of dry gas sampled
DSCFM = Mean volumetric flow of stack gas
C-l
-------
11.9 metric tons of coal is the nominal tonnage
charged per battery and therefore is equivalent
to the coke quenched in any one quench.
Starting with the concentration of a parameter in yg/L
as determined in a water sample,
yg/L x 1900 x 3.785
— a - = grams per metric ton of coal
11.9 x 1,000,000
where :
1900 is the gallons of makeup (either service water
or liquor) added to the quench tower sump during
each quench. Also the quantity of water up the
stack plus any other losses.
3.785 converts gallons to liters
1,000,000 converts ug to g.
Calculations for Particulates (EPA Method 5) :
1. Volume of water vapor (3 70°F and 29.92 In. Hg. - Ft.3
VWstd= 0.0474 XVW
2. Volume of dry gas sampled at Standard Conditions - 70°F,
29.92" Hg. - Ft.3
= 17.71 xV
m
13.6
Tm + 460
3. Percent Moisture in Stack Gas
% M = . wstd x 100.
Vmstd +
4. Average Molecular weight of Dry Stack Gas
MWd = .44 x % C02 + .28 x % N2 + .32 x % 02 +• .28 x % CO
5. Molecular Weight of Stack Gas
MW = MWd x (l. - % iM^ + IS x % M
V 100 / 100
6. Stack Velocity at Stack Conditions at one point - FPS
V3;. = 85. 4S x Fs / (TSi + 460) x A PI
V Ps x MW
C-2
-------
7. Volume Water Collected at one point - ml.
* vw
8. Percent Isokinetic at one point / Pm^ Y\ /Ts-
.00267 x WLiqi + Vmi -V + IO
" CTm,- + 46C)
i = 1.667 - - — 5 - i-2i -
1 Tt± x VSJL x AnjL
9. Stack Velocity at Stack Conditions for one Nozzle - FPS
j x (TS
10. Volume Water Collected for one Nozzle - Ft.3
v;Liqnoz = Vm\~ Vmi x vw
vm
11. Percent Isokinetic for one Nozzle
.00267 x V.V ,-,
= 1.667
x V x
n
12. Stack Velocity at Stack Conditions Average - EPS
VSave = 85.48 x Fs x / s* n { *P± x '
-------
1M-. Percent Isokinetic Average
<^
%IAve = <%Inoz * Vmj ~ Vml
15. Actual Gas Flow Rate @ Stack Conditions - ACFM
Qstk = 60 x VSAve x Area of Stack
16. Gas Flow Rate Dry @ Standard Conditions - SCFMD
Qs = 1062.6 x Vs, x ps x MD x Area of Stack
rt Ts •»-• 460
17. Particulate emissions in grains per standard cubic feet
Dry - (Gr/SCFD)
GRD = .0154- x Wt/Vm . ,
"'std.
18. Particulate emissions in grains per standard cubic feet
Dry adjusted to 12% C02 - (Gr/SCFD ® 12% C02) '
GRDX = GRD x' 12. /% C02
19. Particulate emissions in grains per actual cubic feet - (Gr/ACF)
G = 17.71 x GRD x Ps x MD
(Ts + 460)
20. Particulate emissions in pounds per hour C^bs./Hr.)
P = .008572 x GRD x Qs
21. Particulate emissions in pounds per process unit (Lb/
process unit)
Pm = P/ (process unit)
C-4
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-79-082
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Coke Quench Tower Emission Testing Program
5. REPORT DATE
April 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
A.H. Laube and B.A. Drummond
9. PERFORMING ORGANIZATION NAME AND ADDRESS
York Research Corporation
One Research Drive
Stamford, Connecticut 06906
10. PROGRAM ELEMENT NO.
1AB604
11. CONTRACT/GRANT NO.
68-02-2819, W.A. 1
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
ERED
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES
541-2733.
IERL-RTP project officer is Robert V. Hendriks, MD-62, 919/
is. ABSTRACT The report giveg results Of a field study to further define quench tower
organic emissions, the character and magnitude of which are virtually unknown.
(Limited testing in 1976 indicated that a large quantity of organic material was emit-
ted from quench towers, but these data were inconclusive because so few samples
were analyzed.) Sufficient stack samples were taken under controlled coke and
quench water quality conditions to provide a statistically confident basis for emis-
sion factor determination. The samples were subjected to extensive organic chemi-
cal analysis for identification and quantification of similar functional groups and
selected individual compounds known or expected to be carcinogenic. Fifty-three
organic compounds were found in quench tower emissions; seven in sufficient quan-
tity to be considered potential health hazards. The use of waste water from other
coke plant sources for quenching greatly increases the organic load when compared
to quenching with river water. Although the water itself is the principal source of
organic emissions, the coke also appears to contribute. The majority of organics
detected are either gaseous or associated with small particles, so they will contri-
bute to ambient air loads beyond plant boundaries.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Coking
Quenching
Testing
Waste Water
Organic Compounds
Flue Gases
Carcinogens
Pollution Control
Stationary Sources
13B
13H
14B
07C
2 IB
06E
3. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |