United States Industrial Environmental Research EPA-600 9-80-012
Environmental Protection Laboratory February 1980
Agency Research Triangle Park NC 2771 1
Research and Development
Proceedings:
First Symposium on
Iron and Steel Pollution
Abatement Technology
(Chicago, IL, 10/30
11/1/79)
Interagency
Energy/Environment
R&D Program Report
-------
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 MISCELLANEOUS REPORTS series. This
series is reserved for reports whose content does not fit into one of the other specific
series. Conference proceedings, annual reports, and bibliographies are examples
of miscellaneous reports.
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 Information
Service, Springfield, Virginia 22161.
-------
EPA-600/9-80-012
February 1980
Proceedings: First Symposium on
Iron and Steel Pollution Abatement
Technology (Chicago, IL, 10/30-11/1/79)
Franklin A. Ayer, Compiler
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
Contract No. 68-02-2630
Task No. 6
Program Element No. 1AB604
EPA Project Officer' Robert C. McCrillis
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
The proceedings for the "Iron and Steel Pollution Abatement Technology"
symposium constitute the final report submitted to the Industrial Environmental
Research Laboratory by the Research Triangle Institute in fulfillment of
Contract Number 68-02-2630, Task 006. The symposium was held at the Pick-
Congress Hotel, Chicago, Illinois, October 30, 31, and November 1, 1979.
The purpose of this first EPA-sponsored symposium was to serve as a
forum for the exchange of information on environmental assessment and pollution
abatement technology development and demonstration in the iron and steel industry.
This publication contains the text of all papers presented at this sym-
posium. Papers cover significant multimedia programs conducted by the iron
and steel industry, research organizations, and governmental agencies.
ii
-------
CONTENTS
Page
30 October 1979
Keynote Address 1
Stephen Gage
Opening Session 13
Robert C. McCrillis, General Chairman
Statement of Symposium Objectives 14
Robert C. McCrillis
EPA's Iron and Steel Program 16
Norman Plaks
Environmental R, D&D in the Iron and Steel Industry 25
Earle F. Young, Jr.
Session 1: AIR POLLUTION ABATEMENT 36
Joseph W. Kunz, Session Chairman
Air Pollution Emission Standards 37
Don R. Goodwin
Innovations and Improvements of the Ore Sintering Process
for Air Pollution Control 46
Thomas E. Ban
Environmental Assessment of Coke By-Product Recovery Plants 75
C. C. Allen, Jr.
Coke-Oven Door Seal Demonstration 89
Albert 0. Hoffman" and
Ralph Paul
Environmental Assessment of Coke Quench Towers 112
A. J. Buonicore
Coke Battery Environmental Control Cost-Effectiveness 143
William F. Kemner* and
Steven A. Tomes
"^Indicates speakers
tAbstract only
iii
-------
CONTENTS (continued)
Page
31 October 1979
Volatilized Lubricant Emissions from Steel Rolling
Operations ................... „ ........ 164
Charles Mackus and
Kaushik Joshi
Presented by Jeffrey Wentz*
Emission Factors for Open Dust Sources
Chatten Cowherd, Jr.
Estimating Fugitive Dust Contributions to Ambient Particulate
Levels in the Vicinity of Steel Mills by Use of a Snow
Cover Criterion .......................... 198
Donald C. Lang and
David B. Smith*
Sinter Plant Windbox Gas Recirculation and Gravel Bed
Filtration Demonstration ................ ..... 215
Gene P. Current
Review of Foreign Air Pollution Control Technology for BOF
Fugitive Emissions ........................ 233
David W. Coy* and
Richard Jab 1 in
Fugitive Particulate Emission Factors for BOP Operations ...... 252
Jim Steiner* and
Larry F. Kertcher
Session 2: WATER POLLUTION ABATEMENT ................ 272
Gary A. Amendola, Session Chairman
Steel, Water, Regulations & Etc .......... . ........ 273
Robert B. Schaffer* and
Edward L. Dulaney
Total Recycle of Water in Integrated Steel Plants .......... 279
Harold J. Kohlmann and
Harold Hofstein
Presented by Dominick D. Ruggiero*
Use of Spent Pickle Liquor to Remove the Phosphates
in Municipal Sewage Treatment Plants ............... 300
B. J. Kerecz, Jr.,*
R. T. Mohr, and
W. F. Bailey
IV
-------
CONTENTS (continued)
Page
Physical-Chemical Treatment of Steel Plant Wastewaters
Using Mobile Pilot Units 325
Richard Osantowski and
Anthony Geinopolos*
Study of Non-U.S. Wastewater Treatment Technology
at Blast Furnaces and Coke Plants 341
Harold Hofstein* and
Harold J. Kohlmann
Formation and Structure of Water-Formed Scales 354
George R. St. Pierre" and
Rhonda L. McKimpson
1 November 1979
Session 3: SOLID WASTE POLLUTION ABATEMENT 365
Eugene F. Meyer, Session Chairman
Federal Requirements for Chemical Waste Disposalt 366
Eugene F. Meyer
Environmental and Resource Conservation: Considerations
of Steel Industry Solid Waste 367
M. R. Branscome,*
V. H. Baldwin,
C. C. Allen, and
B. H. Carpenter
Deoiling and Utilization of Mill Scale 398
S. R. Balajee
Characterization and Utilization of Steel Plant Fines 428
Donald R. Fosnacht
International Mineral Recovery, Ltd., Dezincing Process 448
John E. Allen
Closing Remarks 455
Norman Plaks
UNPRESENTED PAPERS 458
Air Pollution Emissions Characterization of a Coal
Preheater 459
A. J. Buonicore,
B. Drummond,
Carl Rechsteiner, and
Julie Rudolf
v
-------
CONTENTS (continued)
Page
Appendix A: ATTENDEES 497
vi
-------
KEYNOTE ADDRESS
PREPARED FOR DELIVERY BY
DR. STEPHEN GAGE
ASSISTANT ADMINISTRATOR
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
AT THE
IRON AND STEEL POLLUTION ABATEMENT TECHNOLOGY SYMPOSIUM
CHICAGO, ILLINOIS
OCTOBER 29, 1979
SPONSORED BY
ENVIRONMENTAL PROTECTION AGENCY
-------
As I flew into Chicago to address this conference on pollution
control in the iron and steel industry, I remembered the first time I
saw Chicago. I was fifteen, coming east for my first trip off the
plains of Nebraska. As we approached Chicago, dusk was falling and the
setting sun illuminated the billowing red clouds of pollution from south
Chicago and Gary like an eruption directly up from hell. Still fairly
fresh from the clean skies of the prairies, I was awed by that overwhelming
scene.
Now from the perspective of a few more years, 1 remain as equally
impressed, as I was those twenty-five years ago, with how far we've come
and how far we still must go. Not all of the conventional air and water
pollution problems have been solved by the iron and steel industry, but
many have and the citizens of Chicago and Pittsburgh are much the better
off for those advances. We must now deal with a tougher problem, the
invisible intrusion of thousands of toxic chemicals into the environment--
a commitment as challenging as the first commitment to rid our environment
of the obvious noxious pollutants. I am very pleased to note that this
is a "multimedia" conference indicating that we in the Environmental
Protection Agency, as well as you in the iron and steel industry, recognize
that potential pollution problems exist in the air, in the water, and on
land. The interrelationship among these three media must be clearly
understood as we attempt to solve pollution problems. For example, we
may solve an air pollution problem only to create a severe water pollution
problem in the process. Also, we are finding more and more alarming
cases where toxic materials improperly disposed of in landfills and
lagoons are leaching out and polluting local surface and ground waters.
-------
I am also pleased to welcome EPA's Offices of Air Quality Planning
and Standards, Water Planning and Standards, Solid Waste Enforcement.
With this broad agency representation, we should give broad visibility
to the breadth of our environmental problems. The role of the Office of
Research and Development in supporting the entire agency is certainly
highlighted by the participation of the regulatory offices.
In general, the participants in this symposium represent a wide
diversity of interests--government, the iron and steel industry, univer-
sities, engineering firms, and equipment vendors. This is a healthy
sign since no single sector has a monopoly on knowledge. To advance
toward a common goal of controlling pollution problems wherever they
exist and to do so in an efficient and cost-effective manner requires a
pooling of our common knowledge and expertise. We, in Washington, know
full well that we cannot by ourselves solve environmental problems; and
we also know that given the nature of our open, pluralistic society,
there's going to be a lot of tugging and pulling over what gets cleaned
up and when. That's why Congress gave up the job of setting and enforcing
national environmental standards and regulations.
Ambient environmental standards and industrial discharge regulations
are established after a long period of research by both industrial and
governmental laboratories; after extensive review of scientific research
and economic data; and after extended public scrutiny by outside experts
in technology, health effects and economics. These standards are,
therefore, the best reflection of what we know about the short-term and
long-term health and environmental effects of the pollutants and the
technology available to abate those pollutants. They are based on the
best data we can find. So, as new data become available, it is essential
-------
that this information be transferred to involved government agencies, to
private scientific groups, and to industries at the earliest possible
moment. We all need to be aware of the latest research data, so that
standards can be established on a common up-to-date base of technical
knowledge.
As you know, the setting of environmental standards and regulations
is a very complex process involving numerous government and private
scientific groups, industries, and the public at large. It is thus
impossible to make decisions that will satisfy everyone. But since
these decisions must be made on a rational, scientific basis, confer-
ences such as this serve as a sounding board for scientific thought,
where industry, universities, and government may exchange ideas and
develop new ones.
Court cases drain the resources of both industry and the taxpayer.
When possible, it is better to attempt to resolve problems and obtain a
consensus at the scientific roundtable than to test the decisions at a
judge's bench.
Despite best intentions on both sides, however, EPA and the iron
and steel industry have been, on many occasions, antagonists. Recently
we have been at loggerheads with the industry over its insistence that
even handwritten drafts of chapters for the sulfur oxides criteria
document are subject to discovery, a claim we felt verged on harassment,
but we'll probably work out a solution to that issue eventually. On the
other side of the ledger, EPA and several major steel companies have
negotiated landmark settlements in air and water pollution cases which
have been dragging for years. Encouraging, too, is the fact that there
have been a number of instances of joint technical effort to develop new
-------
pollution abatement technology. An excellent example of this is the
jointly funded coke oven door seal project. While it is the nature of
our system of government that we shall find ourselves frequently viewing
issues from different vantage points, \,e must continue to work together
whenever feasible so we can proceed down the road towards a cleaner
environment.
I know there is much skepticism on the part of many persons concern-
ing the work of EPA. There have been charges by industry that we are
moving too fast and too soon with too little data on which to base
decisions. We are further accused of putting industry "out of business"
because of excessive pollution control costs.
However, while it does cost money to protect our health and envi-
ronment, the cost of pollution control contributes less than one-half of
one percent to the annual rate of inflation. Pollution control also
creates more jobs than it displaces.
Industry is not the only segment of our society that is dissatisfied
with what we, at EPA, are doing. Environmental groups, by contrast,
accuse EPA of not moving fast enough. They say too little has been
accomplished. Here again, the record shows that'we are making progress:
Fish are returning to our lakes and streams. As an example, Atlantic
salmon are returning to the Connecticut River for the first time in
years. Fish and other aquatic life are returning to the Cuyahoga River,
the Buffalo River, the Detroit River, and the Houston Ship Channel.
There's even a new sport emerging in downtown Washington these days—bass
fishing in the Potomac within the shadow of the Kennedy Center. And the
fishing's good!
-------
Similar improvements have been made in reducing air pollution:
Carbon monoxide levels have fallen, particulate emissions have been
reduced, and SO levels are measurably smaller. Hopefully, these trends
X
will continue.
The role of a governmental regulatory agency--whatever it is regu-
lating—is never one that is cheered and acclaimed from all sides.
However, as our population grows, as our national economy grows, and as
we move to greater use of our domestic energy resources, we must remember
that our air, water, and land resources are not growing. The additional
health and environmental pollutants being generated by such growth must
be adequately controlled. Also, as additional hazards are recognized,
the regulations may have to become more and more stringent. But we are
looking for innovative ways to regulate and control pollution that will
not impede growth or reduce our standard of living.
Our standard of living is, in no small measure, dependent on steel.
There are about 130 steel-making plants in the U.S. centered in the
traditional steel-producing districts such as Pittsburgh and Chicago.
Over 75% of total U.S. steel output is accounted for by the six states
bordering the Great Lakes. U.S. iron and steel plants employ upwards of
a half million persons and produce products whose gross sales amount to
$40 billion annually. Keeping this industry healthy is of vital impor-
tance to this nation. But we must also keep our people healthy. We
must review each phase of industry operations to keep air, water, and
land pollutants within safe limits.
The making of coke fr>m coal, the reduction of iron ore to iron,
and the refining of the iron and melting of scrap to produce steel all
have the potential for emitting vast quantities of air pollutants.
6
-------
Loading, unloading, transfer, and storage of raw materials can also
result in the entraimnent of dust. Careful attention to process varia-
bles, operating practices, and maintenance schedules can, in some cases,
achieve adequate control. Often, however, these efforts must be combined
with the addition of control equipment or significant process modifications.
The downstream forming and finishing operations require vast quantities
of water for cooling, lubrication, and cleaning. Control of water
pollutants requires the addition of control equipment. To do this at
reasonable cost, however, process modifications are needed so that the
net flow of contaminated water to the treatment plant will be reduced.
Recycle appears to be an essential element for good water pollution
control. As we move towards increased control of pollutants, discharge
flows must be reduced by increasing in-plant recycling of water wherever
possible.
All these controls on air and water, of course, also yield solid
wastes. Emphasis is being placed on technology to permit returning more
of these materials back into the steel-making process, but there will
still be some solid wastes which must be disposed of in an environmentally
sound manner.
I will not dwell on the regulations controlling these pollutants
here, because I am certain they will be discussed in detail later in the
program. But I would like to call special attention to the agency's
attempt to control toxic chemicals discharged into the environment by
industry.
In 1976 a consent decree, settling a court suit filed under the
Federal Water Pollution Control Act, was signed by EPA and four environ-
mental groups. EPA agreed to set effluent standards for 65 classes of
-------
compounds for 21 industries—including the iron and steel industry- As a
result, the Effluent Guidelines Division of EPA published a list of 129
organic pollutants and metals on which they proposed to set regulations.
These have since become known as the "priority toxic pollutants." Subse-
quently, amendments to the water pollution legislation, known as the
Clean Water Act of 1977, converted the terms of the consent decree into
law. Specifically, the 1977 Act requires the application of best avail-
able technology economically achievable (BATEA) to control the priority
pollutants by July 1, 1984.
We recognize that these regulations, plus the advent of the Resource
Conservation and Recovery Act, the Toxic Substances Control Act, pre-
treatment standards of the Clean Water Act, and the 1977 amendments to
the Clean Air Act place a great burden on the iron and steel industry.
However, we are also aware of the progress the iron and steel industry
is making toward solving the problems associated with meeting these
standards. Systems for recycling blast furnace scrubber water are
becoming more common, as is biological treatment of coke plant waste
waters. Many plants are implementing staged charging and are placing
pushing-emission-controls on their coke batteries.
We recognize that the iron and steel industry faces economic
problems, such as its low profit margin and the threat of foreign imports,
which are generally more intractable than the economic pressures created
by pollution control requirements.
We are looking to economic regulatory reforms to help ease the
regulatory burden. We believe they will provide positive incentives for
industry to invest in developing new, more efficient forms of control
technology and practices. As many of you are probably aware, Douglas
-------
Castle, EPA's Administrator, has been prominent in this effort not only
at EPA, but across the board in all federal regulatory activities. He
has been appointed Chairman of President Carter's Regulatory Council,
which is comprised of all the federal department agency and independent
regulatory commissions. The purpose of the Council is to search for
alternative approaches to make regulation less burdensome on business--
without reducing benefits to the public.
We at EPA have been in the forefront of this federal effort and
have developed a number of new regulatory reform initiatives. We have
been encouraged by the response to them and the results that have been
achieved. I'd like to close with a brief description of these new
regulatory approaches--the "bubble approach," an offset policy, and
performance standards.
Early this year EPA proposed a new policy to give industry greater
flexibility in meeting air pollution requirements. This policy—known
as the "bubble approach"—would allow a firm to devise its own methods
of cleaning up individual polluting sources within a plant facility.
Picture a giant plastic bubble placed on a steel mill—over all the
buildings and surrounding on-site facilities. At the top of the bubble
is a single hole through which all the plant's pollution escapes.
Up to now to control this pollution what we have been doing is to
put limits on emissions from each source in the facility. Under the
bubble approach, the only pollution measurement would be taken at the
top of the bubble. Plant management could control pollution from indi-
vidual sources within the facility as they saw fit, so long as the air
escaping from the top of the bubble met federal standards. Managers
might choose to omit controls from particularly difficult sources while
installing stricter controls on easier sources.
-------
For example, in Pennsylvania, agreement was reached among U.S.
Steel Corporation, EPA, and the Pennsylvania Department of Environmental
Resources to allow trade-offs in controlling particulates emitted by
their furnaces. In another case last June, E. I. DuPont Company proposed
to clean up its hydrocarbon emissions at one of its chemical plants in
New Jersey. Using the standard controls approach, it would cost $20
million and achieve an 85% reduction in hydrocarbons. Using the bubble
approach, the company said it could remove 90% of its hydrocarbons at a
cost of only $5 million.
As Doug Castle said at the announcement, "This policy will mean
less expensive pollution control—not less pollution control."
This basic approach has also been extended to cover not only a
single plant, but also several facilities in a particular area, where
trade-offs might be possible. In effect, the so-called "offset" policy
treats entire regions as gigantic bubbles, requiring only that overall
air standards be met for the region. This approach allows major modifi-
cations or new facility construction, provided the additional pollution
generated is offset by reductions elsewhere—either within a facility or
at nearby facilities.
In some cases, companies have been willing to pay large sums of
money to control pollution at nearby plants so that they could proceed
with construction of their own new facilities. In others, companies
have advised local government officials that they would invest in building
new large plants in the area if the governmental agencies could find
offsetting reduced pollut-' >n sources. Oklahoma City and Shreveport,
Louisiana, are two examples of where municipal governments have persuaded
other companies to reduce their emissions to make room for new plants.
10
-------
Another variation in this offset polity is a recent amendment that
allows "banking" of air pollution offsets. This has come about because
of the possibility that state agencies might insist on requiring offsets
that would be far greater than the expected pollution generated by new
construction. Under this policy, a company may "bank" any excess emissions
reductions it makes and use them to offset future expansion.
We also see the potential for this concept to be expanded to cover
an entire air pollution region. Local or state governments could set up
offsets banking accounts for companies in their jurisdiction to keep
track of each agreement made to reduce source emissions. A statement of
account balances would be maintained that could be drawn upon for future
industrial expansion. This kind of a centralized accounting would also
facilitate any future searches by industry and local government officials
for offsetting sources.
These approaches we are taking are part of a broader federal regu-
latory policy of setting performance standards, rather than specifying
detailed means to meet the performance goals.
Any of these schemes that substitute economic incentives for specific,
detailed regulations help to stimulate technological innovations. They
also allow industry to select the least costly measures for pollution
abatement. These incentive approaches can result in economic, environ-
mental, and social advantages over the more traditional regulatory
approaches.
The traditional "command and control" regulations are being reviewed
and alternatives are being tested. By enlisting the support and cooper-
ation of industry through these incentives, we can hopefully tie America's
economic genius to its technological and scientific expertise. Through
11
-------
conferences such as this, we are seeking to promote technological inno-
vations for pollution control—to stimulate a freer exchange of scientific
and technical information--and to develop regulatory policies that will
protect the health and environment at minimum cost to industry and the
public.
12
-------
OPENING SESSION
General Chairman: Robert C. McCrillis, Mechanical Engineer
Metallurgical Processes Branch
Industrial Processes Division
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC
13
-------
Statement of Symposium Objectives
Robert C. McCrillis
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
Good morning, ladies and gentlemen:
I am Bob McCrillis, your Symposium General Chairman. Your program
states that this "Symposium is designed to serve as a forum for the
exchange of information on environmental assessment and pollution
abatement technology development and demonstration in the iron and
steel industry." During the next 2-1/2 days you will be subjected
to a barrage of information on iron and steel projects ranging in
scale from laboratory research to full-scale operating units. You
will be presented detailed engineering, emission and cost data
obtained during actual full-scale operation, and you will also hear
the results of engineering paper studies.
The program is subdivided into three sessions with the three media.
Although the air session is much longer than either of the other
two, this should not be construed to mean that EPA places more
emphasis on this media. Rather, it is simply an artifact of the
past when MPB was oriented to air only. We are now multimedia and
as time goes by, our program has become more evenly balanced. Future
symposia will reflect this balance. We are also considering encouraging
more participation from persons outside the EPA-sponsored sphere. In
addition, future symposia may be expanded to include a fourth session
devoted to noise. We will, of course, continue to stress technical
content as you will see over the next 2-1/2 days.
Each author has been allotted 35 minutes of which about 20-25 minutes
is for his presentation and 10-15 minutes for questions from the audience.
I heartily encourage your pat.icipation in the question-and-answer period;
however, in all fairness to subsequent authors, session chairmen will
cut off questions at the end of the allotted time.
14
-------
This symposium is sponsored by the Metallurgical Processes Branch of
EPA under the able leadership of Norman Plaks, whom you will meet
in just a moment. Along with the other members of the Branch who
are Bob Hendriks, John Ruppersberger, and Susan Sharpe, I encourage
you to give us your comments on the symposium itself so that we may
improve our future productions.
15
-------
EPA's Iron and Steel Program
Norman Plaks
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
Presented at the Symposium on Iron and Steel
Pollution Abatement Technology
Chicago, Illinois
October 30 - November 1, 1979
16
-------
BACKGROUND
The Environmental Protection Agency's multi-media research and •
development program for ferrous metallurgical processes is centered at
its Industrial Environmental Research Laboratory at Research Triangle
Park, North Carolina. The concern of the program is for the integrated
iron and steel industry, iron and steel fou .dries, and ferroalloys.
The air portion of the program, which was started in 1968, has its
roots in the National Air Pollution Control Administration of the Department
of Health, Education and Welfare. The water portion of the program,
which was started in 1966, traces its antecedents back to the Federal
Water Pollution Control Administration of the Department of the Interior.
Both became part of the Environmental Protection Agency when the Agency
was established in December 1970. Until the enactment of the Resource
Conservation and Recovery Act of 1976 there had been little basis for
undertaking systematic work in solid waste treatment.
At the time the iron and steel air program was started, the authorizing
legislation did not mandate specific control levels and schedules to be
achieved. However, it did stipulate that R§D was to be done. Prioritization
for the R§D was based upon a study by Battelle entitled, "A Systems
Analysis Study of the Integrated Iron and Steel Industry,"* published in
1969, which indicated that emissions from the manufacture of coke should
have the highest priority for R§D. Other processes identified as having
high priority for R§D, because of their emissions, were charging of
basic oxygen furnaces, sinter plants, blast furnace cast houses, and
both process and open source fugitive emissions in general. In response
to the Battelle prioritization a number of projects were undertaken
between 1968 and 1975. (See Table 1)
Three milestones were established within the general wastewater
treatment program, which included iron and steel, by the Clean Water
Restoration Act of 1966. These were 1) to develop the technology necessary
to achieve the equivalency of secondary treatment (85 percent BOD removal),
2) to achieve state water quality standards (95 percent BOD removal),
and 3) eventually to eliminate all water pollution caused by industry.
Again, much like in the air program, a schedule for achieving these
levels of reductions was absent in the legislation. However, projects
were undertaken to achieve these levels of discharge reduction for the
most significant sources in the iron and steel industry (See Table 2).
Included was work in treating of wastewaters from cokemaking, blast
furnaces, finishing, and other sources from the integrated iron and
steel industry.
*EPA Report No. APTD 1177 (NTIS No. PB 184 576), May 1969
17
-------
In July 1975 a reorganization in the EPA Office of Research and
Development (ORD) resulted in the EPA iron and steel R§D being combined
into a multi-media program. This was accomplished by bringing together,
at Research Triangle Park, North Carolina, ongoing iron and steel wastewater
activities from ORD groups at Grosse He, Michigan, and Edison, New
Jersey, and combining these with the air media portion of the program
already in place at Research Triangle Park.
Program Responsiveness
During the same time period as the 1975 reorganization the regulatory
and enforcement offices were coming up to their full working strength
and efficiency as a result of implementing the mandated programs within
the "Air" and "Water" Acts. To do their work, the Office of Research
and Development had to respond to information and data needs that were
developing. The Office of Research and Development in EPA is much like
the research arm of any organization. If it does not support the product
line of the organization, which in the case of EPA consists of regulations
and enforcement activities, much of its purpose for existing will disappear.
As a consequence, the activities of EPA/IERL-RTF's Metallurgical Processes
Branch in iron and steelmaking pollution control have been aligning
themselves more with the mainstream activities within EPA. Providing
support to the regulatory and enforcement activities within EPA simultaneously
provides support to the iron and steel industry. This statement is not
inconsistent. Even though their motivations may differ, both EPA and
the industry desire to achieve the required level of pollution control
at the lowest cost.
Program Budgets
From Fiscal Year 76 through FY 80, which started October 1, 1979,
EPA's consolidated multi-media iron and steel RSD program has had the
following yearly budgets.
Fiscal
Year
76
77
78
79
80
Air
(I^OOO's)
565
468
574
559.
34Q
Water
($l,000's)
150
442
820
786
530
Solid Waste
($l,000's)
100
200
18
-------
Included in the above listing is funding from EPA's base industrial
program plus some funds from the program offices for tasks specific to
their needs. The FY 80 funding, as shown, is not complete. There will
be additional funds to mount a program in recycle/reuse of waste water
for the iron and steel industry and for projects to assist the program
offices.
Environmental Assessments
Up to 1975 the work that had been done under the program was primarily
concerned with pollution abatement technology development for a relatively
few conventional pollutants. However there were mounting concerns that
there could be other more hazardous pollutants being discharged from
industrial and energy sources to air, water, and land receptors. This
has led to the concept of a comprehensive characterization (called an
environmental assessment) of all materials released to the environment
by a source. Approximately one-third of our budget since Fiscal Year 76
has been for environmental assessments. In general an environmental
assessment contains the following components.
1. An evaluation of the physical, chemical, and biological characteristics
of all input and output streams associated with a process
2. Predictions of the potential effects of those streams on the environment
3. Prioritization of those streams relative to their individual hazard
potential
4. Identification of any necessary pollution control technology.
Such an assessment is complex and technically difficult. The
sampling program must be more extensive than a program acquiring data
for process engineering type activities. Additional complexities are
introduced by the necessity for detecting all materials, posing potential
environmental hazard, above some minimum level of concern.
As would be expected, undertaking a comprehensive and all-encompassing
environmental assessment of an industry or process would be expensive.
To minimize costs, a phased or sequential approach is usually used in
which the assessment proceeds to increasingly more specific and complex
sampling and analysis, once the need and direction are established by a
preliminary stage.
Three phases or sequences of environmental assessment have been
developed. These in generalized form are:
19
-------
Level 1 utilizes semiquantitative sampling and analysis procedures
that yield final analytical results accurate to within a factor of 3 of
the sample. Level 1 is designed to (a] provide preliminary environmental
assessment data, (b) identify problem areas, and (c) formulate the data
needed for the prioritization of energy and industrial processes, streams
within a process, components within a stream, and classes of materials
for further consideration in the overall assessment.
Level 2 is focused by Level 1 results and is designed to provide
additional information that will confirm and expand the information
gathered in Level 1. This information is used to define control technology
needs, and may, in some cases, give the probable or exact cause of a
given problem.
Level 3 involves monitoring the specific problems identified in
Level 2 so that the critical components in a stream can be determined
exactly as a function of time and process variation for control device
development and evaluation.
Technology Development and Demonstration
Considering the low level of funding that was available compared to the
high capital intensiveness of the industry and its processes, projects
that were undertaken had to be judiciously chosen so as to provide
maximum cost effectiveness. Two techniques used when possible, for
development and demonstration projects^ were cost sharing with the
industry and by what could be called multi-use projects.
A major project of the cost shared type is the one to develop
improved coke oven door seals, being done at Battelle Memorial Institute,
in which the industry through the AISI has paid half the cost and EPA
the other half. By this type of arrangement on projects that are
identified as simultaneously being of high priority to both government
and the industry, each of the two groups - EPA and industry - achieves
greater cost-effectiveness for its available funding.
An example of a multi-use project is the mobile waste water treatment
system which has been constructed and is operating at various iron and
steel plants. The system, rated at 5 gallons per minute, consists of
two separate wastewater treatment trains, one physical/chemical and the
other biological, and has been operating on coke plant and blast furnace
scrubber wastewaters. Operation of the system has not only been providing
EPA with information and data that it vitally needs but is also providing
good quality pilot plant data to the host plants to enable them to
control their waste streams. Ultimately if the need exists the system
might be used on other iron and steel industry wastewater streams.
20
-------
From the two major projects which were just described, the projects
range down in size. Included are engineering evaluations and technology
development and studies.
Some of the smaller projects in the p ogram are in the form of
grants to various universities to conduct applied research. This portion
of our program, which is currently in its fourth year, is supported by
about 10 percent of our budget. The majority of the projects in this
group are cofunded with the AISI. Projects such as these will hopefully
provide the advanced technology that will be needed for the future
years. Cofunding of these projects with the industry enables us to fund
more of these projects while also enhancing communication on pollution
abatement technology needs with industry and the academic community.
Forejign Technology Evaluation
A vital portion of the iron and steel program is the identification
and evaluation of pollution control technology used at foreign iron and
steel plants. The objective is to ultimately show how the technology
could be applied to plants in the domestic industry. If a domestic
plant elects to use one of these technologies, it would, of course, have
to do so under the business arrangements that are normal for such
technology transfer.
There are currently two projects underway to evaluate foreign
pollution control technology. The first is evaluating secondary emission
control for the basic oxygen process. Included are emissions from
charging and tapping, desulfurization, reladling, and deslagging.
Preliminary evaluations are being made of plants in Western Europe and
Japan. After completion of the evaluation, and with permission of the
plant owners, sampling will be undertaken at selected plants which
appear to have the most efficient control technology. Generalized engineering
to show how this technology can be applied to the domestic industry is
also included. The second foreign pollution controT technology evaluation
is concerned with wastewater treatment. Consideration is being given to
coke-plant and blast-furnace top gas scrubber wastewater treatment and
also to recycle/reuse. Visits have been made to Western Europe, Japan,
Taiwan, and Australia. Visits have also been made to arid region plants
in Mexico, South America, and South Africa to learn how they achieve
high levels of recycle/reuse.
Another program concerned with foreign technology is the Ferrous
Metallurgy Project under the US/USSR Environmental Agreement which has
been the source of information on pollution control technology within
the Soviet Union. The Soviet Union is now the largest producer of steel
in the world, followed by the US and Japan.
21
-------
Under the US/USSR program pollution control technology at a number
of Soviet iron and steel plants and institutes have been evaluated by
mixed teams of EPA and industry personnel. About 26 tonnes of dry
quenched coke samples were obtained from the Soviets, distributed to
several iron and steel companies, and subjected to a number of tests.
The results were of interest to both the Government and the Industry.
The US/USSR Ferrous Metallurgy Project has now advanced to the
cooperative phase. Projects have been identified and are underway on
coke oven door seals and wastewater treatment. Status reports will be
made as the projects proceed.
Under the foreign pollution control technology evaluation portion
of the iron and steel program, unique technology applications have been
discovered. It is essential that these applications be investigated and
brought to the attention of both EPA and the iron and steel industry.
It is expected that this work will be continued and even expanded where
the opportunity presents itself.
Technology Transfer
The preceding description of the EPA research and development
program for iron and steel presents a brief history of the program, its
budget, responsibilities, and projects. The principal output of all
these activities is technical information gathered on industry emissions
and discharges and their control. Technology transfer is the very
important final step of putting this information in the hands of the
user.
Three types of technology transfers are practiced - reports, the
EPA Ferrous Metallurgical Processes Review, and symposiums.
At the conclusion of our projects, final reports are issued detailing
the results and conclusions, and also explaining the work that was done
leading up to them. These final reports provide the basic documentation
on a project.
Recently we put out our first issue of the EPA Ferrous Metallurgical
Processes Review. This will be continued with approximately four issues
per year, and will provide brief discussions of the work that is being
done by us.
Finally we are sponsoring this, our first iron and steel pollution
abatement symposium. Symposiums are an effective means for achieving
technology transfer if the papers presented are carefully chosen. It has
been felt that a need exists for a symposium dedicated to pollution
abatement in the iron and steel industry. If there is sufficient interest,
these symposiums will be continued and even expanded.
22
-------
Table 1 IRON AND STEEL AIR PROJECTS UNDERTAKEN PRIOR TO FY 76
Cokemaking
Evaluation of Process Alternatives to Improve Control of Air
Pollution from Production of Coke, EPA Report No. APTD 1250 (NTIS
No. PB 189 266), January 1970.
Coke Charging Pollution Control Demonstration,
EPA-650/2-74-022 (NTIS No. PB 234 355), March 1974.
Coke Oven Charging Emission Control Test Program,
EPA-650/2-74-062 (NTIS No. PB 237 628), July 1974.
Emission Testing and Evaluation of Ford/Koppers Coke Pushing
Control System, EPA-600/2-77-187a (NTIS No. PB 273 812), September
1977.
Enclosed Coke Pushing and Quenching System Design Manual, EPA-
650/2-73-028 (NTIS No. PB 226 418), September 1973.
Sampling and Analysis of Coke-Oven Door Emissions,
EPA-600/2-77-213 (NTIS No. PB 276 485), October 1977.
Study of Concepts for Minimizing Emissions from Coke Oven Door
Seals, EPA-650/2-75-064 (NTIS No. PB 245 580), July 1975.
Blast Furnace
Blast Furnace Cast House Emission Control Technology Assessment,
EPA-600/2-77-231 (NTIS No. PB 276 999), November 1977.
Sinter Plant
Sinter Plant Windbox Recirculation System Demonstration: Phase 1.
Engineering and Design, EPA-600/2-75-014 (NTIS PB 249 564), August
1975.
Steelmaking
Development of Technology for Controlling BOP Charging Emissions,
EPA-600/2-77-218 (NTIS No. PB 277 Oil), October 1977.
23
-------
Table 2 IRON AND STEEL WASTEWATER TREATMENT PROJECTS UNDERTAKEN
PRIOR TO FY 76
Cokemaking
Research Study of Coal Preparation Plant and By-Product Coke Plant
Effluents, EPA-660/2-74-050, (NTIS No. PB 252 950) April 1974.
Biological Removal of Carbon and Nitrogen Compounds from Coke Plant
Wastes, EPA-R2-73-167, (NTIS No. PB 221 486) April 1973.
Blast Furnaces
Pollution Control of Blast-Furnace Plant Gas Scrubbers Through
Recirculation, EPA-660/2-74-051 (NTIS No. PB 250 435) July 1974.
Finishing
Treatment of Wastewater/Waste Oil Mixtures, EPA Project No. 12010
EZV (NTIS No. PB 195 161) May 1970.
Closed Loop System for the Treatment of Waste Pickle Liquor,
EPA-600/2-77-127 (NTIS No. PB 270 090), July 1977.
Countercurrent Rinsing on a High Speed Halogen Tinplating Line,
EPA-600/2-77-191 (NTIS No. PB 272 590), September 1977.
Limestone Treatment of Rinse Waters from Hydrochloric Acid Pickling
of Steel, EPA Project No. 12010 DUL, February 1971.
An Electromembrane Process for Regenerating Acid from Spent Pickle
Liquor, EPA Project No. 12010 EQF, November 1970.
Sulfuric Acid and Ferrous Sulfate Recovery from Waste Pickle
Liquor, EPA-660/2-73-032, (NTIS No. PB 233 112) January 1974.
General
Combined Steel Mill and Municipal Wastewaters Treatment, EPA Project
No. 12010 DTQ, (NTIS No. PB 210 198) January 1971.
Water Pollution Control in the Carbon and Alloy Steel Industries;
EPA-600/2-76-193 (NTIS No. PB 252 963), April 1976.
24
-------
ENVIRONMENTAL R, D & D
IN THE IRON AND STEEL INDUSTRY
by
Earle F. Young, Jr.
Assistant Vice President,
Environmental Affairs
American Iron and Steel Institute
at
EPA Symposium on
Iron and Steel Pollution Abatement Technology
October 30, 1979
It has been said in a number of places, and particularly in some
government reports, that the iron and steel industry spends very, very
little on research and development, and particularly on R&D in the envi-
ronmental area. It is my hope today to refute that statement. I believe
that it comes from a lack of understanding of the real nature of the tech-
nical problems of environmental control in the steel industry, and of the
scientific, engineering, and technical work reqxiired to solve those prob-
lems.
My topic today, Environmental Research, Development, and Dem-
onstration in the Iron and Steel Industry, is so broad that a major prob-
lem was deciding what approach to take. Should I make it a personal talk?
Should I talk about AISI applied research through the years? Should I talk
25
-------
about AISI supported research in universities? Should I talk about re-
search activities in company laboratories? Should I talk about research
outside the steel industry itself, by the equipment manufacturers and ven-
dors? Or should I concentrate on what is perhaps the most difficult and
demanding aspect, the development of prototype units in the field? Look-
ing at that list, I decided the right answer was "All of the above. "
I'd like to start with a little personal account. I started in the
steel industry back in 1956, Within the first year, I was involved in two
major assignments which can properly be classed as research and devel-
opment in the environmental aspects of the iron and steel industry. One
of these was for my company, looking for a new process for the removal
of phenolic compounds from coke plant wastes. This research was com-
pleted several years later and culminated in the development of the J&tL
solvent extraction dephenolization process. Initial laboratory studies of
solubilities, equilibria, and basic data of that sort were followed by pilot
plant work on new types of equipment, then design and installation of a
plant, and then an extended period of break-in and modification, which
ultimately resulted in what was, I believe, the best dephenolization pro-
cess then available.
The second assignment, again back in 1956, was working with the
Steel Industry Advisory Committee of ORSANCO. This group was attempt-
ing to evaluate all the technology then being proposed from a number of
different sources as means of recovery and regeneration of sulfuric acid
waste pickle liquor. That study, which was completed about a year later,
26
-------
came to the conclusion that, while in the laboratory one could regenerate
waste pickle liquor, none of the processes then available was economically
viable. Shortly therafter, a consortium of steel companies invested what
was then a lot of money in a pilot plant to test a new technology, the Blaw-
Knox Ruthner process. That work confirmed our conclusion: "Waste pickle
is a lousy raw material for sulfuric acid. " That conclusion has held up-
pretty much correct over 20-some years.
I think that's enough of the personal story to indicate (1) that the
steel industry has been doing research in the environmental area for a
long time, and (2) that I've been familiar with this work for a long time.
The American Iron and Steel Institute is a trade organization which
represents some 63 domestic steel companies accounting for over 93% of
the steel production of the United States. Since 1938, for over 40 years,
the AISI has sustained the Water Resources Fellowship at Mellon Institute,
now properly known as Mellon Institute of Carnegie-Melion University.
For this entire period, the resources of this fellowship have been directed
at basic and applied investigation of pollution problems of the iron and
steel industry. The fellowship has resulted in a great deal of fine techni-
cal work and many technical publications. The first published paper, "In-
dustrial Stream Pollution Problems and Their Solution, " by Richard D.
Hoak appeared in Chemical Industries in August 1941. The most recent
paper published by the fellowship is "Evaluation of EPA Recommended
Treatment and Control Technology for Blast Furnace Wastewater" by S.
C. Caruso and George M. Wong-Chong, published by the American Society
27
-------
of Civil Engineers in April 1978. The fellowship has built and maintained
one of the finest water analytical laboratories in the world. This has been
devoted to development of many of the analytical procedures now in use
throughout the industry and by the EPA, as well as to analysis of special
samples for the industry and providing a referee analytical capability for
the industry. The fellowship has investigated a wide variety of processes,
looking both at the chemistry and the process engineering of industry waste
treatment. The work of this fellowship continues under the direction of
AISI's Technical Committee on Environmental Quality Control. That same
Committee has also sponsored a number of other studies, ranging from
biological treatment of coke plant wastes to gas evolution during slag cool-
ing.
AISI also has a Committee on General Research which has for
many years been supporting basic and fundamental research programs at
universities. In the past decade, this Committee has supported 20 envi-
ronmental research projects, ranging from studies of reaction kinetics
and mechanisms through exploration of new chemical and physical pro-
cesses for pollution abatement. We are proud to note that EPA considers
these projects to be of sufficient value that EPA has joined us in the fund-
ing of about half of these projects in our current program.
A significant proportion of the research activities in the member
steel companies is also devoted to environmental work. Based on a very
informal survey, I find about 15% of the total research expenditures in the
28
-------
steel companies is applied to pollution abatement projects. The scope of
these is very wide-ranging and would have to be reported by the compan-
ies themselves. Several of the papers being presented later in this sym-
posium by steel company representatives will demonstrate the types of
work going on.
In addition to the work done within the steel industry itself, the in-
dustry has been offered and in many eases has taken advantage of exten-
sive research and development by equipment manufacturers and by engin-
eering contractors. The manufacture and sale of pollution abatement
equipment is a very competitive field. Improved technology is one way
to beat your competition. While I haven't seen any better mousetraps, I
have seen improved precipitators, scrubbers, clarifiers, flocculators,
filters, you name it. The R&D that goes into improved equipment is in-
corporated in the steel plant facilities and, I assure you, it's paid for by
the steel companies in the price of the equipment they buy.
f.a*,
I've talked so far about what you generally think of as research
and development, and the many sources of the basic technology for envi-
ronmental cleanup of the steel industry. But there is another aspect of
the development of practicable, workable, plant-scale technology which
far exceeds all of the above in cost to the industry, and yet which is often
neglected in consideration of industrial R&D. It's called "making things
work. "
29
-------
The. steel industry is big. We handle about 300 million tons of
raw materials every year. From this we produce about 100 million tons of
finished steel products, and, remember, steel sells for about 20 cents a
pound. The very size of the operations involved in the steel industry is
such that many units that are installed in the industry represent an ex-
trapolation in size from anything built and installed in other industries.
This presents unique problems.
We work at extreme conditions. The large tonnages of materials
utilized in the industry are heated at temperatures up to the range of
3000°F in the smelting and refining operations. These extreme conditions
impose unique problems.
We work not with pure substances, but with very complex mix-
tures of natural materials. The best example, of course, is coal, which
is broken down in the coking process to a huge variety of compounds rang-
ing from the simple hydrogen and carbon monoxide through the aromatics,
benzene, toulene, and xylene, and on into the more complex tars and tar
acids. This results in purification problems much more complex than
they appear on the surface.
And our operations are cyclic. This is exemplified by the coke
ovens, where coal is charged to the ovens and 17 hours later coke is
pushed from those same oven . Or consider a basic oxygen steel fur-
nace, where in a period less than an hour the vessel is charged with
scrap, then hot metal, then blown with oxygen to achieve the refining
30
-------
reactions necessary to make steel, and then tapped to get this molten
steel from the furnace. Gas evolution varies from essentially none at
room temperature to hundreds of thousands of CFM at 3000-1- degrees,
and back to nothing. This means that steady-state design parameters
just don't work for our systems.
Because of the range and complexity and cyclic nature of the op-
erating conditions that are encountered in steel mills, it turns out that
the technology required for effective working plant systems goes far be--
yond the technological principles that can be developed in laboratory and
pilot plant. Almost every installation in the steel industry is a prototype,
and the biggest and most complex technical jobs are making those proto-
types work in the big, hot, impure, cyclically varying atmosphere of the
.steel industry.
I'd like to illustrate this with some examples.
One relatively simple example was the installation of deep bed
pressure filters for treatment of hot strip mill effluents at a member
company. A large installation -was made based on good engineering de-
sign and vendors' recommendations. But let me read you a few sections
of the initial report presented by the company which installed this sys-
tem. "The week of December 28th, the filters went into 24-hour oper-
ation. The week of January 4th effluent quality began to deteriorate. "
"The plant could not be kept in service for more than two or three days
without going down for an extended backwash; the vendors specified
31
-------
backwash cycle could not be used to properly flush the tanks. " "Finding
the proper combination of flocculant and alum and the proper injection
procedure took months of experimentation. ""Other unforeseen problems
. ..." I could go on, but I think I've made my point. The start-up of a
prototype system in a steel mill proved complex, vexatious, and expen-
sive.
Another steel company installed a complex system for treatment
of wastewaters from its coke plant. That system included two distillation
towers for ammonia removal. To give you an idea of the scale of opera-
tion, the larger tower is 127 feet tall. A paper presented by that company
said, "The total ammonia recovery system had never been operated be-
fore and no operating information was available. This facility is a test
unit.for subsequent plants. " The biological treatment section of this plant
is larger and more complex than most municipal plants. In spite of a
large electronic control system, the operation of this plant boils down to,
and again I quote from the operators of the plant, "This is mostly an op-
erator judgment situation where a visual observation of the color, smell
and consistency of the froth dictate whether changes in operation are re-
quired. " This system cost $27 million to install, yet it was truly a pro-
totype and it took years and millions of dollars to make it operate.
In the steelmaking area, the emissions from the steelmaking pro-
cesses are a long-recognized problem and one that the industry has largely
solved. Today we have installations of electrostatic precipitators, wet
32
-------
scrubbers, and bag houses. I can tell a number of tales of the complexi-
ties of the development work required to make these systems operable.
I could tell of systems which were put in after very careful design and
found not to do the job they were designed to do. They had to be replaced
or upgraded. But this is an area where, I think, we now have mature
technology. We do know how to install equipment to control the process
emissions from steelmaking.
Today, however, the emphasis is changing from control of process
emissions to control of fugitives. After the controls were developed and
installed on the main emissions came a recognition that other operations
-- the charging of the furnace, the tapping of the furnace, the handling of
the materials into the furnace -- also caused emissions, -what we call pro-
cess fugitives. Today a number of systems are being installed attempting
to control charging and tapping emissions. One member company installed
a canopy hood system at a cost of over $2 million attempting to control
these. And just last year they added to that system a second system at
a cost of $4. 6 million to control charging and tapping emissions. This
was the first retrofit of its kind installed anywhere in the country. Re-
ports are that it works very well, although there are still some emissions
from the building.
But installations of this type are pioneering, prototype installa-
tions, and their development represents a major expense which I feel
33
-------
properly is classified as part of the R, D & D effort of the industry, al-
though it does not show up as a research expense.
Turning to the coking process, which presents the major technical
problems of the industry today, I'd like to consider briefly two operations,
charging and pushing. Several years ago, AISI and EPA jointly sponsored
a million dollar research program attempting to develop a charging sys-
tem. The program was not a complete success, but some of the mechan-
ical developments attempted in that operation are included in today's stage
charging operations. Stage charging is a simple process in concept: the
coal is charged to the ovens in such a way that gasses drawn into the by-
product gas system carry with them all the particulate emissions. It's
simple in concept and theory, but making it work in the plants has involved
equipment changes, operating changes, and personnel changes. These are
time consuming and expensive, and they have succeeded. Charges for this
time and expense don't show up as R&D, but truly represent development
necessary to clean up steel mills.
The final operation I'd like to talk about is the control of coke oven
pushing. When coke is pushed from ovens, there are some emissions.
At times, there are copious, black, burning, ugly emissions. In recent
years, the industry has been working hard on the development of systems
for the control of these emiss' >ns. In principle, it's very simple. You
capture the emissions in a stream of air, then you pass that stream of air
through a dust collector and isolate the emissions. In practice, to achieve
34
-------
this simple end is a very difficult task. It can be investigated only with
full-scale installations. A wide variety of systems has been built on an
experimental basis, and yet there is none today that is universally con-
sidered 100% reliable and completely satisfactory from the viewpoint of
performance. Nonetheless, 65 installations of a possible 96 installations
in the industry will have some type of pushing emission control facility
attempting to operate by the end of this year. At an average cost of over
$4 million per installation, this represents an investment of well over a
quarter of a billion dollars on unproven technology. That, I think, is the
outstanding example of research, development, and demonstration in the
iron and steel industry.
I hope this brief presentation has been enough to convince you, as
I am convinced, that the steel industry's major investment in environ-
mental research, development, and demonstration has been an important
part in the transition from the steel plant of yesterday to the steel plant
of today and tomorrow.
35
-------
Session 1: AIR POLLUTION ABATEMENT
Chairman: Joseph W. Kunz, Acting Chief
Special Enforcement Section
Air Enforcement Branch
Enforcement Division
Region III, EPA
Philadelphia, PA
36
-------
AIR POLLUTION MISSION STANDARDS
Don R. Goodwin
U.S. Environmental Protection Agency
Research Triangle Park, NC
Thank you, Mr. Chairman.
1 appreciate the opportunity to review with you the regulatory
actions EPA is considering for the iron and steel industry.
In the half-hour allotted to this discussion, T plan to review for
you the status of each federal regulation or potential regulation relating
to the steel industry. My summary will be restricted to about 20 minutes
to allow time for your questions and comments.
We are currently working in three areas:
(1) Review of existing standards. The Clean Air Act Amendments
(CAA) of 1977 require EPA to review and revise where appropriate
all New Source Performance Standards (NSPS) every 4 years.
(2) New regulations are being considered under Section 112
because the pollutions are either declared hazardous or
potentially hazardous. In this regards I am referring specifically
to benzene and polycyclic organic materials (POM).
(3) Finally, additional regulations that will regulate new or
modified sources (coke-oven charging, top side and door leaks,
by-product plants) but not existing sources. These regulations
are required by the CAA amendments of August 1977.
37
-------
Under the first category, review of existing standards, two sources
in the steel industry have been regulated by new source performance
standards for some time. These are basic oxygen and electric arc furnaces.
Regulations controlling particulate emissions from EOF furnaces were
promulgated on March 8, 1974, and required control of only the primary
**
emissions to a level of 0.02 grains per dry standard cubic foot. Opacity
regulations were not promulgated at that time because our data were
considered insufficient and inconsistent. Secondary or fugitive emissions
were also not regulated. In March, 1977, after the collection of additional
data, opacity regulations were proposed for EOF furnaces under EPA's
Method 9. These regulations were promulgated in April 1978 limiting the
opacity to 10 percent with the provision that opacity as high as 20
percent may occur once per steel-production cycle.
Immediately following this promulgation, legal action was taken by
two environmental groups, NKDC and GASP, who protested that EPA should
also regulate fugitive emissions. We attempted to resolve this litigation
out of court by discussing this with the two environmental groups. EPA
agreed that the fugitive emissions from EOF furnaces are significant but
that we did not include these sources in the initial regulations because
there was no demonstrated technology at that time upon which to base a
regulation. EPA did agree that there was technology now being developed
to control EOF fugitive emissions, both in this country and in Europe
and we proposed a schedule to the environmental groups for the collection
of this data and the preparatir i of the regulation. However, we were
unable to agree with the environmental groups that our schedule calling
38
-------
for initiating the engineering work in October 1979 with standard proposal
in April 1981, was reasonable. We were unsuccessful in convincing the
environmental group that the schedule could not be shortened because BpF
fugitive emissions are difficult to test. In fact, new test methods
must be developed. Briefs were filed with the court and oral arguments
were heard on September 20, 1979. We are awaiting decision from the
court concerning our proposal.
While this litigation was underway, we continued our review of the
existing EOF NSPS as required by the CAA and have published our conclusions.
Briefly these are the following:
1. The best demonstrated system of emission control at the time
the standard for primary emissions was established has not changed in
the past five years. These technologies control emissions to a level
consistent with the current standards. Therefore, revision of the
existing standard is not required if only primary emissions are to be
controlled.
2. The ambiguities in the present standard concerning the definition
of a BOF furnace and the operating cycle during compliance testing need
to be clarified and a project to do so has been initiated.
3. Secondary or fugitive emissions from BOF furnaces represent a
major air pollution emission source. EPA, therefore, intends to initiate
a project to revise the existing standard of performance to cover these
fugitive emissions. This project is scheduled to being in October 1979,
and lead to a proposal 20 months later.
39
-------
The second existing standard which has undergone review is electric
arc furnaces. The new source performance standards to control particulate
emissions from electric arc furnaces was proposed in October of 1974 and
promulgated in September 1975. This standard dealt only with particulate
emissions from the furnace during normal operations, restricting them to
0.005 grains per dscf.
Since the NSPS was promulgated in September of 1975, five new
installations subject to the NSPS have been completed. Although only
one official NSPS compliance test has been carried out since promulgation
of the new source performance standard, some unofficial test data indicate
that certain new furnaces are well below the NSPS for particulates and
visible emissions. However, the effectiveness of pick-up systems for
process and fugitive emissions has greatly improved. Plants are now
using total enclosure in conjunction with hoods and air curtains, to
capture melt downs, charging, tapping, and slagging. It has been recommended
therefore that the standard be modified to give consideration to these
new capture techniques. This recommdndation is made in spite of the fact
that fabric filters, the primary method of emission control when the original
standard was promulgated, are still considered the best control technology.
In addition, it is recommended that argon-oxygen decarbonization furnaces
be included in the revised EAF standard since they are often found in
EAF shops and utilize common control systems. We anticipate approval of
our recommendation by the EPA Administrator and publication of our
findings for your comment late this year. Schedules for actual revision
of the regulation will be developed following a review of the public
comments.
40
-------
The second area of EPA regulation of steel mill emissions is under
Section 112 which deals with hazardous pollutants.
Section 122 of the Clean Air Act specifically requires the Adminis-
trator of the Environmental Protection Agency to make a regulatory
decision regarding control of atmospheric emissions of polycyclic organic
matter (POM) on the basis of a health risk assessment of these emissions
conducted by EPA and supported by similar assessment of coke-oven emissions.
The Administrator is considering POM as a hazardous air pollutant under
Section 112 of the Clean Air Act. In anticipation of this, EPA is in
the process of developing and collecting data upon which regulations
limiting POM emissions from coke-oven production facilities can be
based. We have used data from epidemiological studies and ambient monitoring
to estimate the health risk of POM emissions from coke plants. For our
health risk calculations we have taken into consideration the improved
emission control at coke plants resulting from implementation of State
regulations, consent agreements, Occupational Safety and Health Administration
regulations, and industry initiatives.
The regulations we are considering would, as a minimum, require
emissions control to levels that are attainable with best available
control technology as defined by EPA's recently proposed cancer policy.
The types' of controls upon which the standard would be based will vary
among the sources within the coke plant, but may include revised operating
and maintenance procedures as well as modifications to equipment and
installation of control devices.
41
-------
Health assessment documents are now being prepared. It will be
necessary to obtain concurrence from the Administrator's Science Advisory
Board on the listing of POM under Section 112 of the Clean Air Act as a
hazardous air pollutant. Current schedules call for listing POM in
March of 1980 with public hearings to follow. In anticipation of this
designation, we are looking at sources of POM which we will be required
to regulate in response to the listing of POM as a hazardous pollutant.
The initial sources involved are wet coal charging, topside leaks, and
the door leaks which are all part of coke making. The technical package
for this regulation will be available next summer, with a proposal
anticipated in about a year.
Since this action is under the hazardous pollutant section of the
CAA, any regulation will apply to both new and existing sources.
A second coke-oven source which will probably be regulated under
Section 112 of the Act is dry coal charging, more commonly referred to
as coal preheating. Our preliminary data indicates that the emissions
contain polycyclic organic matter (POM) but the technical work to support
a regulation has just been initiated. Any regulation will be over a year
away.
Another pollutant which may be regulated as a hazardous pollutant
is benzene. Benzene emissions were officially listed by EPA as a hazar-
dous pollutant under Section 112 of the Act on June 8, 1977, and regula-
tory packages are being prepared for the major benzene emitting sources.
42
-------
Our highest priority of sources emitting benzene are:
1. Maleic anhydride production.
2. Coke by-product plants.
3. Benzene storage and handling.
4. Gasoline marketing.
With regard to the by-product plants, we initiated the preliminary
analysis of the process streams to determine which streams contain the
major quantities of benzene. This was an attempt to determine the
primary sources of benzene emissions from a coke oven by-product plant.
This analytical work has proven to be very difficult, and I anticipate
some delay in categorizing the sources which will require control. In
spite of these delays, I anticipate having a technical package for
review by an internal EPA working group in mid-1980, and, of course, we
will make it available to the industry for your comment at that time.
The final area which we should cover is the development of standards
of performance for new or modified sources for which there is no existing
NSPS.
During the 1977 hearings on the Clean Air Act, Congress received
testimony on the need for more rapid development of new source performance
standards. There was concern that not all sources which had the potential
to endanger public health or welfare were controlled by NSPS and that
the potential existed for environmental blackmail from source categories
not subject to NSPS. These concerns were reflected in the Clean Air Act
amendments of 1977, specifically in lll(f). We were required to use the
three criteria listed by Congress: (a) the quantity of emissions, (b)
the extent to which each pollutant endangers public health or welfare,
43
-------
(c) and the mobility and competitive nature of the sources, to prioritize
all sources which should be subject to new source performance standards.
This has been accomplished, and EPA has undertaken: a program to promul-
gate new source performance standards for source categories on this
priority list by August 1982. Development of standards has already been
initiated for nearly two-thirds of the source categories on the priority
list. Work on the remaining source categories will be initiated within
the next six months. The following will be a brief summary of the steel
industry sources to be covered:
1. Coke-Oven Battery Stacks. Source testing was completed at
Kaiser Steel in Fontana, California, during September. Benzene soluable
organics, particulates and benz-a-pyrene were measured. We expect the
technical report to be completed in early 1980. Proposal is scheduled
for about October 1980. Data confirms our decision to develop NSPS
rather than NESHAP for battery stacks.
2. Wet Coke Quenching. This source was originally thought to be
a high emitter of POM. However, data now available indicate this is not
the case and it is now considered a candidate for NSPS for particulate
rather than NESHAP. As you will see later in the program, testing is
complicated and expensive. Our current thinking is that clean water and
good design baffles will be the approved technology. However, we are
watching closely a source test being conducted this week at Lone Star
Steel. Our schedule is not clear but a proposed regulation is at least
a year away.
44
-------
3- Sinter Plants. This has just been initiated as a Section 111
NSPS project.
4. Desulfurization of Coke Oven Gas. EPA will begin engineering
work in February 1980.
5. Coke-oven Pushing. We do not currently have a schedule for
coke pushing. It is not considered a POM source but could be a
particulate source.
45
-------
INNOVATIONS AND IMPROVEMENTS
OF THE ORE SINTERING PROCESS
FOR AIR POLLUTION CONTROL
By: Thomas E. Ban
Vice President
Research and Development
McDowell-Wellman Company
Cleveland, Ohio 44114
Presentation: Symposium on Iron and Steel
Pollution Abatement Technology
U. S. Environmental Protection Agency
Chicago, Illinois
October 30 - November 1, 1979
46
-------
INNOVATIONS AND IMPROVEMENTS
OF THE ORE SINTERING PROCESS
FOR AIR POLLUTION CONTROL
Thomas E. Ban
The McDowell-Wellman Company
ABSTRACT
An improved sintering process for agglomerating
ores has been developed at the Dwight-Lloyd Research
Laboratories. This embodies strand cooling with hot
and cold recycle draft streams. The new process
minimizes draft exhaust quantities, oxidizes entrained
hydrocarbons, refilters solid particulates, and con-
ditions the final sinter plant exhaust for ultimate
cleaning.
4?
-------
INNOVATIONS AND IMPROVEMENTS
OF THE ORE SINTERING PROCESS
FOR AIR POLLUTION CONTROL
INTRODUCTION
The continuous sintering process for agglomerating
ores is perhaps the highest capacity processing system
in the metallurgical industries. On a worldwide basis
the production of sinter in ferrous metallurgy exceeds
500 million tons per year and the nonferrous and non-
metallic industries produce about 40 million tons per
year. The large individual Dwight-Lloyd machines con-
sume more than 15,000 tons per day of raw materials and
have exhausts of combustion draft which exceed 1,000,000
CFM.
In extractive metallurgical practices sintering is
used to beneficiate ores through a higji^.temperature
agglomeration process. This makes a "furnaceable" pro-
duct from raw ore which would otherwise have substantial
penalties because of fines, nonuniform size consist, poor
handling characteristics, and detrimental chemistry. Ad-
ditionally, sintering is used to recycle plant waste dusts
48
-------
and to pre-blend and pre-calcine fluxstone as a con-
stituent with ore charge to provide a self-fluxing or
superfluxed sinter.
Beneficiation is an all encompassing term to cover
the improvement of the physical and chemical properties
of an ore by lowering the cost requirements for smelting.
Conversion of ores, recycle fines and fluxes into bene-
ficiated charge of sinter product is noted by the fol-
lowing factors.
1. The fines are agglomerated into hardened
semi-fused and enlarged particles
which are porous and durable for gas-
solid smelting reactions.
2. The raw materials are dried and cal-
cined for removal of sulfides and
hydrocarbons as well as carbonates,
sulfates, and hydrates which cause
endothermal smelting reactions.
3. Undesirable constituents as volatiles
and sublimates are removed from the
product by high temperature
reactions of sintering.
49
-------
Sinter product is characterized as hardened,
cellular masses which are specifically crushed and
graded for smelting furnace operations, i.e., ideally,
-l"+l/4" structure resembling "popcorn" in shape and
size.
PRINCIPLES
The sintering process is carried out by incipient
fusion of bedded particles in the quiescent state.
Combustion of solid fuel dispersed within the bed and
induced draft are used to sustain the sintering phe-
nomena. In continuous practices raw materials such as
ore fines, fluxes, fuels and recycle return fines are
proportioned, blended and tumbled with moisture to
create a nodular textured burden as a ribbon of porous
charge for a series of traveling grates. The leveled
charge supported on moving grates is surface ignited
by flame ignition and a combustion draft is induced
through the bed by suction fans which perpetuate a
frontal firing wave downward through the charge as it
is transported horizontally by the traveling grates.
This gives rise to an inclined firing plane throughout
the moving sinter bed as illustrated in Figure 1. The
50
-------
high temperature firing zone causes incipient fusion
at the solidus temperature of the blended raw mate-
rials. In the case of iron ore sintering this peaks
at about 2500 °F.
The mechanics of sintering is typified as a series
of periodic thermal gradients or a firing wave front
advancing through the charge as exemplified in Figure 2
which illustrates the gradients at initiation of sin-
tering and 257o, 5070, 75% and completion. The peak tem-
perature of the wave is the firing zone and it is pre-
ceded downstream by the lower temperature calcining
and drying zones respectively as the wave advances
toward the unreacted charge. A cooling zone is upstream
of the firing zone and it serves to transfer and recu-
perate sensible heat for the draft of combustion.
Figure 1 further illustrates the specific zones
within the moving sinter bed as unreacted charge is
converted into cooled sinter through coincidental com-
bustion, calcination and vitrification reactions.
The quiescent state of the sinter bed charge and
the filter bed nature of wet packed nodules in the lower
layers are features which minimize entrainment of solids
in the draft stream. Sintering is remarkably efficient
51
-------
in this respect. It is shown that less than 0.1% of
the charged materials become air entrained or stripped
from the bed despite the severe temperature changes,
high velocities of evacuated draft and use of raw
materials comprised of 100% minus 200 mesh particles.
Contrarily, sintering gives rise to emissions because
one of its functions is to pre-remove the deleterious
volatile constituents from the ore. The high tempera-
ture firing phenomena cause constituents such as alka-
lies, chlorides, heavy metal salts, and other subli-
mates to fume from the firing zone and be entrained in
the draft stream. Also, higher draft temperatures
which precede and follow the drying front cause vola-
tilization of inherent hydrocarbonaceous materials
within the sinter burden.
Recycle steel plant materials such as flue dust
and sludge contain high quantities of salts and sub-
limates because the original plant dusts nucleate
their precipitation. These dusts also contain high
percentages of condensables such as hydrocarbons.
Other prominent sources of condensables as potential
particulates are the lubricants inherent with recycle
mill scales, flotation reagents of ores and the vola-
52
-------
tile matter constituents of the coke breeze and
other types of solid carbonaceous sinter fuels.
SINTERING PRACTICES
A generalized flowsheet as a sinter plant layout
is shown in Figure 3 which illustrates a conventional
plant for converting iron ore fines, miscellaneous
plant dusts, and limestone into self-fluxing sinter.
Raw materials of a wide variety of physical character-
istics are ordinarily delivered to rail, truck or yard
hoppers for admission to the sinter plant. These mate-
rials consist of (1) several varieties of iron ore fines
as minus 3/8 in. screenings in the damp natural state,
(2) miscellaneous steel plant recycle materials such as
dry or conditioned blast furnace flue dust, moist sludge
and mill scale, (3) fine limestone, and. (4) ground coke
breeze for sinter fuel. After charging these into the
plant they are transferred by belt conveyers to separate
day bins which surge and proportion the raw materials.
Table feeders or large belt feeders are used to with-
draw materials as controlled by belt scales to a col-
lecting conveyer and withdrawal rates are pre-designed
to produce controlled carbon input of approximately 470
53
-------
of the charge. Limestone is controlled to make the
required basicity ratio with gangue constituents.
A nodulizing operation is used to blend and pre-
pare the raw materials as a uniform and permeable
charge with a nodular texture. While the raw materials
are in transit to the nodulizing operations they accu-
mulate recycle fines from dust collectors and hot and
a*,
cold sinter screening operations which supply sinter
returns. The composite blend is nodulized with mois-
ture in rotary drums and nodules are charged to the
Dwight-Lloyd® traveling grate. In many cases a recycle
hearth layer comprised of 3/4xl/4-in. sinter returns
is applied as an initial 1%-in. layer directly upon
the grates to serve as a charge filter, heat sink and a
parting plane for facilitating the discharging. Raw
nodulized sinter machine charge is applied directly
on the hearth layer and it is leveled by a strike-off
gate to an even 12- to 18-in. layer which passes di-
rectly beneath an ignition furnace. Open flame for
about 60 to 90 seconds duration ignites the surface
of the bed and this initiates the sintering reactions
which are maintained by a continual downdraft as
induced by a sintering fan. Dust collection equipment
54
-------
.is used downstream of the fan and the exhaust is di-
rected to a stack. Sinter production is controlled to
terminate when the fired line reaches the grates at the
discharge end. A consolidated sinter cake which is
relatively cool in the upper layers and hot, semi-fused
near the bottom layers is discharged as a mass the
approximate size of the sintering machine pallet. In
large installations the cake is approximately 10-ft.
wide by 4-ft. long by 12-in. thick. Rotary breakers
and stationary grizzlies serve as primary crushers for
the cake and a hot screen is used for immediate rejec-
tion and recirculation of hot sinter returns.
The gravity flow arrangement allows the broken,
hot sinter to be continuously applied to a circular
traveling grate and forced draft fans are used to air
cool product by convection. Cooled product is then
transferred to a final screening station for recovery
of cold returns, hearth layer, and final disposition
of the sinter product to storage piles. In the more
modern sinter plants the final sinter product is care-
fully crushed and graded to the preferred size of about
lxl/4-in. sizes and further sinter returns are generated
by this operation.
55
-------
The conventional sinter plant and sintering system
have disadvantages which impose high capital and high
processing costs as well as environmental problems.
Some of these are listed as follows:
1. The belt transfer of unconditioned,
dusty raw materials and returns on
conveyer systems creates high quanti-
ties of in-plant airborne dusts.
2. The gravity transfer of hot sinter
through the discharge end - hot
screening - cooling series imposes
capital and operating expenses inher-
ent with multi-story operations.
3. The hot sinter discharge imposes
expense for hot crushing, hot
screening, and hot de-dusting
because of excessive maintenance,
downtime periods, and plant returns
with lower sinter product qualities.
4. The separate sinter cooler machine as
charged with a wide range of broken
sinter sizes requires excessive draft
exhaust and machinery because of parti-
56
-------
cle segregation and ncnuniform dis-
tribution of cooling media.
5. A very large volume of sinter machine
exhaust as collected by windboxes
along the machine necessitates extensive
and massive air pollution control equip-
ment for arresting fine entrained par-
ticulates.
Three prominent stack sources for sinter plant
emissions can be noted in the flowsheet of Figure 3
and each of these are specifically identified in the
following outline:
1. The main sinter machine exhaust
This source contains all of the
products from the high temperature
sintering and combustion operations
as composited from all windboxes along
the sintering machine. The high tem-
perature sintering phenomena cause
entrainment of fine dusts, sublimates
and condensable hydrocarbons. Table 1
presents some general compositions and
characteristics of iron ore sintering
57
-------
Stack emissions which ordinarily
escape arrestment from low energy
"cyclone" dust collectors.
2. Discharge end - screening exhaust
This source of emissions is caused
by unconsolidated particulate matter
in the product which becomes en-
trained through dumping, crushing,
and screening operations.
3. Cooler exhaust
Emissions in the massive exhaust of
cooling operations are largely caused
by entrainment of unconsolidated
particles in the product which are a
result of screening inefficiencies
and improper sintering operations.
IMPROVED SINTERING PROCESS
A new sintering process has been developed at the
Dwight-Lloyd Research Laboratories which uses recycle
draft with strand cooling principles for overcoming air
pollution. A combination of cold recycle draft from
initial windboxes and hot recycle from terminal wind-
boxes re-uses sinter flue gases for the following pur-
poses:
58
-------
1. Settling partials of solid partic-
ulates within the wet nodular filter
bed of the sinter charge.
2. Incinerating partials of entrained
condensables and combustible gases
3. Diminishing the volumes of exhaust
gases.
4. Conditioning exhaust gases for
ultimate cleaning.
The elements of this new patented process have been
proven within several continuous pilot plant sintering cam-
paigns and currently the principles are being carried
out in commercial design and operations. Essentials
of the process are illustrated in Figure 4 which shows
the draft flow arrangement with a sinter bed section
on the Dwight-Lloyd traveling grate. The sintering is
performed by terminating the fire line of combustion
approximately midway within the third zone which com-
pletes the combustion and performs primary cooling.
Terminal cooling is carried out in the fourth zone and
the preheated air is recycled for initiating sintering
immediately after ignition. The second and third zones
use sequences of cold recycle draft and the final
59
-------
exhaust is vented from the third zone. Through this
arrangement sintering is maintained with its flue gases
containing H20, C02, and 02, each within the range of
about 8-16%. This media has been proven adequate if
not beneficial for the maintenance of all sintering
phenomena. Figure 4 illustrates the use of low energy
inertial collectors such as cyclones for protection of
the recycle blowers and two stages of cleaning pre-
ceding the final exhaust blower in the third zone.
These stages are used to indicate sequences of ultimate
incineration or quench stages followed by dry or
scrubbed particulate removal. Table 2 provides some
pilot plant data which compares sinter machine emis-
sions of the conventional process with the improved
process. From these data it is apparent that the
emissions can be reduced by a factor of two to four
through use of the Improved Sintering Process™. Because
of the recirculatory draft system the C02 and H20 con-
tents of the draft stream increase two- to threefold
and the exhaust draft temperatures increase from about
250 OF to 650 °F. The draft of the Improved Sintering
Process is thereby conditioned with humidity and tem-
perature for dry electrostatic precipitation. Also,
60
-------
the diminished volume enables a .ligh velocity energy scrubber
to be economically applied for arresting the diminished
loading of both solid and condensable particulates for
compliance of sinter plant environmental requirements.
BENEFICIAL ASPECTS
The adaptation of strand cooling principles of the
sintering processes has finally been accepted in Euro-
pean iron ore practices wherein nine very large scale
plants have been constructed since 1973 by Delattre-
Levivier, licensee of the Dwight-Lloyd sintering
process of the McDowell-Wellman Company^^.
Though pioneering efforts were carried out by Jones &
Laughlin Steel Corporation during the early 1950"s
using strand cooling principles of the Dwight-Lloyd
sintering process, at Benson mines, more than 25 years
elapsed before it was accepted in the very large scale
(23)
iron ore sinter plant construction ' . Strand cooling
plant designs have shown capital and operating cost
benefits over the conventional sinter plant design
illustrated in Figure 1. These benefits have been
attained principally because of the low profile plant
construction, single machine utilization, and mainte-
nance-fuel process savings.
61
-------
Use of draft recycle has been practiced in pel-
letizing operations for converting fly ash and iron
ore concentrate wherein the pellets were fired with
solid fuel. Future developments of sintering utilizing
the liquid sealed circular sintering Dwight-Lloyd ma-
chine will show further cost savings and low cost envi-
ronmental control through the totally sealed features
of the circula*r machine.
Comparison of the materials balance flowsheet for
conventional sintering and the Improved Sintering Pro-
cess are illustrated in Figure 5 which shows the marked
(4)
diminution of exhaust volumesv . In capsule, the
Improved Sintering Process offers the following advan-
tages over conventional sintering practices.
1. Quantity of total exhaust draft:
25-35% of that in conventional
sintering practices.
2. Quantity of solid particulate
materials in exhaust stream:
20-50% of that in conventional
sintering practices.
3. Quantity of condensable particu-
lates in exh ,ust stream:
62
-------
15-457o of that in conventional
sintering practices.
Pilot plant experience from sintering iron ore
with blast furnace flue dust and mill scale have shown
organics as condensable hydrocarbon emissions to be
lower than 0.03 gr./dry std. cu. ft. ahead of particu-
late arresting equipment. The foregoing improvements
may meet compliance without extensive further cleaning
of draft streams. However, if further cleaning is
required it is more economically attained. This is
apparent because of the volume diminution and condi-
tioned nature of gases for further treatment by incin-
eration, dry precipitation, or wet scrubber applications
COMMERCIAL APPLICATIONS
After detailed pilot plant and engineering design
work by McDowell-Wellman Company, the Jones & Laughlin
Steel Corporation constructed a single pass cold recycle
sintering system at their 6,000 tons/day Aliquippa iron
ore sinter plant. The commercial data essentially
verified the pilot plant test data which were used as
bases of design for retrofit and which illustrated:
63
-------
1. Maintenance of the high sinter
rates despite the use of recycled
flue gas as sintering media.
2. Considerable diminution of exhaust
gas volumes.
3. Marked diminution of solid and
condensable particulates.
A large pilot plant project was recently completed
for Lone Star Steel Company using the Improved Sintering
Process - Multipass Recycle and Strand Cooling. On
the bases of very favorable results, Lone Star Steel
Company is constructing a large scale sinter plant at
their ore beneficiation facilities at Lone Star, Texas,
which will use this new technology. As a result of
separate field studies by Lone Star Steel Company, the
Improved Sintering Process will embody a new design of
the Lone Star Steel Company proprietary Hydro-Sonic®
scrubber (-*). This will operate within the depleted
t
volume and depleted particulate exhaust stream of the
sintering process. All test data confirmed that the
Lone Star Steel Company sinter plant will economically
and environmentally conform to the air pollution code
requirements.
-------
SUMMARY
Conventional sintering practices for agglomerating
ores have high air pollution concerns because of massive
total draft exhaust volumes containing fine solid fume
and condensable particulates. These disadvantages can
be overcome to a considerable extent by use of an Im-
proved Sintering Process which uses recycle draft and
strand cooling principles. This innovation combines
the multiple machine applications of the conventional
practices into a single Dwight-Lloyd sintering machine.
The overall benefits are cost savings inherent with
•
strand cooling and marked diminution of (1) draft ex-
haust, (2) solid particulates, and (3) condensable
particulates.
On the basis of favorable continuous pilot plant
operational data, the commercial applications have been
committed and initiated.
65
-------
LITERATURE CITED
1. Dorville, Robert, N. Meysson, J. Astier, J. A.
Didier, 36th Ironmaking Proceedings, The Iron
and Steel Society of AIME, Vol. 36 (Pittsburgh,
April 1977).
2. Simon, Francois, Robert Nicholas, "No. 2 Sintering
Plant at Sidmar," The Iron and Steel Engineer,
July 1973.
3. Rowen, Harold, "Modern On-Strand Sinter Cooling,"
Society of Mining Engineers Fall Meeting, AIME,
Pittsburgh, Pa. (September 1973).
4. Ban, Thomas, "An Improved Sintering Process to
Overcome Environmental Problems in the Sinter
Plants," Society of Mining Engineers Annual
Meeting, AIME, (New York, N.Y., 1975).
5. Ewan, T. K., J. S. Master, "Fine Particle Scrubbing
with Lone Star Hydro-Sonic Cleaners - The Coalescer,"
Second Symposium on Fine Particulate Scrubbing,
sponsored by US-EPA, (New Orleans, La., May 1977).
66
-------
TABLE 1
IRON ORE SINTER MACHINE EMISSIONS
GENERAL RANGE OF ANALYSES*
Chemical
Analyses
%
Fe203
Si02
A1203
CaO
MgO
K20
Na20
Cl"
so4=
Organics
Grain Loading of
10-40
2-8
2-5
7-15
2-5
2-10
1-6
3-10
4-14
10-40
Particulates:
Size
Analyses
Cum. % Retained
+40 Micron 10-50
+20 Micron 15-65
+10 Micron 20-80
+3 Micron 60-90
+1 Micron 90-95
-1 Micron 5-10
Solids: 0.1 to 0.5 gr./dry std. cu. ft.
Condensables: 0.05 to 0.3 gr./dry std. cu. ft.
* Multiple source of information as general
range data from six different plants after
primary cleaning by cyclones.
67
-------
TABLE 2
EMISSIONS MEASURED DURING COMPARATIVE
PILOT PLANT SINTER TEST RUNS*
Emissions Emissions Percent of
Conventional Improved Conventional
Sintering Sintering Sintering
Process Process Process
Ibs/hr Ibs/hr 70
13.4
21.5
4.1
2.6
27.2
23.7
2.6
3.4
4.5
4.7
0.6
0.6
11.1
11.8
1.1
1.0
33.6
21.8
15.8
25.0
40.8
49.8
44.3
33.2
MATERIALS BALANCES
Nos. 3 & 4
Solid
particulates - A
- B
Condensable
particulates - A
- B
MATERIALS BALANCES
Nos. 5 & 6
Solid
particulates - A
- B
Condensable
particulates - A
- B
* Results from comparable continuous pilot plant
runs at charge rates of 7,000 to 8,000 Ibs/hr
using raw materials comprised of iron ore,
blast furnace flue dust, blast furnace sludge,
BOF fume, mill scale, and limestone.
68
-------
LIST OF CAPTIONS
Figure 1 - Isometric View of Sinter Bed Illustrating
Zones of Reaction.
Figure 2 - Periodic Sinter Bed Temperature Gradients.
Figure 3 - Perspective View of Iron Ore Sinter Plant.
Figure 4 - Improved Sintering Process Draft Flow
Arrangement.
Figure 5 - Materials Balance Diagrams Comparing
Conventional and Improved Sintering
Process.
-------
-J
o
Figure 1 - Isometric View of Sinter Bed Illustrating
Zones of Reaction.
-------
3000
2500
2000
<
Of.
SURFACE CENTER
LOCATION WITHIN BED
Figure 2 - Periodic Sinter Bed Temperature Gradients.
BOTTOM
-------
-J
Isi
INTERNATIONAL
ENGINEERING SERVICES
DE LATTRE LEVIVIEH, PARIS FRANCE
FLOW SHEET
ALTOS HOHNOS DE MEXICO S A
DWIGHT-LLOYO SINTERING PLANT
LICENSED BY
McOOWELL-WELLMAN
ENGINEERING CO
I.E.S. 63-116
Figure 3 - Perspective View of Iron Ore Sinter Plant.
-------
ZONE
Feed
N /Discharge
Figure 4 - Improved Sintering Process Draft Flow
Arrangement.
-------
ORE, FLUX, FUEL
SINTERING AIR
\ s
DISCHARGE
EXHAUST
COOLING AIR
V /
RECYCLE RETURNS
AND HEARTH LAYER
Materials Flow Diagram
Conventional Sintering
and Cooling Process.
ORE, FLUX, FUEL
\ 1.28 >y
&>
MOISTURE INJECTION ^
COOLING AND
SINTERING AIR
RECYCLE RETURNS AND HEARTH LAYER
TOTAL EXHAUST
Materials Flow Diagram Recycle DRAFT
Sintering-Cooling Process.
Figure 5 - Materials Balance Diagrams Comparing
Conventional and Improved Sintering
Process.
74
-------
RTI/1736/03-01S September 1979
ENVIRONMENTAL ASSESSMENT OF COKE BY-PRODUCT RECOVERY PLANTS
by
C. C. Allen, Jr.
Research Triangle Institute
P. 0. Box 12194
Research Triangle Park, North Carolina 27709
This paper identifies potential air pollution sources of environmental
concern in coke by-product recovery plants. Data concerning the design and
operation of existing plants and processes were collected. Since many process
variations exist, a survey of the industry was carried out to determine the
most common processes. Following this, the processes at a representative
plant were sampled, using EPA's Industrial Environmental Research Laboratory
RTF Level 1 protocol. Air pollutants of concern included benzene, cyanide,
and polynuclear aromatic hydrocarbons. The air was sampled at suspected
pollution sources, primarily storage tanks. The largest emission source was
the final cooler tower where concentrations of aromatics at >50 g/Mg coke and
cyanide at 278 g/Mg coke were found.
75
-------
ENVIRONMENTAL ASSESSMENT OF COKE BY-PRODUCT RECOVERY PLANTS
This report discusses the findings of a screening study of the multimedia
environmental effects of U.S. coke by-product recovery plants and their related
pollution control technologies. The purpose of the study was to analyze
relevant background data, to acquire new data by sampling and testing, and |o
draw conclusions concerning the environmental acceptability of the process.
There are 60 coke by-product plants in the country; these processed gases
from an estimated 44 million tonnes of coal in 1978. There are many potential
air pollution sources for the by-product recovery plant. These are related to
seven major operations: (1) tar processing, (2) ammonia processing, (3) final
cooling-naphthalene handling, (4) light oil recovery, (5) desulfurization,
(6) cyanide handling, and (7) water handling. For each operation there are
alternative technologies, and existing plants employ only a few of the thousands
of combinations of operations available.
Except for still vents and forced drafts (e.g., final cooler cooling
tower), emissions to air are fugitives—tank breathing and working losses,
open decanters, and basins. Fugitives are also due to faulty equipment, such
as pump seal leaks and flange leaks. Light aromatics and polynuclear aromatics
were positively identified as air pollutants, and polynuclear aromatics were
indicated for other media.
One pollutant species of particular interest is benzene. Benzene has
been listed as a hazardous air pollutant under Section 112 of the Clean Air
Act (National Emission Standards for Hazardous Air Pollutants). To protect
the public health from unreasonable risks associated with exposure to airborne
benzene and polycyclic organic materials, a standard is being developed to
decrease benzene emissions from the by-product plants. Emissions of carcinogenic
air pollutants such as benzene are of particular concern since no safe level
of exposure has been identified. Although the generation of benzene is an
unavoidable consequence of coke production, most of the benzene which is
produced in coke ovens is recovered, leaving typically only a small percentage
that is emitted into the air. Potentially significant amounts of pollutants
are emitted, however, due to the volume of plant throughput. The emissions of
hazardous air pollutants are highly dependent on the type of processes used in
the by-product plant.
PROCESS OVERVIEW
Although there are many possible process combinations, there are generally
only two or three ways widely used in U.S. coke by-product recovery operations.
The gases leaving a coke oven are generally at around 700°C and of course
contain all of the material to be processed in the by-product plant. Coke
ovens are maintained at a slight positive pressure (1 mm water) to prevent air
infiltration. As the gas leaves the oven, it is immediately subjected to
spray cooling, both to cool the gas and to introduce a collecting medium for
the tar as it condenses. After a short duct run the gas passes through a
valve and enters a suction main, remaining below atmospheric pressure. At
this point, the gas which has generally been cooled to the 100°C range con-
denses much of the water, tar, ammonia, and other low boiling point compounds.
76
-------
Further removal by condensation is accomplished in the primary cooler. The
tar and the water soluble compounds are separated by decantation. The tar is
generally further dewatered before sale. If phenol is recovered from the
ammonia liquor, it is often absorbed in an organic solvent before the ammonia
recovery step. The ammonia liquor is traditionally steam-stripped to put the
ammonia back into the gas stream. The waste mmonia liquor requires addition
of a base to release some chemically bound ammonia.
Looking again at the gas stream, the exhauster is the fan which provides
motive power for the gas. Both scrubbers and electrostatic precipitators are
used in the industry for nearly complete recovery of the remaining tar in the
gas. After the ammonia stripped from the waste ammonia liquor rejoins the gas
stream, the ammonia can be scrubbed from the gas with a dilute sulfuric acid
solution. Ammonium sulfate crystals form and are separated from the saturated
liquor. The final cooler is a pretreatment step for light oil (benzene)
recovery. In the process which generally uses contact cooling with water,
naphthalene is condensed from the gas. The naphthalene may be removed from
the water by absorption in organics or by flotation. Light oil is usually
recovered by absorption in a petroleum fraction (wash oil). The light oil is
steam stripped from the wash oil and recovered, and the wash oil recirculated.
Desulfurization, if practiced, is intended to make coke oven gas a more acceptable
fuel. Only a few large plants practice desulfurization. No process is in
widespread use today.
At least three powerful influences militate against any single process
description being widely applicable: (1) today's by-product plants have often
evolved over 20-50 years of maintenance, design, and operational changes;
(2) the technology is mature, and there are many proven alternate ways to
recover chemicals; and (3) the market for coal chemicals is uncertain, and
economic pressure has led to changes in operating philosophy.
EXPERIMENTAL
Based upon a review of the more common process combinations currently in
use for by-product recovery, a representative plant was selected for Level 1
environmental assessment. The purpose of a Level 1 environmental assessment
is to provide a screening or survey look at emissions from an industry, high-
lighting potential areas of environmental concern. The test work was done at
the Fairfield Works of U.S. Steel Corporation near Birmingham, Alabama.
The Level 1 assessment protocol recommends that all identified emissions
to all media be sampled and analyzed, as well as the feeds to and products
from the process. All of the samples are grab samples, and the intended
accuracy is to be within a factor of 2 or 3 of the actual emissions. Exami-
nation of the process flow of a by-product plant showed that most air emissions
were fugitive and primarily composed of organic compounds. The potential for
these fugitive emissions to contain significant amounts of aromatics and high
molecular weight polynuclear aromatics (PNA's) was apparent. Hydrogen cyanide
was also identified as a potentially significant component.
The sampling program developed for this study was centered on organic
vapor emissions from tank vents and a cooling tower. Three types of sampling
were used for the organic vapors: (1) glass bulb grab samples, (2) evacuated
canister grab samples, and (3) 1 to 4 hour samples drawing the gas through an
77
-------
adsorbent resin, XAD-2. The glass bulbs were analyzed for light (C^-C-,)
hydrocarbons and volatile sulfur species using an on-site gas chromatograph
(GC). Benzene and toluene were also quantitated with this GC. The evacuated
canister samples were returned to the laboratory for analysis to identify and
quantitate benzene, toluene, the xylenes, and ethylbenzene. The adsorbent
resin was intended to adsorb hydrocarbons with carbon numbers greater than 7,
or boiling points above about 100°C. The resin was extracted with a solvent
and the extract analyzed in three ways:
(1) Total Chromatographable Organics (TCO), which is nominally the
mass of organic compounds with boiling points between 200°C
and 300°C.
(2) Gravimetric Analysis (GRAV), which is nominally the mass of
organics with boiling points above 300°C; and
(3) Liquid Chromatography, LC, which is used to divide an extract
into seven fractions (or cuts) which are graded by their
polarity.
The analysis generally proceeds with a TCO and GRAV analysis of the
original sample extract (preliminary), a concentration step to achieve a
specified organic concentration (GRAV and TCO are also run on this concen-
trate), and then the LC work, with a GRAV and TCO determination on each LC
cut.
The data for the plant site tested were developed utilizing methodologies
based on the Environmental Protection Agency's Level 1 protocols, with limited
gas chromatograph-mass spectrometer identification of specific pollutants.
The PNA's shown are the quantities of residual organics obtained upon evapora-
tion of the solvent used for extraction, which are nominally those organics
with boiling points above 300°C. The PNA's emission factors are subject to
the additional uncertainty inherent in this method of estimation. Specific
PNA's were not identified except for three sources considered most likely to
involve them: the tar decanter, the dewatering and storage tanks, and final
cooler cooling tower. Sulfur compounds are reported as hydrogen sulfide;
cyanides, as hydrogen cyanide.
In addition to the Level 1 sampling and analysis, samples for hydrogen
cyanide were taken at the final cooler cooling tower and 24-hour integrated
samples were collected at three points around the plant boundary. The gas was
bubbled through a sodium hydroxide solution for cyanide absorption and analyzed
by wet chemistry.
To assist in the determination of the potential environmental signifi-
cance of the various pollutant concentrations, comparison was made to values
derived from the MEG's charts. The most toxic compound in each of the 17
compound classes was identified, and its minimum acute toxic effluent (MATE)
concentration was used for comi .risen. It must be kept in mind that the
resulting ratio is biased. If it is well below unity, there would appear to
be no concern for compounds in this class; if, however, the ratio is above
unity, it is merely a signal for potential environmental concern.
78
-------
TAR PROCESS EMISSIONS
The major potential emission sources for benzene and polycyclic organic
materials are the tar decanting, dewatering, and storage operations.
The flushing liquor from the collecting mains is separated from the
collected tar in a decanter. Tar decanters are often elongated, multi-
compartment, rectangular tanks, the tar collecting on the bottom of the tank
and flushing liquor being removed at the top. If covered, the tanks have
hatches to allow access to the inside of the decanter. The temperature of the
flushing liquor in the decanters is around 80°C. Another 20 percent of the
tar is removed in the primary cooler which may be either a direct (spray type)
or an indirect (shell and tube) facility. Both systems require cooling towers
to cool the flushing liquor (direct) or cooling water (indirect).
In the tar decanting process vapor emissions occur because the flushing
liquor is highly contaminated with organics, ammonia, hydrogen sulfide and
other volatile compounds, the liquor is hot, and the decanters are vented.
The available data indicate an emission rate of about 2.2 son/metric ton coke
produced, the vapor containing significant amounts of benzene and sulfur
compounds. The rate of benzene emission from the tar decanter was 15.6 g/Mg
coke production.
Flushing liquor holding tanks and primary cooler condensate tanks are
similar in function. Both are potential fugitive vapor emission sources. The
available data for the source indicates a total organic vapor emission rate of
1.7 scm/metric ton coke for all the tanks in this secondary decanting service.
The rate of benzene emissions from these tanks was estimated as 9 g/Mg coke
production. The emission rate is probably more dependent on the number of
vents and tanks than on the volume of coke produced.
Tar dewatering is the process of reducing the water content of the tar
below that which can be achieved in the decanters. Two types of dewatering
equipment could be used: (1) mechanical, such as centrifuges, or (2) heating
to elevated temperatures to drive off the water. The use of dewatering equip-
ment depends on the requirements of the tar end-use.
In many existing plants, the coal tar is not refined on site but is sold
to tar refiners. It is a common specification that this tar sold should
contain no more than 2 percent water, but the tar underflow from the tar
decanter generally contains much more than this. Accordingly, the crude coal
tar is dewatered, ordinarily by heating it to reduce its viscosity and pro-
viding residence time for water droplets to coalesce and rise to the surface
of the denser tar. If, as is commonly the case, the temperature is maintained
above 90°C, then the combined vapor pressures of hydrocarbons over the tar
phase and water over the aqueous phase can exceed one atmosphere. The result
is a plume of steam and hydrocarbons issuing from the vent. There are no data
available for dewatering by heating the tar. If the tanks are vented to the
atmosphere, however, the emission would be significant due to the elevated
temperatures and the presence of some relatively light organic materials, such
as benzene in the tar.
Tar is commonly stored in heated tanks in order to facilitate handling.
General industry practice has not been determined, but the available data
indicate a storage temperature of about 80°C. In cases where tar storage is
-------
used to dewater the tar, an aqueous liquor will be present on top of the tar.
The emission rate would be at least 0.14 scm/metric ton of coke based on
working losses from the tanks.
Tar refining as a whole has not been tested as an emission source. Data
are available on one potential emission source, tar distillation product
storage, in particular "chemical oil" storage. Chemical oil is the light ends
of the tar recovered after single stage flash distillation to separate the
pitch. The organic vapor emission was fugitive from a tank vent, and was
estimated at 0.024 scm/metric ton coke, again based only on working losses,
with no data available on total emissions. For the remainder of the tar
refining operation, most emissions could be expected to be fugitive vapors
from tank vents, with potential emissions from vacuum distillation if the
vacuum is drawn with steam jets.
The emissions from tar processing are essentially all fugitive vapor
emissions from vented tanks. The primary sources are tar decanters, flushing
liquor holding tanks, indirect primary cooler condensate holding tanks, tar
dewatering, tar storage, tar refining, and tar distillation products storage.
FINAL COOLER
The emission sources identified for the contact, recirculating water type
final cooler are those associated with the naphthalene separation from the
water and emissions from the cooling tower. Naphthalene handling by melting/
drying in vented tanks was another significant emission source.
After separation of the naphthalene, the water is commonly cooled in an
atmospheric cooling tower and then recirculated to the final cooler. The
basic function of the final cooler is to cool the coke oven gas from around
60°C to about 25°C in order to improve light oil absorption in the light oil
scrubber.
Three forms of final cooler and naphthalene recovery technology are in
use:
1. Cooling with water and naphthalene recovery by physical
separation; or
2. Cooling with water and naphthalene recovery into tar in a tar
bottom final cooler; or
3. Cooling with a wash oil which also absorbs napthalene.
About 25 percent of the plants utilize direct water cooling and physical
naphthalene recovery, 60 percent utilize tar bottom final coolers, a couple of
plants utilize wash oil cooling, and technology at the other plants is not
known.
The choice of final cooler type will have a significant impact on the
distribution of some pollutants in a by-product plant, cyanide being a good
example. If the cyanide is not removed in the final cooler water, it remains
in the gas and causes problems downstream. If it is not stripped out of
recirculating cooling water, the blowdown will be high in cyanide and the
wastewater plant will be more heavily loaded.
80
-------
Naphthalene condenses in the final cooler water and is collected as a
dirty brown slurry. Naphthalene will separate by gravity in a sump, or the
separation may be enhanced with a froth flotation separator or similar equip-
ment. The naphthalene may then be skimmed from the surface of the water.
The naphthalene crystals are wet with a ilm of mixed hydrocarbons, often
of a brownish color, suggesting that some tar fog bleeds through the electro-
static precipitation and the ammonia saturator. This liquid hydrocarbon
medium, of unpredictable amount, is a solvent for benzene. At these conditions
(say 45°C, 5 psig) a liquid hydrocarbon would contain about three moles of
benzene per 100 moles of liquid, perhaps 6 percent by weight, much of which
would be evaporated during naphthalene handling and processing. (The naphtha-
lene is, for example, conveyed some distance in open troughs, heated and
"melted"--more precisely, dissolved in the accompanying hydrocarbon—to dis-
engage water, then stored hot for convenience in handling.)
Since crude naphthalene has little market value, it is small wonder that
more than half of all plants obviate the nuisance by some variant of the
tar-bottom cooler. About one-fourth of the plants, however, handle naphthalene
in some manner. Since there is more naphthalene in the tar than is recovered
at the final cooler, and some of this naphthalene can be recovered in the
course of tar refining, the tar-bottom final cooler does not rule out the pro-
duction and sale of naphthalene.
The plant sampled began the separation with a froth flotation operation.
No vent stream was at a rate sufficient to be measured, although there were
visible wisps of vapor. The vapor directly above this liquid surface was
sampled. The aromatic hydrocarbons are again of potential environmental
concern. The naphthalene slurry which floated to the top of the water was
sKimmed and collected in open sumps, and the water was passed through a series
of small basins to allow additional naphthalene separation.
Naphthalene collected as described above is impure and in roughly a 60
percent water slurry. Ths naphthalene slurry was pumped into a horizontal
cylindrical tank. Once the tank was full, the water was decanted. Steam
coils within the vessel were then utilized to dry and melt the naphthalene.
This operation continued for one to two days. The vent rate was estimated to
be 2.9 son vapor/Mg coke by measuring the rate of air entering the vessel due
to the chimney effect. The temperature in the tank was 101°C. Naphthalene
was sampled at a concentration of 533 g/scm, which amounts to 1.56 kg
naphthalene per Mg coke.
The final cooler unit, regardless if the tower is a direct-water once
through or recycle, direct water-tar bottom, or wash oil operation, should not
generate air emissions because it is designed as a closed system. Air
emissions may be emitted, however, from the induced-draft cooling tower used
in conjunction with the direct water and tar bottom final coolers.
The final cooler cooling tower has for some time been recognized as a
potential source of cyanide emissions, and was sampled both for cyanide and
organics. The level of cyanide in the water depends on the degree of cyanide
stripping which is accomplished in the ammonia stills, along with final cooler
operations and coal composition. At the site sampled, hydrogen cyanide was
present in the gas leaving the cooling tower at an average concentration of
76.5 ppm, which corresponds to a mass emission of 0.28 kg/Mg coke (0.56 Ib/ton)
81
-------
This source accounts for more than half of the cyanide generated. The gas
flow rate was estimated by assuming that the gas mass flow was equal to the
known liquid circulation rate. The benzene emissions from the cooling tower
were estimated as 51.6 g/Mg coke production from the estimated gas flow rate
and the measured benzene concentration. The extent to which these components
are dissolved in the water and then stripped into the air is dependent on the
operation of the final cooler and cooling tower.
The second common way of handling the final cooler water is to pass the
water through tar in the bottom of the final cooler and allow the naphthalene
to dissolve in the tar. The naphthalene is then included with the tar in any
additional refining operations. The efficiency with which naphthalene is
removed by the tar was not available in the literature although it is apparently
fairly high. The final cooler water is cooled in a cooling tower and recircu-
lated to the top of the tower. Again, air stripping of light components in
the water occurs to some extent in the cooling tower.
LIGHT OIL RECOVERY
Light oil is a clear yellow-brown oil, with a specific gravity of about
0.86. It is the coke oven gas fraction in which the more than 100 constituents
with boiling points between 0°C and 200°C or so reside. Benzene is generally
60 to 85 percent of light oil, with toluene (6 to 17 percent), xylene (1 to 7
percent), and solvent naphtha (0.5 to 3 percent) being the more important of
the lesser constituents. With few exceptions, light oil is recovered from
coke oven gas by wash oil absorption in the United States. Light oil refining
capability is present at about 35 percent of the plants. The products of the
refining operations are mostly benzene, toluene, xylene, and solvent naphtha.
The emissions identified with light oil recovery include several decanted
water streams, fugitive tank emissons, and a vent from the light oil condenser.
Several wastewater streams are decanted in the light oil plant. The
primary source of the water is the live steam used to strip light oil from
wash oil, and water must be separated from all the hydrocarbon liquids condensed
from the still vapor as well as from wash oil. The waste could be avoided by
using reboilers for non-contact heating with steam. They are commonly collected
in the "intercepting sump" and treated in the combined wastewater treatment
plant. The rate has been estimated at between 100 and 500 1/Mg coke depending
on the ability of the operator to tightly recycle the water.
Fugitive emissions occur from wash oil storage, wash oil decanters, and
light oil storage. Only the light oil storage tank was sampled, as it was
amenable to data reduction by the tank working loss equation. The rate of
benzene emissions was estimated as 3 g/Mg coke production. No emissions with
measurable rates were present.
The noncondensibles vent off the light oil condenser was not accessible
under Level 1 constraints and v s not sampled. No data are available in the
literature. This stream probab_y consists of the fraction of the coke oven
gas which dissolved in the wash oil, as well as light oil vapor. This stream
is thought to be quite small, appropriate for the 2-inch pipe used to vent the
condenser.
82
-------
MISCELLANEOUS EMISSION SOURCES
Leaking pumps, valves, and flanges are sources of benzene emissions in a
by-product recovery plant when they are handling benzene containing materials.
Some of the benzene containing liquids that are pumped include tar, naphtha-
lene, crude light oil, and refined benzene. No data is available that speci-
fically quantifies benzene emissions from the pumps, valves, and flanges of a
by-product recovery plant. However, data is available on vapor emission ,
factors from refinery processes that can be used for comparing the sources.
Leaking pumps are the most significant sources, and the emissions from
flanges are several orders of magnitude lower than from pumps. Emissions from
valves handling light liquids are one order of magnitude lower than leaking
pumps on a per source basis.
Wastewater handling and treatment is a necessary part of the coking
operations, as raw coke oven gas contains water vapor driven from the coal in
the coke oven. This water vapor is due to both surface moisture on the coal
and bound water. Depending on coal type and coking practice, the flow of
wastewater originating in the coke is around 100 to 200 1/Mg coke.
Wastewaters from other sources within the by-product plant are often
combined with the waste ammonia liquor for treatment. Some of them are
unavoidable; others can be either greatly reduced or eliminated by proper
choice of process technique. The major secondary sources of wastewater are:
1. Barometric condenser water from steam jets used to draw vacuum
on the ammonia crystallizer;
2. Steam stripping waste from wash oil and light oil decanters;
3. Slowdown from the final cooler.
Wastewaters are often carried in open sewers and may be held in open holding
tanks, aeration basins, and biological treatment plants. Various organics
that are partially dissolved or floating on the water may evaporate before the
final effluent treatment. The wastewater may contain phenol, ammonia, cyanide,
sulfur compounds, and hydrocarbons that can be emitted as fugitive air pollutants.
No data are available on these emissions.
The transfer of crude light, oil or refined benzene is a fugitive source
of benzene emissions. These emissions are due to the displacement of air con-
taining benzene by the liquid as the tank is filled. The problem of
hydrocarbon vapor loss has been dealt with extensively by the petroleum
refinery industry.
Filter cakes, tar and oil sludges, and slurries result from water treat-
ment, sedimentation in tanks, and cleaning of pipelines and process vessels.
These wastes are a potential source of fugitive emissions due to the evaporative
loss of volatile organics they may contain. Handling procedures for these
wastes include transport in open containers and trucks with subsequent disposal
in pits or landfills.
83
-------
MULTIMEDIA EFFLUENTS
In addition to air emissions, water and solid waste emissions also occur
in by-product plants. Estimated emission factors for these pollutants, derived
from sample data from one plant, are given in Table 1. Nine sources were
investigated, seven by sampling. For the sources investigated, the daily
total emissions were estimated from processing 1.8 million cubic meters per
day of coke oven gas. The uncertainty in these values is anticipated to be a
factor of 2-3.
TABLE 1. MULTIMEDIA ENVIRONMENTAL EMISSIONS
Light aromatics (mostly benzene), 372 kg/day
Polynuclear and high boiling aromatics (PNA) 57.6 kg/day
Sulfur compounds 108 kg/day
Cyanides 1,162 kg/day
Ammonia 334 kg/day
Phenols 5.6 kg/day
Light oils 40.1 kg/day
Light aromatics, the predominant emissions, were found in the highest
concentration in emissions from the tar decanter, the primary cooler conden-
sate tank, the naphthalene separator, the light oil storage tanks, and the
distillation product storage tanks. PNA's (as total non-evaporables) concen-
trations were highest at the following sources: wastewater treatment sludge
tar decanter, tar dewatering and storage, tar distillation products, naphtha-
lene separator, final cooler cooling tower, and water from the biological
treatment plant. Cyanide concentrations were highest at the final cooler
cooling tower and in the aqueous effluent from the biological treatment plant.
Sulfur compound concentrations were highest at the tar decanter, the primary
cooler condensate tank, the naphthalene separator, the light oil storage
tanks, and in the plant wastewater effluent.
The data suggest that the PNA's accumulate as a concentrate in the liquid
streams (tars, flushing liquor, tar products, wash and wastewaters). PNA's
accumulated in the water from the final cooler reentered the air as recycled
water passed through the open cooling tower.
Ambient air samples, taken upwind and downwind of the by-product plant,
showed increases in both benzene and cyanide concentrations. The following
results were obtained:
TABLE 2. AMBIENT AIR CONCENTRATIONS AT A BY-PRODUCT PLANT
Hydrogen Cyanide Benzene
(volume ppm) (volume ppm)
Downwind 0.062 0.8
Upwind 0.006 0.6
Gain 0.056 0.2
Toxic units/son toxic unit/scm toxic units/son
Downwind 0.0062 0.9
Upwind 0.0006 0.7
84
-------
These results indicate that cyanide concentrations downwind of the
by-product plant were well below levels for environmental concern. Cyanides
of this plant were more a problem in wastewaters than in the air. Downwind
benzene concentrations, on the other hand, were of potential environmental
concern.
Some of the remaining unstudied sources include the wash oil storage,
decanters, and condenser vents of the light oil recovery operations, the
decanted water from the naphthalene dryer, the wastewaters from dephenoli-
zation, the tar-topping barometric condenser, and the cyanide handling pro-
cesses, some of which are an inherent part of desulfurization operations. The
wastewater streams involved in these operations were sent to a combined waste-
water treatment plant at the study site.
The relative overall environmental impact of some of the pollutant
sources within the by-product coke plant is addressed in Table 3. The
biological treatment plant effluent is the most significant of the by-product
plant sources. This was due to a combination of a large effluent rate and the
sensitivity of the impact measurement to organic pollutant concentrations.
The other major sources are the cooling tower for the contact final cooler and
the biological treatment plant sludge.
TABLE 3. RELATIVE HAZARD OF BY-PRODUCT COKE PLANT POLLUTANT SOURCES
Normalized Relative Hazard
Tar Decanter Vapor 0.036
Tar Dewatering/Storage Vapor 0
Primary Cooler Condensate 0.017
Tank Vapor
Distillation Product Storage 0.001
Cooling Tower for Contact 0.349
Final Cooler
Light Oil Storage Vapor 0.028
Biotreatment Plant Effluent 0.434
Biotreatment Plant Sludge 0.135
The procedure used to arrive at Table 3 uses a weighting process which
considers pollutant concentration, hazard in the proper media, and emission
rate. The ratios of the amount of pollutant emitted to the MATE values for
each category were summed. The normalized relative hazard is the ratio of the
summation for a source to the total for all of the sources. The MATE weighting
factors were obtained from the Multimedia Environmental Goals. For the
by-product plant, the weighting factors reflecting the great hazard of certain
PNA's essentially controlled the results- Weighting factors were obtained
from the Multimedia Environmental Goals.
85
-------
This study is a limited-scope first look at the by-product plant from the
environmental point of view- As such, it points to a need for control of
light aromatics and PNA's. Control may be most likely achieved through
techniques that essentially eliminate the sources: venting tanks back to the
gas mains; blanketing with coke oven gas.
EMISSIONS CONTROL TECHNIQUES
A large proportion of the emissions from a by-product plant originate in
the various vents in the plant. Recovery of vapor from these sources will ,.
generally be complicated by the presence of naphthalene. Wilputte Corporation
has installed water sprays on some tar decanters, and the techniques might be
extended to other vents. Vapor recovery from these sources to the suction
side of the exhausters, probably ahead of the primary coolers, might be possible,
Venting to the gas main upstream of the askania valve is another alternative.
The blanketing of holding tanks with coke oven gas which was originally used
in the light oil recovery process to exclude air and prevent the buildup of
sludges, eliminates the tank vents as an emission source. The blanketing gas
is vented back into the main gas stream. This technique could perhaps be
applied to many sources, even to refined benzene tanks, if the gases were
first desulfurized to prevent deteriorization of the product. Problems to be
addressed in considering the broader use of blanketing include making provision
to admit the flammable gas into the various operating areas, and to prevent
the condensation of naphthalene. Naphthalene condensation would require that
the vents be heated, and the corrosive nature of the vapor (perhaps including
chlorides) would cause materials problems. The system might be designed to
float on coke oven gas at slightly above atmospheric pressure. This system
when compared to internal floating covers offers an attractive alternative for
small volume tanks, less than half a million gallons, and plants with multiple
storage tanks which may be yoked together in a system.
The method for controlling emissions from the tar decanter consists of
providing a water seal and a vent to the gas main at low pressure. The water
seal is a heavy plate structure which is suspended from the roof of the tar
decanter near the sludge discharge end. The major portion of the liquid
surface is vented to the gas main near the coke oven battery. The tar decanter
can be vented to the coke oven gas main to operate at a pressure of
approximately 4 to 6 mm of water. The installation of the water-seal plate
will be a comparatively difficult job. For retrofit conditions it will be
necessary to take the tar decanter out of service, clean it thoroughly, and
then make the installation after inspecting and repairing any hole that would
allow the infiltration of air.
Naphthalene collection in open vessels inherently causes naphthalene
emissions. Avoidance of exposed naphthalene by the use of a tar bottom final
cooler and keeping the naphthalene in the tar are proven and should be pre-
ferable. A wash oil final cooler also collects naphthalene, but the naphtha-
lene must eventually be removed from the wash oil. The final cooler cooling
tower with a tar bottom final c Dler would still have about the same level of
cyanide emissions, although hydiocarbons emissions might be down. A wash oil
final cooler should avoid the cyanide emissions, although the cyanide must go
somewhere.
86
-------
Some of the emission control systems for sources such as light oil storage
are relatively inexpensive since the credits from the recovered benzene could
pay for the system in a relatively short time. Other process modifications
such as conversion to a wash oil final cooler would be more expensive, but
could substantially reduce the overall by-product plant emissions.
87
-------
REFERENCES
1. VanOsdell, D. W., et. al. "Environmental Assessment of Coke By-Product
Recovery Plants." EPA-600/2-79-016, January 1979.
2. Hamersma, J. W. , et al. IERL-RTP Procedures Manual: Level I Environmental
Assessment. EPA-600/2-76-l60a, NTIS PB 257 850.
3. Cleland, J. G., and G. L. Kingsbury. Multimedia Environmental Goals for
Environmental Assessment. EPA-600/7-77-136a and EPA-600/7-77-136b, November 1977.
4. Wetherold, R. and L. Provost. Emission Factors and Frequency of Leak
Occurrence for Fittings in Refinery Process Units. EPA-600/2-79-044, February
1979.
5. Private communication with John Crosby, Wilputte Corporation.
88
-------
COKE-OVEN DOOR SEAL DEMONSTRATION
by
Albert 0. Hoffman and Ralph Paul
Battelle-Columbus Laboratories
Coke making is a batch process for producing blast furnace and
cupola coke through the procedure of carbonizing coals in narrow, vertical
ovens. Each oven has a full-height door on either end. These doors
are removed (for product discharging) and replaced each coking cycle
(14 to 30 hours). There are about 25,000 coke-oven doors in operation
in North America, ranging in height from 10 feet to over 20 feet. With an
assumed average coking cycle of 20 hours, about 1,250 doors are handled
each hour of the year.
Figure 1 is a photograph of one side of a coke battery showing:
a) the narrow, vertical doors between the vertical supports of a battery, and
b) the emissions that leak past the existing seals on these doors. The control
of emissions leaking past door seals represents a difficult problem to
many different groups, including the operators of coke batteries.
Vertical coke~oven doors
FIGURE 1. PHOTOGRAPH OF ONE SIDE OF A COKE BATTERY
89
-------
The goal of this paper is to present the important elements of
a research/development/demonstration funded by the EPA and the American
Iron and Steel Institute (AISI), aimed at developing an improved and
retrofittable system for sealing coke-oven doors. The Battelle-Columbus
Laboratories were awarded the initial research contract in competitive
response to the original Request for Proposal by the EPA. The contract
was followed by funding for a development and a demonstration project.
Overall Objective of the Coke-Oven Door Sealing Program
As stated by the Sponsors, "the specific objective will be
to innovate and to develop at least one system that will be proven
in the field to be retrofittable to existing coke ovens, mechanically and
physically suitable for commercial use in steel plants, and highly
effective in containing and controlling emissions from ends of ovens."
We were given to understand that the Sponsors wanted an
analytical, technical approach working within a developed set of
criteria, leading to firm recommendations. There was to be no empirical
testing of new seal ideas at coke batteries prior to making our
recommendation.
Project.Organization Within the Program
The overall program consists of three projects:
1. Study of Concepts for Minimizing Coke-Oven Door Emissions ^
2. Development of Concepts for Minimizing Coke-Oven Door Emissions^
3. Demonstration of the Recommended Retrofittable Coke-Oven Door Seal
90
-------
In the first project, 45 different concepts for new coke-oven
door seals were evaluated. These concepts were obtained from every
possible source, including surveys of sealing methods used in industries
other than the steel industry. The major conclusion of this project
was that upgraded metal seals have the best potential for being developed
into more effective seals. It was recommended that these upgraded
seals should have increased flexibility and improved resistance to
heat distortion.
The development project has been completed and has resulted in
a recommendation of seal design and material. Details of the development
(3)
project have been described in reports and a paper and will not be
covered here. Rather, a discussion of the recommended seal design
and material, and the first results of the demonstration project will
be presented.
Definition of Terms
Retrofit is usually defined as furnishing new parts or equipment not
available at the time of manufacture. This definition fits the program
but was expanded to include the criterion that the retrofit be accomplished
with a minimum of equipment modification or replacement of parts.
Effective refers to the elimination or virtual elimination of seal
emissions. In addition, "effective" includes:
a) a long successful life and lowest possible "life cost",and
b) adaptability to all or almost all batteries, including those in which
the doors and jambs have been warped in service.
Considered a bonus would be an effective operation in which
91
-------
the effectiveness depends more on the characteristics of the design rather
than on the skill and conscientiousness of the workers in making seal
adjustments.
Brief Background Description of Existing Seal Designs
Existing seal designs,, in general, can be classified into
two groups. They are either a spring design or of the stiff "fixed-
edge" designs.
In the United States about 75 percent of the operating coke
ovens that need retrofit door seals have the S-shaped spring design shown
in a cross section view in Figure 2. This is a venerable design that
we have seen operate completely emission-free when new and when given what
appeared to be special adjustment attention. This design is about 30 years
old.
Jamb
Plunger and
Spring ,,:••
FIGURE 2. Side View of S-Shaped Seal System
92
-------
The fixed-edge designs are used in about 25 percent of the
operating batteries. Views of these designs are given in Figures 3 and
4. Such designs are inordinately stiff in the vertical direction and
we have never seen emission-free performance from seals with these
designs, regardless of the attention given to adjustments of the bolts.
It is a fact that a knowledgeable worker can minimize leaks with this
design if given enough time between movements of machinery; but people
willing to develop and exercise such skills are now becoming an
endangered species.
FIGURE 3. Side-View of Adjustment Equipment on Fixed-Edge Seals
93
-------
FIGURE 4. Cross-Section of a Typical "Knock-Type" Seal
It is worthy to note that all existing designs produced by
engineering companies and others require periodic adjustments along
the edge of the seal. A coke-oven door has one shape or profile when it
is cold. However, on being heated to operating temperature there is a
slight but significant change in profile. In fact, there is a change
in the profile of the door and the attached seal with every change in the
temperature differential across the door, e.g., during a rain storm. In
this respect, a spring seal pr^ vides a resilient element that can better
accommodate itself to temperature-induced changes in the distance between
the jamb and the doorback. In our development work we concentrated
on developing a spring seal with increased flexibility and decreased
94
-------
susceptibility to heat distortion.
The concept of having a multitude of adjustment bolts and
springs along the 100-plus miles of coke-oven seal-edge length in day-
to-day operation has advantages and disadvantages. Some of these will
be discussed in the presentation of the recommended seal. There is,
however, an apparent and an accepted logistics problem in finding the
time between machinery moves to make these adjustments. In many plants
it is not a question of having available seal-adjustment labor or skills,
or possessing appropriate portable elevator equipment to move men
into location; rather, it is a question of the limited amount of access
time during daylight hours. Based on our observations, one of the
criterion of the development project was to incorporate automatic gap-
closure capability in the design, i.e., to develop a method of eliminating
or minimizing the need for manual adjustments to achieve sealing effectiveness.
Recommended, Retrofittable Seal
Within the design requirements (and restraints), the seal
design that evolved in the development project is shown in Figure 5. A
list of descriptive elements and expected advantages is given as follows:
95
-------
Mounting Angle /- Scaling Ring
FIGURE 5. Side View of Recommended EPA/AISI
Retrofittable Seal
1. The entire seal consists of the carbon-steel mounting angle and the
simple, two-bend main seal with one back-up or reinforcing leaf. The
back-up leaf can be used to absorb some of the high levels of
latching force now being used in the coke industry. All seals are
deflected outward when doors are latched in place with force
levels of about 20 tons. There are indications in the demonstration
project that, as expected, the back-up leaf may not be required.
96
-------
2. The recommended alloys for the main seal are high-performance, high-
temperature alloys known to have a long operating life in heated
spring applications. Taking into consideration fabricability,
thermal expansion coefficients, and other factors, our recommended
first choice in the demonstration project was a nickel alloy (Inconel
X-750). Subsequent evaluations should be completed using lower-
cost alloys. Inconel X-750 can be heat-treated at 1400 F to give
yield strengths of up to 8 times higher than those for alloys
presently used in seal fabrication. This increased yield strength
and the alloy's high resistance to creep at 800 F and lower make it
possible to safely increase the seal deflection about three-fold
over that of existing spring seals without distortion.
3. The height of the seal lip that contacts the jamb has been lowered
to 3/8-inch. This increases the seal flexibility along the height
of the jamb by a factor of 10 over the flexibility of existing
seals. As proven in laboratory testing,the combination of increased
seal deflection during door latching, along with increased seal
flexibility along the length of the jamb will result in automatic
closing of minor gaps caused by "unevenness" in the jamb contour, and
it will also "iron out" minor irregularities in the as-manufactured
straightness of the seal edge. For various reasons some operators
of coke batteries would like the seal lip to be wider than 3/8-inch.
The performance of wider-lip widths is expected to be evaluated once
the recommended seal design has become accepted.
97
-------
4. Part of the plan to use high-strength alloy material in this design
was to eliminate the need for adjustment bolts and adjustable
plunger-spring seal supports on the recommended design. The seal
alone should conform to the requirements without any assistance
or adjustments. This approach will also eliminate the distortions
seen on standard spring seals caused by the point loading on the
back of these seals.
The use of high-performance alloys will result in a more
expensive seal because, of the higher price per pound for alloy, plus
the need for heat-treatment following fabrication. To counter this
unfavorable situation, the seal weight was lowered and the design was
adjusted for high-volume fabrication methods. A single-leaf seal for
a 6-meter-tall oven weighs only 84 pounds but it is more resistant
to physical damage than the existing spring seals. In high volume
production of these seals, the forming can be done by low cost hydroforming
of the corners and roll forming of the sides of the seals. With roll
forming it is expected that all machining of the seal edges may be
eliminated,
Retrofit Conditions and Procedures
As previously stated, the recommended seal design is to be
more flexible, it is not to have adjustment bolts or spring plungers on
98
-------
the end of the seals, and yet it is to be retrofittable to all end
closures including those that exhibit extreme jamb and door warpage
conditions. In many circumstances the recommended seal is expected
to be bolted into place without any special considerations, and the
inherent flexibility of the seal must accommodate any minor variations
in the profile relationships between the jamb and door. This remains
to be established. However, one interest in the EPA/AISI seal shown by
the coke plant operators is its potential for effectiveness where the
jamb and door profiles are very incongruent. If this potential can
be realized, it would not be necessary to replace jambs. For incongruent
profiles, a retrofit procedure had to be developed.
As background to this effort, it should be appreciated that all
door frames or doorbacks have bowed profiles attributable to temperature
differentials across the doors. All jamb and door profile combinations
are incongruent to some degree; otherwise, there would be no need
for adjustment bolts and springs in existing seals. Figure 6 presents an
exaggerated example of variations that exist for relative contours of jambs and
doors. This figure indicates the required buildup under the seal (shaded
areas)to achieve congruency of seal edge with jamb. It is this buildup
approach under the seal holder that became the recommended retrofit
procedure.
99
-------
lamb
Seal
Example A. Horizontal distances
between jamb and door
ace all the same.
Iamb
Seal
Door
Away irom Oven
|amb
Door
Seal
Example B. Door has left curvature
than {amb. Iamb corner! are
clour lo door than at Hi*
middle ol iamb/door.
Iamb
Seal
Door
FIGURE 6.
Example C. Door is curved inward and Example D. Jamb and door bow towards
k lacing either a stiaighl each other.
jamb or distance between
)amb and door is greater
al the corners.
Exaggerated Examples of Variations That Exist in Terms
Of The Relative Contours of Jambs and Door Frames
At the first demonstration site, a new type of incongruency
was encountered in which the jamb and the door are bowed outward in the
center, away from the coke oven. The measurement data that were obtained
are plotted in Figure 7 by using different scales for the presentation of
the horizontal and vertical measurements. . In the retrofit procedure
used for this incongruent situation, only the variation in the horizontal
distance between the jamb and the inboard side of the doorback is of
100
-------
FIGURE 7. JAMB AND DOOR PROFILES.
OVEN 973, PUSHER SIDE
101
-------
tnterest because this is the space that the seal must "fit" and be
effective. These are the A to B distances along the length of the
doors shown in Figure 7.
At first, considerable effort was expended in developing
measuring methods to obtain the A to B measurements. However, in this
instance these measurements were taken by merely (a) turning in the
adjusting bolts to positively set the screws on the seal while the door
was on an oven,(Figure 4), and (b) measuring the height of the adjusting screws
above the doorback in the repair station. Any minor variations in
measurement accuracy are countered by the high flexibility of the seal.
The recommended retrofit procedure to remedy the incongruent
profile situation is to substitute a one-time measurement and seal-
setting procedure for the periodic adjusting-bolt manipulation.
As an example, for the jamb and door profiles shown in Figure 7,
the height of the seal edge at the latch levels must be close to 2.3 inches
in order to permit door mounting. From these reference points and
the A to B measurements obtained, it is not difficult to determine the
amount of seal-height adjustment along the height or length of the door
to obtain seal edge-to-jamb congruency. These determinations for one
case are plotted in Figure 8. Note that the range in height above
the doorback is about 0.4 inch. This may be well within the flexibility
range of the seal to maintain contact with the jamb without making any
effort to set the seal height. However, for the early stages of the
demonstration project we elei :ed to set the seal height. The
recommended method for setting seal height is to use metal spacers
under the seal holder.
102
-------
29
FIGURE 8. REQUIRED HEIGHT OF THE SEAL EDGE ABOVE THE DOORBACK
TO GET THE SEAL CONGRUENT WITH THE JAMB WHEN AT
OPERATING TEMPERATURES. OVEN 973
103
-------
Taking measurements at operating doors and then setting a
variable seal height to obtain near-perfect congruency is not desirable
from the viewpoint of some coke plant operators. One expressed criticism
is that this procedure will prevent the movement or shifting of
doors between ovens inasmuch as the seal on one door is custom fitted
to one jamb. However, the practice of keeping ovens and doors together
is already in operation at several plants. In the future it is
expected that there will be workable compromises. For example, it is
anticipated that the majority of doors and jambs on a battery will have
similar congruency patterns. It may,- therefore, be possible to shift
doors within this regime without problems. Also, any problem doors
can be removed for straightening or even bending to shape. This approach is
is much preferred to replacing jambs. There are other alternatives
that can be considered depending on the overall operating results of
the retrofit seals.
The Demonst r a tion Project
The first problem that arose while planning arid scheduling
the demonstration project was that the established fabricators to the
steel industry were overloaded with work, and for a long time (over
6 months) no fabricator showed any interest in producing prototype
seals of an unfamiliar alloy. Also, fabricators having experience
with high-strength alloys did not want the entire contract consisting
of purchasing, tool-making, forming, and heat-treating. To encourage
bidding and to convince the fabricators of the importance of tin's
project, the AISI Project Officer obta ined commitments from .six steel
104
-------
companies to (a) purchase 32 seals for demonstration at 8 batteries,
and (b) divide the cost of the fabrication tooling required. This
approach, coupled with personal visits, did not result in any response.
This log jamb was broken when Bethlehem Steel Corporation (a) ordered
the sheet alloy, (b) requested Bethlehem Engineering to adapt the
recommended seal to a fixed-edge battery, (c) ordered the fabrication
tooling required, and (d) placed a priority order for the fabrication
of four seals for the 6-meter battery (fixed-edge seals) in Lackawanna
(Buffalo) New York. As per agreement with enforcement agencies, a
schedule was set for engineering, fabrication , and demonsUr.ition.
The adaptation of the recommended seal to a battery having
fixed-edge seals is a complex mutter because these batteries exhibit a
design distance of only 2.3r inches betwet'n the jambs and the doorbacks,-
as compared to a distance of ').(> inches in the majority of designs-
having a spring seal. Bethlehem Engineerhig followed the conceptual
design outlined in our reports .-ind Figure ') shows the resulting
seal attachment method.
There were some potential problems in this compromise design.
In adjusting the height of the seal it was necessary Lo release and
move the seal only, rather than shift the seal and its holder as a
unit. This design was accepted by Bethlehem/Mattel l,e based on faith in
the increased flexibility of the recommended se;i 1 .
Prototype testing is rather straightforward in description, i.e.
put the prototype into its intended service and evaluate the results
relevant to the expectations. After delays by the fabricator, the
four 21-foot-l.ong seals bec.-ime available for mounting on doors. The
first two seals were equipped with back-up leave..1? and the seal heights
105
-------
FORCE
X Ig
TO HANGER. BAR-
FIGURE 9. First Seal Attachment Method,
(Lackawanna Ovens) a Fixed-
Seal Type of Door.
106
-------
were set on the doors using the method previously described. These
demonstration seal doors were placed on the selected oven the next time
this oven was discharged in daylight hours. In a few minutes we were
aware that our joint efforts with Bethlehem Steel were being only
partially rewarded. In both seals, the 84 total feet of vertical seals
were emission-free, but the 8 feet of seal across the ends of the doors
(top and bottom) released emissions. The conformity of the seal to the
vertical portions of the warped jamb was as expected and this conformity
was maintained as the door profile changed in response to heating. The
emission control at the top and bottom ends of the seals was, however,
ineffective.
Because of pressure to get the four demonstration seals into
operation, a third seal was installed, but in this instance the back-up
leaves were omitted to increase the flexibility of the seal ends. There
was some measureable improvement in gap-closing ability on the seal
ends but in general the emission pattern was the same, i.e., rated
ineffective at top and bottom.
In the development effort, special technical attention was
directed toward the design of the corners of the seal. 'The study of
the problem centered on improving the geometry of the existing S-shaped
seal corners. The development effort included measurement of the
deflection and gap-closing capability of a full-scale stainless model
of the recommended seal-end (two corners attached to a cross piece).
These development efforts increased the measured deflection capability of
the corners. For example, the corners were given an inside radius to
minimize stress concentration in the corners, and welds were displaced
from the corners as shown in Figure 10.
107
-------
I
Welds Are Disploced •
From Corner
Mounting Surfoee
FIGURE 10. Completed Corner of AISI/EPA Seal
With the localized seal leakage problem that developed, an investigation
was started working with tools built for measuring the unevenness of the
seal edges on the ends. Also evaluated was the gap-closing capability
of the demonstration seals. The results obtained are shown in Figure 11.
When the seals were originally fabricated, the edges Were machined
flat while being clamped to the holder. The end profiles are shown
in Figure lla. When released from the bars some unevenness developed
in the end edges, probably caused by the residual stresses introduced
during the final machining operation. (Figure lib). However, when
the bar was bolted to the door and the seal was bolted to the bar, the
unevenness of the seal in the critical end of the seal became pronounced
(figure lie). Pressing a plane surface against the end seal to deflect
the seal 0.3 inch did not completely close the gaps (Figure lid) and
108
-------
11-
12
I
13
17
I
22
U
REFERENCE
NUMBERS
II
11
I
12 13
i i
17
I
22 23 2H
i I i
11A
11B
11C
11D
AS MEASURED
ATTACHED TO
HANGER BAR
FREE
ON DOOR
. .051
0.3 DEFLECTION
(AMBIENT)
.< .005
FIGURE 11. FIELD DATA SHOWING THE UNEVENNESS OF THE SEAL EDGE ON
THE END OF THE DEMONSTRATION SEALS UNDER VARIOUS
CONDITIONS. DISTANCES SHOWN ARE GAPS REMAINING
BETWEEN THE SEAL EDGE AND A PLANE SURFACE
109
-------
this seal leaked emissions while in operation. It is apparent that the
unevenness of the seal edges as they now exist is beyond the gap-
closure capability of the seal design, and it is inferred that the seal
unevenness is related to the method of attaching the seal to the door.
All efforts to decrease the level of unevenness in the end seal edges
in the field were unsuccessful.
We have two plans in action to correct the problem. The
fourth demonstration seal will be remounted by using a new attachment method
to start with a relatively flat edge on the two ends of the seal.
Bethlehem Steel coke plant operators indicated that they could increase
the space between the jamb and the doorback by machining the latch
bars on the door. This should make it possible to (a) attach the
seal by using the original angle-iron attachment method, (b) grind or
machine the fourth seal flat, and (c) mount the (seal plus holder) on the
fourth door. This should be completed early in November, 1979,
We are also investigating the stress patterns introduced in
the seal corners during deflection under hot operating condition. This is
being done because unappreciated problems are suspected in the corners
of all existing coke-oven door seals. A design change in the corners might
increase the gap-closing capability of spring seals.
It is our judgment that the ends of the recommended seals can be
made to perform in a manner as effective as the vertical portions of the
demonstration seals. Unfortunately , this need for a problem-solving step
in the demonstration project has made it impossible to conform to
the schedule given to us.
110
-------
References
(1) "Study of Concepts for Minimizing Emissions From Coke-Oven Door Seals".
This report may be obtained from the National Technical Information
Service as Report Number PB 245580.
(2) "Development and Demonstration of Concepts for Improving Coke-
Oven Door Seals: Interim Report". This report may be obtained from
the National Technical Information Service as Report Number PB 286628.
(3) Hoffman, A. 0., Paul, Ralph L., "The Development, Description, and
Status of the Retrofittable AISI/EPA Coke-Door Seal". Presented
at the AIME Ironmaking Conference in March 1979 in Detroit.
Ill
-------
ENVIRONMENTAL ASSESSMENT OF
COKE QUENCH TOWERS
by
A.J. Buonicore, P.E.
YORK RESEARCH CORPORATION
One Research Drive
Stamford, CT 06906
presented at the
SYMPOSIUM ON IRON AND STEEL POLLUTION TECHNOLOGY
PICK-CONGRESS HOTEL
CHICAGO, ILLINOIS
OCTOBER 30, 31 and NOVEMBER 1, 1979
ABSTRACT
The body of information presented in this paper is directed to
those individuals concerned with quantifying gaseous and particulate
emissions from the coke quenching process. Both dry and wet quench
systems will be discussed.
Results of testing programs at the Lorain Plant of U.S. Steel,
DOFASCO's No. 2 Coke Plant in Hamilton, Ontario, and an Eschweiler
Bergwerks-Verein (EBV) plant in Erin, W. Germany, will be presented.
Testing methodology will be reviewed and emission results compared.
112
-------
SUMMARY
An environmental assessment of air emissions from various types
of quench towers was made based upon data collected by York
Research Corporation at
• U.S. Steel Corporation Lorain Works
• Dominion Foundries and Steel (DOFASCO) Hamilton, Ontario
Works
• Eschweiler Bergwerks - Verein (EBV) Works, Erin, West
Germany
Estimates of particulate emissions rates from the natural draft
quench towers at Lorain and DOFASCO, and the pressure quench
system at Erin are presented. Organic emission rates from the
Lorain quench tower are reviewed and the environmental impact of
dry quench towers also discussed.
The quench tower exhaust at Lorain contained 3.5 pounds of particu-
late per ton of coal when contaminated water was used for quenching,
In addition, gaseous emissions such as sulfur dioxide, ammonia,
phenol, and cyanide contributed 0.5, 0.3, 0.2, and 0.01 pounds
per ton of coal respectively. When clean (river) water was used
as makeup to the quench tower, particulate emissions were reduced
by more than one-half and gaseous emissions were reduced by one
order of magnitude.
Fifty three different organic compounds were found in the Lorain
quench tower emissions; seven of these, including benzo(a)pyrene,
were found in sufficient quantity to be considered a potential
health hazard. 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 a principle source of organic emissions, the coke also
appears to contribute to the emission of organic pollutants.
Since the majority of organics detected are either gaseous or
associated with small particles, they will contribute to ambient
air contamination beyond plant boundaries.
Particulate emission tests conducted at DOFASCO's No. 2 Coke
Plant Quench Tower using river water averaged 0.466 Ib/ton of
coal. Particle size data indicated that the majority of the
particulate emissions were less than 10 microns in diameter.
The DOFASCO particulate data could be deduced from the Lorain
data when taking differences in quench water quality and baffle
design into consideration. Particulate emission rates from the
EBV pressure quench system cyclone averaged 0.612 Ib/ton of coal
using river water.
Dry quenching has the advantage of virtual elimination of air
and water pollution problems associated with conventional wet
quenching. Dry quenching is performed in a closed-cycle system
using continuously recirculated inert gases to cool the hot coke.
113
-------
INTRODUCTION
Coke Manufacturing
The majority of coke manufacturing in the United States is
performed to supply the steel industry with blast furnace coke.
There are two generally accepted methods for manufacturing coke;
the beehive process (nonrecovery) and the by-product, or chemical
recovery process. Today, the by-product process produces about
ninety-nine (99) percent of all metallurgical coke.-
In a by-product coke plant bituminous coal is heated, out of con-
tact with air, to drive off the volatile components. The residue
remaining in the ovens is coke; the volatile components are re-
covered and processed in the by-product plant to produce tar,
light oils, and other materials of potential value, including
coke oven gas. This process is accomplished in narrow, rectangular,
silica brick ovens arranged side by side in groups called batteries.
Coal is charged through holes into the tops of the ovens from
hopper bottom cars which run on tracks over the top of the battery.
During the coking period, the gases and volatile materials distilled
from the coal pass into the collection main(s) which run the length
of the battery. At the end of the coking period, the doors are
removed from each end of an oven and a long arm pushes the incandes-
cent coke into the quench car. The car then moves to the quenching
tower where the coke is cooled by water sprays. The cooled coke
is delivered to handling equipment for subsequent use as a basic
raw material for the blast furnaces.
Coke Quenching
The common practice in the U.S. steel industry for cooling in-
candescent coke is wet quenching, an operation which produces
billowing clouds of steam, water droplets and air contaminants.
The operation involves the direct application of water sprays
in towers which are so constructed as to draft much of the re-
sulting quench plume up from the immediate work area. The water
used for quenching may be once-through river water, river water
recycled from prior use in quenching, or contaminated waste
ammonia liquor and other coke plant wastewaters.
Typically, 10 to 20 ton loads of hot (2000 + °F) coke are quenched
by 6,000 to 12,000 gallons of water. Each quench takes about
2 to 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 in the tower .bove the quench car to reduce the
amount of large diameter particulates emitted and the amount of
make-up water needed. Water not expelled from the tower is recycled
back to the sump.
114
-------
Due to the high visibility of the quench tower plume, the quenching
operation has been a subject of intense interest to those who
promulgate and enforce environmental regulations. Although the
quench tower has been historically viewed primarily as a source
of particulates, organic emissions have recently been of concern
as evidenced by the fact that quenching is being considered by
EPA as the subject of a National Emission Standard for Hazardous
Air Pollutants.
Quench Tower Emissions
Due to the short, intermittent bursts of emissions, high moisture
content, presence of droplets, flow variability, and other factors,
an accurate assessment of the magnitude and nature of air pollu-
tion from quench towers has not been well established. Many
pollution control engineers have theorized, however, that significant
amounts of coke plant water contaminants (particularly dissolved
solids) directed to quench tower sumps for disposal were trans-
formed by the quenching process into air pollutants which escaped
the baffles. Quench water contaminants could also leave the
quench tower, however, by being deposited on the cooled coke or
by being dredged after settling (or possibly precipitating) to
the bottom of the sump.
PROCESS DESCRIPTIONS ,
Lorain Works of U.S. Steel Corporation
At Lorain, the coke ovens are about 30 feet long and average 18
inches wide, being wider at the car end than at the pusher end.
The recovered coke gas is burned in flues located between the
walls of adjacent ovens; typically, coke plants use about 40
percent of the gas to heat the ovens, with the remainder being
used elsewhere in the steel mill. At this plant, there are 413
coke ovens in use in seven batteries, each battery consisting of
59 ovens. These ovens are normally operated three shifts per
day, seven days per week, and the plant converts upwards of
7,000 tons of coal per day to coke. Normal coking time is 17
to 18 hours. However, this varies with the condition of the
oven.
The quenching operation at Lorain utilizes approximately 9,000
gallons of water per quench and is described in Figure 1. Con-
taminated quench water is made up of process wastewater which
includes tar still wastewater, waste ammonia liquor from the
primary cooler, ammonia absorber and crystallizer blowdown, light
oil (benzol) plant wastewater, and gas desulfurizer and cyanide
stripper wastewater.
Quench Tower No. 1 at Lorain was equipped with a baffle arrange-
ment as shown in Figures 2, 3, and 4. The 45 baffles were con-
structed in such a way as to allow a vertical area of one inch
115
-------
between the bottom of one baffle and the top of another. This
vertical open area permitted a portion of the stack flow to pass
straight through the baffles with little opportunity for impinge-
ment on the baffle surfaces. Furthermore, several baffles had
been removed (or destroyed) and this provided additional open
area.
DOFASCO, Hamilton, Ontario
DOFASCO's Hamilton, Ontario plant is a fully integrated basic
steel producer. The specifications of the three coke plants
which supply metallurgical coke to the four blast furnaces are
given in Figure No- 5. The emission sampling program was con-
ducted at DOFASCO's No. 2 Coke Plant Quench Tower. This plant
consists of two batteries of 53 ovens each. Approximately 16
tons of coal is charged into each oven. After a coking time
of 16-17 hours, the coke is transported to a wooden rectangular
shaped quench tower. In the quench tower, approximately 5000
USG of water is sprayed directly on the 11 tons of coke, cooling
it from approximately 2000°F to 180°F. This cooling water is
supplied from a 10,000 gallon head tank through eight spray
nozzles. The various phases of the quenching cycle are indicated
in Figure 6. The water flows into a 60,000 USG sump located
adjacent to the tower. It then flows into a clean well through
a stainless steel screen. Bay water is added to the clean well
by float control to make up evaporative losses.
The quench tower is constructed of treated wood, is 56.5 ft. high
and has an exit area of 652.5 ft2. Two rows of 1" x 4" wooden
baffles are located 9' 9V below the tower outlet. Each row is
inclined 20° to the vertical with a 2" spacing between rows and
3-5/16" spacing between the baffles. Eighteen water sprays are
located above the baffles. These are turned on once a shift in
order to wash down any solids which have accumulated on the baffles,
Eschweiler Bergwerks - Verein, Erin, W. Germany
The EBV plant at Erin has five coke batteries approximately 20
years old. The ovens are 4.2 meters high, with a capacity of
15.6 metric tons of coal. Typical coking time was 16 hours and
oven temperatures were approximately 1350°C.
The EBV Pressure Quencher consists of a special quench car with a
coke container designed to receive the entire load of incandescent
coke without moving. A narrow hood is attached to the guides and
the quenching station. At the quench station a lid is clamped
on the coke car and a measured amount of city water is sprayed
on the coke through nozzles in the lid. The amount of quench
water (city water) and the flow rates are controlled. During
the YRC tests these were fixed at a predetermined optimum setting
of 570 gallons/min for 3.5 minutes (or approximately 2,000 gallons
per quench). When the water contacts the hot coke it cools the
coke by evaporation. The steam is superheated by the hot coke
and is vented into a large cyclone. A typical pressure quench
cycle is tabulated below:
116
-------
TYPICAL PRESSURE QUENCH CYCLE
Push starts
Push
Transfer quench car to
quench station
Lower lid
Lock lid (scrubber water
on)
Quench water-on (test
period)
Unlock lid
Raise lid
Move car to wharf-dump
and return next oven
Lower hood ready for
next push
Cumulative
time
sec
0"0
35
00
2
2
6
7
7
9
10
05
45
15
45
00
30
30
00
Time
Interval
min sec
00
00
35
30
40
30
30
15
30
00
30
TEST METHODOLOGY
Previous test methodologies to determine the magnitude and nature
of air pollutants from quench towers were often flawed by sampling
difficulties such as the exclusion of certain size particles or
limited in the measurement of necessary process parameters.
Attempts by state and local agencies to measure particulate
emissions utilizing standard EPA sampling equipment were also
limited in their scope and encountered a number of sampling
difficulties. (1,2) Major problems occurred because the short,
violent, rush of steam did not allow accurate reading and adjust-
ment of the sampling equipment to maintain isokinetic conditions
(the EPA method contemplates measuring velocities and making ad-
justments every 3 to 5 minutes—more than the total time of the
quench). Also, droplets in the exhaust stream plugged filters
and made determinations of the molecular weight of the gases
nearly impossible. In addition, the short duration of the quenches
made capturing the prescribed volume of gases extremely difficult.
The shape of most quench towers, and the use of internal partitions
probably compounds these problems by causing uneven flows across
the tower cross-section.
In order to minimize sampling error, a number of modifications
were made to the standard EPA particulate sampling train. A
high volume sampler was used to minimize the time required to
collect sufficient gas volume and sample weights. A cyclone was
fitted on the front of the probe to collect droplets and large
diameter solids. This addition allowed more accurate determination
117
-------
of moisture in the stack gases, prevented filter plugging, and
provided a means of particle and droplet sizing. Also, the use
of a Hastings-Raydist velocity meter with continuous recording
and continuous purge enabled more precise evaluation of isokinetic
sampling rates. Details of YRC's sampling methodology have been
discussed previously.(3-6)
RESULTS
Lorain
In order to quantify emissions from the quenching process and
to determine the nature and source of these emissions, YRC was
contracted by the EPA to evaluate the process. The site selected
for the research program was U.S. Steel Corporation's Lorain Works.
Two separate test programs were conducted on the No. 1 Quench
Tower, the first in 1976 (for particulate and general gaseous
emissions) and the -second in 1977 (for specific organic emissions).
Particulate Emissions
The particulate emissions data obtained during the 1976 tests are
summarized in Table 1.^ The plume exiting the stack at Lorain
contained an average of 3.5 pounds of particulate per ton of coal
when contaminated water was used for quenching. This was con-
siderably larger than the quench tower emission factor of 0.9
pounds of particulate per ton of coal. Particulate emissions
were drastically reduced when clean water was used as make-up
for the coke quench tower. Overall reduction was by. more than
one-half, to approximately 1.6 pounds of particulate per ton of
coal.
Size distribution information for the particulate (and droplet)
emissions was provided by the structural design of the sampling
train. Particles (or droplets) within specific size ranges tend
either to be collected by, or to settle out in, certain structural
elements of the train. The cyclone collected particles and drop-
lets predominantly ^10 ym. Particles and droplets between 0.3 ym
and 10 vim (averaging 3-4 ym) were deposited in the probe and the
filter collected particles ^0.3 ym. The impingers (back half)
collected SOX and particles <0.3 ym.
A substantial portion of the large size particulate (>10 ym) is
believed to be due to thermal shattering of rapidly quenched coke
and erosion by turbulent water sprays.
The quantity of large sized particles and droplets collected at
Lorain is comparable to tt j findings of earlier researchers.
However, the quantity per quench of smaller particles (<10 ym)
collected at Lorain was an order of magnitude greater than the
quantity of larger particles. The average particle size of the
total particulate emission was found to be approximately 3-4 ym.
118
-------
Particulate concentrations in quench water were also determined
and are presented in Table 2. '(5).
Figure 7 shows the total amount of particulate matter measured
by the sampling train versus the total dissolved solids content
of the quench water lost up the stack. (5) Each is expressed in
pounds per ton of coal. This graph suggests a strong relationship
between quench tower emissions and contamination in the quench
water. However, it also shows that not all of the emissions
were related to water contamination levels. For example, the
graph indicates that even if water with no dissolved solids was
used to quench, approximately 1.4 pounds of particulate per ton
of coal would be emitted from the Lorain tower.
Gaseous Emissions
Table 3 presents the results of gaseous emission tests as average
gaseous emission rates (Ib/ton of coal) for four compounds using
both clean and contaminated quench water.(5) It is evident that
emission rates for these gases are strongly dependent on quench
water quality.
The gaseous contaminants found in inlet quench water and subse-
quently in stack emissions, are evolved from several sources,
including coke oven flushing liquors, and the incandescent coke
itself. (See Table 4.)
Organic Emissions - Lorain (1976)
Three stack tests were conducted to quantify organic emissions.
Table 5 presents the results of these tests. ($> The data are
grouped into several categories of organic compounds. The aromatic
heterocarbons are clearly predominant. Polynuclear Aromatic
Hydrocarbons (PAH) were also investigated and these results appear
in Table 6. (5) Some of the lower molecular weight PAH were pre-
sent in equal or higher amounts in the stack emissions collected
than in the quench water.
So few samples were analyzed that no accuracy could be assigned
to these tests and the results were inconclusive. However, the
data suggest several interesting characteristics of quench tower
organic pollutants. A substantial quantity of organics, 1 to
8 Ib/ton of coal, may be emitted. These emissions are bulk
aromatic in nature and a small part of the total organics is
represented by Polynuclear Aromatic Hydrocarbons.
Organic Emissions - Lorain (1977)
Fifty-three (53) different organic species were identified and
quantified in these quench tower emission tests, many found in
significant amounts. Many of these organics were part of the
class of Polynuclear Aromatic Hydrocarbons and others were polar
compounds (heterocyclic oxygen and nitrogen compounds).
119
-------
The PAH test results for the three different test conditions
(considering water quality and coke greenness) are shown in
Table 7. (7> It is clear that the concentration of organics in
the quench tower plume increases greatly when contaminated water
is used and, to a lesser degree, when green coke is being quenched.
This trend was also revealed by statistical analysis of the data.
Three sample trains representing the three test conditions were
analyzed in detail by sampling train component to help determine
the physical characteristics of the organic emissions from quench
towers. These results indicate that most of the PAH found in each
test was in the organic adsorber unit, which means this material
is either gaseous or associated with particles smaller than 0.3
micron. These respirable particles can remain airborne over
long distances and can penetrate deep into the lungs.
Samples of clean and contaminated quench water were taken and
analyzed for PAH and polar compounds (Table 8). The contaminated
water was the major source of organic emissions as evidenced by
the high organic content of the water.'''
Because of the interest in using benzo(a)pyrene (BaP) as a surrogate
for PAH emissions, a separate analysis, highly specific for BaP,
was conducted. Table 9 shows significant amounts to be present.
BaP was not detected in every test; however, during those tests
in which BaP was detected, its concentration exceeded the Minimum
Acute Toxicity Effluent Value. (8>9) There appears to be no corre-
lation between concentrations of BaP and other organics, hence,
it does not seem to be a good indicator of the concentration of
other organics in the quench tower emissions.(7)
A comparison of the organic emissions data with various rating
systems showed that the following compounds considered to be
toxic, hazardous, or carcinogenic (8,9,10,11) are present in the
quench tower plume:
• Benzo(a) pyrene
9 3-Methyl cholanthrene
• 7,12-Dimethyl benz(a)anthracene
• Dibenz(a,h)anthracene
9 Dibenzo(a,h) pyrene
• Dibenzo(a,i) pyrene
• Benz(a)anthracene(s)
• Pyridine
• Indenod, 2, 3-cd) pyrene
• Phenanthrene
• Phenol
• Cresol
• Quinoline 120
-------
DOFASCO
Emission sampling was conducted by YRC at Dofasco's No. 2 Quench
Tower between August 15 and September 1, 1977. Prior to the
actual sampling, however, extensive , reliminary testing was re-
quired. At a meeting held with the Ministry of Environment, it
was determined that 12 points in a horizontal cross-section would
be tested. Two levels were suggested, 4' 6" and 2" above the
baffles. The 4' 6" level was ultimately selected since flow
testing indicated a more uniform gas flow at this level (refer
to Figure 8). Velocity profiles were performed at each of the
12 sampling points. A typical velocity profile is shown in
Figure 9. Gas velocities and gas temperatures during the first
water spray period were consistently higher than during the second
water spray period. (Fig. 9 & 10). The average gas velocity
and temperature range during emission sampling was 10.1 ft/second
and 155-170°F, respectively. The overall average stack gas flow
was 395, 400 ACFM. <12) Preliminary tests indicated that the
moisture in the gas leaving the quench tower averaged 27% and
peaked at 38% during the water spray periods of the quenching
cycle.
In all, 18 separate particulate tests were performed. Each test
consisted of three individual quenches at two sample points for
a total of six quenches per test. A total of 108 quenches were
tested over the sampling program.
Wind speed much greater than 10 mph would cause significant
shearing of the quench plume. Tests were conducted, therefore,
only during periods of low wind velocity. Wind conditions were
found to be most suitable in the early evening. The average wind
speed during testing was 5 mph.
Total particulate emissions (front plus back half) for all tests
averaged 0.466 Ib/ton of coal. Three modes of quenching operation
were tested during the sampling program. Nine particulate tests
conducted under normal operating practice using recycled quench
water averaged 0.464 Ib/ton of coal. In an attempt to achieve
better particulate removal, the baffle sprays were turned on and
recycled water was used during two tests. The particulate
emissions for these tests averaged 0.368 Ib/ton of coal. Six
tests carried out using once-through Bay water for quenching
averaged 0.573 Ib/ton of coal. Composite samples were collected
of water entering the quencher sprays. A typical analysis of
recycled quench water and Bay water is shown in Figure 11. Test
results indicate that particulate emissions do not vary signifi-
cantly for the three modes of operation tested.
Various process parameters, including coking time, coking temp-
erature, and emissions during pushing (amount of green coke) were
monitored during the emission testing in order to determine their
effect on particulate emissions. The pushing emissions were
ranked by visual observation, prior to quenching, on a scale of
121
-------
one to five. A ranking of one denoted very little particulate
emission. The rankings ranged from one to four with an average
of 2.6. Coking times averaged 14.8 hours ranging from 14.2 -
15 5 hours. Oven temperatures (average of coke side and pusher
side) averaged 2340°F, ranging from 2280 - 241QQF. Correlations
of particulate emissions with greenness, coking time, coking
temperature and sample point location could not be found.
Six particulate samples were saved from the cyclone catch for
particle size analysis, microscopic analysis and Coulter Counter
analysis. The particulate caught in the 10 micron cut size
cyclone accounted for 9.4% of the front half particulate catch.
Microscopic analysis indicated an average particle size of 63.1
microns and a range of 4.86 - 294 microns. The Coulter Counter
analysis indicated a median particle diameter of 64.5 microns.
The particulate caught in the probe accounted for 75.9 wt % of
the particulate catch. The remainder, 14.7 wt %, was found
on the filter paper. Since more than 90% of the particulate
passed through the 10 micron cut size knockout cyclone, it may
be concluded that the majority of the particulate was less than
10 microns in size.
ERIN
The pressure quencher at the Erin Plant of EBV was tested by YRC
in October, 1978. (13^ The objective of the program was to obtain
preliminary particulate emission data on the quencher cyclone ex-
haust and a pilot 2500 ACFM venturi scrubber. The venturi scrubber
operated at approximately 17-18 in. W.C. with a liquid-to-gas
ratio of 10-11 gallons per 1000 ACFM. Quencher water and scrubber
inlet water was relatively clean, containing 253 mg/1 of dissolved
solids and 3 mg/1 of suspended solids.
Total particulate emission results (cyclone, front half wash,
filter, back half) from the cyclone exhaust averaged 0.612 pounds
per ton of coal with 90% less than 10 microns in size. Emissions
from the scrubber exhaust are not presented due to questions con-
cerning the "representativeness" of the slip stream.
PARTICULATE COMPARISON
Particulate data from Lorain, DOFASCO and EBV are compared in
Table 10. The EPA Method 5 particulate emission rate (back half
excluded) was found to correlate closely with the water quality,
in particular, the total dissolved solids content. The DOFASCO
river water quench emission data can be deduced from the Lorain
data by comparing the TDS levels in the quench water and noting
that none of the baffles j~i the DOFASCO tower arrangement were
missing. The considerably higher back half catch on the EBV
pressure quench is hypothesized to be attributed to differences
in the physical-chemical nature of the processes, i.e., super-
heated steam aerosol (no air infiltration) at 280-400°F in the
pressure quencher versus an air-steam exhaust at approximately
170 F in ambient quench systems. Coal sulfur and ash contents
122
-------
during each of the test programs were comparable, averaging
approximately 1% and 7%, respectively.
DRY QUENCHING AS AN ALTERNATIVE PROCESS
Many complex problems are generated by most of the wet quenching
systems. In addition to air and water pollution problems, coke
quality degeneration and loss of usable coke are also associated
with the operation. Furthermore, substantial energy is wasted
during the process. Dry quenching provides a different approach
to the problems associated with wet quenching (see Figure 12).
Dry quenching of incandescent coke after it has been pushed from
the coking ovens is a proven, reliable process that is presently
being used in several industrialized countries. Foremost among
dry quenching's advantages are: (1) virtual elimination of air
pollutants emitted during quenching; (2) elimination of potential
water pollution associated with wet quenching; (3) improvements
in the working environment; (4) saving substantial amounts of
energy in usable forms; (5) producing more usable coke that is
superior to wet-quenched coke. By continuously circulating inert
gases through a cooling chamber that contains hot coke, dry
quenching recovers waste-heat energy that can be used to produce
steam, to produce electricity, to preheat combustion air, to
preheat coal, to dry coal, and to preheat feed water supplied
to fuel-fired boilers.
Dry quenching uses inert gases as a medium to collect heat from
incandescent coke and to transport that heat to waste-heat boilers.
Heat transfer from the coke to the inert gases is achieved by
direct contact of the gases with the coke. Inert gases used to
cool the coke and to transfer the heat are formed from an initial
intake of air. Oxygen contained in the initial intake of air
reacts with the hot coke to form C02. Except for the periodic
introduction of hot coke, dry quenching is a closed cycle operation,
The same gases are continuously circulated through the system.
Consequently, the composition of the circulating gases does not
fluctuate widely. Because oxygen is almost completely absent
from the quenching medium, the danger of explosion is minimized.
Nevertheless, explosion precaution procedures (e.g., pressure
relief doors) are taken and the composition of the circulating
gases continuously monitored.
The disadvantages of dry quenching include:
1. New equipment for dry quenching - with no reduction for
equivalent conventional steam producing equipment - has a
higher capital cost than wet quenching equipment.
2. Dry quenching facilities present potential explosion hazards
somewhat comparable to those associated with pulverized coal
boilers or with gas or oil steam generators. Reliance on
automatic sequencing, monitoring, feedback, and purging
systems help to minimize the hazards.
123
-------
3 Dry quenched coke is inherently dustier at screening stations
and during transport than wet quenched coke. This dustiness
may require the installation of additional fugitive dust con-
trol equipment or measures.
CONCLUSIONS
The research performed thus far on coke quench tower emissions
has yielded information concerning the physical and chemical
composition of the emissions and the source of contaminants.
Contaminants found to be present in the quench tower plume include:
• Particulate
• Various gases (sulfur dioxide, phenol, ammonia, cyanide)
• Organic compounds (including 53 identified species,
7 of which were found in significant quantities to be
considered a potential health hazard.)
The physical characteristics of these substances are as follows:
• Most of the particulate is relatively small in size
(less than 10 microns)
• Most of the organic species are either gaseous or
associated with particulates below 0.3 micron.
The primary source of the particulate, gases, and organics in
quench tower emissions is contaminated quench water. Specifically:
• Particulate, gaseous, and organic matter emissions
increase substantially when contaminated quench
make-up water is used.
• The quantity of organic matter increases when green
coke is being quenched.
• Pollutant reduction can also be achieved by minimizing
or eliminating the oil used for oven density controls.
Total particulate emission levels on the order of 0.3 - 0.5 Ib/ton
of coal (front plus back half catches) are believed to be the
best achievable from a conventional, well designed, properly
baffled quench tower using relatively clean make-up xvater. By operating
within the constraints of a natural draft system, only limited
increases in collection efficiency are believed possible. To
achieve any additional significant emission reductions, the
utilization of expensive add-on hardware will be required.
The use of a pressure quench system has a number of advantages:
1. The hot car is deep enough to accept the full load of
incandescent coke without moving. This allows a sealed
connection to control push emissions.
124
-------
2. Coke moisture is uniform and can be precisely adjusted.
3. The quench emissions are carried in a controlled flow
of steam to whatever control devices are required.
4. Energy recovery from this pressu ized steam flow is
potentially available and could lead to a closed system
with minimal emissions.
5. The volume of quench emissions is substantially less
than in a conventional quench tower, because no air dilutes
the steam from the coke bed.
6. No significant water pollution.
In particular, should quench tower emissions be severely restricted
by EPA, advantages (3) and (5) become most attractive from a con-
trol viewpoint.
Other alternatives for emission reduction would require signifi-
cant changes in the quench process. Dry quenching is one such
example. Characteristics of the dry quenching process that are
attractive include: substantial reduction of air pollution
caused by coke quenching; curtailment of air pollution - especially
soots and tarry smoke particles - during the hot coke push and
during transportation to the quench tower; elimination of water
pollution associated with wet quenching and improvements in the
working environment and the surrounding community.
ACKNOWLEDGEMENTS
The author wishes to express his appreciation to Bernard Bloom,
Carl Edlund, Robert V. Hendriks, and R.M. Statnick of the U.S.
Environmental Protection Agency for their role in the execution
of the Lorain test programs. The Lorain work was supported by
U.S. EPA Office of Enforcement under Contracts 68-01-3161 and
68-01-4138 and Industrial Environmental Research Laboratory
(IERL) under Contracts 68-02-1401 and 68-02-2819.
Appreciation is also expressed to Gerry Ertel of DOFASCO and
J.G. Riecker of Hartung, Kuhn Company for information presented
in this paper.
The assistance of Julie L. Rudolph and Carl E. Reichsteiner of
Arthur D. Little, Inc. in providing the organic analytical data
for the Lorain work is gratefully acknowledged.
125
-------
REFERENCES
1. Sidlow, A.P., "Source Test Report, Kaiser Steel Plant,
Fontana, California", San Bernardino County Air Pollution
Control District, February 29, 1972.
2. Memorandum from Robert A. Armbrust, Region IX, New York
State Department of Environmental Conservation, to Bernard
Bloom, Division of Stationary Source Enforcement, Office
of Enforcement, U.S. E.P.A., November 6, 1975.
3. Buonicore, A.J. et. al., "Characterization Program for Coke
Quench Tower Emissions," APCA Specialty Conference, Control
of Air Emissions from Coke Plants, Pittsburgh, Pennsylvania,
Proceedings, 126-143 (April 1979) .
4. Laube, A.H., Jeffrey, J., Edlund, C., Particulate Sampling
Techniques for a Coke Quench Tower, APCA Annual Meeting,
Toronto, Canada, June 1977.
5. Laube, A.H., Jeffrey, J., and Sommerer, D., Evaluation of
Quench Tower Emission Part II - Draft Report, Contract No.
68-01-4138, Task No. 3, U.S. Environmental Protection Agency,
Research Triangle Park, NC, December 1977.
6. Laube, A.H., Drummond, B.A., Coke Quench Tower Emission
Testing Program - Draft Report, Contract No. 68-02-2819,
Tast No. 1, U.S. Environmental Protection Agency, Research
Triangle Park, NC, February, 1979.
7. Hendriks, R.V., et.al., "Organic Air Emissions from Coke
Quench Towers," Paper No. 79-39.1, 72nd Annual Meeting of
the Air Pollution Control Association, Cincinnati, Ohio
(June 1979).
8. Rudolph, J.L. and Rechsteiner, C.E., Analysis of Samples
from Coke Quench Tower Emissions, Cambridge, Massachusetts,
Arthur D. Little, Inc., November 1978.
9. 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,
NC, November 1977.
10. Schalit, L.M. and Wolfe, K.J., SAM/IA;A Rapid Screening
Method for Environmental Assessment of Fossil Energy Process
Effluents, EPA-600/7- 8-Q15, U.S. Environmental Protection
Agency, Research Triangle Park, NC, February 1978.
11. Particulate Polycyclic Organic Matter, National Academy of
Sciences, Washington, DC, 1972, pp 4-14.
126
-------
REFERENCES (continued)
12. Ertel, G.L., "Quench Tower Particulate Emissions," APCA
Specialty Conference, Control of Air Emissions From Coke
Plants, Pittsburgh, Pennsylvania, Proceedings, 114-125
(April 1979) .
13. Buonicore, A.J., et.al., "Pressurized Quenching of Coke
at Erin, Germany", Sixth National Conference on Energy
and the Environment, American Institute of Chemical
Engineers, Pittsburgh, Pennsylvania (May 1979).
14. Linsky, B. et.al., "Dry Coke Quenching, Air Pollution
and Energy: A Status Report", J. APCA, Vol. 25, No. 9,
918-924 (1975).
127
-------
®\""'
I ~
^6^
41
•S/S/M
SS/S/Sf
/\i\ f\ A A /\/\
D
•*v
L
(D
r
. (D ,
• 1'
,.. V
*•
o
.J
rt
<*<
r.
O
t
rt
01-
II
©
CD
(9)
INCANDESCENT COKE
QUENCHED COKE
EXHAUST GASES
CONTAMINATED WATER
SERVICE WATER
HEAD TANK - OVERFLOW
HEAD TANK- STAND PIPE
NOZZLE HEADER
NOZZLE HEADER DRAIN
EMISSIONS TESTING STATION
SUMP
RETURN DRAIN DITCH
3AFFLES
Quench Tower
Figure 1
128
-------
FIGURE 2
BAFFLED SECTION OF QUENCH TOWER
^L
XVWWsg^^S
ffl
1—'
L
J
x>x^^ow^o
BAFFLED
SECTION
to
VO
0
1
1
1
o
B
BAFFLES
ooo 600
COKE QUENCH
CAR
S
SPRAY
NOZZLES
xF-
/ /
d
/ /
o
!
1
1
, i
BAFFLES
V
I i
I I
I I
I r
/
-------
Figure 3
DETAILS OF BAFFLES
(10 BafOes)
7'-6"
Six sections installed
each section 6'-3Y' * 9'-0" overall
6'-Oy x 9 '-3" inside
-------
I'M GURU 4-
Oi'hlN AHIi'.A BliTWEliN
9"
Typical
6-3/4"
8"
-------
DOFASCQ'S
COKE PLANT QUENCH TOWERS DATA
NO. OF OVENS
OVEN SIZE
(Length x width
x height)
COAL CHARGED
PER OVEN -TONS
COKING TIME -HR.
COKE QUENCHED
PER OVEN -TONS
COKE QUENCHED
PER DAY -TONS
QUENCH TOWER
HEIGHT - FT.
QUENCH TOWER
OUTLET AREA - FT2
MIST ELIMINATOR
WATER/QUENCH
USG
COKE PLANT NO. 1
BATTERIES NO. 1,2,3
10§
40'-8"x 17"x 13'
16
17.9
11
1.560
60
867
PACKING
18 -24" OF 3"
PLASTIC PALL
RINGS
5.000
COKE PLANT NO. 2
BATTERIES NO. 4,5
108
40'-6" x 17"x 13'
16
17.81
11
1,591
56.51
652
BAFFLES
1 x 4" SLATS
2 ROWS AT
20° INCLINED TO
VERTICAL
5,000
COKE PLANT NO. 3
BATTERY NO. 6
35
48'-5Vi"x 16 54' x 20'
30
14.0(Min]
21
1,343
110.25
1,800
BAFFLES
1 x 4" SLATS
4 ROWS AT 20°
INCLINED TO
VERTICAL
9.000
FIGURE HO 5
DOFASCO'S NO. 2 COKE PLANT
QUENCHING CYCLE
L STEPS
(1) QUENCH CAR IN TOWER
(2) WATER SPRAYS ON
(3) WATER SPRAYS OFF
(4) WATER SPRAYS ON
(5) WATER SPRAYS OFF
(6) QUENCH CAR OUT OF TOWER
TOTAL TIME FROM START OF CYCLE
isecj
0
10
55
75
105
130
FIGURE NO 6
132
-------
6.0 -
FIGURE?
TOTAL QUENCH TOWER PARTICULATE EMISSIONS
VERSUS
DISSOLVED SOLIDS IN QUENCH WATER
(95% CONFIDENCE LIMITS SHOWN)
V=0.18x •*• 1.4
D
QUENCHING WITH CONTAMINATED MAKE-UP HATER
QUENCHING WITH "CLEAN" MAKE-UP WATER
JL I I
I i I L_L
JL
0 1.0 2.0
4.0 6.0 8.0 10.0 12.0 .. 14.0 16.0
QUENCH TDS FACTOR *( POUNDS OF TDS PER TON COAL)
*PRODUCT OF TDS CONCENTRATION AND AMOUNT OF WATER EJECTED FROM THE TOWER
18.0 20.0
-------
COKE PIAMT QUEMCH TOWER
SAMPLE PORT LOCATIONS
SIDE VIEW
SAMPLE POIWT LOCATIONS
TOPVKW
POUT A
w-r-
-«MT-
-®-
POWB
-
-------
TEST RESULTS
QUENCH TOWER QAS TEMPERATURE
OVER QUENCH
* 14 1.2
TIMC.mln
Figure 10
TEST RESULTS
QUENCH SPRAY WATER ANALYSIS
ANALYSIS ppm
pH
SUSPENDED SOLIDS
TOTAL DISSOLVED SOLIDS
TOTAL ORGANIC CARBON
AMMONIA
CHLORIDE
SULFATE
RECYCLED WATER
7.6
265
327
8
.6
170
53
BAY WATER
8.0
19
313
7
.6
55
53
Figure 11
135
-------
TYPICAL DRY QUENCHING PROCESS
CLEANED COOLER GAS
FINE DUST CYCLONE
AUTO-
MATIC
HARGIN£
COVER
COKE
BUCKET
STANDARD
STEAM
BOILER
DRY
QUENCH
TOWER
COARSE
' DUST
SEPERATOR
COOLED GAS
DISTRIBUTOR
COKE
GATE
QUENCH
GAS
FAN
SUPER
HEATED
STEAM
320JF
590 PSI
WATER IN
TRANSPORT
CAR
(reprinted from: Linsky,B., et al;
Dry Coke Quenching, Air Pollution
and Energy: A Status Report. JAPCA
Vol.25,No.9, Sent. 1975)
\
QUENCHED COKE CONVEYOR
Figure 12
136
-------
Table 1 - PARTICDLATE TEST RESULTS (LORAIN 1976)
Parameter Clean Water Contaminated Water
Flow Conditions (Average)
% Isokinetic 99.0 99.4
Flow Rate, DSCFM 190,700 183,300
% Moisture by Wt. 25.0 24.7
Particulate Emissions, Ib/ton
of coal
Cyclone 0.28 1.17
Probe 1.16 1.15
Filter 0.02 0.45
Total Front Half 1.46 2.77
Total Back Half 0.16 0.31
Totals 1.62 3.58
-------
Table 2 - QUENCH WATER CONTAMINANT CONCENTRATIONS (LORAI2I 1976)
CLEAN WATER
CONTAMINANT
TOTAL. DISSOLVED SOLIDS
Sprayed on Coke
Returned to Sump
Make-Up Water
TOTAL SUSPENDED SOLIDS
Sprayed on Coke
Returned to Sump
Make-Up Water
mg/1 Ib/ton of coal
CONTAMINATED WATER
mg/1 Ib/ton of coal
1050
1040
520
60
.270
2.5
6.8
5.5
0.54
0.38
1.4
0.003
985C
9670
5370
165
340
140
64
52
6.6
1.0
1.3
0.17
Table 3 - GASEOUS TEST RESULTS (LORAIN 1976) LB/TON OF COAL
COMPOUND rr.v\K WATER CONTAMINATED WATER
Sulfur Dioxide
Phenol
Cyanide
0.0053
0.0038
0.026
0.0023
0.46
0.22
0.33
0.01
Table 4 - COMPARISON WITH CONTAMINANT LEVEL IN STAOC EMISSIONS
AND QUENCH WATER-LB/TON OF COAL (LORAIN 1976)
QUENCH WATER
CONDITION
Clean
dean
Clean
Contaminated
Contaminated
Contaminated
COMPOUND
Phenol
Cyanide
Phenol
Ammonia
Cyanide
GASEOUS
EMISSIONS
0.0038
0.026
0.0023
0.22
0.33
0.01
QUENCH WATER
CONTAMINANT LEVEL
0.0045
0.13
0.0015
0.21
0.7B
0.055
138
-------
Table 5 - TOTAL ORGANIC EMISSIONS (LORAIN
TYPES OP ORGANIC
COMPOUNDS
ALIPHATIC
HYDROCARBONS
AROMATIC HYDROCARBONS
NON-AROMATIC
HETEROCARBONS
AROMATIC HETEROCARBONS
ALDEHYDES, ESTERS,
CARBOXILIC ACIDS,
ACRYLATE POLYMERS
KETONES, AMINE
SALTS, PHOSPIIINES,
ISOTHIOCYANATE
TOTALS
(1) Average of 2 tests
- Not Detected
N/A Detected but weight not available
CLEAN WATER
CONTAMINATED WATER
STACK EMISSIONS (1) WATER (COMPOSITE)
Ib/ton of coal
N/A
S 5.63
0.504
0.107
7.96
mg/m3 Ib/ton of
735 0.008
N/A
N/A
2408 0.03
213 0.02
44 0.008
3400 0.07
coal nig/1
1.2
N/A
5-4
3.9
0,3
10.8
STACK
Ib/ton of
0.04
0.008
0.68
0.25
0.08
1.06
EMISSIONS
coal mg/raj
14
2
259
95
33
403
INLET WATER
Ib/ton of coal rag/I
0.008 Q.5
0.008 0.9
0,57 98,3*
0,29 50,4
o.ooa 1,1
o.aa J.5J..2
-------
Table 6 - POLYNUCLEAR AROMATIC HYDROCARBON DETERMINATIONS (LORAIN .1976)
•c-
o
Contaminated Water
Stack
Ibs/ton of
(xio~5)
Anthracene/Phenanthrene
Methyl Anthracenes
Fluoranthene
Pyrene
Methyl Pyrene/
Fluoranthene
Benzo (c) phenanthrene
Chryseae/Benz (a) -
anthracene
Methyl chrysenes
Dimethyl benz(a)-
anthracene
Benzo Fluoranthenes
Benz (a)pyrene
Benz (e)pyrene
Indene (1,2,3-cd) pyrene
Total
0.24
Q.10
0.12
0.09
0.07
0.07
0.10
0.79
Emissions
coal
yg/m3
0.9
0.4
0.5
0.4
0.3
0.3
0.4
3.2
Inlet Water
Ibs/ton of coal
(xlO-5)
0.26
0.07
0.28
0.16
0.10
0.06
0.17
0.09
0.12
0.02
0.04
0.04
1.4
Stack
Ibs/ton >of
yg/1 (xlO~5
0.5
0.1
0.5
0.3
0.2
0.1
0.3
0.2
0.2
0.04
0.08
0.06
2.58
0.13
0.07
0.07
0.05
0.04
0.04
0.05
Clean Water
Emissions*
coal
) yg/m3
0.6
0.3
0.3
0.2
0.2
0.2
0.2
Water (Composite)
Ibs/ton of
\xi6-5)
0.03
0.02
O.O3
0.02
0.02
0.03
coal
yg/i
0.05
0.04.
0.05
0.04
0.04
0.05
0.004 0.2
0.05
0.50
0.2
2.4
0.15
0.27
* Average of two tests
-------
Table 7 - PQLraTOCLEAR AROMATIC H!TOEQGMBONS (LOBAUT1977)
LB/TOH OF COAL)
dean. Water
* Low Greenness
« High Greenness
Contaminated. Water
0.000838
0.00234
0.0642.
Table 8 - QUENCH WATER ANALYSIS - LB/TON OF COAL (LOBA1N 1977)
total PAH
total Polar
Compounds
!not including
phthalates)
Clean.
Service
Water
0.000162
0.0000036
Water
Inlet, to
Nozzle
0.000065
Contaminated
Flashing
Liquor
0.0334
0.0981
Water
Inlet to
Nozzle
0.0122
0.2074
Table 9 - BEHZO
-------
TABLE 10
PARTICULATE EMISSION DATA COMPARISON
Ni
Particulate Emission Data (Ib/ton
Lorain-1976 Lorain-1977
Cyclone
Front Half
Washings
Filter
Subtotal
Beck Half
Subtotal
Total
Quench Water
Quality
(mg/1)
TDS
TSS
River
Water
0.28
1.168
0.016
1.464
0.156
0.156
1.620
1050
60
Contaminated River Contaminated
Water Water Water
1.17 0'.669 1.774
1.15 0.148 0.193
0.45 0.634 0.314
2.77 1.451 2.281
0.81
0.81
3.58 —
9850
165
of coal)
DOFASCO
River
Water
0.028
0.180
0.052
0.260
0.204
0.204
0.464
327
265
River Water +.
Baffle Sprays
0.022
0.181
0.010
0.213
0.155
0.155
0.368
327
265
EBV
City
Water
0.076
0.046
0.070
0.192
0.420
0.420
0.612
253
3
-------
COKE BATTERY ENVIRONMENTAL CONTROL
COST-EFFECTIVENESS
by
William F. Kemner
and
Steven A. Tomes
PEDCo Environmental, Inc.
Cincinnati, Ohio 45246
ABSTRACT
A computerized optimization model has been developed to examine the cost-
effectiveness of alternative emission control strategies for coke plants. The
model calculates the lowest cost mix of controls to meet a given overall level
of emissions for a given air pollutant, and also calculates the lowest overall
emissions that can be achieved for a given cost. The data base is uncoupled
from the model so that it can be updated as new or improved data become avail-
able. The present emission data base contains emission factors for four air
pollutants—particulate matter, benzene soluble organics, benzene, and benzo-
a-pyrene—for 14 coke plant sources. The plant data base encompasses 216
batteries in 58 plants. The cost data base contains capital and annualized
cost functions for 41 control techniques, but as many as 8 control options can
be accommodated for each source. The data base can be subdivided to enable
examination of other factors, such as old versus new batteries or large
versus small batteries. The optimization can be focused on either capital
cost or annualized cost.
143
-------
COKE BATTERY ENVIRONMENTAL CONTROL
COST-EFFECTIVENESS
INTRODUCTION
The characterization and control of coke oven emissions have been of
intense interest and study for over 10 years. Originally, focus was directed
primarily toward visible emissions because most coke oven emissions are fugi-
tive in nature. As data became available on the complex chemical structure
and health effects of the emissions, attention shifted to the organic com-
ponents. Because environmental control of coke oven operations requires the
evaluation of numerous options, and because technology and new information are
continually developing, the U.S. Environmental Protection Agency (EPA) con-
tracted PEDCo Environmental, Inc., to develop a computer model that could
calculate the cost and emission levels for any combination of controls. Even
more important, the model should be able to calculate the lowest cost mix of
controls to meet a given overall level of emissions, or alternately the lowest
overall emission level that can be attained at a given total cost. Either
capital cost or annualized cost can be the focus of optimization.
Intended as an engineering tool for evaluating control strategies on a
continuing basis, the model operates on the EPA computer at the National
Computation Center in Research Triangle Park, North Carolina. It accommodates
data on four pollutants: particulate matter (defined herein as front-half-
Method 5), benzene soluble organics (BSO), benzo-a-pyrene (BaP), and benzene.
The model includes the coke oven battery, the coal storage and preparation
steps, the quenching and coke screening operations, and the byproduct plant.
It addresses both conventional batteries and preheated coal batteries. The
model contains very limited information regarding byproduct operations, but
studies are under way by the EPA to characterize the emissions from this
source.
144
-------
MODEL STRUCTURE
The model consists of two distinct parts, the mathematical structure and
the data base. It must be recognized that tie present data base is only in-
tended to represent a starting point. It is designed to be easily updated
because new data, especially for emission factors, are continually being
developed.
Dataset 1: Emission Factors
Fourteen air emission sources and four pollutants are presently con-
sidered, as shown in Table 1. (Note that two sources have alternate factors.)
The term "uncontrolled" is not easily defined for coke ovens, but for our
purposes it represents the conditions existing at the majority of batteries in
the late 1960's. Although this definition still leaves much room for judgment,
it eliminates the totally uncontrolled conditions that would prevail if a
coking process were operated with no concern whatever for emissions.
Emissions from most of the sources are fugitive in nature, and do not
lend themselves readily to Method 5 sampling techniques. Two commonly used
techniques are single-point sampling with a Method 5 train and High-Vol samp-
ling. Isokinetic conditions are often difficult to achieve, and sample results
are generally corrected to reflect anisokinetic sampling.
The use of test data to develop emission factors requires several assump-
tions. The first is that the results of the test are representative of the
emissions found at the entire battery; the second is that the emissions from
the tested battery are representative of the industry as a whole. Clearly,
notable exceptions will be found to these assumptions, and the emission data
used herein must be treated with appropriate caution. In the case of door
emission tests, for example, results obtained on only one door do not indicate
that all the doors leak similarly. Further, the sampling results from a
single green push do not necessarily represent all the coke pushes at that
battery.
The definition of particulate matter is important in dealing with coke
oven emissions. Generally, a reference to particulate catch means the front
half or filterable portion of a Method 5 sample train (or its equivalent).
Total suspended particulate (TSP), however, implies both the front and back
145
-------
TABLE 1. SUMMARY OF UNCONTROLLED EMISSION FACTORS
(Ib/ton of coal)
Source
code No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Emission source
Larry car charge (wet coal)
Coke pushing
Quench, clean water
Doors
Topside leaks
Combustion stack (old)h
Coke handling
Coal preheat
Coal preparation
Coal storage
Pipeline charge (dry coal)
Redler conveyor (dry coal)
Hot larry car (dry coal)
Byproduct
Combustion stack (new}h
Quench, dirty water
Pollu
' TSPa
1.0"
2.0f
1.7f'9
0.4b
0.2d
1.3d
1.0d
7.05b
U.5d
0.15d
O..016d
0.010d
0.017d
0C
0.13d
3.2f'9
BSD
0.08f
1.7xlO-3b
0.5b
0.25d
0.006d
Od
1.05b
Od
Od
0.019d
0.006d
0.019d
0.3C
6xlO-4d
6.4xl
-------
half of a Method 5 test. The definition used must be consistent with the
predicted control efficiencies, particularly when discussing TSP. It is
likely that many of the particulate control devices would not "see" the
organic particulate captured in the back half of the sampling train because it
would be gaseous in form as it passed through the control device and would
condense some time later.
Governed by these restraints, emission factors were developed for each of
the four pollutants. To the extent possible, results from emission tests were
used. When explicit data were unavailable, an attempt was made to derive an
emission rate from other available information and assumptions. If an emis-
sion factor could not be developed by either of these approaches, an engineer-
ing estimate was made. The entire emission matrix had to be provided so that
initial runs of the model could be completed. Where estimates have a broad
confidence range, model runs can be made for various values to examine sensi-
tivity.
Dataset 2: Cost Functions
D
All cost functions are expressed as Y = AX , where Y is annualized cost
or capital cost in fourth quarter 1978 dollars, X is tons of annual coke
capacity, and A and B are constants for the specific control technique.
Capital and annualized cost functions are provided for both new and retrofit
installations.
Table 2 lists the control options by source. The efficiencies shown are
initial estimates only and are subject to change as new data become available.
The model will accommodate as many as 8 control options for each source, but
only a total of 41 are presently considered. Although control efficiency is
discrete in some cases and continuous in others, discrete levels have been
used in the model for simplification. As new control options are added or
existing ones modified, the appropriate cost functions are added to Dataset 2.
The A and B coefficients for some of the cost functions presently used in the
model are shown in Table 3. Further examination of the cost functions for
small batteries (below 100,000 tons of coke per year) is necessary to deter-
mine if the functions are accurate in this low range.
147
-------
TABLE 2. CONTROL OPTIONS BY SOURCE
00
Source
Larry car charging
Coke pushing
Quenching clean water
Doors
Topside
Combustion stack-old
Coke handling
Coal preheater
Coal preparation
Coal storage yard
Pipeline charging
Redler charging
Hot larry car charging
Byproduct plant
Combustion stack-new
Quenching-dirty water
Control option
Modified car
New car
Retrofit second main + new car
Controlled coking
Shed + ESP 95%
Shed t scrubber 95%
Enclosed car
Shed + ESP 99%
Shed +• scrubber 99%
Baffles
Diverted flow baffles
Dry quenching
Cleaning and maintenance
High pressure water cleaning
Door hood and scrubber 95%
Door hood + scrubber 98%
Luting and cleaning
Luting and maintenance
New lids and castings * luting and
cleaning
Oven patching
Dry ESP 90%
Dry ESP 98%
Baghouse 98%
Enclosures + baghouse 99%
Scrubber-15 in.
Dry ESP 95%
Scrubber-30 in.
Ury ESP 99%
Enclosure and baghouse-99%
Water truck
Unload sprays and water truck
Coal pile sprays
Operation and maintenance
Operation and maintenance
Operation and maintenance
Maintenance
Oven patching
Baffles
Clean water + baffles
Diverted flow baffles + clean water
Dry quenching
Efficiencv
TSP
30.0
99.0
99.5
60.0
85.5
85.5
88.2
89.1
89.1
70.0
90.0
98.0
60.0
80.0
88.5
93.3
90.0
95.0
97.0
80.0
90.0
98.0
98.0
89.1
95.0
95.0
98.0
99.0
97.0
60.0
75.0
90.0
99.0
99.0
99.0
NA
80.0
70.0
85.0
95.0
99.0
BSD 1 BaP
80.0
99.0
99.5
60.0
45.0
49.5
54.0
45.0
54.0
70.0
90.0
99.0
60.0
80.0
78.0
83.8
90.0
95.0
97.0
80.0
50.0
60.0
50.0
NA
60.0
45.0
60.0
50.0
NA
NA
NA
NA
99.0
99.0
99.0
80.0
80.0
35.0
75.0
85.0
80.0
99.0
99.5
60.0
45.0
49.5
36.0
45.0
54.0
70.0
90.0
99.0
60.0
80.0
78.0
83.8
90.0
95.0
97.0
80.0
50.0
60.0
50.0
NA
60.0
45.0
60.0
50.0
NA
NA
NA
NA
99.0
99.0
99.0
80.0
BO.O
35.0
80.0
85.0
99.0 99.0
Benzene
80.0
99.0
99.5
60.0
45.0
49.5
54.0
45.0
54.0
0.0
0.0
99.0
60.0
80.0
65.0
72.0
90.0
95.0
97.0
80.0
50.0
60.0
50.0
NA
50.0
45.0
50.0
50.0
NA
NA
NA
NA
99.0
99.0
99.0
80.0
80.0
0.0
75.0
75.0
99.0
Remarks
19.0 h avg. coking time vs. 17.5 h base
Includes option 5 on source Z
Includes door cleaning machine
Not applicable to pipeline batteries
Includes option 5 on source 2
NA - not applicable
-------
TABLE 3. COST FUNCTION COEFFICIENTS FOR CONTROL OPTIONS0
(cost in fourth quarter 1978 dollars)
Source control option
Wet coal charging
Modified larry car
New larry car
New larry car & second main
Coke pushing
Controlled coking
Shed and ESP, 95%
Shed and scrubber, 95%
Enclosed hot car
Shed and ESP, 99%
Shed and scrubber, 99%
Quenching, clean water
Conventional baffles
Diverted flow baffles
Dry quenchingb
Doors
Cleaning and maintenance
High pressure water cleaning
Door hoods, scrubber, 95%
Door hoods, scrubber, 98%
Door hoods, scrubber, 95i one side
Door hoods, scrubber, 98% one side
Topside
Luting and cleaning
Luting, cleaning, and maintenance
New lids, luting, and cleaning
Combustion stack, old
Oven patching
Dry ESP, 90%
Dry ESP, 98%
Fabric filter, 98%
Basis
for
X value
Battery
Battery
Plant
Battery
Battery
Battery
New Installation
Capital cost
A
290,539.2
2,784.7
326,894.4
0
17,498.8
25,028.5
423,778.8
14,710.7
23,357.7
1.7
82.6
771.8
0
414,499.8
18,558.0
21,562.1
13,431.0
11,682.0
0
0
81,100.0
0
2,534.3
2,609.3
418.3
B
0.0250
0. 4882
0.1935
0
0.4228
0.4223
0.1938
0.4422
0.4310
0.8412
0.7119
0. 7065
0
0
0.3453
0.3441
0. 3409
0. 3620
0
0
0
0
0. 5283
0.5484
0.6518
Annuali zed cost
coefficient
A
227,751.9
68,022.8
341,620.2
25.7
3,177.5
7,752.5
70,214.2
2,907.6
5,243.8
0.4
9.5
87.6
804,801.2
876,701.5
954,615.0
879,789.9
106,371.0
863,212.0
264,900.1
503,300.4
300,799.9
503,300.4
5,373.4
3,989.6
551.5
B
0.0762
0.1934
0.1117
0.9593
0.4737
0.4439
0.2354
0.4836
0.4823
0.8750
0.7714
0.7683
0
0
0.0624
0.0715
0.0431
0.0614
0
0
0
0
0.3994
0.4333
0.5543
Retrofit Installation
Capital cost
coefficient
A
319.187.0
3,064.5
293,877.1
0
22,052.2
30,128.9
466,163.3
18,452.7
28,061.2
2.1
107.5
848.9
0
414,499.8
21,613.0
25,298.2
15.000.0
13,625.0
0
0
105,399.9
0
2.976.1
3,085.1
515.9
B
0.0251
0.4882
0.2046
0
0.4141
0.4166
0.1938
0.4337
0.4254
0.8412
0.7119
0.7065
0
0
0. 3487
0. 3465
0.3443
0.3638
0
0
0
0
0.5306
0.5500
0.6504
Annual! zed cost
coefficient
A
230,194.8
63,015.2
320,512.5
25.7
3.659.2
8,519.4
76.212.7
3,345.8
5,825.5
0.4
12.7
96.8
804,801.2
876,701.5
880,426.2
812,304.1
998,158.0
808.202.0
264,900.1
503.300.4
304.299.8
503.300.4
5.061.8
3,800.7
546.8
B
0.0761
0.2014
0.1178
0.9593
0.4670
0.4403
0.2332
0.4771
0.4776
0.8707
0.7621
0.7651
0
0
0.0708
0.0800
0.0496
0.0682
0
0
0
0
0.4101
0.4440
0.5616
-------
Dataset 3: Coke Oven Battery Data
This dataset provides a record of the following for each battery in the
United States:
Company code
Date installed or date of last major rebuild
Number of ovens
Capacity* tons of coke/year
Type of charging used
Oven height
Number of collecting mains
Control equipment in place
Input of the battery data is arranged so that they can easily be updated.
Although some of the data are estimated, most of the industry is correctly
represented in the census; the aggregate costs calculated from the current
data base should be representative exclusive of those cost increments that are
highly site-specific.
Capacity is used as the variable in determining control costs. Certain
costs, however, are not strictly a function of capacity. Shed cost, for
example, is a function of oven height and number of ovens. In the case of
quench towers, cost is proportional to the number of towers. It is assumed
that one quench tower can handle up to 2500 tons of coke per day. For coal
yards, coal preparation, coke processing, and byproduct plants, the model
calculates costs for entire plants rather than individual batteries. Table 4
shows the relationship between key oven parameters used to translate capacity
data into the physical size data needed to determine certain costs. For
example, oven volume is the key parameter for sizing larry cars, and tons of
coke per push is the key parameter for determining enclosed hot car cost.
These relationships were used in calculating the cost functions for model
input. Because most batteries fall into one of the three categories shown in
Table 4, the use of capacity as the cost variable is reasonable.
Site-specific factors ce jainly affect cost. For example, the economy of
scale gained by combining toro or more adjacent batteries under a common con-
trol device has not been considered. Such site-specific factors are not
150
-------
TABLE 4. RELATIONSHIPS OF SIZE AND OTHER PARAMETERS,
COKE OVEN BATTERY
Basis: 50 ovens
Oven height, m
Oven volume, ft
Tons coke/push
Without preheat
Coking time, h
Pushes/day
Tons coke/yra
With preheat
Coking time, h
Pushes/day
Tons coke/yra
3
540
8.5
17.5
68.6
213,000
12.5
96
296,000
4
720
12.0
17.5
68.6
300,000
12.5
96
420,000
6
1390
25.0
17.5
68.6
626,000
12.5
96
876,000
aDirectly proportional to number of ovens and inversely proportional to
coking time.
151
-------
considered at this time, but the model could be refined to do so. Additional
factors irore specific to individual batteries, such as the need to relocate
existing facilities to make room for control equipment, could also be added to
the model if site-specific information of this detail were available.
To evaluate projected growth (or decline) in the industry, the user may
add battery data corresponding to the projected growth or delete battery data
for projected retirements. The primary distinction between "new" and "exist-
ing" in the model is the use of new or retrofit cost functions.
In representing existing control at a plant, the exact control equipment
may not be the same as that designated in the available control options. The
user must, therefore, select the control option that most closely corresponds
to the existing equipment or control program on the basis of efficiency. Al-
ternatively, additional control options can be added to the model.
Costs are adjusted in the model in one of two ways:
1. Correcting total industry costs for those plants where some control
is already in use.
2. Allowing the use of a control baseline (e.g., SIP) and considering
only costs above this baseline.
As an example of the first case, assume Battery 1 has already installed a
new larry car for stage charging and has good stage charging practice. The
data for this battery would indicate these conditions, and the only costs
calculated would be those for more efficient controls (if such exist). Tear-
out costs for removal of existing controls are not considered in the present
model.
If a control system is installed and does not correspond to one of the
options in the model, it must be considered comparable to achieving a given
level of control efficiency rather than equivalent to the specific hardware
configuration. For this reason, the control options are ordered according to
degree of efficiency. This permits the user to select the option that achieves
the highest level considered appropriate.
152
-------
In the control baseline case, the user establishes a base level of control
below which costs are not considered. For example, if the baseline for charg-
ing control is modified larry cars for stage charging, costs will be con-
sidered only for options exceeding this level
Two "interactions" are recognized in the model:
The use of shed control affects control of coke-side door emissions.
If a shed and door hoods are selected, the door hoods are used only
on the push side.
The use of dry quenching affects control of pushing emissions be-
cause dry quenching utilizes an enclosed hot car. The cost of dry
quenching also includes the cost of water treatment at those plants
that would otherwise use dirty water for quenching.
MODEL FORMULATION
The first step in the model is the calculation of total industry control
costs and emissions. Mathematical nomenclature has been defined as follows:
i = the emission source (i = 1 . . . I)
j = the pollutant (j = 1 . . . J)
k = the control technology (k = 1 . . . K)
n = the specific battery (n = 1 . . . N)
E = the annual emissions, in tons/year
C = the annualized cost (or capital cost)
e... = the control efficiency
X = the capacity of battery n in tons of coke/year
U.. = the uncontrolled emission factor in Ib/ton of coal
Note that for some sources, such as coke handling, X actually represents
capacity for all batteries in a given plant.
n
Then C... represents a specific dollar value calculated from the general
i J Kfi
cost function:
Cijkn " AijkXnijk
where A = y intercept, B = slope
153
-------
Note that Cilkn = Ci2kn = Ci3kn = C14kp
In other words, the cost of a specific control system does not vary by pollutant.
(100 - e... )
Similarly, Eijkn - [ ^T^ ] '' 2°°°
but Eilkn + Ei2kn + Ei3kn ' Ei4kn
The C and E matrices are calculated from the input datasets. Note that
k = 1 will represent no additional control. Consequently, C.. .-|n = zero by
definition and
Eijln ' WijMV.7)] * 200°
The total emission restriction (Mode 2) is entered as an overall percent
efficiency, represented by R.. This will be converted to pj, an annual quan-
tity, by the equation:
(n^ 1f1 Eijn' k " ])
The total cost restriction (Mode 3) will be entered as a total dollar
amount, T.
Finally, note that not all i exist for every n. The C and E matrices,
therefore, are not full.
The next step is to compress the total C and E matrices to a total indus-
try basis, i.e., to eliminate the n dimension. Let C1 and E' represent the
total industry:
N
r' = v r
ijk n=1 ijkn
Again noting that Cjlk = c:2k = c;3k = c:4k
Similarly,
N
Eijk =nf] Eijkn
and Eilk ' Ei2k " Ei3k ' Ei4k
154
-------
Furthermore, E' can be treated as four separate matrices, E!. for j = 1,
Elk for j = 2, etc.
The data are now reduced to two simple matrices,
C-k and Eik for each j
To find the optimum combination of controls, consider Mode 2, the restric-
tion being total emissions, p., and the objective being to find lowest cost.
J
Another matrix, Y, must now be introduced. The values of Y will be either one
or zero. A one will indicate that a control option, k, is selected, and a
zero will indicate that the control option is not selected. The Y matrix is a
mathematical device to solve the optimization and has no significance from an
engineering standpoint. For example,
let Y,k . Yu - 1
This means that Control Option 4 on Emission Source 1 is part of the optimum
solution.
If this is so, then by definition all other Y's for source 1 are zero:
V =Y =Y =Y =Y =Y =0
Yll Y12 Y13 Y15 Y16 Ylk u
That is, a source can only be controlled by one option at a time. This
integer device avoids the potential problem of finding optimum solutions that
fall between discrete control systems. If all control systems had continuous
cost-efficiency functions, a noninteger optimization could be used.
The statement of the problem in matrix form is, therefore:
Minimize ZCY for a given j
subject to SY = 1 for all i
and EEY £ p for a given j
and Y >^ 0 for all i
In expanded form:
minimize CY + CY + CY
r Y
L2kY2k
CikYik
155
-------
subject to Yn + Y12 + Y13 + Ylk = ]
YO-, + Y99 + Y97 + Y?. = 1, etc., for each
21 ^ ^J ^ source
i— \ /
and:
F Y H
^rzi
r L12'12
h F Y +
L21T21
Ik Ik
2k 2k
E,,,Y.. < PJ
and Y.. > 0 for every i and k
1 K
The optimal solution to this problem will be determination of the Y
matrix, which in turn defines a k value for each i (i.e., a control option for
each emission source). Note that any given k may equal 1, i.e., no control.
Controls for any given emission source can be fixed at a predetermined level
and the source can thus be removed from the optimization procedure. In
general, the program will select those alternatives that reduce emissions the
most and cost the least.
After the optimum solution is found, the Y matrix is superimposed onto
the E matrices for the other pollutants to determine the emissions of the
pollutants that were not restricted. To the optimum totals the program adds
the costs and emissions for those sources excluded from the optimization to
obtain total industry costs and emissions.
The statement for obtaining minimum emissions for a given cost, T, is
very similar:
Minimize IEY
subject to: ZCY <_ T
and ZY = 1
and Y _> 0
The approach described above is used in this case also. Operation of the
model in either case is identical whether annualized cost or capital cost is
the subject of optimization.
The advantage of the model is the ability to rapidly answer "what if"
questions. The user can examine sensitivity to variations in the emission
factors and control costs by varying the input data.
156
-------
RESULTS
An example best illustrates the logic of the model. The data base for
this example consists of all 216 coke oven batteries, the uncontrolled emis-
sion factors presented in Table 1, and the control cost functions as presented
in Table 3.
Figure 1 is a graphic presentation of the capital cost functions for the
three control options applicable to Source 1, wet coal charging. Every cost
function could, of course, be similarly plotted.
Figure 2 shows the annualized cost per pound of particulate removed for
the same three options. The spacing of the curves is related to both the rel-
ative costs of the options and the relative efficiencies. Although Option 3
is more costly than Option 2, the curves are very close because Option 3 is
considerably more efficient than Option 2. Option 4, on the other hand,
represents an efficiency improvement of only 0.5 percent over Option 3 at a
high cost. Each option for each source and each pollutant could be analyzed
in a similar manner.
The function of the model is to analyze all such curves and find the
optimum combinations. Figure 3 is the model output to meet a restriction
calling for an overall particulate control of at least 95 percent efficiency,
which requires the highest possible control level on every source except dry
coal charging and the byproducts plant. The latter are excluded because their
contributions to particulate emissions are very low; in fact, particulate
emissions from the byproducts plant are assumed to be zero in the model. As
noted previously, no cost is assigned to the enclosed car option for pushing
emissions because the equivalent of an enclosed car is included in the cost of
the dry quenching system. Total annualized costs for the industry are
$1,396,000,000 and total capital costs (retrofit) are $2,887,000,000. In
these examples, the model seeks to minimize annualized costs, not total capi-
tal costs. These values do not take into account any existing control, as
indicated in the table. Based on similar runs for 80, 85, and 90 percent
control, Figure 4 shows the optimum composite cost-effectiveness function for
this hypothetical case of no existing controls.
157
-------
4 x 10'
3 x TO1
& 2 x 106
8
LU
tt
1 x 106
OPTION 4
OPTION 3
OPTION 2 ••
BASIS: 1 BATTERY
60 OVENS
4TH QUARTER
1978 DOLLARS
MODIFY LARRY CAR + STEAM SUPPLY + SMOKE BOOT
OPTION 1 - UNCONTROLLED, COST • 0
I i
100.000
200,000 300.000
ANNUAL COKE PRODUCTION, tons
400,000
Figure 1. Capital cost of control options
for wet coal charaina.
158
-------
lOi
o
Tl
*
O
UJ
i
uj
oc
UJ
8
O
O
O
Q.
to
O
O
i
100,000
Figure 2.
200,000 300,000
ANNUAL COKE PRODUCTION, tons
Cost per pound of particulate removed
options for wet coal charging.
400,000
for control
-------
COKE QVCN BPTINIZATION
IBJE£Tm« MINIBUS AiNifALIZEB CNT RESTRICTION!
9A8ELii£i A88WUN8 HO SIS' 01 EXI8T1N8 CONTROLS
93.01 OVERALL EFFICIENCY POLLUTANTlTBP
M8C TEM 1T79
CONTROLLEB EMISSIONS
(LIB/TON COAL) (TONS/YEAR)
CMfTMUCI CMT
(NILLION BOLLAftS)
ISttftCE
IA0RV FAB rHsJ>(SSS»»
-------
g 2000
1600
8
^ 1200
IK.
01
800
400
-•^B
I I I I L I 1
10 20 30 40 50 60 70 80
OVERALL PARTICULATE REMOVAL EFFICIENCY, %
I
90 100
Figure 4. Total annualized cost as a function of overall efficiency.
-------
Cost is presently overstated in some cases because each battery is treated
independently. An example would be a plant with four batteries, at which both
door hoods and a wastewater recirculation system are provided for each battery.
In such a scheme it is likely that one common water system could be installed
to serve all four batteries for less than the cost of four separate water
systems. Similarly, if a shed and scrubber were installed to control pushing
emissions and coke-side door emissions, and a hood system were installed for
push-side doors, the water system (and perhaps the scrubber itself) could be
designed to handle both sources. The existing model can be modified to
address these issues, but specific assumptions will be required.
When a different pollutant is examined, quite different results are seen.
Figure 5 illustrates an example of minimizing BaP emissions for an annualized
cost restriction of $575,000,000. The resultant control efficiency for BaP is
89 percent, although particulate control drops to about 50 percent. In this
example, existing control was taken into account. As seen in Figure 4, the
annualized cost for 89 percent particulate control is nearly double that for
equivalent BaP control, and capital cost is proportionately even more. Capi-
tal cost is a somewhat misleading parameter because many of the control options
do not use capital equipment but rather operating and maintenance programs.
These examples clearly represent only a few of the cases that can be
evaluated. The results for other pollutants and emission factors, different
control functions, and battery subsets have not been examined. These are
among the many possibilities that remain for the users of the model.
Consideration is now being given to expanding the model concept to in-
clude wastewater, solid waste, and energy impact, and to expanding the scope
to include other processes.
162
-------
COKE OVEN OPTIMIZATION
OIJECTIVEl MINIMUM EMISSIONS RESTRICTION! I 573. MILLION ANNUALIZEB COST
BASELINE! COST AIJUSTED FOR EXISTING CONTROLS
POLLUTANT IIAP
BASE YEAR 1979
CONTROLLED EHI8BIONS
(LIS/TON COAL) (TONS/YEAR)
CONTROLLED COST
(MILLION IOLLARS)
CO
SOURCE
LARRY CAR CHAR6INS
COKE PUSHING
QUENCHING - CLEAN UATER
IOORS
TOPSI BE
COHIUSTION STACK - OLI
COKE HANDLING
COAL PREHEATER
COAL PREPARATION
COAL STORAGE YARD
PIPELINE CHARGING
REDLER CHARGING
HOT LARRY CAR CHARGING
IT-PRODUCTS PLANT
COHIUSTION STACK - NEU
WENCHING - DIRTY HATER
TSP
.01
2.00
.17
.08
.01
.24
1.00
.35
.30
.13
.00
.00
.00
.00
.03
.94
DSO
.0110
.0600
.0002
.1000
.0073
.0012
.0000
.4200
.0000
.0000
.0002
.0001
.0002
.3000
.0001
.0019
DAP
.0000
.0000
.0000
.0004
.0000
.0000
.0000
.0002
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0001
IEN
.0030
.0040
.0000
.0040
.0002
.0000
.0000
.0070
.0000
.0000
.0001
.0000
.0001
.2000
.0000
.0003
TSP
494
109443
4104
4378
328
9748
34731
1794
27343
8209
0
0
0
0
444
29341
ISO
543
4378
4
3473
410
43
0
2140
0
0
0
0
0
14419
2
58
DAP
0
2
0
32
1
0
0
0
0
0
0
0
0
0
0
2
IEN CONTROL SCHEME
248 NEU CAR, STEAN, BOOT
328 UNCONTROLLE8
0 IIVERTEI FLOU IAFFLE8
218 CLEAN AFTER EVERY CHARGE
8 NEU LIDS I CASTINGS
0 OVEN PATCHING
0 UNCONTROLLED
35 SCRUBBER, 95Z
0 UNCONTROLLED
0 UNCONTROLLED
0 OPERATION I MINT.
0 OPERATION > HAINT.
0 OPERATION I HAINT.
10944 UNCONTROLLED
0 OVEN PATCHING
7 BAFFLES
CAPITAL
303.4
.0
S.I
84.2
21.B
.0
.0
2.3
.0
.0
.0
.0
.0
.0
.0
9.3
AMNUALIZED
144.7
.0
14.0
182.4
43.0
88.4
.0
2.1
.0
.0
3.2
.7
.3
.0
20.1
3.3
TOTAL UNC. 18.7 3.33V .007 .744 308238 122349 337 97377
TOTAL CONTROLLED 3.3 .922 .001 .223 230443 29474 37 11790
429.3
342.3
EXISTING
TOTAL BATTERIES 214 TOTAL OVENS 12221
TOTAL CAPACITY 1094*4247 TONS COAL
74423000 TONS COKE
I iOT IN OPTIMIZATION
NEU
TOTAL DATTERIES 0
TOTAL CAPACITY
TOTAL OVENS 0
0 TONS COAL
0 TONS COKE
Figure 5. Example of minimizing BaP emissions for an annualized cost restriction.
-------
VOLATILIZED LUBRICANT EMISSIONS FROM STEEL ROLLING OPERATIONS
Written By
Charles Mackus
Kaushik Joshi
Presented at
Symposium on Iron and Steel
Pollution Abatement Technology
October 30 - November 1, 1979
Pick-Congress Hotel
Chicago, Illinois
164
-------
VOLATILIZED LUBRICANT EMISSIONS FROM STEEL ROLLING OPERATIONS
The paper presents the findings of tv.j reports prepared by Pacific
Environmental Services, Inc. (PES) for the Industrial Environmental
Research Laboratory of the U.S. Environmental Protection Agency (IERL,
EPA). The first report entitled "The Use and Fate of Lubricants,
Oils, Greases, and Hydraulic Fluids in the Iron and Steel Industry",
was the first study of lubricant usage within the steel industry and
its overall environmental impact. The report was concentrated in two
areas: (1) the development of correlations between lubricant usage
rates and the steel production capacity and the types of products
made; and (2) the preparation of total oil, grease, and hydraulic
fluid material balances for specific as well as typical integrated
steel plant.
Lubricant-type materials in the steel industry can generally be
broken down as greases, lubricating oils, process oils, and hydraulic
fluids. Due to the variety of operation and the amount and size of
the equipment used, no other industry employs a more complete range of
lubricants. High and low temperatures, shock loading, exposure to
water, and abrasive materials are some of the adverse conditions nor-
mally encountered. All types of motors, turbines, bearings, gears,
and drives must be properly lubricated to keep the machinery func-
tioning. A typical steel mill may have over 400,000 lubrication
points using as many as 150 different specifications for lubricants.
Grease is generally used under the following conditions: (1)
where there is no way to retain oil for the part being lubricated; (2)
when the lubricant must act as a seal to prevent the entrance of dirt;
(3) when a lubricant is seldom added; and (4) where speeds are low and
pressures are high. Oils are used for a wide variety of purposes.
Very simply, oils are used in all cases where operating conditions do
not dictate the use of grease. A basic list of common oils include:
165
-------
engine oils, extreme pressure oils, rust and oxidation inhibiting (R &
0) oils, rolling oils, and protective coating oils. Hydraulic fluids
are generally oil-based products and come in slightly less variety
than lubricating oils and greases. Some typical products in use in-
clude: invert emulsions, phosphate ester fluids, water-glycol fluids,
rust and oxidation inhibiting hydraulic oil, and extra-duty anti-wear
hydraulic oil.
The prime objective of the initial PES study was to analyze data
obtained from the steel industry, and to use this information to de-
velop material balance estimates identifying the usage and fate of
lubricants, oils,- greases, and hydraulic fluids. The major factors
which influence lubricant usage (and their fate) are the size, design,
and age of the steel mill equipment. Generally speaking, mill equip-
ment which is large and old requires substantially more lubricants
than would a smaller or newer piece of equipment. The lubrication
practices and application methods also vary widely from mill to mill.
The last major factor (the type of steel product produced) determines
the amount of lubricant used within an entire mill since the majority
of lubricant used (between 80 and 90 percent) in an integrated mill,
is used in hot and cold rolling processes.
These factors, as well as maintenance and housekeeping practices,
influence the amounts and types of lubricants applied, and to some
extent, the fate of these lubricants. Figure 1 depicts this
relationship by using a product parameter developed by PES. To obtain
a relative percentage of rolled strip and coiled products, the sum of
all hot strip, cold rolling, cold finishing and temper mill capacities
at each plant was divided by the sum of all shaping, rolling, and
finishing mill capacities. This parameter is expressed as a
percentage. Plants with large capacities to roll hot strip or cold
rolled products have product parameters of around 60 percent.
To develop a material balance over an integrated mill, in addition
to lubricant usage terms, output or loss terms must be identified and
166
-------
Figure 1. TOTAL USAGE RATE VS. PRODUCT PARAMETER
10.0
-a
(B
4->
O
9.0
8.0
7.0
6.0
E
O
5.0
4.0
3.0
C)
2.0
QUSSC-S.
1.0
HICAGO
J&L
O
ISER
OMILL A
BETHLEJHEM
O
INTERLAKE
USSC-GARY
O
S & T
O INLAND
10
20 30 40
PERCENT
PRODUCT PARAMETER
167
50
60
70
Least Squares Line
Y=0.057Xf2.53
r=0.580
N=10
p=90 to 95%
-------
quantified. Figure 2 illustrates both sides of a lubricant mass bal-
ance over an entire mill. Two input terms, virgin make-up, and re-
claimed or recycled lubricants, and hydraulic fluids enter the mate-
rial balance. Several output or loss terms are identified in the
figure, including: lubricant losses on the steel products shipped
from plant; oils and greases attached to mill scales which are stock-
piled or recycled to the sinter plant; lubricants, especially greases,
left in containers or lost during storage and handling; oils and
greases on trash and debris that are collected and disposed; lubri-
cants volatilized, burned, or consumed in various steel making and
shaping processes; and oils, greases, and hydraulic fluids in the
wastewater streams which are either discharged to waterways, recovered
and disposed, or are reclaimed. Oils, greases, and hydraulic fluids
are also collected by oil skimmers, in scale pits and wastewater
treatment facilities, and are treated on- or off-site in waste oil
reclamation facilities for re-use as lubricants, fuels, or road oils.
Quantifying each of these input and loss terms for each oil,
grease, or hydraulic fluid was determined to be an impossible task. A
given lubricant may be used in several areas or pieces of equipment
and may appear in several wastewater circuits. The wastewaters from
different areas of the plant are often combined, resulting in a blend
of oils, and greases which cannot be separated and traced back to
their source. The wastewater sampling and analysis methods that are
currently being used only determine the total quantity of oil, grease,
and hydraulic fluids. Several of the output or loss terms, although
recognized by the steel industry, have not been investigated to date
and quantitative data of any type is unavailable.
PES obtained sufficient lubrication data to develop balance
scenarios for several mills. Summaries are presented in Tables 1 and
2. As seen in Table 1, overall lubricant usage ranged from a low of
1.63 pounds of lubricant/ton of product, to a high of 9.99 Ib/ton with
the average being 4.81 Ib/ton. Also, note that the two mills with
168
-------
INPUT TERMS
purchased oils,
greases and
hydraulic fluids
reclaimed &
recycled
OUTPUT OR LOSS TERMS
-> on products
-> on mill scales
->to scale pile
->to sinter plant
left in containers or lost
in storage and handling
leaks and spills onto ground,
generally cleaned up and disposed
-> volatilized, burned, or
consumed in process
-> in sludges, trash and debris-
-> in wastewaters-
wastewater
treatment
->• drained, collected or skimmed-
reclaimed
lubricants
waste oil
reclamation
fuel
to disposal
discharged to
waterways
sludge disposal
road oils
Figure 2. LUBRICANT, OIL, GREASE, AND HYDRAULIC FLUID MATERIAL BALANCE ESTIMATE
-------
Table 1. USAGE AND PRODUCT DATA
STEEL MILL
USSC, Gary
USSC, South Chicago
Inland Steel
Youngstown Sheet & Tube
Bethlehem Steel
Jones & Laughlin
Republic Steel
Interlake
Kaiser Steel
Mill A
Units
PRODUCT
PARAMETER*
59
0
62
61
49
35
0
49
39
42
TOTAL OIL, GREASE &
HYDRAULIC FLUID USAGE
1.653
0.816
2.120
4.997
2.126
2.981
1.377
2.566
2.642
2.815
(3.306)
(1.632)
(4.240)
(9.993)
(4.252)
(5.961)
(2.754)
(5.131)
(5.284)
(5.630)
TOTAL OIL USAGE
1,473
0.146
2.069
3.671
1.381
1.677
0.780
2.674
2.311
2.920
(0.353.)
(0.035)
(0.496)
(0.880)
(0.331)
(0.402)
(0.187)
(0.641)
(0.554)
(0.700)
kg/1000 kg (Ib/ton) 1/1000 kg (gal/ton)
* The sum of all hot strip, cold rolling, cold finishing, and temper mill
capacities at each plant divided by the sum of all shaping, rolling, and
finishing mill capacities.
170
-------
Table 2. SUMMARY OF MATERIAL BALANCE LOSS TERMS
LOSS TERMS
On Products
On Mill Scales
Discharged to Waterways
In Sludge, Trash & Debris
To Reclaimers
Left in Containers
Leaks & Spills
Volatilized or Consumed
TOTAL
UNACCOUNTED FOR
USCC
GARY
3.6
1.1
7.1
28.5
53.8
0.9
0.7
3.7
99.4
0.6
INLAND
3.4
13.7
3.7
21.8
14.1
?
0.4+ ?
?
57.1
42.9
YS & T
1.5
3.6
11.7
46.8
12.1
?
0.2+ ?
?
75.9
24.1
BETH.
4.7
2.3
14.1
55.6
14.5
5.0
3.7
2.0
101.9
-
J & L
19.4
20.1
9.8
52.8
0
?
?
?
102.1
-
REP.
5.0
9.3
7.0
54.7
0
?
?
?
76.0
24.0
KAIS.
0.9
34.3
0.1
35.0
8.5
5.0
5.0
12.6
101.4
-
MILL
A
7.6
6.9
10.7
14.5
46.9
0.3
1.3
9.7
97.9
2.1
AVG.
5.8
11.4
8.0
38.7
18.7
2.8
1.9
7.0
RANGE
0.9 - 19.4
1.1 - 34.3
0.1 - 14.1
14.5 - 55.6
0.0 - 53.8
0.3 - 5.0
0.2 - 5.0
2.0 - 12.6
"TYPICAL
MILL"
6
11
8
39
19
3
2
7
95
5
Notes: 1. All numbers represent percentage of total lubricant, oil, grease, and hydraulic fluid input to the
steel mill.
2. Whole number values were selected for the "typical mill", reflecting the accuracy of available data
and estimates.
-------
product parameters of zero (No rolling mills) have the two lowest
lubricant usage figures. Table 2 summarizes the Material Balance Loss
Terms of the mills surveyed. Note that many of the mills did not have
data for potentially large loss term factors such as the quantities
volatilized or consumed.
The loss term percentages for a typical steel mill shown in the
last column of Table 2 were applied to the total oil, grease, and
hydraulic fluid usage rate estimated for a 4.0 x 10 ton/year capac-
ity plant. This usage rate was estimated to be about 1,200,000 lb/
month. Multiplying this usage rate by the loss term percentages, a
material balance estimate was derived for a typical mill. Table 3
summarized the potential pollution which could be generated by a typi-
cal 4.0 x 10 ton/year integrated mill. This estimate revealed that
of the total quantity of oil, grease, and hydraulic fluid used at a
typical plant, approximately 10 percent enters the environment as air
pollution, 9 percent as water pollution, and 44 percent as solid waste.
Since the first study showed that the majority of lubricant used
in the steel industry is used in steel rolling operations, a new study
was funded to refine the lubrication data and the above emission
estimates at these locations. As in the first study, data from nine
steel rolling operations were used to: (1) define the volatilized
portion of lubricants used in rolling operations; and (2) prepare
total oil, grease, and hydraulic fluid material balances for actual
and typical cold and hot rolling processes from data acquired by
questionnaires, mill visits, and emission source sampling. The major
difference between the first and second study was the addition of
actual source sampling. The remainder of this paper will focus on
that aspect. Sampling was conducted at three rolling operations, two
at Inland Steel Indiana Harbor Works, and one at United States Steel
Gary Works.
172
-------
Table 3. POLLUTION SUMMARY
LOSS TERM
On Products
On Mill Scales
Discharged to Waterways
In Sludge, Trash &
Debris
To Reclaimers
Left in Containers
Leaks & Spills
Volatilized or Consumed
TOTAL
PERCENT
6
' 6
<5
8
39
19
3
2
7
95
AIR
POLLUTION
_
4
-
-
-
*•
-
6
10
WATER
POLLUTION
_
1
8
-
-
-
-
-
9.
SOLID
WASTE
_
1
-
39.
1
1
2
-
44
— — •• - i
ASSUMPTIONS
No resulting pollution.
Of the 6% of oil in stockpiled mill
scale 1% is washed off and becomes water
pollution; 1% is permanently "stored".
Of the 5% of oil in recycled mill scale
4% is volatilized and the remainder is
combusted or captured
The fate of oil in landfills was not
considered
One percent sludge from the reclamation
process. The potential air pollution j
from fuel combustion was not considered.
One percent is in containers which are
discarded in a landfill.
All leaks and spills are "'eaned up
and disposed of in landfills.
One percent is combusted.
32% does not result in air or water
pollution or solid waste and 5% is
unaccounted for.
/
-------
The two mills sampled at Inland Steel were the 56 inch four-stand
tandem cold strip mill, and the 80 inch five-stand tandem cold strip
mill. Both of these mills had a central air evacuation system which
exhausted to a central stack (one stack for each mill). The evacua-
tion system is necessary since large amounts of water and oil would
condense on the rolling train if they were not removed. Also, cold
mill operators must work very near the rolling train, and the evacua-
tion system prevents oil and water from hindering the operator's abil-
ity to perform their assignments. Mist eliminators are located in
underground exhaust tunnels which aid in the elimination of any aero-
sols which may reach the evacuation systems induction (I.D.) fans.
The collection efficiency of the evacuation system was estimated by
PES to be 95 percent with the remaining five percent escaping to the
mill building atmosphere.
The sampling train used to sample the hydrocarbons in the exhaust
stack was basically a method five train with a Beckman 400 hydrocarbon
analyzer placed in series with the train. Continuous HC stack sam-
pling was conducted and measured on the HC analyzer for approximately
90 minutes in each of the cold mills. The actual average total hy-
drocarbon concentrations for 56 inch and 80 inch mills were determined
to be 60 and 57 ppm (as methane) respectively.
Grab samples were taken from each of the mills' stacks during the
period when continuous stack hydrocarbon sampling was being conducted.
The chemical analysis of these samples revealed that the majority of
hydrocarbons were aliphatic in nature. The aliphatic hydrocarbons
were further identified to be in the C-14 through C-38 range. The
majority of the aliphatic hydrocarbons emitted through the 56 inch
mill stack were in the C-16 to C-18 range while those of the 80 inch
mill were concentrated in the C-14 to C-28 range. These ranges cor-
respond to the chain lengths of the rolling solution prior to rolling
so it appears th ; little chemical decomposition occurs during the
rolling process.
174
-------
The relative hydrocarbon emissions concentrations appear to be
quite low. However, when you incorporate the large air volumes and
the approximate operating schedules of the 56 inch and 80 inch cold
strip mills, the total yearly volatilized hydrocarbon emission rate
becomes 50 and 64 ton/year respectively. For the 56 inch cold strip
mill 60 ton/year of volatilized hydrocarbons reduces to 20.8 percent
of all lubricants used on the mill exclusive of hydraulic oil. The 64
ton/year emission rate for the 80 inch cold strip is 26.9 percent of
the total mill lubricant usage exclusive of hydraulic oil. In both
cases, the volatilization term is the second largest term in the cold
strip mill lubrication mass balance, with waste oil drained and col-
lected being the single largest term (approximately 45 percent for
both mills).
The testing conducted at United States Steels' 84 inch hot strip
mill at the Gary Works facility was not as straight, forward as the
tests used on the cold strip mills. This hot strip mill has the capa-
bility to use rolling oil at the finishing stands. Unfortunately, no
rolling oil was used during the hot reduction operation due to winter
damage to the mills' coolant/lubricant application system.
A portable organic vapor analyzer (Century Systems Model OVA-108)
was used for the hydrocarbon emission monitoring. The exhaust from
the hot roll operations area vents directly to the ambient (building)
atmosphere. The hot mill complex is not equipped with any central air
evacuation system, and no control stack duct exists through which the
hot mill and exhaust air can be discharged into the outer atmosphere.
All vapors generated by the process eventually exit through roof moni-
tors located directly over each of the mill stands. Since there was
no central location to measure the generated vapors, readings were
taken at the top of each of the mills' finishing stands. The hydro-
carbon concentrations ranged between 7 ppm and 23 ppm as methane, with
the average ppm being 15.
175
-------
In an effort to quantify the total volatilization emissions from
this mill the following approach was taken. The rate of evaporation
of oil and water in the surface film on the strip depends on the fol-
lowing principle factors:
1. Surface area of strip.
2. Speed of the strip.
3. Temperature of the steel strip.
4. Thickness of the surface film.
5. Specific heat of the steel strip and rolling oil film.
6. Ambient air temperature.
PES engineers assumed the exit rate of the volatilized vapors was
approximately equal to the speed of the strip at each stand. This as-
sumption was made because the following conditions exist:
1. No central duct - since this is the case, a stack duct
which was the width of the.strip and the length of the
finishing stand was imagined.
2. Dilution air would tend to dilute the concentration.
However, this has a net negative effect on overall emis-
sions.
3. Hydrocarbons which escape the imagined duct could not be
accounted for. Again, this has a net negative effect on
overall emissions.
The actual speed of the hydrocarbon vapors is probably somewhat
less than the actual speed of the strip which would yield a net posi-
tive effect on overall hydrocarbon emissions. However, since the
dilution air and the vapor escape factors tend to reduce the hydrocar-
bon emissions at the point of sampling, the assumption of an escape
velocity equal to the speed of the strip at a concentration of 15 ppm
at the height of the strip was deemed acceptable. By using this as-
sumption, an annual hydrocarbon emission rate of 42 ton/year was ob-
tained for the 84 inch hot strip mill.
As a last exercise, emission factors for typical cold and hot
strip mills were developed. It should be stated that the designation
176
-------
of a typical mill is somewhat loose since mill designs and operating
practices vary greatly. The cold mill emission factors were estimated
to be accurate within a range of + 20 perce t.
In the case of hot strip mills, the amount of data was signifi-
cantly less than that available for cold strip mills. Therefore, the
emission factors may vary as much as an order of magnitude.
It was determined that for a typical cold strip mill approximately
0.34 Ib/ton of hydrocarbons (as methane) will be emitted under normal
operation. The emission factor for a hot strip mill which does not
utilize rolling oils was determined to be 0.019 Ib/ton of product.
These factors were expanded to show that for an average rolling opera-
tion (either hot or cold), approximately 3.7 percent of the lubricant
purchased for that operation is volatilized. The above calculation
assumes only one rolling operation is required to obtain a final pro-
duct. If one assumes each final product undergoes three rolling oper-
ations, then the total amount of lubricant volatilized is approximate-
ly 11 percent of the lubricants purchased for rolling operations.
177
-------
EMISSION FACTORS FOR OPEN DUST SOURCES
By
Chatten Cowherd, Jr.*
ABSTRACT
The information presented in this paper is directed to those inter-
ested in inventorying particulate emissions from open dust sources within
integrated iron and steel plants. The results of a recent field testing
program are used to refine previously developed emission factor equations
for generic source operations including vehicular traffic on unpaved and
paved surfaces, batch drop operations, continuous drop operations, and
wind erosion. Significant improvements in emission factor precision are
achieved through the following modifications: (a) addition of a wheel
number correction term to the equation for unpaved roads; (b) addition of
road shoulder and lane number correction terms to the equation for paved
roads; and (c) addition of a drop distance correction term to the equation
for continuous drop operations. Wind erosion test results indicate that
natural surface crusts are very effective in mitigating dust emissions, an
effect which is not taken into account in the previously developed emission
factor equation. Limited testing of chemical dust suppressants applied to
unpaved roads indicates a high initial control efficiency (exceeding 90%)
which decreases by more than 10% with the passage of 200 to 300 vehicles.
Consistent with the emission factor equation, the lowering of emissions is
reflected by the reduced silt (fines) content of the road surface material
after the application of chemical dust suppressants.
Midwest Research Institute, 425 Volker Boulevard, Kansas City,
Missouri 64110.
178
-------
INTRODUCTION
Iron- and steel-making processes, which are characteristically
batch or semicontinuous operations, generate substantial quantities of
fugitive (nonducted) emissions at numerous points in the process cycle.
There are numerous materials handling steps in the storage and preparation
of raw materials and in the disposal of process wastes. Additionally,
fugitive emissions escape from reactor vessels during charging, process
heating, and tapping.
Fugitive emissions in the iron and steel industry can be generally
divided into two classes—process fugitive emissions and open dust
source fugitive emissions. Process fugitive emissions include uncaptured
particulates and gases that are generated by steel-making furnaces,
sinter machines, and metal forming and finishing equipment, and that are
discharged to the atmosphere through building ventilation systems. Open
dust sources of fugitive emissions include such sources as raw material
storage piles, from which emissions are generated by the forces of wind
and machinery acting on exposed aggregate materials.
Fugitive emissions are especially difficult to characterize for the
following reasons.
1. Emission rates have a high degree of temporal variability.
2. Emissions are discharged from a wide variety of source con-
figurations.
3. Emissions are comprised of a wide range of particle sizes,
including coarse particles which deposit immediately adjacent to the
source.
The scheme for quantification of emission factors must effectively deal
with these complications.
In a recent study of fugitive emissions from integrated iron and
steel plants, Midwest Research Institute determined that open dust
sources (specifically, vehicular traffic on unpaved and paved roads and
storage pile activities) ranked with steel-making furnaces and sinter
machines as sources which emit the largest quantities of fine and suspended
particulate, taking into account typically applied control measures. ±J
It became evident that open dust sources should occupy a prime position
in control strategy development for fugitive particulate emissions
within integrated iron and steel plants. Moreover, preliminary analysis
of promising control options for both process sources of fugitive emissions
and open dust sources indicated that control of open dust sources has a
highly favorable cost-effectiveness ratio for particulate.
179
-------
The technical soundness of these conclusions and the foundation for
more detailed investigation rest on the availability of reliable particu-
late emission factors and particle size distributions for the sources
under consideration. This paper presents refined emission factors for
open dust sources developed from an expanded data base incorporating the
results of additional field testing within integraded iron and steel
plants.
Previously Developed Emission Factors
Since 1973, MRI has been engaged in a series of field testing
programs to develop emission factors for open dust sources associated
with agriculture and industry. To provide for the requirement that the
emission factors would be applicable on a national basis, at the outset
MRI analyzed the physical principles of fugitive dust generation to
ascertain the parameters which would cause emissions to vary from one
location to another. These parameters were found to be grouped into
three categories:
1. Measures of source activity or energy expended (for example,
the speed and weight of a vehicle traveling on an unpaved road).
2. Properties of the material being disturbed (for example, the
content of silt in the surface material on an unpaved road).
3. Climatic parameters (for example, number of precipitation-free
days per year on which emissions tend to be at a maximum).
By constructing the emission factors as mathematical equations with
multiplicative correction terms, the factors developed by MRI became
applicable to a range of source conditions limited only by the extent of
experimental verification. —*—/
The use of the silt content as a measure of the dust generation
potential of a material acted on by the forces of wind or machinery
was an important step in extending the applicability of the emission
factor equations to the wide variety of aggregate materials of industrial
importance. The upper size limit of silt particles (75 /Ltm in diameter)
is the smallest particle size for which size analysis by dry sieving is
practical, and this particle size is also a reasonable upper limit for
particulates which can become airborne. Analyses of atmospheric samples
of fugitive dust indicate a consistency in size distribution so that
particles in specific size ranges exhibit fairly constant mass ratios.^>-V
180
-------
In order to quantify source-specific emission factors for open dust
sources, MRI developed the "exposure profiling" technique, which uses the
isokinetic profiling concept that is the basis for conventional source
testing..^/ Exposure profiling consists of the direct measurement of the
passage of airborne pollutant immediately dov wind of the source by means
of simultaneous multipoint sampling over the effective cross section of
the fugitive emissions plume. This technique uses a mass-balance calcu-
lation scheme similar to EPA Method 5 stack testing rather than requiring
indirect calculation through the application of a generalized atmospheric
dispersion model.
The emission factors developed by MRI have been made specific to
particles smaller than 30 ^tn in Stokes diameter, so that emissions may
be related to ambient concentrations of total suspended particulate.
The upper size limit of 30 urn for suspended particulate is the approximate
effective cutoff diameter for capture of fugitive dust by a standard
high volume particulate sampler (based on a typical particle density of
2 to 2.5 g/cra)-—' It should be noted, however, that analysis of parameters
affecting the atmospheric transport of fugitive dust indicates that only
the portion smaller than about 5 /im in diameter will be transported over
distances greater than 5 to 10 km from the source.-5/
Test Results from Steel Plants
In 1977, as noted above, MRI performed field testing of open dust
sources at two integrated iron and steel plants (designated as Plants A
and E) in order to extend the applicability of the previously developed
emissions factor equations to open dust sources in the iron and steel
industry—' The sources tested were: (a) light-duty vehicular traffic on
unpaved roads; (b) heavy-duty vehicular traffic on unpaved roads; (c) mixed
vehicular.traffic on paved roads; (d) mobile stacking of lump iron ore;
(e) mobile stacking of pelletized iron ore; and (f) load-out of processed
slag into a truck with a front-end loader. These sources involved
materials handling equipment of a scale significantly larger than had
been tested previously. However, as indicated in Table 1, the expanded
data base incorporating the results of these tests still was not adequate
to provide emission factor equations with a high degree of quality as-
surance.
Therefore in 1978 a follow-up field testing program was conducted at
three integrated iron and steel plants (designated as Plants F, G, and H).
The following tests were performed:
Six tests of medium-duty vehicular traffic on untreated, unpaved
roads.
181
-------
TABLE 1. EMISSION FACTOR QUALITY ASSURANCE LIMITATIONS
(Effective March 1978)
Source
Typical
percent of
total controlled
TSP emissions^
Emission
factor QA
rating^'
Test data
limitations^/
Vehicular traffic on unpaved surfaces
Unpaved roads
Storage pile maintenance
Vehicular traffic on paved surfaces
Batch drop operations (loaders,
railcars, trucks, gantry/clamshell
buckets)
Continuous drop operations (stackers,
conveyor transfer stations, bucket
wheels)
Wind erosion
Storage piles
Exposed areas
65
13
14
B - dry conditions
C - annual conditions
B - urban traffic
C - industrial traffic
B - loaders
C - other
B - stackers
C - other
I; questionable measurement
accuracy
I; testing under dry condi-
tions only
I; probable effect of dust
resuspension from
underbodies
I
I; questionable measurement
accuracy
I; testing limited to dry
uncrusted surfaces
a/ Quality assurance rating scheme:
A » formulation based on statistically representative number of accurate field measurements (emissions, meteorology
and process data) spanning expected parameter ranges
B « formulation based on limited number of accurate field measurements
C » formulation or specific value based on limited number of measurements of undetermined accuracy or extrapolation
of B-rated data from similar processes
0 m estimate made by knowledgeable personnel
E - assumed value
b/ I » insufficient number of tests
182
-------
. Three tests of light-duty vehicular traffic on untreated, unpaved
roads.
. Two tests of light-duty vehicular traffic on treated unpaved roads.
. Six tests of mixed vehicular traffic on paved roads.
. Five tests of storage pile stacking of pelletized iron ore and coal.
Eight tests of wind erosion from coal storage piles.
Four tests of wind erosion from bare ground areas.
This mix of tests was derived from statistical analysis of source contribu-
tions and the relative reliabilities of previously developed emission factor
equations. To the extent possible, testing was restricted to periods with
moderate winds (5 to 15 mph) of constant mean direction, 3 or more days after
significant rainfall (accumulation exceeding 0.5 in.).
The primary tool for measuring fugitive dust generated from open dust
sources was the MRI Exposure Profiler. A vertical line grid of samplers was
used for measurement of dust emissions from vehicular traffic on unpaved and
paved roads. For the remaining sources, a cross-arm supporting additional
samplers was added so that both the vertical and lateral dimension of the
dust plume could be defined. At all times the MRI Exposure Profiler was
positioned within 5 m of the source with air samplers covering the effective
cross section of the fugitive dust plume.
Other equipment utilized in the testing included (a) cascade impactors
with cyclone preseparators for particle sizing, (b) high-volume air samplers
for determining upwind particulate concentrations, and (c) recording wind in-
struments utilized to determine mean wind speed and direction for adjusting
the MRI Exposure Profiler to isokinetic sampling conditions. A detailed
presentation of the testing methodology is provided elsewhere.—'
In order to determine the properties of aggregate materials being dis-
turbed by the action of machinery or wind, representative samples of the
materials were obtained for analysis in the laboratory. Unpaved and paved
roads were sampled by removing loose material (by means of vacuuming and/or
broom sweeping) from lateral strips of road surface extending across the
traveled portion. Storage piles were sampled to a depth exceeding the size
of the largest aggregate pieces.
Moisture contents of samples were determined in the laboratory by weight
loss after oven drying at 110°C, and texture was determined by standard dry
sieving techniques. The moisture content of an exposed aggregate material is
183
-------
dependent on its initial moisture content and on the precipitation and evap-
oration which occurs while the material is in place. Thornthwaite's P-E
Index—'' is a useful approximate measure of average surface soil moisture,
but is not suitable for freely draining aggregate stored in open piles.
REFINEMENT OF EMISSION FACTOR EQUATIONS
This section presents refined emission factor equations for:
(a) vehicular traffic on unpaved roads; (b) vehicular traffic on paved
roads; (c) storage pile formation by continuous load-in or stacking; and
(d) wind erosion of storage piles and bare ground areas. Refinements to
previously developed equations have been adopted as necessary to extend
the predictive capability of the equations to the expanded test data
bases without loss in precision. In this way, the quality assurance
(QA) ratings, as given in Figure 1, may be improved.
Vehicular Traffic on Unpaved Roads
Figure 1 shows the predictive emission factor equation for vehicular
traffic on unpaved roads, as derived by multiple regression analysis of
the test data shown in Table 2. The coefficient and the first two correc-
tion terms in Figure 1 are identical to the expression given in AP-42 as
follows:—'
0.6 (0.81 s)
which describes the emissions of particles smaller than 30 /^m in Stokes
diameter generated by light duty vehicles traveling on unpaved roads.
The weight correction term in Figure 1 was developed on the basis of
prior testing; however, the term was formerly raised to the 0.8 power.—/
Table 2 compares measured emissions with predicted emissions as
calculated from the equation given in Figure 1. In addition to the
test results from iron and steel plants, the results of testing of
traffic on haul roads at a taconite mine (Test Series I) , which was
performed as part of another study, have also been added to the data
base.—' As shown in Table 2, in the tests conducted on a previously
inactive road (Runs 1-1 through 1-5), emissions approached the predicted
values with successive tests. The test truck was loaded between Runs
1-3 and 1-4. Also, measured emissions for Runs 1-7 and 1-8 were signifi-
cantly lower than predicted, presumably because of the considerable rain-
fall on the days prior to testing.
184
-------
OPEN DUST SOURCE: Vehicular Traffic on Unpaved Roads
QA RATING: B for Dry Conditions
C for Annual Average O iditions
cc — 1
tl 1 .
7 f S \l M
7 \12Jl48J
/ W 1
\2Ji
| 0.7, ,0.5
l4J
I , 1 L r, /l/o!n Lrrr,
I 365 j k9/veh-km
EF=5.9 ^ ^j-
| |
^
Determined by profiling
of emissions from light-
duty vehicles on gravel
and dirt roads under
dry conditions.
I
Estimated factor to
account for mitigating
effects of precipitation
over period of one
year.
Determined by profiling of emissions from
medium-and heavy-duty vehicles on gravel
and dirt roads under dry conditions.
EF = suspended particulate emissions
s = silt content of road surface material
S = average vehicle speed
w = average number of wheels per vehicle
W = average vehicle weight
d = dry days per year
metrjc
non-metric
kg/veh-km Ib/veh-mi
% %
km/hr mph
tonnes
tons
Figure 1. Predictive emission factor equation for vehicular traffic
on unpaved roads.
185
-------
TABLE 2. PREDICTED VERSUS ACTUAL EMISSIONS (UNPAVED ROADS)
Rnad surface
Run
R-l
R-2
R-3
Type
Crushed
Limestone
R-8 |
R-10 Dirt
R-13
A- 14 ( Crushed
A-15 slag
E-l
E-2
Dirt
E-3J
F-21
Dirt/
F-22 crushed
F-23 1 slag
F-24 ^ Dirt/slag
F-25 (Cohere^lE./
G-27
G-28
G-29
G-30
G-31
G-32
I-l
1-2
1-3
1-4
1-5
Crushed
slag
Crushed
rock and
I/ glacial
till
1-7 ( Crushed
1-8 i'i/ rock
(taconite/
waste)
1-9
1-10
1-H
Crushed
rock
(TREX)£/
Silt
(%)
12
13
13
20
5
68
4.8
4.8
8.7
8.7
8.7
9.0
9.0
9.0
0.03
0.02
5.3
5.3
5.3
4.3
4.3
4.3
4.7
4.7
4.7
4.7
4.7
6.1
6.1
1.3
1.8
Average
vehicle speed
(km/hr)
48
48
64
48
64
48
48
48
23
26
26
24
24 ~
24
24
24
35
37
39
40
47
35
24
24
24
24
24
22
22
21
21
23
(mph)
30
30
40
30
40
30
30
30
14
16
16
15
15
15
15
15
22
23
24
25
29
22
15
15
15
15
15
13.5
13.5
13
13
24
Average
Emission factor—'
vehicle weight Average No. of Predicted^'
(tonnes)(tons) vehicle wheel
3
3
3
3
3
1
64
64
31
31
21
3
3
4
3
3
15
11
8
13
7
27
61
61
61
142
142
107
106
100
102
115
3
3
3
3
3
3
70
70
34
34
23
3
3
4
3
3
17
12
9
14
8
30
67
67
67
157
157
118
117
110
112
127
4.0
4.0
4.0
4.5
4.0
4.0
4.0
4.0
9.4
8.3
6.4
4.0
4.0
4.1
4.0
4.0
11.0
9.5
7.8
8.5
6.2
13.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
Actual
s (kg/VKT)(lb/VMT)(kg/VKT)(lb/VMT)
1.7
1.8
2.4
2.9
0.93
9.3
6.0
b.O
4.7
5.1
3.4
0.62
0.62
U.76
±1
d/
3.0
2.3
1.8
2.1
1.4
3.9
3.5
3.5
3.5
6.4
6.4
6.1
6.1
d/
d/
d/
5.9
6.4
8.5
10.4
3.3
33.0
21.4
21.4
16.7
18.0
12.0
2.2
2.2
2.7
d/
d/
10.7
8.1
6.3
7.5
6.1
14.0
12.4
12.4
12.4
22.6
22.6
21.6
21.5
d/
d/
d/
1.7
1.9
2.2
2.3
1.1
9.0
6.0
6.5
3.8
3.4
4. 1
0.84
0.48
0.65
0.021
0.10
3.4
2.0
1.6
2.4
1.4
4.5
1.0
2.1
4.1
5.1
7.0
3.3
3.3
0.56
0.65
1.0
6.0
6.8
7.9
8.1
3.9
32.0
21.5
23.0
13.6
12.2
14.5
3.0
1.7
2.3
0.073
0.36
12.0
7.2
5.6
8.7
5.1
16.0
3.7
7.5
14.5
18.1
25.0
11.6
11.0
2.0
2.3
3.6
Predicted
-r actual
0.98
0.94
1.08
1.29
0.85
1.03
1.00
0.93
1.23
1.47
0.83
0.73
1.29
1.19
0.84
1.13
1. 1:
0.87
0.99
0.88
3.3b
1.66
0.86
1.25
0.90
1.86
1.85
-
&l Particles smaller than 30 urn in Stokes diameter, based on actual density of silt particles.
b_/ Based on revised MRI emission factor equation.
£/ Tests performed on treated road (see text).
d_/ Equation not applicable.
e/ Test Series I-l through 1-5 performed on previously inactive road.
jj Tests performed on day following 2 days of rain totaling 1.13 in.
£/ Assumed value.
186
-------
The wheel correction term appears in the emission factor equation
for the first time. The need for this term was indicated by the fact
that for Test Series E and G, the emission factor equation without a
wheel correction term consistently underpredicted the measured factors.
This appeared to be due to the effect of 10- and 18-wheel trucks, which
comprised a substantial number of the passes in those tests. In all
other test series, the vehicle mix was dominated by four- and six-wheel
vehicles.
Excluding Test Series I except for Run Nos. 1-3 and 1-5, the revised
emission factor equation presented in Figure 1 predicts actual test re-
sults with a precision factor of 1.48. (The precision factor (f) is de-
fined such that the 95% confidence interval for a predicted emission fac-
tor value (P) extends from P/f to Pf.) By comparison, the precision fac-
tor for the unrevised equation is 1.66.
As stated above, limited testing of the effects of a chemical dust
suppressant was also conducted. Coherex®(a petroleum-based emulsion)
was used to treat a dirt/slag surfaced service road traveled by light-
and medium-duty vehicles at an integrated iron and steel plant. Coherex®
was applied at 10% strength in water.
Figure 2 shows a plot of measured dust control efficiency as a
function of the number of vehicle passes following application of the
road dust suppressant. Control efficiency was calculated by comparing
controlled emissions with uncontrolled emissions measured prior to road
surface treatment. As indicated, the effectiveness of the road dust
suppressant was initially high but began to decay with road usage. It
should also be noted that the apparent performance of Coherex® was
negatively affected by tracking of material from the untreated road
surface connected to the 100-m treated segment.
Figure 2 also shows the results obtained from the similar testing
of another chemical dust suppressant at a taconite mine.— TREX (ammonium)
lignin sulfonate—a water soluble by-product of papermaking) was applied
to the waste rock aggregate comprising the surface of a haul road. A 20
to 25% solution of TREX in water was sprayed on the road at a rate of
0.08 gal./sq yard of road surface.
Once again the effectiveness of the dust suppressant was found to
be initially high, but decayed with road usage. According to taconite
mine personnel, the binding effect of TREX can be partially restored by
the addition of water to the road surface.
187
-------
EFFECTIVENESS OF ROAD DUST SUPPRESSANTS
oo
oo
100
^ 90
E
u 80
Z
LLJ
U
2: 70
O
C£.
\—
Z
O
60
50
40
VEHICLE TYPE
DUST SUPPRESSANT
O Haul Truck
A Light- Duty Vehicles
Lignin Sulfonate
Co he rex
100 120 140 160 180 200 220 240 260
VEHICLE PASSES FOLLOWING TREATMENT
280 300
Figure 2. Effectiveness of road dust suppressants.
-------
With regard to the effects of natural mitigation of road dust
emissions, the final term in the emission factor equation for traffic on
unpaved roads (Figure 1) is used to reduce emissions from dry condi-
tions to annual average conditions. The simple assumption is made that
emissions are negligible on days with measurable precipitation and are
at a maximum on the rest of the days. Obvio sly, neither assumption is
defendable alone; but there is a reasonable balancing effect. On the
one hand, 0.01 in. of rain would have a negligible effect in reducing
emissions on an otherwise dry, sunny day. On the other hand, even on
dry days, emissions during early morning hours are reduced because of
overnight condensation and upward migration of subsurface moisture; and
on cloudy, humid days, road surface material tends to retain moisture.
Further natural mitigation occurs because of snow cover and frozen
surface conditions. In any case, further experimentation is needed to
verify and refine this factor.
Vehicular Traffic on Paved Roads
Figure 3 shows the predictive emission factor formula for vehicular
traffic on paved roads. As indicated, the coefficient and the first- two
correction terms were determined by field testing of emissions from
traffic consisting primarily of light-duty vehicles on urban arterial
roadways and on a test strip that was artificially loaded with surface
dust in excess of normal levels.— The vehicle weight correction term was
added by analogy to the experimentally determined factor for unpaved
roadways, and more testing is needed to confirm the validity of this
correction term. The number of lanes comprising the traveled portion of
the road and over which the surface dust loading is distributed was
added as a correction term to account for the fact that emissions increase
in proportion to surface dust loading.^-'
The industrial road correction factor was added Jo the emission
factor equation because measured emissions from medium-duty and heavy-
duty vehicles traveling on paved roadways at both Plant E (tested previously)
and plant F were substantially in excess of the predicted levels without
such a correction term. There are several plausible explanations for
the increase in dust emissions from paved roads within integrated iron
and steel plants as compared to urban roads. Paved roads within inte-
grated iron and steel plants are typically bordered by unpaved surfaces
and there are no curbings to prevent traffic from traveling on these
surfaces. Therefore, additional dust generation may result from: (a) re-
suspension from vehicle underbodies of dust accumulated during travel over
unpaved surfaces; (b) emissions from unpaved shoulders generated by the
wakes of large vehicles; and (c) emissions from unpaved shoulders during
passage of two large vehicles.
189
-------
OPEN DUST SOURCE: Vehicular Traffic on Paved Roads
QA RATING; B for Normal Urban Traffic
C for Industrial Plant Traffic
EF = 0.026IH
M s 1
-JlToj
/-M
\ 280 1
/W \
(TJ)
0.7
kg/veh-km
EF = 0.090I ( —
IUI
Determined by profiling of
emissions from traffic (mostly
light-duty) on arterial road-
ways with values for s and L
assumed.
Determined by profiling of emissions
from industrial plant traffic yielding
higher than predicted emissions,
presumably due to resuspension of
dust from vehicle underbodies and
from unpaved road shoulders.
W
3
W IL / L
-z-1 Ib/veh-mi
1
L
Assumed by analogy
to experimentally
determined factor
for unpaved roads.
Determined by profiling of emissions from
light-duty vehicles on roadway which was
artificially loaded with known quantities
of gravel fines and pulverized topsoil.
EF = suspended particulate emissions
I = industrial road augmentation factor (see text)
n = number of traffic lanes
s = silt content of road surface material
L = surface dust loading on traveled portion of road
W = average vehicle weight
metric
non-metric
kg/veh-km Ib/veh-mi
kg/km
tonnes
Ib/mi
tons
Figure 3. Predictive emission factor equation for vehicular traffic
on paved roads.
190
-------
Also, there may be a wheel effect similar to that indicated for
unpaved roads. Resuspension of dust from vehicle underbodies was visually
evident at Plant E as the heavy-duty vehicles traveled from an unpaved
area onto the paved roadway.
Quantification of these phenomena would require substantial additional
testing with detailed analysis of site conditions and traffic patterns
at test sites. For now, it is suggested that a multiplier of 7 be used
with the emission factor equation when Item 1 above is readily observed
and that a multiplier of 3.5 be used when the paved road (usually without
curbs) is bordered by unpaved and unvegetated shoulders. These factors
were determined by regression analysis of the test data for Plants E and F.
Table 3 compares measured emissions with predicted emissions as
calculated from the equations given in Figure 3. The revised emission
factor equation predicts actual test results with a precision factor of
3.31. This is a marked improvement over the precision factor of 14.1
associated with the unrevised equation.
It should be noted that the emission factor for re-entrained dust
from paved roadways contains no correction term for precipitation.
Although emissions from wet pavement are reduced, increased carryover of
surface material by vehicles occurs during wet periods, and emissions
reach a maximum when the pavement dries. More testing would be helpful
in analyzing the net effects of precipitation on re-entrained dust
emissions.
Storage Pile Formation by Continuous Load-In (Stacking)
Figure 4 gives the predictive emission factor equation for storage
pile formation (load-in) by means of a translating conveyor stacker.
The equation was originally developed from the results of field testing
of emissions from the stacking of pelletized and lump iron ore at Plant A.—
The effect of wind speed on emissions occurs presumably because of the
increased atmospheric exposure of suspendable particles during the drop
from the stacker to the pile.
An additional adjustment term containing drop distance has been
added to the emission factor equation. It was assumed that emissions are
proportional to drop distance, accounting for the additional energy
released on impact and the greater time of exposure during the drop.
191
-------
TABLE 3. PREDICTED VERSUS ACTUAL EMISSIONS (PAVED ROADS)
Road
surface dust
Loading No. of
excluding curbs?/ traffic
Run
P-9
P-10
P-14
E-7
E-8
P-3,
P-5,
P-6
P-15, ^
P-16
F-13 ^
/
F-14 >
F-15 J
F-16 \
F-17 >
(
F-18 f
Type
Pulver-
ized
topsoil^
Gravel^
( Iron and
steel)
Plant E
Urban
arterial
site IS-1
Urban
arterial
site 2£/
Iron and
steel
plant
Iron and
steel
plant
(kg/km)
1,990
809
1,890
225
225
45.1-
42.0-
57.2
57.2
57.2
629
629
629
a/ Loading distributed over
b/ Particles smallo
c/ Based on revised
r than 30
(Ib/mlle)
7,060
2,870
6,700
800
800
' 1*P
' wf-'
203
203
203
2,230
2,230
2,230
lanea
4
4
4
2
2
4
4
2
2
2
2
2
2
traveled portion of
HTH in Stoke
Silt
m
45
92
23
5.1
5.1
10*'
10*'
13.2
13.2
13.2
6.8
6.8
6.8
Average
I vchlc le
(Industrial weight
multiplier) (tonnes)
1
1
1
7
7
I
1
1
1
1
3.5
3.5
1
road, I.e., traffic
s diameter based
on actual
3
3
3
6
7
3
3
7
5
5
12
11
5
lanes .
density of
Emission factors—
Predicted^-/ Actual
(tons) (kg/VKT) (Ib/VMT) (kg/VKT) (lb/VMD
3
3
3
7
8
3
3
8
5
5
13
12
5
silt
0
0
0
0
0
0
0
0
0
0
0
0
0
82 2.9 1.0 3.7
.68 2.4 0.59 2.1
.39 1.4 0.13 0.46
.26 0.93 0.21 0.76
.29 1.02 0.28 1.0
.0039 0.014 0.0042 0.015
.0037 0.013 0.0037 0.0130
.096 0.34 0.16 0.58
.068 0.24 0.056 0.20
.068 0.24 0.045 0.16
.76 2.7 0.70 2.5
.70 2.5 0.48 1.7
.11 0.39 0.14 0.48
Predicted
-T Actual
0.78
1.14
3.04
1.22
1.02
0.93
1.00
0.59
1.20
1.50
I. OS
1.47
0.81
particles.
MRI emission factor equation.
oadvay artificially loaded.
c/ Four-lane roadway with traffic count of about 10,000 vehicles per day, mostly light-duty.
f/ Est imated valUP.
-------
OPEN DUST SOURCE: Storage Pile R motion by Means of
Conveyor Stacker
QA RATING: B
cc -
tr -
1 1 / 1 9 2/1 ^ J
r\ nrvr\ort •^'>*-«i/« «/
-0.00090 , 2
kg/ tonne
EF = 0.0018
I
Ib/ton
I
Determined by profiling of emissions
from pile stacking of pel let! zed and
lump iron ore and coal.
EF = suspended particulate emissions
s = silt content of aggregate
M = moisture content of aggregate
U = mean wind speed
H = drop height
metric
kg/tonne of
material transferred
m/sec
m
non-metric
Ib/ton of
material transferred
mph
ft
Figure 4. Predictive emission factor equation for storage pile
formations by means of conveyor stacker.
193
-------
Table 4 compares measured emissions with predicted emissions as
calculated from the equation given in Figure 4. The revised emission
factor equation predicts actual test results with an improved precision.
However, the sample size remains too small for meaningful statistical de-
termination of the precision factor.
Addition of the drop distance correction term aids significantly in
predicting the results of Runs H-10 through H-12 although a large discrepancy
remains for the first two of these runs. This may be due to lack of
representativeness of the pellet moisture values for these runs. The
pellets stacked during these runs comprised the last portion of a barge
shipment, and moisture variations may have been substantial if water had
collected in the bottom of the ship hold. The pellets were observed to
be unusually wet when the samples were taken.
Wind Erosion
Twelve tests of wind erosion emissions were performed utilizing a
portable wind tunnel with a specially designed isokinetic sampling system.
Eight tests were performed on the upper flat surface of an inactive coal
storage pile—three tests of one section of undisturbed (crusted) surface
and five tests of a disturbed section. This was followed by two tests of
the flat ground surface (undisturbed) adjacent to a dolomite storage pile
and two tests of disturbed prairie soil in the same area. Both mass emis-
sion rates and particle size distributions were measured as a function of
tunnel wind velocity. The wind erosion testing procedures are described
elsewhere.—'
The results of the wind erosion testing indicated that natural surface
crusts are very effective in mitigating suspended dust emissions. In ad-
dition, test data showed that a given surface has a finite potential for
erosion prior to additional mechanical disturbance. As expected, erosion
rates increase with wind velocity and decrease with erosion time as erodible
material is depleted. Good agreement between measured emissions from an
uncrusted coal surface and the previously developed emission factor equation
only after the equation was modified to reflect the wind speed dependence of
emission rate.— Although the mitigating effect of surface crusted should
be reflected in the reduced silt content of the surface material, which is
taken into account in the previously developed emission factor equation, it
is impossible to measure silt without destroying the surface crust. There-
fore, an alternative approach is needed in modifying the equation to apply
to crusted surfaces.
CONCLUSIONS
Based on an expanded data set of 24 tests, the revised MRI emission
factor equation for traffic-entrained dust from unpaved roads predicts
194
-------
TABLE 4. PREDICTED VERSUS ACTUAL EMISSIONS (LOAD-IN BY STACKER)
vo
Emission factor—' x 10
Aggregate
Run
Type
A- 8 ] Iron
ore
pellets
A-ll \
A-12
1 Lump
• iron
1 ore
A-13 f
H-10\
H-ll
Iron
ore
pellets
H-12
F-19 Coal
Silt Moisture
(%) (%)
4.
4.
2.
11.
19.
1.
1.
1.
5.
8 0.64
8 0.64
8 2.^
9 4.3
1 4.3
4 2.6
8 3.5
7 3.4
9 4.8
Drop
distance
(m)
3.0
1.5
4.5
3.0
3.5
9.0
11.0
12.0
5.0
Predicted—'
Wind speed
(m/sec)
1.0
2.0
0.8
0.8
1.0
0.7
1.8
2.7
1.3
(mph)
2.3
4.5
1.8
1.8
2.2
1.5
4.0
6.0
3.0
(kg/
tonne)
3.9
3.7
0.27
0.16
0.38
0.13
0.31
0.50
0.18
(lb/
ton)
7.8
7.5
0.54
0.33
0.76
0.27
0.62
1.0
0.37
Actual
(kg/
tonne)
1.1
25.0
0.26
0.19
0.12
1.1
1.4
1.1
0.070
(lb/
ton)
2.3
5.0
0.53
0.38
0.25
2.3
2.9
2.3
0.14
Predicted
-4- actual
3.39
1.50
1.02
0.87
3.04
0.12
0.21
0.43
2.64
a/ Particles smaller than 30 jum in Stokes diameter based on an adjusted density of 2.5 g/cm^;
O
multiply emission factor values by 10 to obtain units given.
b_/ Based on revised MRI emission factor equation.
c_/ Estimated value.
-------
measured emission factors with a precision factor of 1.48 as compared to
a precision factor of 1.66 for the unrevised equation. The addition of
a correction term related to the average number of wheels per vehicle
reduced the mean prediction error, as suggested by the clear tendency
of the unrevised equation to underpredict measured emission factors when
the test road was traveled by a substantial portion of 10- and 18-wheel
vehicles rather than 4- and 6-wheel vehicles.
Limited testing of chemical dust suppressants for industrial unpaved
roads indicates a high initial control efficiency (exceeding 90%) which
decreases by more than 10% with the passage of 200 to 300 vehicles.
Consistent with the emission factor equation, the lowering of emissions
is reflected by the reduced silt content of the road surface material
after the application of chemical dust suppressants. Additional testing
is needed to better quantify the performance of road dust suppressants.
Testing is also needed to verify and/or refine the emission factor
adjustment term which accounts for climatic mitigation.
The expanded test data set for traffic-generated dust from paved
roads indicates that the unrevised MRI emission factor equation consistently
underpredicts emissions (by up to a factor of 7) for industrial paved
roads. This is thought to be due to the additional dust generation from
unpaved areas adjacent to the paved surface. Incorporation of emission
factor correction terms which account for emissions from unpaved shoulders
and for the number of traffic lanes, improves the precision factor from
14.1 to 3.31.
Modification of the MRI emission factor equation for continuous
drop operations (translating conveyor stacker) by the addition of a
linear correction term involving drop distance aids in improvidng the
predictive capability of the equation. However, predictive errors
remain significant, which indicates effects of complex physical phenomena
not accounted for in the emission factor equation.
The results of the wind erosion testing indicate that natural surface
crusts are very effective in mitigating suspended dust emissions. This
effect is not taken into account in the previously developed emission fac-
tor equation because the measurement of silt destroys the surface crust.
In addition, test data show that a given surface has a finite potential for
erosion prior to additional mechanical disturbance. As expected, erosion
rates increase with wind velocity and decrease with erosion time, as erodible
material is depleted.
ACKNOWLEDGEMENT
The work upon which this paper is based was performed in part pursuant
to Contract No. 68-02-2609 with the U.S. Environmental Protection Agency
(Robert V. Hendriks, Project Officer).
196
-------
REFERENCES
1. Bohn, R., T. Cuscino, Jr., and C. Cowherd, Jr. Fugitive Emissions From
Integrated Iron and Steel Plants. Final Report, Midwest Research Insti-
tute for U.S. Environmental Protection Agency, Publication No. EPA-600/
2-78-050, March 1978.
2. Cowherd, C., Jr., K. Axetell, Jr., C. M. Guenther, and G. Jutze. Develop-
ment of Emission Factors for Fugitive Dust Sources. Final Report, Midwest
Research Institute for U.S. Environmental Protection Agency, Publication
No. EPA-450/3-74-037, NTIS No. PB 238262/AS, June 1974.
3. Cowherd, C., Jr., C. M. Maxwell, and D. W. Nelson. Quantification of
Dust Entrainment From Paved Roadways. Final Report, Midwest Research
Institute for U.S. Environmental Protection Agency, Publication No. EPA-
450/3-77-027, July 1977.
4. Cowherd, C., Jr. Measurement of Fugitive Particulate. Second Symposium
on Fugitive Emissions Measurement and Control, Houston, Texas, May 1977.
5. Cowherd, C., Jr., and C. M. Guenther. Development of a Methodology and
Emission Inventory for Fugitive Dust for the Regional Air Pollution
Study. Final Report, Midwest Research Institute for U.S. Environmental
Protection Agency, Publication No. EPA-450/3-76-003, January 1976.
6. Cowherd, C., Jr., R. Bohn, and T. Cuscino, Jr. Iron and Steel Plant Open
Source Fugitive Emission Evaluation, Final Report, Midwest Research In-
stitute for U.S. Environmental Protection Agency, Publication No. EPA-
600/ 2-7 9-103, May 1979.
7. Mann, C. 0., and C. Cowherd, Jr. Fugitive Dust Sources, Compilation of
Air Pollution Emission Factors. Section 11.2, U.S. Environmental Protec-
tion Agency, Publication AP-42, NTIS No. PB 254274/AS, December 1975-
8. Cuscino, T., Jr. Taconite Mining Fugitive Emissions Study. Final
Report, Midwest Research Institute for Minnesota Pollution Control
Agency, June 7, 1979.
9. Cowherd, C., Jr. Development of Emission Factors for Wind Erosion of
Aggregate Storage Piles. Paper No. 79-11.1 presented at the Annual
Meeting of the Air Pollution Control Association, Cincinnati, Ohio,
June 1979.
197
-------
ESTIMATING FUGITIVE DUST CONTRIBUTIONS TO
AMBIENT PARTICULATE LEVELS IN THE VICINITY OF
STEEL MILLS BY USE OF A SNOW COVER CRITERION
By
Donald C. Lang
Inland Steel Company
3210 Watling Street
East Chicago, Indiana 46312
And
David B. Smith
Equitable Environmental Health, Inc.
1420 Renaissance Drive
Park Ridge, Illinois 60068
198
-------
ABSTRACT
Suspended particulate measurements in Lake County Indiana
were analyzed to determine the impact of two different source categories
common in steel mills. These are ground-based non-traditional fugitive
dust sources such as unpaved road traffic, material handling operations,
storage piles, construction activity; and the more traditional sources
such as stack emissions and industrial process fugitive particulate
emissions.
The climatology and geography of the study area offer a perfect
opportunity to evaluate the significance of the non-traditional sources.
Three major integrated steel mills are situated along the southern shore
of Lake Michigan with an open water fetch to the north and northeast.
Under persistent onshore wind conditions, incoming background particulate
levels are relatively low and levels measured near the mills are largely
attributable to mill emissions.
Snow cover is common in Lake County winters and it suppresses
fugitive dust emissions. Accordingly, under persistent onshore wind
conditions, presence of snow cover was used as a criterion to assess the
significance of the fugitive dust component of particulate emissions.
Suspended particulate data for Lake County were statistically
analyzed first using broad meteorological and seasonal categories;
concentrations on days with snow cover were found to be significantly
lower than on days with bare ground. Limiting the analysis to cold
season days with winds confined to a narrow onshore sector and
moderate speeds clearly showed that concentrations were stratified by
ground conditions.
Results of the narrowly-focused analysis indicate that, without
snow cover and immediately downwind of the steel mills under the range
of conditions studied, fugitive dust sources account for about half of
measured levels, traditional sources account for about a third and
background is responsible for the remaining one-fifth of total measured
levels. The average contribution of traditional sources varies with
location from 20 to 50 percent of total levels or from about 25 to 100
micrograms per cubic meter while the average contribution of fugitive
dust sources varies with location from about 20 to 70 percent of total
levels or from about 15 to 150 micrograms per cubic meter. These
results strongly support air quality simulation modeling of mill
emissions which indicates a comparable relationship between the
contributions of traditional and fugitive dust sources.
199
-------
ESTIMATING FUGITIVE DUST CONTRIBUTIONS TO
AMBIENT PARTICULATE LEVELS IN THE VICINITY OF
STEEL MILLS BY USE OF A SNOW COVER CRITERION
Fugitive dust has been generally acknowledged as a source of
ambient particulate levels recorded by hi-vol samplers (References 1, 2).
Fugitive dusts in industrial areas become airborne by processes such as
(1) vehicles traveling over roads; (2) material handling operations such as
loading and unloading of stock piles; (3) construction activity; and (4) tur-
bulent eddy entrainment of dust over exposed ground.
The industrial geography and climatology of Lake County
Indiana offer a unique opportunity to assess the significance of these sources
on ambient total suspended particulate (TSP). Three major integrated steel
mills are situated along the south shore of Lake Michigan in Lake County
as shown on Figure 1. Local pollution control agencies operate hi-vol
samplers on an every-sixth-day schedule at sites plotted in Figure 1. A
detailed layout of storage piles and unpaved roads in one of the mills is
shown on Figure 2. Advantage was taken of the local climatology and
source-receptor geometry to establish the contribution of steel mill
fugitive dust sources.
Traditional sources of suspended particulate in the steel mills
can be expected to emit for the most part 24 hours a day, 365 days a year
consistent with continuous operations characteristic of major steel mills.
However, fugitive dust emissions are greatly affected by the weather.
Snow cover effectively suppresses fugitive dust emissions.
Furthermore, existence of snow cover typically is an indication that
exposed surfaces are wet or frozen--a state which also helps suppress
fugitive dust emissions. Studies conducted by EPA Region V support this
effect in showing a reduction of up to 90 percent in the concentration of
certain constituents of typical ambient particulate levels (Reference 3),
when the ground is snow covered. But while snow cover reduces or
eliminates fugitive dust emissions, it has little effect on the emission rate
of traditional sources.
Because snow cover eliminates or at least inhibits fugitive dust
emissions and is frequent in the winter months in northern Lake County,
analyses of TSP data were made using snow cover as an indicator for
200
-------
conditions under which the fugitive dust component of particulate emissions
would be suppressed or eliminated. A statistical test metholology was
established to determine whether significant ' ifferences in concentration
data could be found on days when the ground was snow covered in comparison
to levels found on days when the ground was bare.
The data base for this analysis consisted of measurements of
24-hour average TSP concentrations dating back to 1974 from a network of
hi-vol samplers in northern Lake County and meteorological data from
Midway Airport in Chicago. The principle meteorological data from Midway
Airport was snow cover reported at 0600 CST and hourly winds and weather
conditions. Days on which precipitation other than snow was recorded were
eliminated from the analysis because precipitation might not only restrict
fugitive dust emissions but also wash out categories of particulate emissions
from the atmosphere and hence distort the results.
The analytical methodology consisted of constructing a null
hypothesis in conventional statistical terms. Table 1 summarizes the
hypothesis formulation. The question can be simply stated as follows: Is
the average TSP concentration on days when the ground is snow covered
equal to the average TSP concentration on days when the ground is bare?
Again, the basic assumption is made that traditional particulate sources
emit and affect ambient levels regardless of ground cover while, on the
other hand, fugitive dust emissions are suppressed or eliminated when
the ground is snow covered.
The analysis was restricted to colder months because of the
reduced likelihood of lake breeze circulation influences and because back-
ground particulate levels in winter are inherently lower than those in summer
for several reasons. First, photochemical and other atmospheric trans-
formation processes are more effective in producing particulate levels from
conversion of gaseous pollutants to particulate in summer compared to winter
and, second, more frequent precipitation in winter reduces particulate levels
by washout and rainout.
A statistical test was conducted on a data sample assembled
under the following procedure:
1. sampling dates were restricted to the months of October
through May;
2. resultant wind direction on the sampling date had to range
201
-------
from north through northeast in which cases winds travel
off Lake Michigan across the steel mills towards the
samplers; and,
3. a date with snow cover was defined as one on which more
than 1 inch of snow depth was reported at Midway at 0600
(CST) on the days before, during and after a TSP measure-
ment.
The results summarized in Table 2 clearly indicate that
recorded TSP levels near the steel mills were lower when the ground was
snow covered compared to days when the ground was bare, apparently as a
result of suppression or elimination of fugitive dust emissions by snow
cover and/or frozen surfaces.
The analysis was taken one step further to assess the magnitude
of contribution to recorded levels of the traditional and fugitive dust emissions
from steel mills along the lakeshore by comparing in greater detail levels
recorded at the nearest monitors on days with snow cover and days with
bare ground.
The refined analysis was based on the key assumption that, at
monitors near the lakeshore steel mills during the colder season days with
north to northeast winds, concentrations recorded with snow cover resulted
exclusively from traditional sources plus background while, with bare ground,
concentrations were comprised of an identical contribution from traditional
sources plus background and an additional contribution from fugitive dust
emissions. The exact seasonal and meteorological criteria for selecting
dates for this refined analysis were:
1. time period of October through May;
2. all wind directions during the entire day restricted to
the range of 360° through 060° (north through northeast);
3. mean wind speed during the 24-hour period restricted to
the range of 10 to 20 miles per hour (to avoid wide
disparity in dispersion conditions); and,
4. snow cove^ restrictively defined as a minimum snow
depth of 3 jiches plus an increase in depth within 2 days
of the TSP measurement to insure fresh snow cover.
202
-------
This highly restrictive set of conditions for selecting a sample
narrowed the number of days down to 4 with snow cover and 7 with bare
ground. These data are summarized in Table 3.
Background on these days was estimated on the basis of
recorded levels at monitoring stations at the South Water Filtration Plant
in Chicago, which has an open exposure to Lake Michigan, and the Wirt
School which lies at the extreme east end of Gary a half-mile south of the
lakeshore with no significant sources between the monitor and the lakeshore.
Again, the analytical methodology applied here is based on the following
assumptions:
(1) on snow cover days, each recorded TSP concentration
consisted of contributions only from traditional sources
and background;
(2) on bare ground days, each recorded TSP level consisted
of a contribution from traditional sources assumed equal
to the average level on snow cover days, plus background
and an increment from fugitive dust sources.
Background was assumed to be the arithmetic average of the
concentrations recorded at the two background stations along the lakeshore.
The calculation procedure is outlined in Table 4 and results are summarized
in Table 5. The results of this analysis showed that, averaged over all
monitoring sites, under the meteorological and seasonal restrictions
imposed and conditioned on the assumptions made, traditional particulate
emission sources account for about 50 micrograms per cubic meter and
fugitive dust emission sources account for about 70 micrograms per cubic
meter of total ambient TSP levels. In other words, fugitive dust sources
contribute roughly a third more to ambient levels than traditional particulate
sources. These results are consistent with air quality simulation modeling
results which indicate similar relative contributions of traditional industrial
particulate and fugitive dust sources.
Closer to the steel mills, the fugitive dust contribution increases
relative to other sources while, farther away, it drops. At the site of
highest relative impact, fugitive dusts comprise about two-thirds of average
levels recorded when downwind of the mills.
DCL:mb
203
-------
References
1. Control of Reentrained Dust from Paved Streets, Technical
Report Data No. 907/9-77-007, National Technical Information
Service, U. S. Department of Commerce (Springfield, VA.,
July 1977).
2. Guideline for Development of Control Strategies in Areas in
with Fugitive Dust Problems, Technical Report Data No.
EPA-450/2-77-029, National Technical Information Service,
U. S. Department of Commerce (Springfield, VA., October
1977).
3. Air Programs Branch, Particulate Newsletter No. 5,
United States Environmental Protection Agency, Region V
(Chicago, ILL, 1979).
204
-------
Reference 1
TECHNICAL REPORT DATA
(Please read fasjfttetions on the revene before completing)
1 pEPORTNO. 2.
907/9-77-007
4/TITLE AND SUBTITLE
Control of reentrained dust from paved streets
7. AUTHOR(S)
Kenneth Axe tell and Joan Zell
9. PERFORMING ORGANIZATION NAMH AND ADDRESS
PEDCo-Environraental, Inc.
2480 Per ah ing Road
Kansas City, MO 64108
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Region VII - Air Support Branch
1735 Baltimore
Kansas City, MO 64108
"Wi t_v»
Region VII project officer for this report was
Kansas City, Missouri, 64108
Dewayne E. Durst, 1735 Baltimore,
IB/ABSTRACT
The report is a comprehensive review of the problem of reentrained dust from paved
streets. Information was obtained from literature review, collection of unpublished
data from traffic-related air and water pollution studies, survey of comments from
public works officials, and five different field studies to evaluate the effective-
ness of specific reentrained dust control treasures. All of the assembled data
show agreement that the effect of traffic-related particulate emissions in the form
of reentrained dust is one of the most important sources of particulate matter in
metropolitan areas. The results of the field studies conducted in the project and
analysis of six other studies were inconclusive with regard to the effectiveness of
improved street cleaning as a roe ana of improving air quality. None of the street.
cleaning methods proved to be effective in all studies in which they were evaluated.
In addition, fhe data generated in the field studies did not show a consistent
relationship between street surface loadings and nearby particulate concentrations.
The report gives information on the costs for conventional street cleaning operations
and possible oodificationa to improve air quality. Examples of existing state and
local regulations and ordinances for controlling reentrained dust are presented.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
air pollution traffic volume
suspended particulates
urban '
fugitive dust
reentrained dust
street cleaning
Regulations
IH-PftOCUCH) BY
NATIONAL TECHNICAL
INFORMATION SERVICE
U.8. DCPAHTatENT Of COHHtttt
SPKIhSf IE10, VA. ZZ161
13. DISTRIBUTION STATEMENT
Release unlimited
b.lDENTIFIERS/OPEN ENDED TERMS
air pollution control
xaobil sources
fugitive emissions
particulates
19. SECURITY CLASS (Thit Report}
Unclassified
20. SECURITY CLASS (Thispage)
Unclassified
c. COS AT I Field/Group
22. PRICE
A-M-Aol
EPA Form 2220-1 (3-73)
205
-------
Reference 2
TECHNICAL REPORT DATA
(Please rtilJ JrittntCllons on the itvtrse before coinplelin&)
1. REPORT NO.
EPA-450/2-77-029
4. TITLE AND SUBTITLE
Guideline for Development of Control Strategies in
Areas in with Fugitive Dust Problems -
7. AUTHOHtS)
George Richard, TRW
Dallas Safriet, EPA
3. PERFORMING ORGANIZATION CODE
9. PERFORMING ORGANIZATION NAME AND ADDRESS
TRW
Environment Engineering Division
One Space Park
Redondo Beach, California
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-01-3152
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The document outlines a methodology- for development of control strategies for areas
experiencing non-attainment problems due to fugitive dust emissions.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Particulate Matter
Total Suspended Particulate
Emission Sources
Control Methods
Fugitive Dust
Air Quality Measurement
Air Quality Modeling
'o. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (THiiReport)
Unclassified
21. NO. OF PAGES
r»,m 2220-1 (R.y. 4_7
20. SECURITY CLASS (TMipage)
Unclassified
PRBV10U, E01T10N „
22. PR.di
P£ MS
206
-------
United States
Environmental Protection
Agency
Region V
230 South Dearborn
Chicago, Illinois 60604
Reference 3
RECEIVED APR 1 31979
Air Programs Branch
General USEPA Malfunction Policy
(312) 353-2205
The USEPA malfunction policy was stated in the Utah Implementation Plan
for Sulfur Dioxide relating to Kennecott Copper Corporation's smelter
located in Salt Lake City, Utah, (42 FP. 21472). Basically, the policy
is that all excess emissions are violations of the applicable emission
standards with no automatic exemptions. However, the USEPA doesn't
desire to penalize the prudent operator for emissions that are beyond
the operator's control but would use enforcement discretion depending on
the circumstances. USEPA recognizes that the frequency of incidences
and the quantity of the excess emissions could be reduced by good operation
and maintenance including a comprehensive preventive maintenance program.
With such a program, the existing air pollution control equipment will
provide its maximum benefit and return on investment Onsgard
Particulate Controls for Stoker Fired Boilers
On January 31, 1979, Messrs; Ed Piasecki and Dave Skivens of General
Motors Environmental Activities Staff gave a presentation on various
operating practices and controls used in various GM facilities to a
group of engineers with the Engineering Unit. The following is a
summary of their remarks:
1. Uncontrolled emissions increase as the percentage of fines (i.e.
size below 10 microns) in the coal increase. GM facilities try to
limit this to no more than 20%.
2. Uncontrolled emissions increase as excess air in the boiler increases.
This should be as low as possible to support complete combustion.
Uncontrolled emissions increase as % over-fire air decreases.
Vlith newer types of multicyclones, i.e. with two sets of vanes and
higher pressure drops, GM facilities can meet 0.3 pounds per
million BTU heat input with the above operating practices and if
coal quality is such that higher heating value and ash content are
12500 BTU/lb. and 5 to 3:5, respectively.
5. GM has been working on a side stream separator (see following
story) which should be able to reduce emissions to below 0.2 pounds
per million BTU heat input at a fraction of the cost it takes to
install baghouses or precipitators Roklany
*0ur issues to date have been: Nov. 1973 - Number 1
Dec. 1978 - Number 2
Jan. 197? - Number 4
207
3.
4.
-------
'SOUTH WATER
FILTRATION PLANT
-COMMONWEALTH
EDISON FIGURE 1. INDUSTRY AND MONITORING LOCATIONS USED FOR THE SNOW COVER ANALYSIS
4615
LAKE MICHIGAN
4610
MONITORS IN THE IMPACT AREA
MONITORS USED FOR BACKGROUND
o
00
4605
4600
UNION CARBIDE
AMERICAN
MAIZE
MARKTOW
AMOCO
AMERICAN STEEL
IELD SCHOOL
UNIVERSAL ATLAS CEMENT
MARBLEHEAD LIME
GYPSUM
UNION CARBIDE
U S REDUCTION
NORTHERN INDIANA PUBLIC
SERVICE
BLAW KNOX
U S S LEAD
BIHt-MAN ASPHALT
AST CHICAGO MUNICIPLE
INCINERATOR
GENERAL REFRACTORIES
WIRT SCHOOL
KAISER ALUMINUM
FEDERALBLDG.
OJIVANHOE SCHOOL
— 4615
— 4610
I
JGOLDBLATT'S
I©
I
I
I
-I
I
8I8
X 1 Ul
O 1 ^
815
1
I
1
1
1
~. R OX AIM A
--©
HAMMOND [TJ
CITY HALL /
STOUFFER
CHEMICAL
Q3-ASHLAND
DU PONT
HARBISON
WALKER
4605
460
-AMERICAN BRICK
465
|—4600
-------
N
1mile
FIGURE 2. MAP OF UNPAVED ROADS AND STORAGE PILES IN EMISSION INVENTORY
-------
TABLE 1
HYPOTHESIS FORMULATION
Null Hypothesis, HQ = The mean of TSP concentrations recorded on
snow cover days is equal to the mean of TSP con-
centrations recorded on bare ground days.
Level of significance,
-------
TABLE 2
HYPOTHESIS TEST RESULTS*
X,
132.8 74.1 64,8 35.$ 66 37 5.93
Conclusion:
Reject null hypothesis; TSP concentrations on
snow cover daystare significantly lower than
TSP concentrations on bare ground days.
See Table 1 for explanation of symbols.
211
-------
TABLE 3
LISTING OF DATA FOR SNOW COVER AND BARE GROUND DAYS UNDER RESTRICTIVE METEOROLOGICAL CONDITIONS
Snow Cover Days
DATE
Resultant
Wind (degrees)
Direction
Average Wind
Speed (MPH)
Day of Week
Monitor Site
South Water
Filter Plant
Wirt School
Average
Background
Marktown
Gary Airport
Federal Bldg.
Gary APCO
Field School
Roxanna
Hammond City
Hall
Goldblatts
Ivanhoe School
1/25/76
50
10.1
Sun
Wi/m3
-
39
39
127
59
102
79
t04
89
64
65
82
1/14/78
10
14.7
Sat
21
45
33
94
54
113
68
118
49
72
56
64
1/20/78
20
16.4
Fri
22
34
28
63
68
187
61
92
100
85
68
71
2/13/78
50
16.0
Mon
23
56
39.5
124
-
164
90
47
-
-
106
94
3/1 3/75
10
13.2
Thu
23
21
22
124
120
- ,'
89
184
-
71
71
72
4/24/75
30
13.4
Thu
39
44
41.5
-
•
-
85
271
135
49
•
108
Bare
5/16/75
40
9.8
Tue
22
27
24.5
97
-
-
47
226
83
123
74
- 83
Ground
Days
5/24/76 5/7/77
10
14.4
Mon
-
15
15
68
258
206
100
420
78
72
69
342
20
12.4
Sat
-
35
35
46
79
242
69
292
67
61
193
88
10/29/77
40
10.4
Fri
-
46
46
170
79
174
98
•
271
136
247
72
4/26I
20
13.
Wet
25
30
27.!
115
147
525
117
286
244
118
212
-------
TABLE 4
CALCULATION PROCEDURE FOR ESTIMATING SOURCE
CONTRIBUTIONS AT EACH MONITORING SITE
T- (CSi - bi>
Cs - recorded ambient TSP concentration on snow cover day, i t/ig/m3)
b = average background TSP concentration on snow cover day, i (ptg/m )
NS - number of days of both C$ and b available in pairs
«
T = estimated contribution of traditional sources (jig/m )
£ (cBi
CD = recorded ambient TSP concentration on bare ground day, i
b = average background TSP concentration on bare ground day, i (jug/m3)
Np = number of days of both CR and b available in pairs
o
F = estimated contribution of fugitive dust sources (jig/m )
«j
B = estimated contribution of background (jig/m )
213
-------
TABLE 5
CALCULATION OF ESTIMATED TRADITIONAL SOURCE
FUGITIVE DUST SOURCE AND BACKGROUND PARTICULATE LEVELS
UNDER RESTRICTIVE METEOROLOGICAL CONDITIONS
Site/Parameter
Marktown
Gary Airport
Federal Building
Gary APCO
Field School
Roxanna
Hammond Gity Hall
Goldblatt's
I vanhoe School
Group Mean of Monitor Sites
Closest to Industrial Areas
Group Mean of Monitor Sites
Farthest from Industrial Areas
Average of All Sites
Estimated Contribution of Source**
Traditional
Sources
Fugitive Dust
Sources
Background
T Per- F Per- B Per-
(jug/m ) cent (jiig/m3) cent (pg/m ) cent
67
27
107
40
55
46
40
39
43
52
19
37
45
23
31
47
29
34
30
84
148
1,6
157
68
16
65
52
23
59
51
18
65
46
19
48
41
32
31
33
32
30
33
30
31
32
25
22
12
37
12
23
34
23
25
59
42
52
33
34
33
87
50
71
49
40
46
32
32
32
18
26
21
See Table 4 for explanation of letter symbols.
214
-------
SINTER PLANT WINDBUX GAS RECIRCULATION
AND
GRAVEL BED FILTRATION DEMONSTRATION
BY
GENE P. CURRENT
MANAGER, ENVIRONMENTAL CONTROL
WEIRTON STEEL DIVISON
NATIONAL STEEL CORPORATION
WEIRTON, WEST VIRGINIA
26062
ABSTRACT
This research program was initiated with the overall objective of develop-
ing new technology for the reduction of exhaust gas volume and the control
of emissions from the sintering process in the steel industry- This paper
documents the operating problems, as well as the environmental} energy and
economic relationships associated with the Sinter Plant windbox gas recir-
culation atid gravel bed filtration. At this point in time, an overall
evaluation indicates both strong advantages and serious disadvantages in
these technologies. It has been demonstrated that the pollutant mass
emission reduction achieved by windbox gas recirculation is principally
a function of the percentage of waste gas" volume recycled. It has also
been demonstrated that windbox gas recirculation and gravel bed filtra-
tion are compatible technologies. However, recycle limitations in the
windbox gas recirculation system, and filter media support screen
blinding in the gravel bed' filter system are problems not yet totally
resolved.' The relative merits of these technologies cannot be estab-
lished until total optimization of these facilities is achieved. For
this reason, recommendations concerning the acceptability of these
technologies are reserved until a later date.
' ' •' : '"'"/*" •? \ .
PRESENTED AT THE EPA SYMPOSIUM ON IRON AND STEEL
POLLUTION ABATEMENT TECHNOLOGY
Pick-Congress Hotel
Chicago, Illinois
October 30, 31 and November 1, 1979
SPONSORED BY
EPA'S INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, N. C. 27711
215
-------
SINTER PLANT WINDBOX GAS RECIRCULATION
AND
GRAVEL BED FILTRATION DEMONSTRATION
INTRODUCTION
The sintering process is utilized by the steel industry to agglomerate
large quantities of iron ore fines and steel mill waste iron oxides
into a suitable raw material for the production of iron in a blast
furnace. In essence, a sintering plant is a solid waste recovery fac-
ility which conserves our natural resources and eliminates a solid
waste disposal problem. The first step in the production of sinter
involves the mixing of ore fines, thickener sludge, iron scale, and
other iron-bearing waste material with coke breeze and limestone to form
a mass which can be ignited to produce an aggregate. The coke breeze is
added to provide the required fuel for downdraft combustion in the sin-
tering process, while the limestone provides the necessary flux for the
sinter when it is subsequently processed in the blast furnace. These
materials, which make up the burden to the sinter machine, are passed
through a balling drum to blend and agglomerate the constituents into
a permeable mixture which will result in rapid and uniform sintering.
After blending, the mixture is charged onto a traveling grate (sintering
machine). Near the entry or feed end of the machine, the bed is ignited
on the surface by gas burners in a furnace, and as the mixture moves
along on the traveling grate, air is continuously drawn through the bed
to support the combustion of the material. The combustion process pro-
gresses downward through the bed at a temperature of approximately
1500° C. until the entire depth of the charged material is sintered^
The moving grate then discharges the sintered material for further pro-
cessing and subsequent charging into the blast furnace.
Excess air plus the combustion products and unagglomerated particulate
matter are drawn through the sintering bed by induced draft fans and
enter large chambers located under the moving grates. These chambers,
which run the length of the machine are referred to as windboxes.
Typically a dual arrangement of twelve to sixteen windboxes is lo-
cated under the sinter machine. Each windbox is equipped with facili-
ties to discharge and return to the plant the larger particulate matter
which has been drawn through the grates and collected in the windboxes.
However, approximately twenty pounds of particulate matter per ton of
sinter strand feed, remains airborne and passes through the windbox
system.
216
-------
Control of these particulate emissions from the main windbox system
is a difficult problem. While technically feasible control processes
exist, improvements to present technology are needed to achieve more
effective, reliable and economical control of these emissions. Due to
the high percentage of windbox particulate emissions under ten microns
in size, cyclone and multiclone installations have been proven inade-
quate in this application. High moisture, acid salts and condensable
hydrocarbons in the windbox discharge can cause serious operating prob-
lems in baghouse, wet scrubber and dry and wet electrostatic precipi-
tator installations. Further disadvantages are associated with the
wet-type control technologies due to the related corrosion and water
pollution control problems.
Preliminary investigation by National Steel Corporation had indicated
that recirculation of a portion of the gases generated in the windbox
system of the sintering process may significantly reduce particulate
and hydrocarbon emissions to the atmosphere. Further work completed
by the Company, under the partial sponsorship of the Federal Environ-
mental Protection Agency, developed the engineering and design for the
installation of this technology on the Weirton Steel Division No. 2
Sinter Machine. This information was contained in a report titled
"Sinter Plant Windbox Gas Recirculation System Engineering and Design"
which was published by the Environmental Protection Agency in August
of 1975 (available from NTIS as Report No. PB249-546A5K
Preliminary investigation by the Company had also indicated, on a
pilot scale, that gravel bed filtration can remove main windbox particu-
late emissions with the presence of condensable hydrocarbons in the gas
stream. Based on these investigations, the technologies of windbox gas
recirculation and gravel bed filtration were recommended for full scale
evaluation.
WINDBOX GAS RECIRCULATION SYSTEM
Description of Recirculation System
Windbox gas recirculation, as applied to a sinter plant, involves the
return of a portion of the windbox exhaust gas to a hood above the
sinter bed. The objective is to reduce the volume of waste gas to be
cleaned and to conserve a part of the sensible heat in the windbox ex-
haust gas. The portion of gas recycled to the machine must pass through
the heat zone of the sinter bed thus providing the potential for the
reduction of hydrocarbons through carbonization.
217
-------
Figure No. 1 illustrates the general arrangement of the Weirton Steel
Division Sinter Plant Gas Recirculation System. The effluent gases and
particulate matter exhausted from the sinter bed enter the dual arrange-
mentof fourteen windboxes under the sinter strand, and pass through down-
comers to two parallel waste gas mains on the east and west sides of the
machine. Each waste gas main transports approximately fifty percent of
the sinter machine windbox effluent to a series of four cyclone dust col-
lectors for the removal of the larger particulate matter. From the
cyclones, the windbox effluent re-combines in a plenum chamber for dis-
tribution to the two induced draft fans operating parallel. Each fan was
designed to exhaust approximately 11,000 cubic meters per minute at a
temperature of 200° C. and a static pressure of 1300 millimeters water
column.
The waste gas fan exhausts fifty percent of the total gas volume from the
plenum chamber and delivers it to the gravel bed filter system. The re-
cycle gas fan exhausts the remaining fifty percent of the gas volume from
the plenum chamber, and recirculates a portion of the total gas volume to
the sinter machine via an insulated recycle gas main, six distribution
ducts, and a modular recycle hood which extends from the ignition furnace
to the sinter breaker. The modular recycle hood is designed for quick
removal of individual sections to facilitate maintenance. The system was
designed to recirculate 39 percent of the total windbox exhaust volume.
Windbox Gas Recirculation System Operating Problems
The installation of a recirculation system for a sinter machine can result
in serious operational problems in that the system becomes an integrated
part of the sintering process. Although concessions were required with
regard to the maintenance and operation of the sinter machine, satisfactory
performance has been achieved in all but one area which has prevented the
achievement of the design recirculation rate. This problem area concerns
the expulsion of particulate and gases from under the sinter machine re-
cycle hood.
During the design phase of this system, it was anticipated that the per-
centage of waste gas returned to the machine would be limited by its
oxygen content. However, in actual practice it was found that recycle
rates higher than twenty-five percent resulted in ambient air quality
problems in the operating area. At thirty percent recycle, significant
quantities of particulate and gas were discharged into the plant at
several locations along the perimeter of the hood system.
218
-------
RECYCLE HOOD
RECYCLE GAS
CONTROL HOUSE
WASTE GAS
CONTROL HOUSE
Figure 1. General arrangement of system
WEIRTON STEEL DIVISION SINTER PLANT GAS RECIRCULATION SYSTEM
-------
The reasons for the recirculation limitation of twenty-five percent are
not totally known at this time, but the major contributing factor is
excessive leakage at the windbox system. It is estimated that excess air
at the rate of fifty percent of the total effluent volume is entering the
windbox system via routes other than through the sinter bed. This condi-
tion is not uncommon in older sinter machines, since the economic advantage
of minimal leakage was not apparent until the 1970's. In an effort to de-
crease the leakage rate, increase the windbox recirculation rate and fur-
ther reduce the requirements for the final air pollution control device,
the Company has initiated programs for improved sealing facilities at the
pallet train and windbox junctures of the sinter machine. In addition, an
aggressive maintenance program has been established for the repair of
leakage in the windbox, collecting main and cyclone systems. Repair of
leakage from erosion in these areas is a never-ending, but essential, task.
In order to determine any significant differences in the operating para-
meters and product quality derived from the utilization of windbox gas
recirculation, tests were conducted under both conventional operation and
maximum permissible waste gas recirculation to the sinter strand. Although
the oxygen content of the windbox gas decreased from sixteen percent to
fourteen percent, no significant problems resulted while operating in the
recirculation mode. Sinter quality was identical during the two periods,
and it was concluded that the utilization of waste gas recirculation at a
rate of twenty-five percent has no significant effect on the chemical or
physical properties of the sinter.
Windbox Gas Recirculation System Environmental Concepts
An environmental testing program was conducted by Weirton Steel Division
to determine the effectiveness of the windbox gas recirculation system
(see Table 1). The test program was designed to document a direct com-
parison of non-recycle and recycle operation of the sinter machine. Due
to the variability of the raw material charge to the sinter machine, it
was concluded that the non-recycle and recycle modes should be sampled on
the same day. Sufficient time was permitted after the change-over of each
mode of operation to permit stabilization of the process. The Environ-
mental Protection Agency approved methods of sampling were utilized in all
cases except the measurement of condensable hydrocarbons, where there is no
approved method. In this case, the Environmental Protection Agency draft
method was employed.
The testing program indicated that a twenty-five percent recirculation
rate results in a 32.2 percent mass reduction in particulate emissions, a
17.4 percent mass reduction in sulfur dioxide emissions and a 31.6 percent
mass reduction in condensable hydrocarbon emissions. An analysis of the
test results reveals that the reductions in mass emissions for the para-
meters are principally a function of the percentage of waste gas recycled.
Only minor changes in the concentration of these parameters were noted be-
tween the non-recycle and recycle operational modes. A 7.4 percent de-
crease in particulate concentration was documented under recycle conditions,
220
-------
TABLE NO. 1
SINTER PLANT WINDBOX GAS RECIRCULATION TEST RESULTS
Parameter
Particulars Matter
Concentration (mg/Nrn^ Gas)
Mass Emission Rate
(g/kg Sinter Strand Feed)
Without Windbox
Recirculation
1182
3.20
With 25% Windbox
Recirculation
1094
2.17
Percent Change
- 7.4
-32.2
N3
Sulfur Dioxide
Concentration (mg/Nm^ Gas)
Mass Emission Rate
(g/kg Sinter Strand Feed)
320
0.86
350
0.71
+ 9.4
-17.4
Condensible Hydrocarbons
Concentration (mg/Nm^ Gas)
Mass Emission Rate
(g/kg Sinter Strand Feed)
7.0
0.019
6.4
0.013
- 8.6
-31.6
-------
however, this difference is considered to be within the normal deviation
for Method 5 stack sampling determinations. The 9.4 percent increase in
sulfur dioxide concentration, along with the lower 17.4 percent mass re-
duction, indicates a build-up of this gas within the system which is to
be expected. The 8.6 percent reduction in condensable hydrocarbon concen-
tration was somewhat disappointing. Attempts to perform material balances
by analyzing feed and product samples during the non-recycle and recycle
modes of operation produced mixed results. It is interesting to note that
under either mode of operation, more than 85 percent of the hydrocarbon in-
put loading was destroyed by the sintering process. However, in view of
the questions concerning condensable hydrocarbon sampling and analyses
procedures, and the inconsistencies in the data generated in the test pro-
gram, it must be emphasized that any judgements concerning this parameter
must be made cautiously.
Windbox Gas Recirculation System Energy Consumption
The replacement of two existing 2500 horsepower fans with two new 4500
horsepower fans is a major consideration in the analysis of the energy
aspects of this system. The new fans were installed to accommodate (1)
the increased windbox gas volume due to the anticipated elevation in wind-
box gas temperature, and (2) the additional resistance of the new ductwork
from the recycle fan to the sinter machine hoods.
In actual operating practice, the anticipated elevation in windbox gas
temperature did not occur. The design calculations were based on a wind-
box gas temperature of 200° C., whereas a temperature averaging approxim-
ately 120° C. was documented after the system was commissioned. This
lower temperature was a result of an adjustment in the quantity of coke
breeze in the sinter burden, which was necessitated by excessive sinter
bed temperature. To alleviate the excessive bed temperature and associ-
ated operating and maintenance problems, the quantity of coke breeze in
the burden was reduced by seven percent. Thus, during actual operation,
the temperature below the strand burden is essentially the same with or
without recirculation.
Energy data for the Sinter Plant were tabulated from the Company's monthly
energy reports for reference periods before and after the installation of
the windbox gas recirculation system (see Table 2). As a result of the
fan replacement, electrical power consumption increased 77 joules/m.t. of
sinter or fifty-four percent. As expected, the coke-oven gas consumption
remained constant before and after the installation. With recirculation,
the coke breeze consumption decreased 146 joules/m.t. of sinter or seven
percent. As stated previously, this adjustment was necessary to eliminate
the problems associated with excessive sinter temperature. The total
energy consumption per unit of sinter production decreased 69 joules/m.t.
or three percent from 2510 joules/m.t. prior to recirculation to 2441
joules/m.t. after recirculation. The decrease in the energy requirement
can be attributed to the recovery of waste heat from the recirculated
waste gas and the reduction in coke breeze consumption. These data can
be very misleading if taken out of context. While it is true that the
windbox gas recirculation permits the recovery of waste heat, the fact
remains that an inexpensive source of energy (coke breeze) is being
222
-------
TABLE NO. 2
SINTER PLANT ENERGY CONSUMPTION PER UNIT OF SINTER PRODUCTION
Coke Oven/
Natural Coke
Gas Breeze
Sinter Plant Energy Consumption
Without Windbox Recirculation and
Gravel Bed Filtration
(joules/m.t.)
Electricity
143
221
2146
TOTAL
2510
Sinter Plant Energy Consumption
with Windbox Recirculation
(joules/m.t.)
220
221
2000
2441
ho
to
U>
Energy Consumption by Windbox
Gas Recirculation System
(joules/m.t.)
% Change
77
54%
0
0
-146
- 1%
- 69
- 3%
Sinter Plant Energy Consumption
with Windbox Gas Recirculation and
Gravel Bed Filtration
(joules/m.t. )
266
268
2000
2534
Energy Consumption of Gravel Bed
Filter System (with 60% of Effluent
Volume)
(joules/m.t.)
% Change
46
32%
47
21%
0
0
93
4%
Energy Consumption of Combined
Systems
(joules/m.t.)
% Change
123
47
21%
-146
(- 7%)
24
1%
-------
replaced by an expensive source of energy (electricity). It must be
emphasized that the trade-off between decreased coke breeze consumption
and increased electrical power consumption represents an extremely un-
favorable economic balance.
Windbox Gas Recirculation System Capital and Operating Costs
A Summary of the projected capital and operating costs for the windbox gas
recirculation system is tabulated in Table 3. The capital cost for the
windbox gas recirculation system was $5,334,000 when escalated to 1978
market conditions. Depreciation cost for the recirculation system was
calculated utilizing this capital cost and an estimated useful life of
eighteen years. Utility cost for the operation of the recirculation
system was limited to electrical power consumption. It was necessary to
prorate the electrical cost for fan operation since only a portion of the
power consumed is attributed to recirculation of the gas. It was also
necessary to develop a credit for the seven percent reduction in coke
breeze consumption which occurred as a result of the optimization of
the windbox gas recirculation system. The operating cost for this facili-
ity averaged $864,000 per year which is equivalent to $.79 per metric ton
of sinter produced.
GRAVEL BED FILTER SYSTEM
Description of Gravel Bed Filter System
Prior to 1957, the gas cleaning applications for gravel bed filtration
were limited to batch type operations. In 1957, technology was developed
in Germany which permitted the continuous gravel bed filtration of a gas
stream. Since that time variations in this design were developed and
utilized as air pollution control devices in the cement and lime industries.
However, the application of gravel bed filtration technology at a sinter
plant did not occur until 1976 when a full scale system was commissioned
at Weirton Steel Division.
The gravel bed filter system installed on the No. 2 Sinter Machine at
Weirton Steel Division is a special unit specifically designed for sinter
plant application. The system consists of a parallel arrangement of 24
filter modules of equal size. The filter modules are assembled into
groups of four and are stacked vertically to utilize space more efficiently.
There are two such vertically stacked towers, each containing twelve
modules (see Figure No. 2).
224
-------
TABLE NO. 3
TABULATION OF CAPITAL AND OPERATING COSTS
FOR THE
SINTER PLANT WINDBOX GAS RECIRCULATION AND GRAVEL BED FILTER SYSTEMS
Windbox Gas Gravel Bed Combined
Recirculation System Filter System Systems
A. Capital (1978 Dollars) $ 5,334,000 $ 5,101,000 $ 10,435,000
B. Projected Annual Operating
Cost (1978 Dollars)
Repair Labor 70,000 110,000 180,000
Repair Material 96,000 103.000 199,000
Natural Gas - 150,000 150,000
Electricity 560,000 552,000 1,112,000
Depreciation 296,000 283,000 579,000
Credit for Coke Breeze ( 158,000) - ( 158,000)
Reduction
TOTAL 864,000 1,198,000 2,062,000
C. Cost per Unit of Sinter
Production ,
Cost per Metric Ton $ 0.79 $ 1.10 $ 1.89
-------
RECYCLE
DOWNCOMERS-
BACKFLUSH
RECYCLE TO
RAW GAS
ACKFLUSH
CYCLONE
Figure 2
GRAVEL BED FILTER
FLOW SHEET
-------
The operation of each filter module is identical; 22 modules are in the
forward flow cleaning mode of collecting and filtering sinter dust onto
the minute surfaces of the media while two modules (one in each tower)
are isolated in a backflushing mode for removal of the collected dust
from the filter beds. The system is designed to accept dust laden gas
at a flux rate of 0.6 cubic meters per second per square meter of bed
area with a flange-to-flange pressure drop of approximately 330 milli-
meters of water column.
During the forward flow mode (refer to Figure 3) waste gas from the
sinter machine enteus the filter module through the raw gas duct (1) and
passes down through the filter media (6) where the entrained dust is
captured. The gas continues through the media support screen (7), and
into the clean air duct (9). Note that this flow path is determined by
having the downcomer valves (3) and the backflush valve (10) in the down
position. The cleaned gas then exits through the gravel bed filter I.D.
fan to the stack.
The above "forward flow" or "cleaning mode" continues until the automatic
cycle timer in the main control panel calls upon that module to enter the
"backflush mode" of operation (refer to Figure 4). At this time, the
valve operators (2 and 16) lift the downcomer and backflush valves
(3 and 10) to isolate the module from the gravel bed filter I.D. fan. The
rake mechanism is activated to agitate the bed, and ambient air, preheated
to 150° C. , is drawn up through the isolated beds to remove the agglomer-
ated dust. The dust laden gas then passes through a cyclone system and
on to the backflush fan which discharges it to the inlet of the gravel
bed filter system. The dust collected in the cyclones is recycled back
to the sinter machine. After the pre-set backflush time interval has
elapsed, the valves are again repositioned to place the module back in the
forward or "cleaning mode" position,
Gravel Bed Filter System Operating Problems
Aside from many mechanical problems which occurred in the early stages of
debugging, the main obstacle in achieving optimized operation of the
gravel bed filter system has been the blinding of the filter media support
screens. Shortly after the commissioning of the system, an inspection of
the filter modules revealed a severe problem of blinding and adhesion of
sinter dust to form a one to ten millimeter cake on the support screens.
Since clearance between the rakes and the support screens is maintained
at approximately 13 millimeters to avoid screen damage, the media layer
next to the screen is not disturbed during the backflush cleaning cycle.
Once the screens start to blind, an uneven, high velocity backflushing
action occurs and sections of the support screens are left without media
cover after the completion of the cycle. This phenomenon is of a very
serious nature since it causes short-circuiting of the system and results
in reduced particulate removal efficiency.
227
-------
DOWNCOMER
VALVE
RAW GAS
4
RAKE DRIVE
CLEAN AIR
CHAMBER
12
UPPER BED
RAKE 5
MEDIA 6
)rFINE MESH SCREEN 7
^SCREEN SUPPORT GRATE
16
AIR a
BACKFLUSH
VALVE
LOWER BED
-DOWNCOMER
DUCT
Figure No. 3
FILTER MODULE
FORWARD FLOW MODE
228
-------
4
AKE DRIVE
DOWNCOMER VALVE
UPPER BED |4
RAKE 5
MEDIA 6
FINE MESH SCREEN 7
SCREEN SUPPORT GRATE
CLEAN AIR a
BACKFLUSH VALVE-
II
PREHEATED
BACKFLUSH
AIR
^-RECYCLE
DOWNCOMER
DUCT
13
Figure No . 4
FILTER MODULE
BACKFLUSH MODE
229
-------
It has been concluded that the principal cause of the screen blinding
problem is the condensation of moisture on the screen and the combin-
ation of this moisture with the lime-rich sinter dust in the waste gas.
Although many programs have been implemented to minimize this problem,
its principle cause is inherent in the sintering operation as a result
of the frequent start-up and shut-down modes of the machine, and the
associated fluctuations in incoming waste gas temperatures to the gravel
bed filter system.
The only immediate alternative for obtaining sustained operation of the
system is the implementation of a preventative maintenance program in
screen cleaning during the scheduled weekly sinter machine down turns.
Although costly, this program is being utilized until a more economical
alternative is available.
Gravel Bed Filter Environmental Aspects
During the gravel bed filter demonstration period, stack testing was
conducted while maintaining an average recycle rate of twenty-five percent
in the windbox gas recirculation system. The results of this testing
indicated that the average inlet particulate concentration to the gravel
bed filter system was 1565 mg/Nm^. Stack gas testing simultaneously
taken at the outlet of the gravel bed filter system indicated an average
particulate concentration of 156 mg/Nm^. The average particulate removal
efficiency during this mode of operation was ninety percent with a range
from 81 percent to 95 percent.
The detrimental effects of the screen blinding problem on the operating
reliability and performance of the gravel bed filter system cannot be
over-emphasized. Under optimized conditions, the gravel bed filter • system
has demonstrated the capability of producing an effluent concentration of
less than 70 mg/Nm^ of particulate. During periods of malfunction with
partially blinded screens, the effluent concentration is more than tripled.
Testing indicated that gravel bed filtration reduced condensable hydro-
carbon emissions by approximately twenty-eight percent. It is presumed
that this reduction occurred due to condensation and deposition in the
system at temperatures below 120° C. The benefit of this condition is
questionable since it could be contributing to the screen blinding problem.
Again, it must be emphasized that due to the uncertainities involved in
the method of condensable hydrocarbon determinations, data generated for
this parameter should be viewed accordingly.
230
-------
Gravel Bed Filter Energy Consumption
During the demonstration period the flux rate through the gravel bed
filter system was well below the design due to the limitations of the
induced draft fan. This fan limited the filtration process to forty
percent of the total sinter machine windbox gas voluma. Recently re-
visions to the gravel bed filter system were implemented to permit the
treatment of sixty percent of the total sinter machine windbox gas volume.
With these revisions, the total electrical requirements, including auxiliary
motors for backflush fans and hydraulic pumps increased to 3200 horsepower.
In addition, natural gas is utilized at a rate of 3.5 cubic meters per
minute to preheat the backflush air from ambient temperature to 150° C.
The Sinter Plant energy consumption at that time increased 46 joules/m.t.
or 32 percent for electricity and 47 joules/m.t. or ?.l percent for coke
oven/natural gas (see Table 2). The total Sinter Plant energy consumption
increased 93 joules/m.t. or four percent from 2441 joules/m.t. to 2534
joules/m.t. of sinter production.
Gravel Bed Filter Capital and Operating Costs
A summary of the capital and operating cost for the gravel bed filter
system is tabulated in Table 3.
The capital cost for the revised gravel bed filter system is $5,101,000
when equated to 1978 market conditions. Depreciation cost for the gravel
bed filter system was calculated utilizing this capital cost and an estim-
ated useful life of eighteen years. Costs for electrical power and natural
gas consumption were derived utilizing design data. The projected operating
costs for the system is $1,198,000 per year which is equivalent to $1.10 per
metric ton of sinter produced.
Combined Windbox Gas Recirculation and Gravel Bed Filter System
An analyses of the environmental capabilities of the combined windbox gas
recirculation and gravel bed filter systems cannot be accurately documented
until the systems are totally optimized. As stated previously the Company
is aggressively pursuing programs to reduce the total effluent volume from
the sinter machine and improve environmental performance.
Table 2 indicates the change in Sinter Plant energy comsumption due to
the combined windbox gas recirculation and gravel bed filter systems.
This projection indicates an increase in electrical power consumption of
86 percent of 123 joules/m.t. of sinter production, an increase in coke
oven/natural gas consumption of 21 percent of 47 joules/m.t., and a decrease
in coke breeze consumption of 7 percent or 146 joules/m.t. of sinter pro-
duction. The projection also indicates that the total energy consumption
231
-------
has increased one percent or 24 joules/m.t. of sinter. Again, it must
be emphasized that the trade-off between the decreased coke breeze con-
sumption and the increased electrical power consumption presents an ex-
tremely unfavorable economic balance.
The total capital cost for the combined system is $10,435,000 based on
1978 market conditions. The projected operating cost for the combined
systems is $2,062,000 per year or $1.89 per metric ton of sinter produced.
The capital and operating costs will escalate depending upon the future
expenditures necessary to optimize the system.
It must be emphasized that this windbox gas recirculation and gravel bed
filter demonstration is a full-scale research and development program
which has not yet reached a conclusion. At this point in time, an over-
all evaluation of these facilities indicates both strong advantages and
serious disadvantages. However, the relative merits of these technologies
cannot be established until the optimization of the facilities is achieved.
For this reason, recommendations concerning the acceptability of these
technologies are reserved until a later date.
232
-------
RTI1547/Conference Paper OCTOBER, 1979
REVIEW OF FOREIGN AIR POLLUTION CONTROL TECHNOLOGY FOR BOF FUGITIVE EMISSIONS
David W. Coy
Research Triangle Institute
Research Triangle Park, North Carolina
and
Richard Jablin
Richard Jablin and Associates
Wrightsville Beach, North Carolina
ABSTRACT
This paper reviews BOF fugitive emission control technology in use in foreign
steel plants. The discussion presented is based on literature review and
visits to twelve iron and steel plants in Western Europe and Japan. Basically
two categories of fugitive emissions control 'technology have been seen during
the foreign plant visits. One category relies strictly on local hooding; the
other category combines local hooding and partial building evacuation. Both
are described in general terms in this paper. Then one plant, one of the best
visited in Western Europe that relies strictly on local hooding is described
in more detail.
233
-------
1.0 INTRODUCTION
This paper reviews BOF fugitive emission control technology in use in
foreign steel plants. The discussion presented is based on literature review
and visits to twelve iron and steel plants in Western Europe and Japan. This
survey has been performed to provide the basis for comparison of foreign
technology with that in use in the United States, and where the foreign tech-
nology is superior, show how it may be applied to domestic steelmaking prac-
tice. Control of fugitive particulate emissions from BOF steelmaking and de-
sulfurization of iron outside the blast furnace are being emphasized in this
study.
Basically two categories of fugitive emissions control technology have
been seen during the foreign plant visits. One category relies strictly on
local hooding; the other category combines local hooding and partial building
evacuation. Both are described in general terms in this paper. One plant, in
Western Europe that relies strictly on local hooding is described in more
detail.
This paper also serves as a status report on the EPA project acknowledged
in Section 4.0. A further goal of this study is to conduct tests in two for-
eign plants to document the capture and control capability of technology that
is determined to be superior. The applicability to domestic steelmaking prac-
tice will be determined on the basis of an engineering study using data ob-
tained during the foreign plant visits.
Additional reports will be issued on these subjects.
234
-------
2.0 GENERAL DESCRIPTION OF BOF FUGITIVE EMISSION CONTROL TECHNOLOGY
The fugitive or secondary emission sources of interest in this study are
those from furnace charging and tapping^ hot metal transfer, desulfurization
of iron in torpedo cars or ladles, deslagging of torpedo cars or ladles,
deskulling of ladles, and furnace puffing in the case of suppressed combustion
primary fume control systems. At the time this project was begun there were
virtually no plants in the United States that had attempted to capture essiis-
sions from all or almost all of the above sources. A few plants had installed
control systems for capturing furnace charging, puffing, and tapping emis-
sions. Others had installed only hot metal transfer emission control systems.
Since 1977, the number of U.S. plants with some furnace charging control has
010re than doubled to nine. However, only three plants have installed snore
complete secondary emission control systems, i.e., including such things as
hot metal transfer, deslagging and deskulling, and desulfurization. Two of
these plants retrofitted their secondary emission control systems; the third
was built as part of a new plant.
Only one plant in the United States is known to have attempted secondary
emission control by combining local hooding and partial building evacuations.
One plant combines local hoods and canopy hoods above the charging aisle, and
the remainder only use some fora of local hooding.
The foreign plants selected for visits were chosen on the basis of re-
ports in the technical literature, vendor experience lists, trip reports from
previous EPA foreign plant visits, and discussions with knowledgeable visitors
from foreign countries. The five plants selected in Western Europe were as
follows:
British Steel - Uckenby Works
Krupp - Rheinhausen Works
Hoogovens - Ijmuiden, Netherlands
235
-------
Swedish Steel - Oxelosunds, Sweden
Italsider - Taranto Works
The above plants were visited in March of 1979. Seven Japanese steel plants
were visited in September and October of 1979. They were as follows:
Nippon Steel - Oita Works
Nippon Steel - Yawata Works
Sumitomo Metal Industries - Kashima Works
Kawasaki Steel - Chiba Works
Kawasaki Steel - Mizushima Works
Kobe Steel - Kakogawa Works
Nippon Kokan - Ohgishima Works
Table 1 provides a general description of the emission control technology
in each plant. The two basic approaches used for secondary emission control
are local hooding and local hooding plus partial building evacuation. Figure
1 shows a schematic flow diagram of a local hooding secondary system connected
to a fabric filter. Figure 2 shows a plant arrangement in which a roof wet
ESP is used for partial building evacuation.
Only one of the plants, however, chose to combine local hooding and
partial building evacuation in the initial construction of the plant. The
Chiba Works of Kawasaki did include partial building evacuation through a
roof-mounted wet ESP in the construction of their Q-BOP. All the other plants
have retrofit the partial building evacuation systems, presumably due to inef-
fectiveness of their local hoods alone and changing environmental require-
ments.
Two plants were identified as having some secondary emission capture for
deslagging and dekishing of torpedo cars, and hot metal transfer in separate,
small, enclosed buildings. The basic design principles involved for these,
however, are closer to local hooding than building evacuation.
As is evident from Table 1, the control device of choice is the fabric
filter with a few exceptions. One of the exceptions arises through the use of
one Baumco system design that processes charging emission waste gas through
the primary fume scrubber. This system is best adapted to shops with larger
236
-------
TABLE 1. FOREIGN SECONDARY EMISSION CONTROLS
Company/Plant
Emission Control Technology
Secondary Furnace Hot Metal Transfer, Desul-
Emissions furization, Deslagging,
Etc.
British Steel, Lackenby
Krupp, Rheinhausen
Hoogovens, Ijmuiden
Plant No. 2
Swedish Steel, Oxelosund
Italsider, Taranto
Plant No. 2
Nippon Steel, Oita
Nippon Steel, Yawata
Plant No. 3
Sumitomo, Kashima
Plant No. 2
Kawasaki, Chiba
Plant No. 3
Kawasaki, Mizushima
Plant No. 1
Plant No. 2
Local Hoods + Canopy
hoods - Scrubbers
Local Hoods - Scrubbers
Local Hoods - Scrubbers Local Hoods - Scrubber
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
Local Hoods + Partial
Building Evacuation -
Fabric Filter
Local Hoods - Fabric
Filter
Partial Building
Evacuation -
Roof Wet ESP
Same as Chiba
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter & Scrubber
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter, Enclosed Building
Evacuation - ESP
Local Hoods + Enclosed
Building Evacuation -
Fabric Filter
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
Local Hoods - Fabric
Filter
237
-------
TABLE 1. (Continued)
Company/Plant
Emission Control Technology
Secondary Furnace Hot Metal Transfer, Desul-
Emissions furization., Deslagging,
Etc.
local Hoods - fabric
filter
Local Hoods - Fabric
Filter
Nippon Kokan, Ohgishiroa Local Hoods - Fabric Local Hoods - Fabric
Filter Filter
238
-------
HOT METAL
TRANSFER 2
\
HOT METAL
TRANSFER 1
i
DESULFURIZING CONVERTERS CONVERTER 2 CONVERTER 1 DESULFURI2ING
&DESLAGGING2
(FUTURE)
&DESLAGGING1
DESKULLING
OF LADLES
EXHAUST FANS
BAGHOUSE CAPACITY
1,000,000 m3/hr
MAXIMUM TEMPERATURE
130° C
ACTUAL^ 53° C
I !
| FUTURE BAGHOUSE CAPACITY '
| 500,000 m3/hr
| J
Figure 1. Italsider-Taranto basic oxygen
steelmaking secondary ventilation system.
239
-------
TO GAS RECOVERY
Figure 2. Plant arrangement with partial building evacuation.
240
-------
vessels where the primary fume control systems have larger design gas flows
needed for control of charging emissions. While the obvious advantage of this
arrangement is the reduction of air flow needs in a separate secondary emis-
sions control system, the disadvantage may be poor control of charging emis-
sions depending on scrap quality and the limited gas How available from the
primary system. When charging emissions are treated as part of a larger sec-
ondary emission control system, it is possible to divert gas flow from other
secondary operations long enough to ensure good control of heavy fuming
charges.
Another exception to the fabric filter choice is the roof-mounted wet
ESPs in use in Japan for partial building evacuation. The particular design
in use in the plants visited has no air moving equipment attached to it. Air
movement through them is generated strictly by thermal drafts induced by the
shop heat. The principal advantage of such a system is, therefore, reduced
energy consumption because fan power to move the air through a fabric filter
is not necessary. There are other factors that must be weighed against this
advantage such as reinforcement of building structure to take the roof load,
and wastewater treatment. The wastewater treatment consequence is probably
not significant since most of the recently constructed BOFs use scrubbers for
primary fume control and will have wastewater treatment facilities.
The choice between furnace emission control strictly by local hooding or
in combination with partial building evacuation is expected to vary with the
particular site conditions. For instance, Q-BOP charging may present more
difficulty in achieving good charging fume control by local hoods because of
the necessity for blowing gas through the tuyeres when the vessel is turned
down. In this case partial building evacuation might be a better choice.
Heavy fuming produced by dirty scrap might also provide the incentive for
partial building evacuation. In retrofitting an additional furnace vessel to
an existing shop with inadequate space for good local hooding and ducting,
partial building evacuation could supplement a less efficient local hood
system.
The advantage, in general, of using local hooding instead of partial
building evacuation is that air volumes treated are smaller and hence energy
consumption is lower. The building evacuation system without an air mover
might be an exception. The engineering study to follow in this project is
expected to provide insight into this matter.
-------
Performance Observations
In all of the plants visited, performance observations were made on the
secondary emission control systems. The principal effort was to visually
evaluate the capture capabilities of the various local hoods. Where partial
building evacuation was used, building emissions were observed. Since there
is no standard method for visual determination of capture, the performance
estimates are subjective. In addition, the ultimate determination of which
systems are the best must also consider differences in process conditions
between plants i.e., scrap quality and the relative proportion of scrap versus
hot metal. The overall evaluations are not yet complete, so specific observa-
tions are not reported in this paper. What can be said is that the best
performing technology observed for each of the secondary sources provided
virtually complete capture of the fumes with capture efficiencies estimated in
the range of 90 to 100 percent. As pointed out above, these estimates must
yet be combined with process conditions to relate the system performance to
typical United States plant conditions.
On the basis of our subjective performance estimates and review of the
system design and plant operations, our preliminary evaluation of the Western
Europe plant visits suggests that the Italsider - Taranto Works had the best
overall secondary emission control systems. Because of similarities in design
and performance between several Japanese plants visited and the Taranto Works,
the latter was selected for a more detailed description in this paper.
Specific Plant Description
The Taranto Works secondary emission control system depends entirely on
local hoods for emissions capture. Figure 1 shows a flow diagram for the
secondary ventilation system. As shown, there are eight collection points,
five exhaust fans (one spare) and a dual-section, pressurized baghouse having
a total of 10 compartments. Each compartment has 144 bags, which are 300 mm
diameter by 9120 mm long, yielding a net total cloth area of 12,400 m2. The
cloth is "Terylene", a synthetic fabric suitable for a service temperature of
130°C. Upon startup of the system, Taranto operated one baghouse with five
compartments, later adding the second five compartments. Initially there were
three fans, but two more were added. They are now considering the addition of
five more compartments as well as additional fans in order to raise the system
242
-------
capacity from 1,000,000 m3/hr to 1,500,000 m3/hr. Additional pickup points
would then be added such as ladle repair stations.
Table 2 shows the designed flow and temperature condition at each of the
hoods. In considering the table, it should be noted that desulfurizing and
deslagging Station 1 is out of service because it provides for pneumatic
injection of carbide which is no longer in use. The deskulling station is
also out of service because the dust from that operation was too heavy and
built up in the ductwork, thereby putting the hood out of commission. A
comparison of the total system volume presented in Table 2 with the baghouse
capacity as shown in Figure 1, indicates that time-sharing of the suction
capacity among the various operations is practiced. It also indicates that
gas cooling takes place in the suction duct between the hoods and the bag-
house. In normal operation, four of the existing five fans are operating and
one is in standby condition. The bags are cleaned by reverse air flow which
was originally provided by a small auxiliary blower.
Figure 3 shows a general arrangement of the doghouse and the main suction
(charging) hood at the converter. Not shown are two hinged doors of heavy
plate at the front of the furnace, each about four meters high. The back end
of t-he doghouse ic closed by means of hinged plate doors, however, the front
is open except for partial closure by a pipe curtain at the top which serves
to direct the fumes into the opening of the hood. At the front of the dog-
house, there is the main hood with three opening as shown in Figure 3. At the
rear, there is a smaller auxiliary hood to collect a portion of the emissions
from tapping the furnace. The other portion is directed to the front hood by
means of the sloping sections in the roof and sides of the doghouse at the
rear.
The sketch of Figure 3 was made by visual observation during the visit
without measurements. It is, therefore, to be taken as a rough approximation
of the actual arrangements. The upper hood face (charging hood) was estimated
to be 7.6 meters wide. The upper face opening appeared to be about 60 cm
wider on each side than the outside diameter of the hood skirt.
Figure 4 shows a drawing of the hood at the hot metal transfer station.
As shown, the top of the furnace ladle fits into close configuration with the
bottom of the hood which, therefore, provides a continuation to the exhaust
system. The ventilating slot is approximately ]% to 2 meters wide for admis-
243
-------
TABLE 2. SECONDARY VENTILATION SYSTEM SUCTION VOLUMES
Hood
1.
2.
3.
4.
5.
Description
Converter hood
a. Open - Hot metal charging
b. Closed - Puffing during blow
Deslagging
Desulfurization
Hot metal transfer
Deskulling
TOTAL
Flow, m /min
10,000
2,500
3,000
1,800
3, '000
1,500
21,800
Temperature, °C
480
60
200
130
130
150
294
= 1,300,000 m3/hr
244
-------
EXHAUST TO SECONDARY VENTILATION
1 0.000 m3/min@ 480° C
HOOD OPENINGS
N>
350 TONNE CONVERTER
CHAIN
CURTAIN
FRONT VIEW
OPERATING
FLOOR
SLOPE TOP AND SIDE PLATES
INWARD TOWARD THE CENTER
OF THE FURNACE
SIDE VIEW
•~\
HINGED
DOORS
Figure 3. Italsider-Jaranto basic oxygen steelmaking
--doghouse arrangement.
-------
EXHAUST TOBAGHOUSE
3DOOm3/min@130°C
250 TONNE CAPACITY
TORPEDO CAB
VENTILATING SLOT
Figure 4. Italsider-Taranto basic oxygen steelmaking
— hot metal transfer.
246
-------
si on of the molten iron to the furnace ladle and for fume capture from the
torpedo car. The hood is arranged so that it extends well beyond the center
line of the torpedo car. The entire hot metal transfer station is partially
enclosed by a separate building which is external to the BOF shop. The build-
ing is open on both ends to permit movement of torpedo cars through it by
rai 1.
In the desulfurization building, adjacent to the BOF shop, there are two
fume collection hoods. One hood serves the deslagging station; the other hood
serves the desulfurization staton. Both hoods are mounted above the ladle
position, closely fitted to it, and of essentially the same diameter as the
opening in the ladle. This is made possible by the fact that the ladle is
mounted on a transfer car that moves it into place below the hoods. Clearance
from the top of the ladle to the underside of the circular hood face in the
desulfurization station is approximately 30 cm. The close-fitting deslagging
hood or enclosure is actually rectangular (in spite of previous reference to
diameter) and is open on the side where the slag rake is mounted. The fourth
side of the deslagging hood is effectively provided by the wall of the build-
ing.
A highly desirable feature of the secondary ventilation system is the
methods which are used for controlling the valves to admit suction at the var-
ious hoods. The desirability of the system lies in two factors. One is the
provision for a single control valve at each converter which minimizes the
number of operating conditions and control variables. The other is the simpli-
fied concept of the control scheme as will be described in the following para-
graphs.
At each converter hood, there is one valve that is used in two positions,
fully opened or closed. When it is opened it will pass 10,000 m /min to
accommodate the capture requirements during hot metal charge. When the valve
2
is closed, it fits loosely in the duct and permits the passage of 2,500 m /min
to capture the fumes which result from puffing during the blow, from tapping
and slagging the furnace.
Each furnace pulpit has a panel in which there are manually operated con-
trols for the valves as well as indicating lights to show the position, either
open or closed, of the valves on each of the three converters. The controls
at any vessel are arranged so that the operator can close any of the three
247
-------
valves, but only open the valve pertaining to his vessel. This simple, but
effective, concept permits the operator to obtain maximize suction when his
vessel is being charged with molten iron and permits him to minimize suction
demands from the other two vessels at that time. The concept is made possible
by the fact that only one vessel may be charged with molten iron at any single
instant.
At the hot metal transfer station, the opening of the valves to the suc-
tion hood is interlocked with the control rotating the torpedo vessel in order
to pour the iron. Shutoff is controlled by a timer that closes the valve
after seven minutes of operation.
At the desulfurizing station, the opening of the valve to the suction
hood is initiated automatically upon the descent of the stirring device into
the ladle. When the stirring device is withdrawn, the valve is closed. At
this location, there is also a small exhaust take-off to the carbide handling
system.
At the deslagging station, each movement of the rake opens and closes the
valve for the suction hood. When the rake travels to its extreme position an
electrical interlock opens the valve. When the rake returns to the retract
position, the valve closes. This means that during the raking of the slag,
the valve opens and closes approximately 15 times a minute. When the rake is
in the parked position at the shutdown of the deslagging operation, the valve
is automatically closed and remains so.
At Taranto, the special features of particular note were associated with
the secondary fume control system. Of particular interest was the simple
method of bringing each suction hood on line when needed and insuring that it
was closed when not in operation. The other feature of particular note was
the side-by-side location of desulfurization and deslagging for the molten
iron. This was a very efficient arrangement from the standpoint of operations
and, because the ladle was transported by a transfer car which had a tilting
feature, it was possible to design a very satisfactory fume capture system.
248
-------
3.0 CONCLUSIONS
The two categories of systems for BOF secondary emissions control seen in
the foreign plants were based on
1) local hooding only, and
2) local hooding plus partial building evacuation.
While the partial building evacuation systems for furnace emissions were not
in the majority, there are several situations in which such installations may
offer advantages over the use of only local hooding. In particular, the use
of roof-mounted ESPs with no fans attached, and therefore no air moving--
pressure drop energy costs, may be a cost effective alternative to complete
reliance on local hooding connected to fabric filters.
The particulate collection device in most common use for secondary emis-
sion control systems is the fabric filter. Scrubbers and ESPs are used in
only a few plants of the plants visited.
In respect to performance or capture capability of the local hooding ap-
plications, the best systems observed captured virtually all of the secondary
emissions. The method of estimating capture efficiency was subjective, but
best performance was in the range of 90 to 100 percent capture.
The Italsider - Taranto Works secondary emission control system was simi-
lar in design and performance to several seen during the Japanese plant visits.
A preliminary evaluation suggests the Taranto Works secondary emission control
system was the best of the plants visited in Western Europe.
-------
4.0 ACKNOWLEDGEMENT
The background data supporting the information contained in this paper
was obtained as part of the work effort under Contract 68-02-2651 between
Research Triangle Institute and the U.S. Environmental Protection Agency. The
project is funded by the Ferrous Metallurgical Branch of the Industrial Envi-
ronmental Research Laboratory at Research Triangle Park, North Carolina. Mr.
Norman Plaks is the EPA Project Officer.
250
-------
5.0 REFtRENCES
1. Drabkin, M. and Helfand, R., "A Review of Standards of Performance for New
Stationary Sources-Iron and Steel Plants/Basic Oxygen Furnaces," Mitre
Corporation for U.S. Environmental Protection Agency, EPA-450/3-78-116,
November 1978.
2. Steiner, J., "Trip Report - Presurvey of BOF Shop Republic Steel,
Cleveland, Ohio," Accurex Corporation for U.S. Environmental Protection
Agency, Division of Stationary Source Enforcement, July 27, 1978.
3. Jablin, R. and Coy, D., "Trip Report - Italsider Steel Company, Taranto,
Plant," Research Triangle Institute, March 26, 1979.
251
-------
FUGITIVE PARTICULATE EMISSION
FACTORS FOR BOP OPERATIONS
Jim Steiner
Acurex Corporation
Energy and Environmental Division
485 Clyde Avenue
Mountain View, California 94042
Larry F. Kertcher
U.S. Environmental Protection Agency
230 So. Dearborn Street
Chicago, Illinois 60606
ABSTRACT
Extensive testing was conducted at two steel mills to measure the
amount of fugitive particulate emissions generated during BOP hot metal
addition and tapping operations. Tests were conducted on a new Q-BOP
installation equipped with a doghouse and a single secondary collection
hood; tests were also conducted on an existing BOP installation
retrofitted with a doghouse and multiple secondary collection hoods.
Particulate emission data was correlated with several process parameters,
vessel operating characteristics and visual observations inside and
outside each shop. Particulate emission factors for hot metal addition to
the Q-BOP and the BOP averaged 0.6 Ib per ton (particulate emitted per ton
of hot metal added to the vessel) with a range of 0.2 to 1.2 Ib per ton
and 0.25 Ib per ton with a range of 0.16 to 0.32 Ib per ton,
respectively. For tapping operations, the particulate emission factors
for the Q-BOP and the BOP were 0.92 Ib per ton (particulate emitted per
ton of steel tapped) with a range of 0.15 to 2.28 Ib per ton and 0.16 Ib
per ton with a range of 0.05 to 0.24 Ib per ton, respectively.
252
-------
FUGITIVE PARTICULATE EMISSION
FACTORS FOR BOP OPERATIONS
INTRODUCTION
The Energy and Environmental Division of Acurex Corporation undertook
a series of tests for EPA Region 5 at two steel mills to determine
fugitive particulate emission factors for hot metal addition and tapping.
Tests were conducted on a Q-BOP operation at Republic Steel's Chicago
Works in 1978 and on a BOP operation at Republic Steel's Cleveland Works
in 1979. A concise description of each facility, test program and results
is presented below.
Q-BOP TEST PROGRAM
Facility Operation
Republic Steel Corporation operates a two-vessel Q-BOP shop (225 tons
steel/heat) at its Chicago Works. Each vessel is equipped with a doghouse
enclosure, a secondary collection hood (charge side only) and removal
system design by Pennsylvania Engineering Corporation. Figure 1
illustrates hot metal being charged to the vessel with the doghouse doors
open and the secondary collection hood in operation.
At the start of a heat the doghouse doors were opened and the vessel
was rotated toward the charge side. Suction on the secondary collection
hood (330,887 dscfm) was automatically initiated as soon as the vessel was
displaced more than 20° from the vertical position. Suction on the
primary collection hood (above the vessel) was reduced to approximately
10 percent of full capacity. Figure 2 illustrates the dampers and
ductwork for the primary and secondary collection and gas cleaning systems
necessary to accomplish this flow distribution. Scrap was added to the
vessel via the charge machine. Hot metal was then added to the vessel
using a hot metal ladle and overhead crane. Two ladle additions were
necessary to introduce the proper weight of hot metal for a heat. While
the vessel was accepting a charge of scrap or hot metal, N2 was blown
through the tuyeres in the bottom of the vessel to prevent them from
getting plugged. The N2 blowrate was higher during hot metal addition
than during scrap addition. During the test program the average hot metal
addition time (2 ladles) was 2.9 min and the average amount of hot metal
charged was 203.7 tons (2 ladles). Fumes emitted during scrap and hot
metal addition were captured by the secondary collection hood on the
charge side of the vessel and were ducted to the venturi scrubbers for
removal. Emissions escaping collection by the secondary collection hood
drifted upward and entered the atmosphere via the shop roof monitor.
253
-------
ro
Figure 1. Hot metal addition to Q-BOP at Republic Steel, Chicago.
-------
Isolation
damper
open
Q-Bop No. 1 Furnace Enclosure
•Secondary hood No. 1
Quencher No. 1
80% open
Quencher No. 2
20% open
Secondary hood 1o. 2
Q-Bop Mo. 2
Furnace enclosure
Stack No. 1
Fan No. 2
tack No. 2
Figure 2. Flowchart for primary and secondary collection and
gas cleaning system for Republic Steel Q-BOP shop.
255
-------
After charging was complete, the vessel was turned up to the vertical
position in preparation for Q£ blow. The doghouse doors were closed and
the secondary collection hood was shut off with the primary collection
hood in operation. After completion of $2 blow, the vessel was rotated
to the tap position. Again, as the vessel moved 20° past the vertical
position, the secondary collection hood went into operation. Fugitive
particulate emissions, generated as the steel and additives were poured
into the teeming ladle, were captured and sent to the venturi scrubbers
for removal.
Test Location
The particulate mass and size measurements were made in the ductwork
connecting the secondary collection hood to the pollution control
equipment as illustrated in Figure 3. The single sampling port (for
horizontal traverse of duct) was 1.5 diameters downstream of a bend in the
duct and only 0.5 diameters upstream of the downcomer leading to the
quencher. No traversing was done in the vertical direction because of the
overhead crane clearance requirements. This test location was used for
both the hot metal addition and the tapping tests.
Test Equipment
Since the hot metal addition and tapping portions of the BOP cycle
were of relatively short duration, high volume ( ~5 scfm) particulate mass
and size sampling trains were used to collect large enough samples for
subsequent analysis. The particulate mass tests were conducted using an
EPA Method 5 sampling system as illustrated in Figure 4; the particle size
measurements were made using a Source Assessment Stack Sampler (SASS) as
shown in Figure 5. The Method 5 train was equipped with a 3y cyclone to
remove the large flakes of kish and to prevent tearing of the glass fiber
filter. The particle size train consisted of 10.5y, 3.6y, 1.55y cyclones
in series followed by a glass fiber filter. Both trains had the necessary
components to insure all sampling was done isokinetically.
Test Procedures
Sampling was basically conducted using EPA Method 5 procedures with
appropriate modifications to accommodate the nature of the process being
tested. For safety reasons, the test crew was not allowed to remain at
the test location during hot metal addition. Hence, each test (both mass
and size) was a single point sample. In order to account for
stratification of particulates in the duct, several isokinetic tests were
conducted at each of three points (3 ft, 5 ft, 7 ft) in the 10 ft duct.
All tests were averaged to determine the mass (and size) concentration and
emission rate of particulates generated and captured during hot metal
addition. Volumetric flowrate measurements were made prior to the test
program under conditions simulating hot metal addition and tapping
conditions. The results of these measurements agreed well with the system
design data supplied by PECOR. Crewmembers were allowed on the samplinq
platform during the tapping tests so a total of 12 points were sampled on
a horizontal traverse for each test. During each test observers were
stationed inside and outside the Q-BOP shop. The observer inside the shop
256
-------
Isolation
damper
Sampling
Platform
Downcomer to twin venturi quencher
Single Horizontal Sampling Port
1
Figure 3. Sampling location for hot metal addition and tapping.
257
-------
Figure 4. High volume EPA Method 5 mass sampling train,
-------
i-o
II
Figure 5. High volume size sampling train.
-------
was responsible for collecting all the heat process information (e.g.,
scrap type, weight, amount of hot metal charged, temperature, etc.) for
qualitatively assessing the collection efficiency of the secondary hood
and factors affecting its performance, and for coordinating the entire
test program activities with the process operation. The observer
stationed outside the shop observed the opacity of any emissions escaping
the roof monitor above the vessel being tested (i.e., those fugitive
emissions which were not captured by the secondary collection hood). At
the completion of a test, samples were recovered from the sampling trains
in a room provided by Republic Steel. Both front half and back half
sampling train catches were analyzed using the procedures originally
proposed by EPA in 1971. Emission factors for hot metal addition and
tapping were based on front half catches (nozzle, probe, cyclone, filter)
only.
Test Results
Table 1 presents the results of the hot metal addition tests.
Approximately 50.2 percent of these emissions were >10.5y size; 16.5
percent were between 3.6y and 10.5y in size; 30.2 percent were between
1.55y and 3.6y and 3.1 percent were less than 1.55y in size. In general,
the doghouse and secondary collection hood were capable of collecting
almost all of the fugitive particulate emissions generated during hot
metal addition provided that the crane operator properly positioned the
ladle, and that both ID fans were operating on the secondary hood.
Opacity of the emissions leaving the shop roof monitor (i.e., those
emissions which escaped collection by the hood) were less than 5 percent.
However, when the crane operator was careless or the bell damper (and/or
its instrumentation) failed to operate (one fan on secondary hood and one
fan on primary hood), the secondary hood failed to capture a significant
amount of the fugitive emissions generated which resulted in roof monitor
opacities averaging 17 to 19 percent with individual readings as high as
35 percent. No correlation was evident for measured mass emissions with
such parameters as type of scrap charged, scrap to hot metal ratio, hot
metal charge rate, temperature or composition. High sulfur butts in the
scrap charge did result in significant condensible emissions (back half
catch).
Table 2 presents the results of the tapping tests. It should be noted
that only was ID fan was in operation on the secondary collection hood
(the other was drawing on the primary hood). Approximately 86 to
96 percent of the emissions generated during tapping were >3y in size. In
general, the doghouse and secondary collection hood did not quantitatively
capture tapping fume. Opacities of the roof monitor emissions averaged
9 percent regardless of whether two ID fans or only one ID fan was drawing
on the secondary hood. Since the secondary collection hood is located on
the charge side of the doghouse, it was too far away from the teeming
ladle to effectively capture tapping fume. Hence, the data reported in
Table 2 are underestimated. The type and amount of additives (e.g.,
desulfurization agent, coke, sulfur, etc.) increased condensible emissions
significantly and the measured mass emissions were dependent on tap time
(higher mass emissions with short tap times).
260
-------
TABLE 1. PARTICULATE EMISSION FACTORS FOR HOT METAL ADDITION TO A Q-BOP
Average
Q-BOP
Hot Metal
Charge
(tons)a
201.5
197.5
206.0
202.5
201.5
213.0
203.7
Q-BOP
Hot Metal
Charge Time
(min)
2.2
2.8
2.2
3.4
2.4
4.3
2.9
Q-BOP
Hot Metal
Charge Rate
(tons/min)
91.6
70.5
93.6
59.6
84.0
49.5
74.8
'articulate
Em i s s i on
Rate
(Ib/min)
111.5
50.5
32.0
33.7
17.9
29.5
45.9
Particulate
Emission
Factor
(lb/ton)b
1.2
0.7
0.3
0.6
0.2
0.6
0.6
Particulate
Mass
Concentration
(gr/dscf )
2.3590
1.0689
0.6788
0.7141
0.3790
0.6260
0.9710
aQ-BOP hot metal charge (2 ladles per heat)
bLb of particulate emitted per ton of hot metal charged
TABLE 2. PARTICULATE EMISSION FACTORS FOR TAPPING A Q-BOP
Average
Total Q-BOP
Tap
(tons)a
226
226
226
226
Total Tap
Time
(min)
4.7
6.0
5.2
5.3
Q-BOP
Tap Rate
(tons/min)
48.0
37.6
43.4
43.0
P articulate
Emission
Rate
(Ib/min)
109.4
5.8
14.7
43.3
Particulate
Emission
Factor
(lb/ton)b
2.28
0.15
0.34
0.92
Particulate Mass
Concentration
(gr/dscf)
3.8973
0.3853
0.6848
1.6558
Average metal tapped per heat (R. Kortge Republic Steel)
of particulate emitted per ton of metal tapped
261
-------
BOP TEST PROGRAM
Facility Operation
Republic Steel Corporation also operates a two-vessel suppressed
combustion BOP shop (250 tons steel/heat) at its Cleveland Works. Each
vessel was retrofitted with a doghouse enclosure, and local secondary
collection hoods (charging, tapping, other shop operations) which exhaust
to a common removal system. Figure 6 illustrates hot metal being charged
to the vessel with the doghouse doors open and the secondary collection
hood in operation.
At the start of a heat, the doghouse doors were opened and the vessel
was rotated toward the charge side. After scrap was charged, the vessel
operator initiated the "start charge" mode for the secondary collection
hood system which automatically adjusted the dampers to increase flow from
30 percent to approximately 100 percent of capacity (357,800 dscfm).
Dampers on other hoods were closed to maximize suction on the secondary
charge hood. Figure 7 illustrates the ductwork for the entire secondary
collection and gas cleaning system. Scrap was added to the vessel via a
charge machine but the secondary collection hood was only at 30 percent of
capacity since emissions were minimal. Hot metal (1 ladle) was then added
to the vessel using a hot metal ladle and overhead crane. During the test
program, the average hot metal addition time was 1.74 minutes and the
average amount of hot metal charged was 211.8 tons. Fumes emitted during
scrap and hot metal addition were captured by the secondary collection
hood on the charge side of the vessel and were ducted to the electrostatic
precipitators for removal. Emissions escaping collection by the secondary
collection hood drifted upward and entered the atmosphere via the shop
roof monitor. After charging was complete, the vessel was turned up to
the vertical position for 02 blow. The doghouse doors were closed and
the flow in the secondary system was reduced to 30 percent of capacity to
capture any fume escaping the primary hood. After completion of 02
blow, the vessel was rotated to the tap position. Again, the vessel
operator initiated the "start tap" mode, which automatically adjusted the
dampers to maximize suction on the tap hood (320,700 dscfm). Fugitive
particulate emissions, generated as the steel and additives were poured
into the teeming ladle, were captured and sent to the precipitators for
removal.
Test Location
The particulate size and mass measurements were made in the ductwork
connecting the local secondary collection hoods to the precipitators as
illustrated in Figure 8. Two separate sampling ports (-3 ft apart) were
used to measure volumetric flowrate and particulate mass concentration
simultaneously. These horizontal sampling ports were located
approximately 3 diameters downstream and 2 diameters upstream of any flow
disturbance. No traversing was done in the vertical direction because
extensive and costly modificatic s were required to support the test crew
and equipment. This test location was used for both the hot metal
addition and tapping tests. Figure 9 illustrates the actual test location.
262
-------
Figure 6. Hot metal addition to BOP at Republic Steel, Cleveland,
-------
o
o
TAPPING
CHARGE AND SCAVENGER
GAS MAIN ON ROOF
HOT METAL RELADLING STATION
TEEMING LADLE ADDITIVE
.*,*.*.**.*.*.*.
ELECTRO- I I I I I [ I I I
STATIC 1 I'' * '' I '
PRECIPI- >,T t.t.t.f f >
TATORS p £ (^
TAPPING HOOD
MING
ZLE L
FATION A
T
VESSEL
No. 1
M
"*" .
•:::::±\
ADLE ADDITIVE\
NDDUSTCOLLEC-
ION (CHUTES)
ft-
*
1
I
1
LADLE ADDITIVE AND DUST
COLLECTION (BIN)
\^.
VENT FAN
Y J
INLET
1 SAMPLING
SITE
t >
/VESSEL
No. 2
'I
"*
r°
CHARGING h
SCAVENGER
LADLE ADDITIVE AND
I DUST COLLECTION (CH
•^^WiM
SOUTH 1
LADLE 1
CO3 NEW AND
FEEMING
1IOZZLE C
G STATION
EXISTING
o
MULTICLONES
POURING HOOD
EXIST. STATION
Figure 7- Flowchart for secondary collection and gas cleaning-systrem for- Re/3ot»/»c Steel BOf*
-------
To Precipitators
OS
Charge Duct
From FCE #2
Common Main On Roof
Duct Diameters
Tap Duct
FCE, #2
X X
Proposed Sampling
Sites
Charge Duct
Entry Below
PCE #1
Tap Duct FCE, #1
Figure 8. Sampling location for hot metal addition and tapping tests.
-------
K;
Figure 9. Equipment installed at sampling location.
-------
Test Equipment
The particulate sampling equipment used for this test program was
identical to the equipment described previously. In addition, a
calibrated S-type pitot tube and thermocouple were used to measure the
volumetric flowrate in the duct.
Test Procedures
Sampling was basically conducted using EPA Method 5 procedures with
appropriate modifications to accommodate the nature of the process being
tested. Since hot metal addition and tapping times were relatively short,
it was impossible to traverse the 12 ft duct using the large number of
sampling points recommended by EPA Method 2. Hence, volumetric flowrate
measurements were made using 6 and 22 points on a horizontal traverse for
hot metal addition and tapping respectively. At the same time,
particulate mass and size measurements were made at a single point in the
duct. To account for stratification of the particulates in the duct,
three tests were conducted at each of three points in the duct (3 ft, 6
ft, 9 ft). All tests were averaged to determine the mass (and size)
concentration and emission rate of particulates generated and captured
during hot metal addition and tapping. During each test, observers were
stationed inside and outside the BOP shop. The observer inside the shop
was responsible for collecting all the heat process information (e.g.,
scrap type, weight, amount of hot metal charged, temperature, etc.), for
qualitatively assessing the collection efficiency of the local secondary
hoods and factors affecting their performance, and for coordinating the
entire test program activities with the process operation. The observer
stationed outside the shop observed the opacity of any emissions escaping
the roof monitor above the vessel being tested (i.e., those fugitive
emissions which were not captured by the local secondary collection
hoods). At the completion of a test, samples were recovered from the
sampling trains in a mobile van provided by Acurex. Both front half and
back half sampling train catches were analyzed using the procedures
originally proposed by EPA in 1971. Emission factors for hot metal
addition and tapping were based on front half catches (nozzle, probe,
cyclone, filter) only.
Test Results
Table 3 presents the results of the hot metal addition tests.
Approximately 67.6 percent of these emissions were >10.5y size; 13.7
percent were between 3.6y and 10.5y in size; 7.3 percent were between 1.5y
and 3.6y, and 11.5 percent were less than 1.55y in size. In general, the
doghouse and secondary collection hood were capable of collecting all of
the fugitive particulate emissions generated during hot metal addition
provided the vessel operator initiated the "start charge" mode earlier
enough to insure maximum suction on the charge hood prior to actual hot
metal additions. Opacity of the emissions leaving the shop roof monitor
(i.e., those emissions which escaped collection by the hood) were much
less than 5 percent. However, when the vessel operator did not initiate
the "start charge" mode soon enough, the opacity of roof monitor emissions
reached a maximum of 10 percent. There did not appear to be a correlation
267
-------
lAbLt
PMK I ll/ULrtl L Ll'li JJlUll I r\\j i\ji\*j i \ji\ i iu i riL. i ni_ nuu/ i. i .•.">
Average
Hot
Metal
Charged
(tons)
198.5
218.0
215.0
209.5
208.5
216.5
217.0
222.5
201.5
211.8
Duration
of
Charge
(min)
1.16
1.50
3.21
1.66
2.38
1.31
1.75
1.25
1.40
1.74
Hot Metal
Charge
Rate
(tons/min)
170.1
145.3
66.8
125.7
87.4
164.4
123.9
177.9
143.6
133.9
Particulate
Mass
Emission Rate
(Ib/min)
40.0
23.9
21.1
22.3
22.7
44.1
33.2
56.9
34.3
33.2
Particulate
Emission
Factor
(Ib/ton)
0.24
0.16
0.32
0.18
0.26
0.27
0.27
0.32
0.24
0.25
Particulate
Mass
Concentration
(gr/dscf)
0.698
0.570
0.373
0.466
0.408
0.989
0.606
1.056
0.717
0.654
268
-------
between process parameters and measured mass emissions, although the
presence of oily scrap (e.g., turnings, trimmings) increased condensible
organic emissions significantly. High emission rates of condensible
inorganics and sulfates also occurred but insufficient information on
scrap composition resulted in no definite correlations.
Table 4 presents the results of the tapping tests. Approximately 59.8
percent of these emissions were >10.5y in size; 13 percent were between
3.6y and 10.5y in size; 8.6 percent were between 1.55y and 3.6y, and 18.6
percent were less than 1.55y in size. In general, the tap hood captured
almost all of the fugitive emissions but the unhooded manual charge chute
contributed the majority of emissions observed leaving the roof monitor.
Opacities averaged 2 percent but reached a maximum average of 9 percent on
one occasion because the vessel operator failed to initiate the "start tap"
mode in time. There did not appear to be a correlation of measured mass
emissions with process variables but condensible emissions were influenced
considerably with the addition of such additives as coke or sulfur.
269
-------
TABLE 4. PARTICULATE EMISSION FACTORS FOR TAPPING A BOP
Average
Duration
of Tap
(mins)
10.23
9.00
10.08
8.12
6.57
7.00
6.50
4.58
6.00
7.56
Hot
Metal
Tapped
(tons)
249.0
248.0
270.5
269.0
240.0
259.5
252.0
198.5
257.5
249.3
Hot Metal
Tap Rate
(tons/min)
24.3
27.5
26.8
33.1
36.5
37.1
38.7
43.3
42.9
34.5
Part icu late
Mass
Emission Rate
(Ib/min)
4.3
3.5
3.5
7.3
6.9
8.9
2.1
4.2
7.9
5.4
Particulate
Emission
Factor
(Ib/ton)
0.18
0.13
0.13
0.22
0.19
0.24
0.05
0.10
0.18
0.16
Particulate
Mass
Concentration
(gr/dscf)
0.101
0.082
0.081
0.160
0.147
0.187
0.048
0.102
0.135
0.116
270
-------
REFERENCES
1. Steiner, J., and Knirck, J., "Particulate Matter Emission Factor Tests
for Q-BOP Hot Metal Addition and Tapping Operations at Republic Steel
Chicago, Illinois," Acurex Report TR-78-143, Volume I, November 1978.
2. Steiner, J., and Rape, R., "Particulate Matter Emission Factor Tests
for BOP Hot Metal Charging and Tapping at Republic Steel Cleveland,
Ohio," Acurex Report TR-79-23/EE, Volume I, September 1979.
271
-------
Session 2: WATER POLLUTION ABATEMENT
Chairman: Gary A. Amendola, Chief
Technical Support Section
Region V, EPA
Eastern District Office
Westlake, OH
272
-------
Steel, Hater, Regulations, & Etc.
by
Robert B. Schaffer, Director
Effluent Guidelines Division
and
Edward L. Dulaney, Project Officer
Steel Industry Study
Good afternoon ladies and gentlemen. I am Bob Schaffer, Director of the
Effluent Guidelines Division of the United States Environmental
Protection Agency and I am here to give you a brief rundown on the
activities in our Division, particularly as they relate to the iron and
steel industry.
The steel industry regulations promulgated by the Division in 1974
(Steelmaking segment) and 1976 {Forming, Finishing, and Specialty
segment) in response to the passage of the Clean Water Act of 1972 were
remanded to the Agnecy for further work by the Third Circuit Court of
Appeals in Philadelphia in 1975 and 1976, respectively. At the same
273
-------
time, the Agency was in the process of working but a Consent Agreement
with the Natural Resources Defense Council relative to a more careful
evaluation of toxic pollutants in industrial effluents.
Thus, in response to the remands and the consent agreement, the Division
initiated a new and more thorough study of the steel industry. The
field work of this new study has now been essentially completed although
a few special studies to check specific points in question continue.
We have received roughly two thousand (2000) responses to questionnaires
so, as directed by the Court, we have acquired a great deal of specific
information about steel industry operations. These operations break
down into twenty major subcategories (See Attachment I).
Since we will be writing regulations on the basis of Best Practicable
(BPT), Best Available (BAT), Best Conventional (BCT), Best Demonstrated
(BDT), and Pretreatment for New Source (PSNS) and for existing sources
(PSES) for these 45 subsets we have a great deal of regulation writing
to do.
What we have found from surveying the industry is that about thirteen
percent of the operations discharge to publicly owned treatment works,
although in general these are the smaller operations. In addition,
about twelve percent of the operations reported achieving zero
discharge.
274
-------
Water use rates in this industry are very high but much of that water
use is for non-contact cooling purposes and is not significantly con-
taminated in use except by the addition of heat. Since the impact of
heated discharges is so site specific, this Division has not attempted
to address thermal discharges as a part of the steel industry effluent
regulations on a national basis. To give you some idea of the impact of
effluent regulations on contact or process wastewater volumes though,
let me point out that the remanded regulations were expected to result
in a 96% reduction in the "once-through" applied rate of 5500 MGD.
With the sensitivity of todays testing methods, we find toxic pollu-
tants, conventional pollutants, and "other" pollutants in just about
every waste soruce. But, toxic organics in the milligram per liter
range have been found primarily in coke plant wastes as expected and
toxic metals in this range have been found primarily in the steelmaking
operations discharges. See Attachment No. 2
The treatment technologies that hold greatest promise for this industry
are first of all clarification and flow reduction, i.e., recycle or
cascade rinsing, biological oxidation of coke plant wastes, additional
blowdown treatment for discharge (lime or sulfide precipitation), and
carbon adsorption.
275
-------
We are currently in the process of preparing a draft report in nine
volumes summarizing what we have found so far. Several volumes are
already at the printers and the remainder are to be printed before the
end of the year. We expect to distribute these reports to a limited
list of recipients starting December 1, 1979 with the volumes then a-
vailable. Thirty days after the last distribution we expect to hold an
informal public comments meeting to give interested parties an opportun-
ity to further explain their written comments.
Since our technical study has developed treatment alternatives and the
relative costs and effectiveness of these alternatives, we are now
moving into the phase of selecting the alternative to be used as the
basis for the regulations. In this phase we will be working closely
with the economic evaluations program on two levels. A linear
programming analysis will be utilized to optimize the pollution control
achievable by the application of any given level of capital investment
by comparing the effectiveness of treatment versus its cost between the
various waste sources. Secondly, the overall economic impact on the
industry will be evaluated in the final selection of the treatment
alternatives to be designated as the basis for the regulation. Our
current schedule for this work is to propose a regulation in August of
1980 and to promulgate the regulation in March of 1981.
276
-------
Attachment No. 1
Iron and Steel Industry
Subcategorization
20 Subcategories - 45 sub-subcategories
. By-Product Coke
. Beehive Coke
. Sintering
. Blast Furnace: Iron
: Ferromanganese
• Basic Oxygen Furnace: Semi-wet Air pollution Control Methods
Wet Air Pollution Control Methods:
Open Combustion Systems
Suppressed Combustion Systems
. Open Hearth Furnace: Semi-wet Air pollution Control Methods
Wet Air Pollution Control Methods
. Electric Arc Furnace: Semi-wet Air Pollution Control Methods
Wet Air Pollution Control Methods
Vacuum Degassing
Continuous Casting
Hot Forming - Primary
Hot Forming - Section
Hot Forming - Flat (Plate, and Hot Strip and Sheet)
Pipe and Tube - Hot and Cold Worked
Pickling - Sulfuric Acid (Batch and Continuous)
(Neutralization and Recovery)
Pickling - Hydrochloric Acid (Batch and Continuous)
(Neutralization and Regeneration)
Cold Rolling (Direct Application, Combination, and
Recirculating)
Hot Coatings: Glavanizing
Terne
Other
Combination Acid Pickling - (Batch and Continuous)
Scale Removal: Kolene and Hydride - Batch and Continuous)
Alkaline Cleaning (Batch and Continuous)
277
-------
Attachment No. 2
Iron and Steel Industry
Raw Waste Concentrations
Organ ics
Subcateqory mg/l'-1-)
A
C
D
E
F
G
H
I
o,
M
N
By-Product Coke
1 WAL (41 GPT) 466 (12)
2 FC Bldn (70 GPT) 161 (8)
3 Benzol (73 GPT) 223 (7)
Sinter (3 pi.) *
Blast Furn. (4 pi) 7.5 (3)
BOF ( 4 pi.)
1 Open C. *
Open Hearth *
EAF *
Vac. Degas. *
Cont. Casting *
K, L Hot Forming:
J Primary *
K Section *
L Flat: Plate *
: HS&S *
Pipe & Tube
Hot Worked *
Cold Worked
Cold Rolling:
Direct App *
Recirculated 45 (4)
Metals
mg/lU) Others
87(2) 8000
290
1160
* 2185
40 (1) 2684
22(3) 4007
389(1) 613
109(6) 2761
5(2) 28
* 24
* 157
* 91
* 116
* 60
* 107
*+Fe 2150
7.2 (3)+Fe 42767
NH3, J0H, S,
SCN, O&G, TSS, CN
00H, SCN, NH3, O&G
TSS, S
00H, SCN, NH3, S, (
TSS, CN
TSS, O&G, F
TSS, O&G, F
00H
TSS, F
NOs, F, TSS
TSS
TSS
TSS, O&G
TSS, O&G
TSS, O&G
TSS, O&G
TSS, O&G
TSS, O&G
O&G, TSS
O&G, TSS
U)Sum of concentrations of those present in average at more than
1 mg/l and the number of pollutants found in that amount
278
-------
TOTAL RECYCLE OF WATER IN INTEGRATED STEEL PLANTS
Harold J. Kohlmann, Sr. Vice President
Harold Hofstein, Manager Engineer
Hydrotechnic Corporation
1250 Broadway
New York, N.Y. 10001
Abstract
An engineering study was performed to determine the faci-
lities necessary to achieve total recyle of water for five
integrated U.S. steel plants. The study was in fulfillment of
Contract No. 68-02-2626 issued by the USEPA, Metallurgical
Processes Branch, Industrial Processes Division, IERL, RTP,
North Carolina. Conceptual engineering of facilities required
to reach both BAT and the goal of total recycle was performed.
Capital and operating costs were estimated and energy require-
ments developed. Technologies were compared and the most pro-
mising were selected as being applicable.
This paper summarizes the findings and recommendations
from the study and identifies problems expected to be encoun-
tered. These problems included the necessity for development
and verification of technologies to treat the individual
wastes and combinations of wastes, environmental impacts of
increased off-site power generation, additional fuel require-
ments and the necessity of increased solids disposal.
Estimates of increased steel costs, increased electrical
energy demands and the demand for additional fuels are speci-
fied. The paper presents data that should be contemplated by
industry, environmental groups and regulatory agencies in
their efforts to arrive at environmentally sound and practical
effluent requirements.
279
-------
TOTAL RECYCLE -OF WATER IN INTEGRATED STEEL PLANTS
Harold J. Kohlmann, Sr. Vice President
Harold Hofstein, Manager Engineer
Hydrotechnic Corporation
1250 Broadway
New York, NY 10001
In 1972 Public Law 92-500 set as a national goal zero dis-
charge of pollutants by the year 1985. If this goal were to
be attained many industries would have to upgrade their water
systems drastically and, in some cases, provide systems util-
izing total recycle of water. The steel industry is the single
largest user of water in the United States and installing sys-
tems for total recycle is a tremendous task. To determine the
magnitude of this undertaking the USEPA, Industrial Environ-
mental Research Laboratory, Research Triangle Park, NC commis-
sioned Hydrotechnic Corporation to conduct a study of total
recycle of water in five integrated steel plants with an inter-
mediate step to attain present BAT requirements.
Five plants were selected for study based on a number of
criteria. These plants were:
Inland Steel Corp - Indiana Harbor Works
USSC - Fairfield Works
Kaiser Steel Co. - Fontana Plant
National Steel Corp. - Weirton Steel Division
Youngstown Sheet & Tube Corp. - Indiana Harbor Works
The location of the plants is shown on Figure I.
The plants represent a diverse cross-section of industry
in relation to factors that would influence the installation
280
-------
of wastewater control equipment. The plants varied not only in
production tonnage but also in physical size. Wastewater con-
trol facilities installations ranged from those practically
complying with BAT to those that required facilities to attain
BPT. Some plants were spread out and others built on compact
sites so that the production in tonnage per acre was from low
to very high. Their water sources were from rivers, the Great
Lakes and from wells. We think a fair cross-section of steel
plant conditions was considered.
The plants selected were then visited to obtain informa-
tion on:
water, air and production process flow diagrams of each
production facility;
plot plans of the plants indicating areas that would
be available for the construction of pollution control
facilities;
an indication of facilities the plant has planned for
future installation or deletion;
efficiencies of water pollution and air pollution con-
trol facilities presently installed;
areas of typicality (or atypicality) of the plant;
any constraints that may be placed on the installation
of future pollution control facilities.
After the initial visit, the data collected were analyzed
and process water flow diagrams were prepared for each plant.
281
-------
The five selected integrated steel plants were studied to de-
termine: similarity of wastes and production processes be-
tween integrated steel plants, problems that would be encoun-
tered with respect to site specifics, water uses in various
plants, degrees of treatment currently practiced and'applicabi-
lity of retrofit of treatment processes to plant production
operations and plant waste treatment processes.
POSSIBLE PLANS FOR PLANTS TO MEET BAT AND TOTAL RECYCLE
Plans and designs were prepared for each of the five
plants to achieve the objectives of both BAT and total recycle.
Although the plans take location factors into account, they are
conceptual and contain a number of assumptions about physical
constraints which may exist that will preclude the use of the
suggested systems as presented. In addition, various mixes of
wastes were conceptualized for concurrent treatment which have
not been previously demonstrated. If implementation of any of
the programs presented is planned, comprehensive testing
should be undertaken prior to the design of the systems. After
design and construction, the operators of the facilities should
be of a competence level that will ensure proper operation of
the facilities.
For each of the systems developed seven basic premises
were assumed necessary for satisfactory operation; these are:
1. All non-contact cooling water and storm water must be
segregated from process flows to minimize the amount
of wastewater to be treated.
2. Discharge of non-contact cooling water would be per-
mitted under BAT. For total recycle, except in the
case of Kaiser-Fontana, two steps were used, one
282
-------
allowing the non-contact cooling water to discharge
under total recycle conditions and the other requiring
cooling and total recirculation of all non-contact
cooling water.
3. Storm water runoff from material storage piles would
be collected and stored in lined ponds and gradually
discharged to receiving waters under BAT conditions
and to treatment facilities to condition the water for
recycle under total recycle conditions.
4. Water with high levels of dissolved solids could not
be used to quench coke and slag because of the re-
sulting air pollution problems.
5. Scrubber cars would be utilized at the pushing side
of the coke ovens.
6. The discharge of industrial wastes to municipal
treatment plants would be discontinued necessitating
their treatment at the plant under total recycle con-
ditions.
7. General area runoff and treated or untreated sanitary
wastes would discharge from the plant to either re-
ceiving water or municipal treatment plants.
In the preparation of cost estimates, broad assumptions
had to be made as to the costs of yard piping, both under-
ground and aboveground, since detailed knowledge of inter-
ferences that might be encountered were not available. Ca-
pital and operating costs were based on the use of purchased
electrical power and natural gas as the energy source for the
evaporation of residual waste streams. Equipment costs were
283
-------
obtained from manufacturers and from in-house data.
Since the systems for the five plants are quite compli-
cated this paper will only deal with generalities. Anyone de-
siring a more detailed analysis of the systems proposed for
the five plants should refer to the final report prepared for
this project.
BAT
Figure 2 is a schematic flow diagram of a "typical" inte-
grated steel plant showing the facilities required to attain
the BAT requirements. Since BAT has never been officially pro-
mulgated and is now being developed, a number of assumptions
were made in defining BAT for this report. This diagram shows
a typical integrated plant divided into six separate production
areas for simplicity. These areas and their water systems are
as follows:
1. Coke and By-Products Plant
All water discharged from the coke and by-products
plant would be treated with free and fixed ammonia
stills and a biological treatment plant prior to dis-
charge. In most cases, concurrent treatment of the
blowdown from the blast furnace gas washer system was
recommended since this gas is basically a dilute form
of coke plant wastes. However, prior removal of toxic
metals in the blast furnace wastewater may be re-
quired.
2. Iron Making and Sinter Plant
The sinter plant system should be "bottled-up" as much
as possible and any blowdown should have suspended so-
lids removal. This can be accomplished by separate
284
-------
settling facilities or by directing the blowdown to
the blast furnace thickeners. As discussed under the
coke plant section, the blast furnace blowdown if low
enough in toxic metals 'content, should be compatible
for treatment with the coke and by-products plant
wastes in a biological treatment plant prior to dis-
charge. If this proves impossible, separate treat-
ment consisting of alkaline-chlorination and/or lime
precipitation will be required.
3. Steel Making
The gas washing wastewater would be recycled through
a thickener for solids removal and a portion dis-
charged as blowdown.
4. Primary Hot Forming
Recycle systems utilizing scale pits, settling basins,
filters or clarifiers and cooling towers prior to re-
cycle were selected. A treated blowdown would be dis-
charged to the receiving water.
5. Secondary Hot Forming
The BAT systems envisioned for secondary hot forming
utilize a minimum blowdown after treatment in scale
pits, settling basins, filters or clarifiers and cool-
ing towers.
6. Cold Finishing
There are varied facilities included under the cold
finishing subcategory and most produce wastewater dis-
charges that cannot be recycled further due to high
dissolved solids concentrations. The classical method
of treatment is coagulant addition, neutralization,
aeration, polymer addition followed by settling in
285
-------
flocculator clarifiers. This method was adopted for
treatment of mixtures of cold finishing mill waste-
waters.
The original electroplating facility regulations be-
fore being remanded call for zero discharge of waste-
water. Since these regulations were based on small
plating shops the application of this standard to the
much larger integrated plant operations is question-
able. However, facilities have been included to de-
mineralize electroplating facility blowdown for re-
cycle to attain zero discharge.
Total Recycle
Two premises were investigated to arrive at total recycle
systems. The first did not include control of non-contact
cooling'water since, at present, the effluent guidelines do not
regulate these water systems. The second, included control of
non-contact cooling water since regulations could conceivably
be formulated to include the control of this water. Figure 3
is a schematic flow diagram of a "typical" integrated steel
plant showing facilities necessary for total recycle. Because
of the build-up of scale in tight recycle systems due to high
dissolved solids, all blowdown water is collected and conveyed
to one or more demineralization plants for dissolved solids
removal. A dissolved solids level in the product water was
set at 175 mg/1 and in many cases only a portion of the waste
water would require the demineralizator process.
If non-contact cooling water is included the blowdown
would increase and, in turn, the size of the demineralization
facilities.
286
-------
Treatment Processes
We have not explored the various treatment processes for
the treatment of the various wastewaters generated in an inte-
grated steel plant in this paper; however, the process to de-
mineralize the water deserves some discussion since ±his tech-
nology would be a primary technical and cost consideration in
a total recycle scheme.
Dissolved Solids Removal
Various pretreatraent and treatment processes were investi-
gated for use in the removal of dissolved solids. Based on
this investigation four more processes were considered. The
results of costs for the processes are presented on Table I.
Because of extremely high costs of energy requirements,
total evaporation was eliminated. Ion exchange was eliminated
from consideration on the basis of applicability, annual costs
and off-site land requirements. Thus only reverse osmosis and
electrodialysis remained for further consideration. At this
time, reverse osmosis enjoys a broader technological base and
has been used in more applications than electrodialysis. Re-
verse osmosis was, therefore, selected as the possible dis-
solved solids removal treatment unit operation for our analy-
ses, in spite of the higher capital and operating costs.
SUMMARY AND CONCLUSIONS
No simplified solutions can be developed that would be
applicable throughout the entire industry. The atypical na-
ture of the plants studied, and other differences throughout
the entire industry, makes it difficult to assign standard num-
bers to water flows, costs and various other factors that
287
-------
would prove extremely convenient for determining restrictions
on contaminant levels and the cost of complying with these res-
trictions. For this reason, five plants were examined to quan-
tify the various considerations.
The total capacity of the five plants studied was approx-
imately 19.3 kkg (21.2 million tons) per year which represents
13.5 per cent of the total present integrated steel plant ca-
pacity in the United States. The diversified nature of the
integrated steel plants would probably be more pointed if ad-
ditional plant studies were conducted.
COSTS
Cost estimates were prepared for the proposed systems to
accomplish total recycle with the interim step of reaching the
BAT requirements. Both capital and annual costs were estimated
using 1978 prices. Since only general designs were prepared,
certain site specific considerations, such as the need for pil-
ing, obstructions, railroad crossing, etc., may not have been
taken into consideration. However, contingency factors were
added in an attempt to compensate for unknown and unforeseen
items which would cause cost increases.
Table II presents the estimated costs for both BAT and
total recycle. Natural gas was assumed as the fuel. In
addition, costs per kkg (ton) of steel produced to achieve
both BAT and total recycle are presented.
It would be expected that the costs to achieve both BAT
and total recycle for each plant on the basis of cost per unit
of production of steel would be approximately the same. How-
ever, noticeable differences are evident.
288
-------
The BAT compliance step presented the most differences
in the facilities needed as well as their construction and
operating costs. This was due to the great variety in the
wastewater treatment and recycle systems presently installed.
These differences are mainly due to the degree of existing
control facilities installed, the availability of water for
use in the plants and, in some cases, the States in which the
plants are located.
BAT Costs
The following costs per unit of production were estimated
to achieve the BAT requirements.
Kaiser-Fontana
Inland-Indiana Harbor
National-Weirton
USSC - Fairfield
Y.S. & T. Indiana Harbor
Costs per kkg (ton)
No costs estimated.
$1.91 (1.73)
$2.63 (2.39)
$2.52 (2.28)
$3.95 (3.58)
The costs for Kaiser-Fontana were not estimated for the
BAT step because this plant has facilities which, with some
modifications, would bring it into compliance. Of the costs
for the four remaining plants Fairfield, Weirton and Y.S. & T.
Indiana Harbor are basically in agreement. The cost for Inland
Steel, however, is approximately half that of the other three
plants and this is probably due to two factors. The main fac-
tor is that Inland does not have tinning facilities which re-
quire high cost treatment facilities and high operating costs,
since zero discharge is required for BAT. Another reason could
be the size of this plant which produces almost twice as much
Steel as the next largest plant studied, namely Y.S. & T. ~
289
-------
Indiana Harbor Works. The large plant would, in all probabili-
ty, have treatment facilities with lower unit capital and ope-
rating costs.
Total Recycle Costs
The following costs per unit of production for facilities
to achieve total recycle, with and without the inclusion of non-
contact cooling water were estimated. These costs include the
costs for the BAT step as shown above.
Cost per kkg (ton)
Without Non- With Non-
Contact Cooling Contact Cooling
Water Water
Kaiser-Fontana . - $ 2.99 ( 2.71)
Inland-Indiana Harbor $ 7.63 (6.92) 14.13 (12.86)
National-Weirton 32.11 (29.13) 33.21 (30.13)
USSC-Fairfield - 31.41 (28.49)
Y.S. & T.-Indiana Harbor 9.87 ( 8.96) 10.77 ( 9.77)
The low cost per unit of production for the Kaiser-Fontana
plant can be attributed to their presently installed system
which produces the lowest blowdown amount per unit of produc-
tion of any of the plants studied and is probably one of the
lowest in the world.
The degree of existing control facilities installed is an
important consideration since the newer plants, due to the tech-
nology not previously available and to recent concerns for pro-
tecting the environment, installed facilities to treat their
wastewater to ,a degree which usually meets the BPT requirements
and, in some cases, even the BAT limitations. Plant locality
also has a great effect since plants located near abundant
290
-------
supplies of water were more apt to exclude facilities for waste-
water treatment and reuse.
The State in which a plant is located also has an effect
since, prior to the formation of the U.S.E.P.A., the States
were the sole governing bodies that determined the extent to
which a particular plant had to reduce its discharge of conta-
minants. In some States the restrictions were stricter, thus
resulting in steel plants with more treatment facilities than
those required in other States.
Increase in the Cost of Steel
In 1978 steel products ranged in cost from approximately
$385 to $440 per kkg ($350 to $400 per ton). This variation
is due basically to the wide range of products offered. A cost
of $413 per kkg ($375 per ton) has been used as an average in
order to calculate the cost of the various steps on the price
of steel. The added cost due to BAT will be approximately
$2.67 per kkg ($2.42/ton). Total recycle, excluding non-contact
cooling water, will cost approximately $13.15 per kkg ($11.93
per ton) and total recycle including non-contact cooling water
will cost approximately $16.91 per kkg ($15.34 per ton). This
represents an increase of 0.65 per cent in the cost of raw
steel produced for BAT, 3.2 per cent for total recycle exclud-
ing non-contact cooling water and 4.1 per cent for total re-
cycle including non-contact cooling water.
IN-PLANT EFFECTS
The goals of BAT and total recycle would result in large
expenditures for the construction of water treatment and reuse
systems. These large construction projects, if implemented,
will most probably have a disrupting effect on the operations
291
-------
of the steel plants during construction and, in some of the more
crowded plants, even after the construction is completed.
The transportation of chemicals, sludges, oils, etc.,
within the plants would increase with inherent increased traffic
problems. Safety requirements would require broadening to en-
compass the use of different chemicals and the use of new types
of water treatment process equipment. Monitoring of water sys-
tems would be expanded so that water qualities of the tightly
"bottled-up" systems are not upset causing outages of produc-
tion facilities.
The management of sophisticated water systems in well di-
versified integrated steel plants would in itself be an extreme-
ly complex problem.
EXTRA-PLANT EFFECTS
Whenever extensive and ambitious projects are undertaken in
an industrial plant or in an industry as a whole, effects of
these projects are felt not only within the plant or industry
itself bu.t also external to the plant.
Power Generation
It has been assumed that the electric power required to
operate the facilities for attaining BAT and total recycle
would be generated off-site. An average of the power required
for BAT and total recycle including NCCW for the four most "ty-
pical" plants is 57.5 x 106 j per kkg (14 .5 kWh per ton) and 262
j per kkg (66 kWh per ton), respectively. If this average is
applied to the total U.S. steel,industry, a total of 260 MWe
and 1,183 MWe of new generating capacity will be required for
BAT and total recycle, respectively. Implementation of total
292
-------
recycle were implemented within the next ten years would re-
quire an increase in expected electrical generation needs of 0.5
per cent over the present predictions for the steel industry.
This additional requirement would account for 0.8 per.cent of
the total industrial use of electricity by the year 1987.
Water Loss
Water will be lost due to evaporation under the require-
ments of total recycle. The loss to the atmosphere of the ad-
ditional amount of water may have detrimental effects on the
meteorology of nearby areas.
SUGGESTED RESEARCH
In the formulation of the various possible means of attain-
ing the BAT and total recycle, wastewater treatment processes
have been shown which have not been tested on a full scale ba-
sis and, in some cases, smaller scale tests have not been per-
formed. Use of these processes, however, was necessary because
existing proven technology within the steel industry to attain
this goal does not exist for total recycle and, although it is
available for BAT in the main, certain areas such as the tin
plating process do not possess this proven technology.
Whenever technology is suggested for application to an
industry where it has no-t been previously proven, there is
great and justified concern expressed. These concerns are
justified by the fact that industry cannot spend large amounts
of money to build facilities which they feel may never operate
successfully. It is, therefore, mandatory that research pro-
grams be initiated prior to any decision to impose the re-
quirement of total recycle. Some areas of needed research
293
-------
are: .multi-step biological treatment of by-product coke plant
wastewaters, treatment of blast furnace gas washer system blow-
down and treatment of wastewaters to remove dissolved solids.
It is assumed that the zero discharge requirement for tinning
operations will be changed in the present review of the guide-
lines. If this is not accomplished, research in this area will
also be needed.
294
-------
to
\o
Cn
YOUNGSTOWN SHEET 8 TUBE'
INDIANA HARBOR WORKS
INLAND STEEL
INDIANA HARBOR WORKS
NATIONAL STEEL
WEIRTON STEEL DIV.
•KAISER STEEL
FONTANA WORKS
UNITED STATES STEEL
FAIRFIELD WORKS
LOCATIONS OF SELECTED INTEGRATED PLANTS
Fig. I
-------
ro
COKE
a
BYPRODUCTS
PLANT
i
1
BIO
TREATMENT
DISCH
ARGE
^
BLO\
j>
NDO\
f
i
r
IRON
MAKING
a
SINTER
PLANT
»
RECYCLE
TREATMENT
VN -
1
SLOWDOWN
TREATMENT
1
DISCH
ARGE
4
k
1
STEELMAKING
•»
RECYCLE
TREATMENT
^
MINI
BLOW
T
DISCH
MUM
DOWN
0
ARGE
i
L
1
PRIMARY
HOT ROLLING
a
CONTINUOUS
CASTING
i
RECYCLE
TREATMENT
i
DISCH
ARGE
'I
SECONDARY
HOT
ROLLING
RECYCLE
TREATMENT
BLOW
COLD
FINISHING
^
TREATMENT
' i
DOWN DISCH
ARGE
TYPICAL INTEGRATED STEEL PLANT
RECOMMENDED SYSTEMS: FOR BAT LIMITATIONS
(DOES NOT SHOW NON - CONTACT COOLING WATER)
FIG. 2
-------
vo
COKE
a
BYPRODUCTS
PLANT
l
>
BIO
TREATMENT
t
r
^
IRON
MAKING
a
SINTER
PLANT
1
RECYCLE
TREATMENT
SLOWDOWN
^
t
i
SLOWDOWN
TREATMENT
1
t
SALTS TO STORAGE <
p
>
STEELMAKING
i
RECYCLE
TREATMENT
i
' — .
4 '
t
1
•»
i
PRIMARY
HOT ROLLING
a
CONTINUOUS
CASTING
i
f
RECYCLE
TREATMENT
DEMORALIZATION
i
A
1
SECONDARY
HOT
ROLLING
i
RECYCLE
TREATMENT
COLD
FINISHING
^
TREATMENT
•• •*> DEMINERALIZED WATER TO REUSE
r
TYPICAL INTEGRATED STEEL PLANT
RECOMMENDED SYSTEMS FOR TOTAL RECYCLE
(DOES NOT SHOW NON - CONTACT COOLING WATER)
FIG. 3
-------
TABLE I
DISSOLVED SOLIDS REMOVAL
N3
vo
Oo
Ion Exchange
Pretrealment Treatment Evaporation Solids Disposal Totnl System
Costs Costs Costs'- Costs-!' Costs
($x!06) ($xl06) ($ x 106) ($xl06) ($ x 106)
Annii.il Enerey
Roqui rcments
Capital Annual Capital Annual Capital Annual Capital Annual Capital Annual J x 10 J x 10
(kWhxIO6)
1.15 0.25 14.0 8.78 12.18 18.99
Reverse Osmosis 9.95 1.83 10.1 2.63 19.12 29.87
Electrodialysis 9.95 1.83 9.0 3.08 15.48 23.53
Total Evaporation
73.29 103
17.6 27.33
45.62 12.24 7. r,35
(34) (7.23)
10.2 39.17 4-1.53
8.45 34.43 36.89
40.8 73.29 143.B
1H.97 12.77f.
(52.7) (12.1)
11.41 10. 18
(31.7) (9.64)
9. 4 511. 104
(26. 1) (4R4)
* Includes cost of flue gas desulfurization.
'•* 'Assumption is that land would not be available on site and that solids would be hauled 5 miles off site,
Annual costs include amortization at 10 percent over 15 years plus operations and maintenance.
-------
TABLE II
SUMMARY OF PLANT COSTS TO MEET BAT AND TOTAL RECYCLE
Plant
Phase
Capital Annual Plant Capacity Addl Annual
Costs ? Ccsts 5 kkg/yr (ton/yr) Cost S/kkg (ton)
Kaiscr-
For.tana
BAT
Total Recycle
y/o NCCW
Total Recycle
w/ NCCK
17,717.000
3,267,000
(3,600,000)
9,762,000
2.99 (2.71)
Inland
Steel
Corp. -
Indiana
Harbor
Korks
BAT
Total Recycle
w/o NCCW
Total Recycle
w/ NCCW
36,300,000 18,823,000 1.91 (1.73)
94,172,000 75,235,000 9,866,000 7.63 (6.92)
(10,877,000)
162,079,000 139,675,000 14.18 (12.86)
National
Steel -
Keirton
Steel
Division
BAT
Total Recycle
w/o NCCW
Total Recycle
w/ NCCW
24,051,000 10,298,000
120,633,000 125,595,000 3,912,000
(4,312,000)
129,814,000 129,933,000
2.63 (2.39)
32.11 (29.13)
33.21 (30.13)
United
States
Steel -
Fairfield
Works
BAT
Total Recycle
w/o NCCW
Total Recycle
w/ NCCW
7,760,000
5,559,000
2,208,000
(2,434,000)
59,192,000 69,344,000
2.52 (2.28)
31.41 (28.49)
Youngstown
Sheet t
Tube -
Indiana
Harbor
Works
BAT
Total Recycle
w/o NCCW
Total Recycle
w/ NCCW
19,580,000 23,648,000 3.95 (3.SB)
65,880,000 59,172,000 5,993,000 9.87 (8.96)
(6,606,000)
74,350,000 64,571,000 10.77 (9.77)
Totals'
BAT* 79,931,000 .52,769,000
Total Recycle
w/o NCCW
Total Recycle
w/ NCCW
366,243,000 334,379,000
2.67 (2.42)
280,685,000 260,002,000 19,771,000 13.15 (11.93)
(21,795,000)
16.91 (15.34)
NOTES: 1. Costs shown for total recycle with and without non-contact cooling
water include costs of BAT
2. 'Totals do not include Kaiser Fontana and USSC-Fairfield.
3. NCCW is non-contact cooling water.
299
-------
USE OF SPENT PICKLE LIQUOR TO REMOVE THE PHOSPHATES
IN MUNICIPAL SEWAGE TREATMENT PLANTS
by
B. J. Kerecz, Jr., Engineer, Environmental Control & Coal Conversion Section,
Research Department, Bethlehem Steel Corp., Bethlehem, Pennsylvania 18016
R. T. Mohr, Plant Manager, Back River Wastewater Treatment Plant,
Baltimore, Maryland 21224
W. F. Bailey, Chief, Secondary Treatment, Wastewater Division,
Blue Plains Wastewater Treatment Plant, Washington, D. C. 20032
ABSTRACT
Spent pickle liquor (a ferrous sulfate waste product) from Bethlehem
Steel's Sparrows Point steel-pickling operations was tested as a source of
iron for removing phosphorus from municipal wastewater at the City of
Baltimore's 185 Mgd Back River Wastewater Treatment Plant and at the
District of Columbia's 300 Mgd Blue Plains Sewage Treatment Plant. The
study demonstrated that the pickle liquor is effective in reducing the total
and soluble phosphorus concentrations in the final effluents from trickling
filter systems and from both high-rate and conventional activated-sludge
systems. Continuing and full-time use of the waste pickle liquor (WPL) at the
two plants would be of mutual benefit. Bethlehem Steel Corporation would be
able to minimize disposal problems, and the wastewater treatment plants would
realize significant operating cost savings by using WPL rather than other
treatment chemicals.
300
-------
USE OF SPENT PICKLE LIQUOR TO REMOVE THE PHOSPHATES
IN MUNICIPAL SEWAGE TREATMENT PLANTS
INTRODUCTION AND SUMMARY
Bethlehem Steel Corporation has a daily supply of about 100,000 gallons
of waste pickle liquor (WPL) available at its Sparrows Point Plant, a large
integrated steelmaking facility, with a production capacity of about 7,000,000
ingot tons of steel per year. The steel plant, located on the Patapsco River
close to where it flows into the Chesapeake Bay, is about ten miles southeast
of Baltimore, Maryland and 50 miles north of Washington, D. C.
The WPL, a dilute sulfuric acid solution which contains ferrous sulfate
(FeSO,), is spent acid from the continuous pickling baths in which steel
products are cleaned. It contains about 5.5% ferrous iron and 9% free
acidity on a weight basis and has a specific gravity of about 1.2 (weighs
about 10 lb/galIon). The iron-rich liquor is chemically clean, i.e., it is
low in heavy metals and suspended solids as compared with some of the other
chemical additives commonly sold to sewage treatment plants for phosphorus
removal and sludge conditioning.
Sparrows Point was interested in finding beneficial uses for all or part
of the WPL rather than dealing with it merely as a waste disposal problem. To
this end, Bethlehem approached Baltimore and Washington, D. C. with the
proposal that the WPL be tested as an agent for phosphorus removal at their
municipal sewage treatment plants. In cooperation with municipal and regulatory
officials, Bethlehem planned extensive trials of the WPL as a phosphorus
precipitant at the 185 Mgd Back River Wastewater Treatment Plant and the
300 Mgd Blue Plains Wastewater Treatment Plant. Trials run during a
3-1/2-month period in 1978 at the Baltimore plant and during the summer and
fall months of 1979 at the D. C. plant proved the effectiveness of the proposed
method of phosphorus removal:
• At Back River the total and soluble phosphorus contents of the activated
sludge final effluent were reduced to less than 0.5 mg/1 and 0.2 mg/1,
respectively; the trickling-filter final effluent total and soluble
phosphorus concentrations were reduced to 3.2 mg/1 and to <1.0 mg/1,
respectively.
• In high-rate activated sludge treatment at Blue Plains the final
effluent total and soluble phosphorus concentrations were as low as
0.33 mg/1 and 0.11 mg/1, respectively.
Trial results show that the addition of WPL did not significantly lower the
pH of the final effluent from either sewage treatment plant. At Baltimore for
every 5 mg/1 of WPL iron tested, there was a decrease of about one-tenth of a
pH unit; at Blue Plains, where ferric chloride was being replaced with equivalent
WPL iron units, pH changes were negligible. Only about 1 gallon of WPL had to
be added to every 5,000 gallons of sewage to obtain iron dosages of 10 mg/1,
and there was sufficient alkalinity in the sewage to maintain final effluent
pH levels around 7.
301
-------
Subsequent to the successful test program at Back River, and through
cooperative efforts among all concerned parties, the City of Baltimore formally
approved a contract on August 16, 1979 for the use of 40,000 gpd of Sparrows
Point pickle liquor at the Back River facility starting in early 1980.
Bethlehem and Washington, D. C. officials are presently negotiating a contract
for utilizing WPL at Blue Plains.
WFL TRIAL AT BACK RIVER WASTEWATER TREATMENT PLANT
The City of Baltimore's Back River Wastewater Treatment Plant has received
considerable publicity during the past ten years because of environmental
problems in Back River caused in part by phosphorus in the final effluent from
the facility. Since the city was interested in developing near- and long-term
methods which combined favorable economics and sound technology for upgrading
the facility and minimizing phosphorus pollution of Back River, it was receptive
to Bethlehem's recommendation for a cooperative study to determine the efficacy
of using WPL in the Back River facility.
As a first step, the city agreed early in 1978 to assist Bethlehem
conduct bench-scale tests to determine optimum ferrous iron dosage rates and
injection locations for the WPL. According to this preliminary study, adding
WPL to the primary effluent at 15 to 20 mg of ferrous iron per liter of sewage
in conjunction with 0.2 mg/1 of polyelectrolyte for solids coagulation would
provide maximum removal of phosphorus and suspended solids in the full-scale
facility. Guided by plant layout and results of this study, Bethlehem, the
City of Baltimore and the Maryland Department of Natural Resources agreed
upon a full-scale WPL demonstration.
Commencing in May 1978 the demonstration was conducted in four phases
in terms of the ferrous iron concentration:
• 5 mg/1 in May
• 10 mg/1 in June
• 15 mg/1 in July
• 20 mg/1 in August
Sparrows Point committed over $300,000 to storage and injection facilities
for WPL and polymer and to purchase polymer and pay trucking costs of the
spent acid to Back River. Baltimore obtained a grant from the EPA for $70,000
for the additional operating and manpower costs the city was to incur during
the trial.
The Back River Plant,_Figure 1, is a biological treatment facility which
handles an average of 185 Mgd of wastewater from Baltimore City, Baltimore
County, and other bordering municipalities. The sewer collection system
serves an industrial-residential area of about 140 square miles supporting a
population of 1,385,000. The sewage contains about 6 to 10 mg of total
phosphorus per liter of sewage and is moderately strong as far as biochemical
oxygen demand (BOD), nitrogen and suspended solids are concerned.
302
-------
Primary treatment facilities at the plant precede the biological treatment
facilities. Primary treatment consists of mechanical screening, grit removal,
and primary settling. After primary treatment the sewage undergoes biological
treatment, final clarification, and chlorination. Biological treatment is
divided between an older trickling filter system which handles about 135 Mgd,
and a newer activated-sludge treatment facility, which treats 50 Mgd. The
treated effluent is clarified in circular tanks following the activated-sludge
system and in rectangular and circular humus tanks following the trickling
filters. Solids retention times of about 3-1/2 days are usually maintained in
the activated system. Average hydraulic retention time for this system is
about six hours. All excess activated sludge is returned to gravity thickener
tanks or to an inlet chamber along with humus-tank solids ahead of the grit
chambers and primary clarifiers.
The Back River Plant discharges its treated effluents at two points:
(a) 55-85 Mgd to the head of Back River (outfall 001), a brackish tidal
river adjacent to the facility, and (b) the remaining 100-130 Mgd is
discharged by gravity to the Sparrows Point facility (outfall 002) through
about 6 miles of Bethlehem-owned pipelines. At Sparrows Point this effluent
serves as multi-purpose industrial water in finishing mills and plant areas.
The treatment of the 185 Mgd of sewage at Back River results in the
daily generation of 300-400 wet tons (60 to 80 dry tons) of filtercake solids
which are disposed of at a private landfill adjacent to Back River property.
The solids-handling process consists of gravity sludge thickeners, high rate
anaerobic digesters, elutriation tanks, and vacuum filters.
The rationale for Back River Plant expansion to meet future effluent
limitations, such as for phosphorus, is being developed under a federally
sponsored 201 Facilities Plan. The Back River effluent limitations for
post-1981 have not been finalized by EPA or Maryland's Department of Natural
Resources. However, Baltimore is aware that a phosphorus limitation of from
2.0 mg/1 to perhaps as low as 0.2 mg/1 may be imposed on the final effluent
from the facility, depending on the technical and economic outcome of the 201
study.
During the 110-day plant trial, about 3.5 M gallons of pickle liquor
was injected under controlled rates into 20 billion gallons of sewage. An
average truckload of WPL shipped from Sparrows Point during the trial contained
4,300 gallons, weighed 44,300 pounds, and averaged about 5.5% ferrous iron and
9% acidity by weight.
WPL delivered to the site was sampled routinely and analyzed for iron and
acidity. The injection rate was dependent upon the analyzed iron concentration,
the primary effluent flow, and the iron concentration being tested. The
primary effluent flow was obtained from a totalizing meter near the injection
site. The daily WPL use was determined from an integrating meter at the
injection site along with a tabulation of shipping manifests and storage tank
readings.
303
-------
The WPL was pumped through a distribution manifold submerged in the
primary effluent channel at rates required to attain 5 to 20 mg/1 iron
concentrations in the sewage going to biological treatment. This manifold was
located immediately upstream of a second manifold through which air was being
sparged for thorough mixing of the WPL and sewage. As primary effluent flows
or the iron concentration changed, the WPL addition rate was varied manually
by valve changes on the pressure side of the injection pumps.
Grab and composite samples, obtained from various locations in the
treatment plant three times daily, were analyzed by Bethlehem Steel personnel
and Back River laboratory personnel for pH, suspended solids, soluble and
total iron, and soluble and total phosphorus. The grab-sampling locations
analyzed by Bethlehem were: primary effluent pre-WPL injection, primary
effluent post-WPL injection, activated-sludge final effluent, and final
effluent from the trickling filter (humus tank). In addition to the grab
samples, 24-hour composite samples were obtained by Back River personnel for
analyses from various locations and were analyzed for pH, TSS, BOD, total
nitrogen, orthophosphorus, total phosphorus, and iron.
During most weekend and holiday periods, supply problems developed at
Sparrows Point because of WPL shortages in the 60,000-gallon storage tank.
Pickling operations generally are shut down for these periods, and WPL consumption
at Back River exceeded production. During these periods it was standard
operating practice at Back River to reduce WPL addition rates until production
at Sparrows Point increased. These supply problems resulted in lower iron
concentrations, especially for July and August weekends and holidays when
large volumes of spent acid were utilized to obtain iron concentrations of 15
and 20 mg/1. Daily WPL injection rates for the 3-1/2 months of the trial
averaged 18,000, 30,000, 36,000, and 49,000 gpd for May, June, July, and
August, respectively* The lowest daily addition rate was 10,500 gallons on
May 28, 1978, and the highest rate was 70,200 gallons on August 15, 1978.
Table I lists the average monthly WPL injection rates; and Table II summarizes
major operational changes during each period.
Results At Back River Sewage Treatment Plant
Phosphorus Removal. Figures 2 and 3 present monthly average total and
soluble phosphorus concentrations for the period April through August 1978 for
the activated-sludge and trickling-filter systems. The data clearly show the
steady drop in effluent phosphorus concentrations as iron concentrations in
the system increased stepwise from 5 up to 20 mg/1.
The average monthly total phosphorus concentrations in the activated-
sludge effluent during 1977 varied from a low of 1.5 mg/1 when only about 30
Mgd of sewage was processed in the activated-sludge system, to a high of
3.5 mg/1 in March of 1978 when about 50 mgd of sewage was processed. During
1977 the average total phosphorus concentration in the activated effluent was
2.2 mg/1.
304
-------
Figure 2 shows the effect on the activated sludge effluent of adding
various levels of ferrous iron to the primary effluent. Total phosphorus was
reduced from 2.1 mg/1 in April 1978 down to a 0.85 mg/1 for May, representing,
respectively, the pre-trial month when no WPL was added and the first month of
the trial when only 5 mg/1 iron was added to the system. The average total
phosphorus concentrations of the activated effluent for June, July, and
August, the 10, 15, and 20 mg/1 ferrous iron addition periods, fell to 0.78,
0.50, and 0.47 mg/1, respectively. The hydraulic feed to the activated system
was maintained at 45 to 50 Mgd for the 3-1/2 month trial.
The soluble phosphorus concentrations of the activated-sludge effluent
were reduced from a 1.24 mg/1 for the pre-trial April period to 0.7, 0.4,
0.18, and 0.14 mg/1 at respective monthly iron addition rates of 5, 10, 15,
and 20 mg/1.
The total phosphorus concentration of the trickling-filter final effluent
averaged 7.1 mg/1 during 1977 when no ferrous iron was added to the system.
As shown in Figure 3, WPL ferrous iron addition to the trickling-filter system
significantly reduced this level with increasing additions of the iron. The
trickling-filter phosphorus data indicate that with more efficient capture of
suspended solids in the humus tanks, total effluent phosphorus concentrations
could be still further reduced because almost all of the phosphorus exiting
the final trickling-filter clarifiers during the latter stages of the WPL
trial was particulate phosphorus•
The total phosphorus effluent concentration was reduced in the trickling-
filter facility from 7.3 mg/1 at the start of the trial to successive monthly
averages of 5.8, 4«3, 3.9 and 3.2 mg/1 at ferrous iron additions of, respectively,
5, 10, 15 and 20 mg/1. The soluble phosphorus levels of the trickling-filter
effluent were reduced even more significantly than the total phosphorus
concentrations. With increasing iron additions, the soluble phosphorus level
was reduced from 4.8 mg/1 at the start down to 0.9 mg/1 at the 20 mg/1 iron
dose.
Biochemical Oxygen Demand, Five-Day Demand (BOD^). Figure 4 presents
monthly average BOD- final effluent concentrations for both the activated-
sludge and trickling-filter systems. The BOD5 concentrations for 1977
averaged 13 and 45 mg/1, respectively. As shown in Figure 4, BOD- removal
for the activated-sludge system improved as the ferrous iron addition levels
were increased from 5 to 20 mg/1 during the 3-1/2 months. The BOD_ average
during the 20 mg/1 iron addition was less than 5 mg/1. While it appears that
iron addition did improve BOD_ removal in the trickling-filter system, the
improvement is not as obvious as that in the activated-sludge system.
305
-------
Figure 4 shows that the effluent BOD5 concentration of the trickling-
filter system remained fairly constant, averaging between 25 and 35 mg/1
during the study. The lowest BOD. average of 25 mg/1 occurred during the
20 mg/1 iron addition and was the lowest monthly average in the two years.
Some of this apparent improvement in BOD_ removal during 1978 may be attributed
to the decreased load on the trickling filters as a result of the start-up of
a new 25 to 30 Mgd activated-sludge facility in January 1978.
Soluble Iron. Figures 5 and 6 give the average monthly soluble-iron
concentrations for April 1978 through August 1978. As shown in Figure 5,
the soluble iron levels in the activated-sludge effluent did not increase
significantly throughout this period, being about 0.13 mg/1 in April and
decreasing to 0.11 mg/1 during May when 5 mg/1 iron was being added to the
system. June and July showed soluble-iron increases to 0.17 and 0.25 mg/1,
respectively, with a decrease to 0.14 mg/1 for August when 20 mg/1 of ferrous
iron was being added*
Figure 6 shows that the soluble-iron concentration of the trickling-filter
effluent increased from a low of 0.3 mg/1 in April to 0.53 mg/1 for May. The
soluble-iron concentration increased to 0.6 mg/1 during June and, with subsequent
iron increases, rose to 0.76 and 0.9 mg/1 for July and August.
While the soluble-iron concentration in the effluent from the humus tank
increased during the trial, this concentration was not at a level that represents
a serious problem for the environment. Considering the age of the trickling-filter
bed, precipitation of better than 95% of the applied iron load was excellent.
Baltimore will renovate the trickling filters if the city decides not to
abandon them as a result of the 201 study.
Suspended Solids. Figure 7 presents monthly average suspended-solids
concentrations in the final effluents during 1977 and 1978. These averages
during 1977 for the effluents from the activated-sludge and trickling-filter
humus tanks were 11 and 46 mg/1, respectively. As compared with the 1977-1978
background periods, there was a slight improvement in suspended-solids removal
for the activated-sludge system, but the performance of the humus tanks
declined. Because of various plant conditions related to suspended solids at
the time of the trial, it is difficult to determine the direct relationship
between ferrous iron addition and the concentrations of effluent suspended
solids in the humus tanks. The suspended solids of the trickling-filter
effluent for April 1978, (the pre-trial background month) was 60 mg/1. The
average for the month of May shows an increase to 87 mg/1 followed by decreases
in June, July and August. The August average was 68 mg/1, i.e., 8 mg/1 above
the April background period.
306
-------
The Back River Plant experiences sludge-thickening and sludge-handling
problems that become acute during spring and summer months, but less severe
during cooler months. These problems could result from increases in sewage
temperature, and the concomitant increased biological activity can impair the
settleability of the sludge by making the solids more buoyant as gas is
produced. During such periods, solids build up within the system. With an
increased load of poorly settling solids, some of which are colloidal, being
returned to the primary settlers as recycle, primary performance deteriorates
and this increases the suspended-solids load on the biological processes. The
result, generally, is that higher suspended-solids levels are discharged from
the humus tanks but not necessarily from the activated-sludge system. Since
this condition was experienced at just the time when the WPL trial was started
in May, it was not possible to document any relation between ferrous iron
addition and the amount of suspended solids in the humus tanks.
During the trial, the addition of WPL did result in the production of
additional sludge. Baltimore is installing four additional vacuum filters,
which should be operational in 1980 and will help remove solids.
Operating data obtained from Back River show that the precipitated WPL
iron solids did have a beneficial effect on sludge-handling processes. Tables
III and IV present performance and cost data for sludge conditioning and
disposal. Sludge-conditioning costs per dry ton of sludge handled during the
trial were significantly reduced as compared with comparable costs in the
background period. Average conditioning cost for the 16-month pre-WPL period
was $8.60/dry ton of sludge handled in contrast with $6.28 during the trial.
Another benefit of the higher concentration of iron in the sludge was a
significant improvement in vacuum filtration. Monthly averages for the
filtercake percent solids in the 16-month pre-WPL period were 18.9% as compared
with 20.3% for the period of the trial. The percent solids in the filtercake,
which did not reach 20% during any of the background months, increased to
20.0, 20.6 and 21.1% in June, July and August, respectively.
The vacuum-filter yields also increased significantly during the trial,
being as high as 5.7 Ib of filtercake/sq ft of filter area/operating hour.
The monthly averages (Table IV) for the pre-WPL period averaged only 4.0 as
compared with 5.1 Ib of filtercake/sq ft of filter area/operating hour.
WPL TRIAL AT BLUE PLAINS WASTEWATER TREATMENT PLANT
The Blue Plains Sewage Treatment Plant in Washington, D. C. handles
an average of 300 Mgd of medium-strength sewage from primarily residential
and office areas. The facility, located on 150 acres along the Potomac River
near the southernmost end of the District of Columbia, serves an area of 725
square miles with a population of about 2,200,000. Less than half the sewage
handled in the Blue Plains system originates in the District of Columbia,
about 50% comes from Maryland and 6% comes from Virginia.
307
-------
The Blue Plains Plant, (Figure 8) consists of primary treatment facilities,
i.e., bar screens, aerated grit chambers and circular primary sedimentation
tanks, as well as biological treatment facilities. The latter consist of six
four-pass rectangular aeration tanks providing about 3.5 M cubic feet of
aeration volume. Sludge ages of about one day are usually maintained in the
activated-sludge system. Hydraulic retention time for aeration averages two
hours. Aeration equipment includes submerged diffused-air aerators operated
in conjunction with centrifugal blowers. Dissolved oxygen levels of at least
2.0 mg/1 are maintained during biological treatment.
Floe formed in biological treatment by the oxidation of carbonaceous
materials and by the agglomeration of colloidal sewage matter is settled in
rectangular clarifiers. The practice at Blue Plains has been to add, prior to
final clarification, ferric chloride and polymer for phosphorus removal and
solids coagulation. Ferric chloride has generally been added at dosages of
15-21 mg/1 of ferric chloride (5-7 mg/1 as iron). The ferric chloride,
purchased from an outside company, costs Blue Plains about 22 cents/lb of
iron. The chemical has a specific gravity of 1.4 (weighs about 11.6 Ib/gal)
and contains 8.5-10% ferric iron by weight. It is added to the aeration basin
effluent in a well-mixed compartment positioned after the basins but ahead of
the final clarifiers. To strengthen the iron floe an anionic polyelectrolyte,
generally at a dose of 0.3 mg/1, is added after iron addition but ahead of
final clarification.
Excess biological and chemical solids from activated-sludge treatment
are wasted, i.e., removed, from clarifier underflows to air-flotation or
gravity thickeners. After thickening the waste activated sludge is blended
with gravity-thickened primary sludge and split into two streams. The major
portion of the sludge is vacuum filtered as is. The remainder is digested and
elutriated before dewatering. Prior to vacuum filtration the sludge is
conditioned with lime and/or ferric chloride and/or polyelectrolyte.
Secondary effluent is chlorinated and then discharged to the Potomac River.
In recent years more than $400,000,000 has been spent on expansion of
the plant. The major construction projects under way are facilities for
nitrification and mixed-media filtration. Air-flotation thickeners and vacuum
filters in the new solids processing building have been in service for about a
year now.
The nitrification facility will be on line by 1980, and the mixed-media
filtration should be operational by 1981. The addition of these advanced
wastewater treatment facilities will result in an estimated daily generation
of 400 wet tons (80 dry tons) of additional sludge which will increase the
sludge production rate to about 2000 wet tons/day (400 dry tons/day).
308
-------
A sludge/wood chip composting program at Blue Plains is handling about
20% of the sludge production, and the plan is to sell the composted mixture
for agricultural purposes. The remaining 80% of the sludge produced, about
1,000 tons/day, is trucked to approved disposal sites in Maryland, where
the material is trenched and covered in landfills by private contractors.
There are two activated-sludge aeration basins on the west side of the
Blue Plains facility and four on the east side. To provide maximum aeration
time for ferrous oxidation, all WPL during the trial was added to the head end
of aeration basin 1 or 2 on the west side. The second aeration basin was used
for control purposes, usually at a ferric iron dose equivalent to the ferrous
iron dose that was being added in the other basin. Iron concentrations of 5,
7, and 10 mg/1 were tested in detail during the May through August periods in
1979. Bethlehem Steel Corp. test work on-going through mid-October is expected
to be completed by mid-November 1979.
Table V compares the WPL and ferric chloride data. During both the
background and trial periods covered, 0.3 mg/1 of polymer helped promote
suspended-solids coagulation.
During April and May, pre-WPL trial data were collected. The primary
effluent for the activated systems on the west side averaged 4.8 mg/1 total
phosphorus, 2.9 mg/1 soluble phosphorus, and 81 mg/1 total suspended solids.
The final effluent from the odd and even final clarifiers, when 5 and 7 mg/1
ferric iron were added to the system, averaged 1.4 mg/1 total phosphorus, 0.7
mg/1 soluble phosphorus, and 22 mg/1 total suspended solids.
May 8 through May 20, 1979, ferrous and ferric iron dosages of 5 mg/1
were tried in the two west-side activated-sludge systems and compared* WPL
was added to aeration basin 1, and ferric chloride was added to basin 2.
Total phosphorus concentrations in the final effluent were the same for both
materials, 1.6 mg/1, but the soluble phosphorus concentration was better for
the WPL-treated basin, i.e., 0.38 mg/1 as compared with 0.58 mg/1 for the
basin treated with ferric chloride.
For the 7 mg/1 iron addition period, total phosphorus concentrations in
the final effluent averaged 1.06 mg/1 for the ferrous iron side and 0.88 mg/1
for the ferric iron side. However, soluble phosphorus concentrations were
0.22 mg/1 for the ferrous iron side and 0.26 mg/1 for the ferric iron side.
The total suspended-solids averages were 24 mg/1 and 23 mg/1, respectively.
There were two 10 mg/1 ferrous iron periods, one in July when ferrous iron
was added to aeration basin 2, and the second in August when the ferrous iron
was added to aeration basin 1. For total phosphorus and suspended solids,
removal by ferrous iron during the first 10 mg/1 iron addition period (July)
was not as good as that during the August period (Table III). Although there
are some possible explanations for the poorer period, the important point is
that the July data are not typical when viewed against the performance of the
various periods to date.
309
-------
For the August period, final-effluent total and soluble phosphorus
concentrations averaged 0.33 mg/1 and 0.11 mg/1 for the ferrous iron addition,
and 0.39 mg/1 and 0.13 mg/1 for the ferric chloride addition. Total suspended-
solids concentrations in the final effluent were also less for the ferrous
iron-treated side, i.e., 6 mg/1 as compared with 9 mg/1 for the ferric chloride-
treated side.
Comparison of the performance of the ferrous iron dose of 6.7 mg/1 with
that by a 7.8 mg/1 ferric iron dose for the August 21-31 period (the last
reported in Table V) shows that on an iron-equivalent basis (Ib of iron
added/lb of P removed) the ferrous iron outperformed the ferric iron.
SUMMARY OF RESULTS
1. Back River Wastewater Treatment Plant,
trickling-filter and activated-sludge systems
• Total phosphorus content of the activated-sludge effluent was reduced
to less than 0.5 mg/1 of phosphorus in the final effluent. Soluble phosphorus
concentrations were reduced to less than 0*2 mg/1.
• Total phosphorus content of the trickling-filter effluent was
reduced to 3.2 mg/1 of phosphorus in the final effluent. Soluble phosphorus
was reduced to less than 1 mg/1.
• The biochemical oxygen demand (BOD.) of final effluents was reduced
to the lowest levels in recent years. BOD- removal was over 90%.
• Chlorine demand of the final effluent was reduced.
• Sludge characteristics were improved, resulting in a higher percent
solids in the filtercake. Percent solids averaged 18.9% for the 16 months
preceding WPL addition and 20.3% for the WPL trial.
• Vacuum-filter yields were improved. Filter yields for the background
period averaged 4.0 Ib of filtercake/sq ft of filter area/operating hour as
compared with 5.1 Ib for the trial.
• Polymer demand for sludge conditioning decreased, thus reducing
operating costs by $2.30/dry ton of sludge handled.
2. Blue Plains Wastewater Treatment Plant,
high-rate activated-sludge system
• WPL ferrous iron at a 10 mg/1 iron addition rate reduced the total and
soluble phosphorus concentrations in the final effluent to an average of 0.33
and 0.11 mg/1, respectively, during the August trial period.
• When oxidized to ferric iron in the aeration basins, ferrous iron
was an effective coagulant.
310
-------
CONCLUSIONS AMD RECOMMENDATIONS
Plant trials at the Back River and Blue Plains Sewage Treatment Plants
demonstrated the effectiveness of waste pickle liquor for complexing and
removing phosphates at both plants, and at Back River where chemicals are not
used in biological treatment, the waste pickle liquor improved the coagulation
and filtration of solids* Thus, it has been shown that steelplant waste
pickle liquor can accomplish those Improvements in municipal wastewater
treatment that have traditionally been realized by the use of other, comparatively
high-priced chemicals. It is therefore evident that the use of waste pickle
liquor can be mutually beneficial — a waste material is utilized that would
otherwise require some sort of treatment and disposal at steel plants, and
municipalities are offered the opportunity to minimize treatment costs.
The test programs discussed in this report exemplify commendable government/
industry cooperation to benefit the public as a whole, as well as a steel
corporation and two municipalities• We recommend that such cooperation should
continue and that the EPA should encourage municipalities to investigate and
utilize this treatment practice.
311
-------
TABLE I. BACK RIVER WASTEWATER TREATMENT PLANT
AVERAGE WPL INJECTION RATES
DATE AMOUNT
1978 GAL/DAY
MAY 18,000
JUNE 30,000
JULY 36,000
AUGUST 49,000
312
-------
TABLE II. BACK RIVER WPL TRIAL
SUMMARY OF OPERATIONAL CHANGES
Date
May 1, 1978
June 1
June 6
June 16
June 17
June 19
July 1
July 2
July 3
July 4
July 5
July 9
July 13
August 6
August 7
August 18
WPL trial began at 5 mg/1 ferrous iron addition.
Iron addition was increased to 7.5 mg/1.
Iron addition was increased to 10 mg/1*
WPL storage tank ruptured at 12 noon. WPL off.
Iron addition was resumed at 10 mg/1.
Polymer was discontinued at activated-sludge clarifier influent.
Iron addition was reduced to 2.5 mg/1 for part of the day.
Iron addition was increased to 5 mg/1.
Iron addition was increased to 10 mg/1.
Iron addition was reduced to 5 mg/1.
Polymer was discontinued in the humus-tank influent.
Iron addition was increased to 12.5 mg/1.
Iron addition was increased to 15 mg/1.
Iron addition was reduced to 5 mg/1.
Iron addition was increased to 20 mg/1.
WPL trial terminated in the afternoon.
313
-------
TABLE III. BACK RIVER SLUDGE CONDITIONING AND DISPOSAL COSTS
PRE-WPL PERIOD
Elutriation Polymer
Month
January 1977
February
March
April
May
June
July
August
September
October
November
December
January 1978
February
March
April
PRE-WPL AVG.
Month
May
June
July
August
Ib/dry ton
5.1
5.5
6.3
6.7
8.2
7.7
4.7
5.5
4.6
4.1
4.8
5.2
5.1
5.3
3.3
3.9
5.4
Elutriation
Ib/dry ton
4.1
4.8
3.4
3.2
$/dry ton
2.26
2.40
2.77
2.96
3.61
3.40
2.06
2.40
2.03
1.79
2.11
2.27
2.22
2.35
1.60
i^ZP.
2.40
Polymer
$/dry ton
1.82
2.13
1.49
1.38
Vacuum
Ib/dry
63.9
66.6
78.8
82.1
80.8
67.7
64.7
71.7
65.5
65.5
66.7
69.2
85.3
69.4
66.5
66.9
70.7
Filter Polymer
ton $/dry ton
5.65
5.88
6.90
7.25
7.13
5.79
5.71
6.33
5.78
5.77
5.89
6.11
7.53
6.13
5.87
5.91
6.23
Conditioning Cost
$/dry ton
7.91
8.28
9.67
10.21
10.74
9.19
7.77
8.73
7.81
7.56
8.00
8.38
9.75
8.48
7.47
7.61
8.60
Disposal Cost
$/dry ton
.
-
-
-
-
-
17.44
17.61
17.05
16.69
16.64
16.65
16.80
16.56
16.53
16.51
16.85
Total Cost
$/dry ton
—
—
—
-
_
25.21
25.47
24.89
24.26
24.51
24.96
26.49
25.03
23.86
24.12
24.88
WPL PERIOD
Vacuum Filter Polymer
Ib/dry
58.1
48.3
46.6
54.0
ton $/dry ton
5.13
4.27
4.12
4.77
Conditioning Cost
$/dry ton
6.96
6.38
5.61
6.16
Disposal Cost
$/dry ton
16.56
16.28
15.80
15.38
Total Cost
$/dry ton
23.52
22.65
21.41
21.53
WPL TEST AVG.
3.9
1.71
51.8
4.57
6.28
16.01
22.28
-------
TABLE IV. BACK RIVER ELUTRIATION AND VACUUM FILTER DATA
PRE-WPL PERIOD
Month
January 1977
February
March
April
May
June
July
August
September
October
November
December
£ January 1978
*"" February
March
April
PRE-WPL AVG.
Month
May
June
July
August
Elutriation Sludge
% solids
5.1
4.6
4.9
5.0
4.7
5.0
4.9
4.4
4.8
4.9
5.0
4.9
5.7
5.0
5.1
5.1
4.9
Elutriation Sludge
% solids
4.8
5.2
5.3
6.1
Filtercake
% solids
19.4
19.7
17.7
18.0
16.4
17.5
18.7
18.5
19.1
19.5
19.5
19.6
19.4
19.6
19.7
19.6
18.9
Filtercake
% solids
19.6
20.0
20.6
21.1
Filter Operation
hours /day
33.45
47.12
48.36
59.23
45.23
47.06
45.18
43.90
46.38
47.30
45.82
39.91
42.74
40.10
42.58
43.61
44.88
WPL PERIOD
Filter Operation
hours /day
43.61
49.46
44.4
45.38
Ib/ft2/hour
4.1
3.7
3.4
3.1
3.9
4.3
4.6
4.1
4.2
4.2
4.4
4.0
3.4
3.8
4.2
4.6
4.0
Ib/ft2/hour
4.1
5.0
5.7
5.5
Filter Yield
dry tons /day
49.0
63.4
59.6
67.6
62.4
72.7
74.2
64.4
69.6
71.5
71.6
57.6
52.0
54.3
66.4
75.85
64.5
Filter Yield
dry tons/day
69.3
84.7
90.3
90.0
wet tons /day
252
322
336
375
383
414
397
349
365
367
366
296
262
276
337
384
343
wet tons /day
353
425
439
440
WPL TEST AVG,
5.4
20.3
45.71
5.1
83.6
414
-------
TABLE V. SUMMARY OF BLUE PLAINS WPL TRIAL
Operating Period
Pre-WPL Trial
(4/1/79 to 5/7/79)
5 mg/1 Iron Addition
(5/8/79 to 5/20/79)
7 mg/1 Iron Addition
(5/22/79 to 6/4/79)
U)
10 mg/1 Iron Addition
(7/11/79 to 7/22/79)
10 mg/1 Iron Addition
(8/6/79 to 8/15/79)
Phosphorus, raj?/
Primary Effluent Aeration
P 4.8
total *'8
Soluble 2'9
total
Soluble 2*6
P 4.7
total
Psoluble 2*4
P 4.5
total
Soluble 2'2
Ptotal 3'9
Soluble 1>6
WPL
.
""
Basin 1
1.6
0.38
Basin 1
1.06
0.22
Basin 2
1.31
0.15
Basin 1
0.33
0.11
1 Total Suspended Solids, mg/1
Basin Effluent Primary Effluent Aeration Basin Effluent
FeCl3 WPL Feci3
Basins 1 & 2 _ Basins 1 & 2
1 A
J. . *f
81 - 22
0.6
Basin 2 Basin 1 Basin 2
1.6
98 39 37
0.58
Basln 2 Basin 1 Basin 2
0.88
91 24 23
0.26
Basin * Basin 2 Basin 1
0.48
94 36 12
0.11
Ba^n 2 Basin 1 Basin 2
0. 39
96 69
0.13
7.8 mg/1 Fe 2 Addition Basin 1
6.7 mg/1 Fe Addition Basin 2 P
(8/21/79 to 8/31/79) ai
Basin 2 Basin 1
0.65 0.59
89
Basin 2 Basin 1
18 15
-------
RAW SEWAGE INFLUENT { 185 M GPO 01Y WEATHER FLOW)
PRIMARY AND
ACTIVATED SOLIDS
METHANE
AS
OVERFLOW
TO PRIMARY
SETTLING TANKS
I
t
FILTER CAKE
TO LANDFILL
SCREENINGS
TO LANDFILL"
GRIT
TO LANDFILL-*
SETTLED SLUDGE _
TO THICKENERS -^ - — - -
TRICKLING FILTER EFFLUENT
, , FILTRATE
(WPL INJECTION MAY-AUGUST 1978)
RECYCLED SLUDGE
I
SLUDGE BLEED
-TO THICKENERS
TREATED EFFLUENT
TO BSCORP. -<-
(OUTFALL 002)
L .
r-
L^
HUMUS TANKS
_
ELUTI
\lt
r
UATION
TER
FINAL
C LARIMERS
1
TREATED EFFLUENT
TO BACK RIVER
(OUTFALL 001)
FIGURE I. FLOW DIAGRAM OF BACK RIVER UASTEWATER TREATMENT PLANT
-------
04
as
(0
ft
a
8 -
w
§
cu
«
i
PU
APRIL
MAY
JUNE
1978
JULY
AUGUST
FIGURE 2. BACK RIVER WASTEWATER TREATMENT PLANT
ACTIVATED SLUDGE EFFLUENT PHOSPHORUS
318
-------
APRIL
MAY
JUNE
1978
JULY
AUGUST
FIGURE 3. BACK RIVER WASTEWATER TREATMENT PLANT
TRICKLING FILTER FINAL EFFLUENT PHOSPHORUS
319
-------
?
o
CO
to
o
TRICKLING FILTER FINAL EFFLUENT
ACTIVATED SLUDGE EFFLUENT
HAY JUN JUL AUG SEP OCT NOW DEC JAN FEB HAR APR MAY JUN JUL
JAN FEB MAR APR
AUG
FIGURE <». BACK RIVER UASTEUATER TREATMENT PLANT BODj ANALYSES
MONTHLY AVERAGES OF DAILY COMPOSITE SAMPLES
-------
i-t
^**
i
» »
30NCENTRAT]
^^
i
M
SOLUBLE
• 1.0
- 0.9
-0.8
-0.7
-0.6
-0.5
-0.4
- 0.3
-0.2
<^—*
-0.1
1
1.0-
0.9 -
0.8 -
0.7 -
0.6-
0.5-
0.4-
0.3-
-A^ o.2-
~*~*Cf*^ o.i-
1 i I 1
APRIL MAY JUNE JULY AUGUST
1978
FIGURE 5. BACK RIVER WASTEWATER TREATMENT PLANT,
SOLUBLE IRON IN ACTIVATED SLUDGE EFFLUENT
321
-------
§
OT
PRE-WPL
-1.0
-0.9
-0.8
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
WPL TRIAL
1.0
0.9
0.8 -
0.7 -
0.6 -
0.5-
0.4 -
0.3
0.2
0.1-
APBIL
MAY
JUNE
1978
JULY
AUGUST
FIGURE 6. BACK RIVER WASTEWATER TREATMENT PLANT,
SOLUBLE IRON IN TRICKLING FILTER EFFLUENT
322
-------
u>
to
OJ
00
3
W
W
PM
s
en
90
80
70
60
. 50
- 40
- 30
- 20
TRICKLING FILTER
FINAL EFFLUENT
ACTIVATED SLUDGE
FINAL EFFLUENT
_L
J L
I I i i « » » I I I 1 L
_L
WPL TRIAL
i I I I I
90
80
70
60
50
40 -
30 -
20 -
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP
1977 1978
FIGURE 7. BACK RIVER WASTEWATER TREATMENT PLANT,
EFFLUENT SUSPENDED SOLIDS
-------
AIR
NITRIFICA
TION
BASINS
(FUTURE)
NITRIFICATION
SEOIMENTATIO
TANKS
(FUTURE)
IIXEO HCDIA
;IITRATION
(FUTURE)
>ISINFECTIOI
1
FINAL EFFLUENT
TO POTOMAC RIVER
SLUDGE COMPOSTING
AND TRENCHING
FIGURE 8. FLOW DIAGRAM OF BLUE PLAINS UASTEUATER TREATMENT PLANT
-------
PHYSICAL-CHEMICAL TREATMENT OF STEEL
PLANT WASTEWATERS USING MOBILE
PILOT UNITS
by
Richard Osantowski and Anthony Geinopolos
Rexnord Inc. Corporate R&D
Milwaukee, Wisconsin
Presented at the EPA Symposium on Iron and Steel Pollution
Abatement Technology, Pick-Congress Hotel, Chicago, Illinois,
October 31, 1979
325
-------
"PHYSICAL-CHEMICAL
TREATMENT OF STEEL PLANT
WASTEWATERS USING MOBILE
PILOT UNITS"
Richard Osantowski
Anthony Geinopolos
Environmental Research Center
Rexnord Inc.
Milwaukee, Wisconsin 53214
In-depth pilot scale evaluations investigating the applicability of advanced
waste treatment methods for upgrading steel mill wastewaters to Best Availa-
ble Technology Economically Achievable (BATEA) levels, which were proposed
in 1974*, were performed. The wastewater tested was a Blast Furnace
Category scrubber blowdown meeting 1977 Effluent Guidelines for Best Practi-
cal Control Technology Currently Available (BPCTCA). The advanced treat-
ment methods, both singularly and in combination which were investigated
on a pilot basis included: alkaline chlorination, clarification, filtration,
(dual media and magnetic), ozonation, activated carbon and reverse osmosis.
The residual conventional parameters monitored included pH, temperature,
suspended solids, BOD, oils and grease, phenol, cyanide, fluoride, ammonia,
sulfide and dissolved solids. Priority pollutant samples were also collected
from each process train investigated.
The studies were performed on-site using mobile pilot equipment designed to
operate at a flow rate of 19 Jl/min (5 gpm). Evaluation and comparison of
the data were performed using the criteria: (1) process and/or treatment
train performance, (2) capital and operating costs and (3) space require-
ments.
The results of the pilot program indicated that alkaline chlorination,
ozonation and reverse osmosis were effective in reducing influent contami-
nants to below future BATEA levels. For all three promising technologies,
proper pretreatment would be required. Costs associated with upgrading,
of the blast furnace scrubber blowdown are currently being finalized. This
project effort was funded by the Environmental Protection Agency - Industrial
Environmental Research Laboratory under contract number 68-02-2671.
The information and conclusions presented in this paper are based on pre-
liminary examination of data which will be subjected to extensive analysis
and interpretation before it is considered final.
* The BATEA limits used hereinafter throughout this paper refer to those
proposed in 1974. New BATEA limits for the Iron & Steel Industry
are expected to be proposed in 1980.
326
-------
PHYSICAL-CHEMICAL TREATMENT OF
STEEL PLANT WASTEWATERS USING
MOBILE PILOT UNITS
INTRODUCTION
This project was initiated to provide an evaluation of the .effectiveness of
existing treatment technology for upgrading steel mill wastewater to Best
Available Technology Economically Achievable (BATEA) limits for Blast
Furnace Category scrubber wastewaters. The wastewater tested was effluent
from an operating steel mill treatment system that met 1977 Effluent Guide-
lines for Best Practical Control Technology Currently Available (BPCTCA).
This wastewater contained residual concentrations of suspended solids, BOD,
phenols, cyanides, fluorides, ammonia compounds, sulfides, and dissolved
solids. The in-depth pilot plant study was performed using mobile facilities
containing physical-chemical treatment equipment.
Treatment processes evaluated during the study included: alkaline chlorina-
tion, chemical treatment, dual media filtration, magnetic filtration,
reverse osmosis, ozonation and activated carbon.
The project objective was achieved through the performance of a program con-
sisting of the three phases outlined below:
PHASE I - Bench Scale Investigation of a Blast Furnace
Scrubber Blowdown Wastewater.
PHASE II - Design and Fabrication of the Mobile Treatment
Facilities to House the Pilot Scale Equipment.
PHASE III - Operation and Evaluation of the Advanced
Waste Treatment Pilot Plant Systems at a
Blast Furnace Site.
The purpose of the first phase (bench-scale work) was to provide information
concerning the treatment methods to be studied for the Phase II design and
the Phase III pilot plant investigation (operation and evaluation). Of
particular interest were such items as the pretreatment requirements, magni-
tude of operating variables, expected magnitude of treatment efficiency and
effluent quality, selection of equipment and media, and pilot plant system
design.
The second phase objective (System Design and Fabrication) was to provide a
mobile pilot testing system for evaluating several advanced waste treatment
technologies. The portable treatment system developed included the
technology needed to remove the residual contaminants in the blast furnace
BPCTCA wastewater to the extent that the wastewater was upgraded to meet
BATEA requirements. Schematic representations of the mobile testing systems
are shown in Figures 1 and 2. Trailer No. 1 housed the alkaline chlorina-
tion, chemical treatment, magnetic filtration and dual media filtration
systems. The ozonator, activated carbon and reverse osmosis technologies
were located in Trailer No. 2. The mobile system contained a high degree of
automation which greatly assisted the operators during the study. All of
327
-------
FLOCCULATOR
FOUR CHAMBER
RAPID MIX TANK
MAGNETIC
FILTER
DUAL MEDIA
FILTER
CHEMICAL TANKS
AIR
COMPRESSOR
TRAILER
45'L x 8'W x 13'-6"H
CLARIFIER
Figure 1. Steel plant mobile treatment
system-trailer No. 1.
328
-------
OZONE
GENERATOR
TRAILER
45'L x 8'W x 13'-6"H
SAMPLE
REFRIGERATOR
CARBON
COLUMNS
REVERSE OSMOSIS
SYSTEM
CLARIFIER
OZONE CONTACT
TANKS
Figure 2. Steel plant mobile treatment
system-trailer No. 2.
329
-------
the treatment technologies were designed to treat a nominal flow of 18.9
£./min (5 gpm).
During the pilot testing phase, the movable waste treatment systems were
operated at the steel mill site to obtain an evaluation of the applicability
of the selected treatment technology to the treatment of the respective
wastewater. The study was conducted on a pilot plant scale of sufficient
size to permit the development of performance and basic design criteria
which were used to scale-up to full size equipment.
The advanced waste treatment methods, both singularly and in combination,
which .were investigated on a pilot basis in Phase III included those presented
below:
1. FIL + 0 + CT KEY
2. ACL + CT + AC AC: activated carbon
3. CT + FIL + RO + 0 (on ROB) JCL: alkaline chlorinatlon
CT: chemical treatment
4. CT + FIL + RO + ACL (on ROB) FIL: filtration-dual media
or magnetic
0: ozonation
RO: reverse osmosis
ROB: reverse osmosis brine
A schematic illustration of the process trains investigated for treatment
of the blast furnace wastewater is shown in Figure 3.
For each treatment train investigated, samples and operational data were
obtained for later use in assessing, evaluating, appraising and comparing
the adequacy of the individual advanced waste treatment methods. Evaluation
and comparison of the data were performed using the criteria:
(1). process and/or treatment train performance
(2). capital and operating costs
(3). space requirements
EXPERIMENTAL RESULTS
General
The blast furnace research site selected was located on the southern shore
of Lake Michigan. The plant began production in 1964 and the hot metal
facilities began production in 1969. The mill utilizes two furnaces. During
the 15 week study, blast furnace production averaged 10,919 metric tons/day
(12,029 tons/day), with a scrubber water blowdown flow of 5,397 m-Vday
(1.426 mgd). Based on BATEA limits, the allowable pollutant concentrations
in the effluent would be:
330
-------
(1) Filtration. Ozonation Clarification
NaOH AIR
POLY
w
MF
r*
-*
i
I
I
^ - » i
> 33
ACID
1
i ,
' fe
Vi
c
\/
1
t
PRODUCT
(2) pH Adjustment, Alkaline Chlorination, Chemical Treatment, Clarification, Filtration. Carbon Adsorption
METAL
NaOH ACID SALT
I NaOCl | NaOCl | POLY
ALKALINE CHLOR I :;AT ION
(3) Chenical Treat-ent, Clarification, Filtration. Reverse Osnosis. Alkaline Chlorination (on brine).
SALT
POLY
01 1
T 1r
\ 1 1
OAU b \ T,._ r
RAW ~9 \ C f
\
Ju \ /
°f> \/
L
+
SLUDGE
CO Chemical Treat-ent, Clarific
METAL
SALT
POLY
Q 1 I
T T
" ' \ 1 1
ft fli f fct \ T L V
Jo \/
O° \/
KtY T
AC: Carbon Adsorption 1
C: Clarification ^
03: Ozonat ion
HO: Reverse usmosis
^ l
^^ JHr 1
"* i
9 MK — — '
""^ A
^~^~^ NaOH ACID X
1 NaOCl 1 POLY
0 j | 0| 1 AC
V \ * 1 f
ef &0 \yS
ALKALI JE CHLOR 1 NAT ION T
^ SLUDGE
ation, Filtration. Reverse Osmosis, Ozonation (On Brine),
Clari f icat ion
fc 1
~ JMI- ^^^^
"* 1
,..fc -
^ Ml- AIR POLY
**^ NaOH I ACID
u 1 4 1
nn.i.r- T . °n » ' F L-h PHnnlirT
BRINE j Q [ ~ HKUUULI
V
w
^ SLUUCt
Figure 3. Process trains investigated for treatment
of the blast furnace waste water.
331
-------
BATEA Level
Reverse osmosis
Parameter Slowdown brine-*"
pH
Suspended solids, mg/2,
Ammonia, mg/5,
Cyanide-A, mg/jl
Phenol, mg/2,
Sulfide, mg/2,
Fluoride, mg/£
6.0-9.0
27
10.85
0.271
0.543
0.334
21.71
6.0-9.0
108
43.4
1.03
2.17
1.34
86.8
1. Assuming reuse of the reverse osmosis product water so
that only 25 % of the original blowdown volume is discharged.
Wastewater Treatment System
Each furnace has its own gas cleaning and cooling system and thickener. Each
system consists of a dust catcher, a primary venturi scrubber, a primary
washer, a secondary venturi scrubber, a gas cooler, and a thickener. A
schematic of the system is presented in Figure 4. Both thickeners are 27.4 m
(90 ft) in diameter, 5 m (16.5 ft) deep with a design overflow rate of 60.3
2,/min/m2 (1.48 gpm/ft2). The thickeners were designed to yield an average
effluent suspended solids of 50 mg/&. The clarified overflow from the
thickeners flows into the hot well of the Blast Furnace Closed Water Pumping
Station. Make-up lake water may be added automatically as required for level
control. The water is pumped from the hot well to the five, spray filled
cooling towers.
The five units operate in series. The cooled water flows by gravity through
a bar screen type trash rack to a cold well. Two stage, fixed speed, high
lift pumps move the water to the booster pump stations located near each
blast furnace. The water is then used for washing and cooling the blast
furnace gas. Water is blown down from the discharge side of the high lift
pumps as required. Blowdown is based on level control. A cyanide destruct
unit is located on the blowdown line. The unit consists of an alkaline-
chlorination system which is operated on an "as needed" basis to destruct
excessive concentrations of cyanide and ammonia in the blowdown. The mobile
pilot plant units' feedwater was taken from the blowdown line before the
cyanide destruct unit.
Pilot Study Results
The results of the pilot program indicated that alkaline chlorination,
ozonation and reverse osmosis were all capable of producing an effluent
acceptable for discharge which meets BATEA limitations. A brief summary of
the pilot findings for each of the above mentioned treatment technologies
is presented below.
Alkaline Chlorination Treatment Train
The alkaline chlorination treatment train was successful in reducing influent
concentrations of all parameters of concern to below BATEA levels. The
332
-------
BLAST FURNACE GAS
US
BLAST FURNACE GAS
WATER
IRON ORE/PELLETS
SCRAP, LIMESTONE,
COKE, SINTER
CLOSED WATER
PUMP STATION
CLEA1 BLAST
FURNACE GAS
ToVlANT
SLUDGE
TO VACUUM
FILTER
BOOSTER
PUMP HOUSE
Figure 4. Blast furnace gas cleaning water recirculation system.
-------
treatment train consisted of elevating the pH of the incoming wastewater to
11.0-11.5 with addition of sodium hypochlorite for oxidation of cyanide.
The waste was then neutralized for ammonia removal. The existing wastewater
was dosed with chemical coagulants prior to clarification and the settled
effluent was dechlorinated on activated carbon.
During the alkaline chlorination study, ammonia was found to be the limiting
parameter for meeting BATEA requirements; that is, if ammonia was reduced
below BATEA levels, all other oxidizable contaminants were also removed below
required future limitations. Shown in Table 1 is a summary of the alkaline
chlorination effluent data for selected runs. The results indicated that
as the chlorine to ammonia ratio was increased, ammonia removal improved.
Ammonia concentrations remained high until Cl2:NH3 ratios exceeded 7.3:1 and
then dropped sharply as this ratio was increased. Chlorine to ammonia ratios
from 6.9:1 to 15.2:1 were studied. Effluent ammonia values as low as 0.48
mg/& were achieved. Breakpoint chlorination occurred at a chlorine to
ammonia ratio of approximately 10:1. Excellent removals of cyanide-A were
achieved at all C12'NH3 ratios investigated. Influent ammonia concentrations
during the study were quite stable, ranging from 17.0 mg/£ to 46.0 mg/£
with an average of 33.4 mg/£.
TABLE 1. BLAST FURNACE WASTEWATER
ALKALINE CHLORINATION DATA
Run
no.
1
2
3
4
5
ratio
15.2:1
10.0:1
8.0:1
7.4:1
6.9:1
Ammonia
Inf.
29.2
32.9
32.4
46.0
30.4
, mg/fc
Eff.
0.67
0.48
5-2
6.95
10.3
Cyanide
Inf.
0.02
0.19
0.02
0.18
0.01
A, mg/8.
Eff.
<0.01
<0.01
<0.01
<0.01
<0.01
Following alkaline chlorination, the wastewater was dosed with 100 mg/JZ,
ferric chloride and 0.5 mg/£ polymer to improve effluent quality. The
suspended solids removal averaged 93% with a range from 88-98%. The settled
effluent was passed through a granular activated carbon system to reduce
residual chlorine. The carbon columns were operated in an upflow-expanded
bed mode to eliminate problems associated with the formation of nitrogen
gas in the system.
In dechlorination of the scrubber blowdown, 88% of the influent available
chlorine was removed with a contact time of 8.5 minutes. With a contact
time of 20 minutes, a total chlorine removal of 98% was achieved. A linear
flow rate of 0.163 nrVm2/day (4 gpm/ft2) was used during the study.
334
-------
Ozonation Treatment Train
Ozonation was a second treatment technology investigated to reduce the blast
furnace scrubber blowdown to BATEA limits. The treatment train consisted of
filtration, pH elevation to 10.5-11.5, ozonation, neutralization and
clarification. Major treatment train components are discussed below-
During this study, both a dual media filter and a magnetic filter were
investigated for suspended solids and turbidity removal prior to ozonation.
1. The filtration media used in the dual media filtration tests
consisted of 39 cm (15 in.) of Red Flint filter sand and 43 cm
(17 in.) of anthracite coal. Filtration removed significant
quantities of suspended solids (83%) and turbidity (75%).
From pilot test data analysis, influent feedwater contained
an average suspended solids concentration of 84 mg/£. Typical
suspended solids feed to the ozone system following filtration
was 14 mg/2,.
2. The magnetic filter reduced influent suspended solids by 70%
and turbidity by 32%. Based on suspended solids and turbidity
removal data, the dual media filter provided a higher quality
feed water to the ozone process element.
Ozonation was found to be a reliable treatment technology for meeting BATEA
guidelines. Operational variables investigated included hydraulic retention
time and ozone dosage. Preliminary studies indicated that the ammonia limi-
tation was the critical factor in meeting discharge levels. Therefore, the
ozone test study was structured toward meeting the ammonia limitation.
For the scrubber blowdown discharge, the prefiltered wastewater was elevated
to 10.5-11.5 prior to entering the ozone contact columns. The total applied
ozone dosage was varied from 0-1,500 mg/2.. Contact times were adjusted
between 60-240 minutes to determine optimum contactor retention times.
Selected results of testing are shown in Table 2. The data indicate that
ozone was effective in meeting BATEA limits. One test run was conducted at
elevated pH using compressed air. Results varified that ozonation rather
than stripping was the mechanism for removing the ammonia. For purposes of
design, a 60 minute contact time was used with an applied ozone dose of
733 mg/fc.
Suspended solids present in the neutralized ozonated effluent were easily
removed by clarification. For the ozonated scrubber blowdown, polymer was
added at a rate of 0.5 mg/Ji to improve settling. Suspended solids removal
consistently averaged 95% or better at all flow rates investigated.
Reverse Osmosis Treatment Train
The reverse osmosis treatment train was the third successful group of
process elements investigated for upgrading the BPCTCA blast furnace scrubber
blowdown to BATEA discharge limitations. Technologies belonging to this
treatment train included chemical treatment, filtration and reverse osmosis.
335
-------
TABLE 2. BLAST FURNACE OZONATED WASTEWATER QUALITY
Date
Run number
Contact time, min.
03 applied, mg/£
03 utilized, mg/£
Parameter
PH
SS, mg/£
VSS, mg/ft
Phenol, mg/£
NH3 as N, mg/£
N03 as N, mg/Jl
TKN as N, mg/A
S, mg/X,
CN, mg/2,
CNA, mg/A
F, mg/«,
TOC, mg/£
BOD, mg/£
Color, APHA units
O&G, mg/£
1/13
1
240
1054
363
7.90
1
<1
0.010
3.1
1.1
<0.2
<0.01
10.8
<2
5/10
<1
1/17
2
120
1205
364
7
3
9.4
34.5
4.8
<0.2
<0.01
<0.01
11
6
<5
0/5
1/20
3
171
1074
480
6.90
15
5
0.005
3.4
0.67
<0.2
<0.01
<0.01
12
5
5
1/22
4
120
739
612
7.60
16
5
0.190
10
12.7
0.1
0.03
0.03
16.2
<1
10
Chemical treatment and clarification was required to remove suspended solids
and turbidity from the scrubber blowdown prior to filtration and application
by the reverse osmosis system. Chemical addition consisted of alum (50-150
mg/£) and polymer (0.5 mg/£). Turbidity removal averaged 68% while a 83%
reduction in suspended solids was also achieved.
Dual media and magnetic filtration were the technologies investigated to
provide final pretreatment to the wastewater before desalting by the reverse
osmosis membranes. Superior suspended solids and turbidity removal was
achieved by the dual media filter and this technology was used to provide
feed to the reverse osmosis system.
The reverse osmosis testing program was performed in two phases. Phase 1
was operated at a 33% product water recovery while a 75% recovery was
utilized in Phase 2. The two phases were used to obtain expected product
336
-------
water quality data and also cleaning information. TFC Model 4600 polyamide
spiral wound membranes were used during the study.
The system was operated continuously over the test period and a total of
409.5 hours were put on the RO membranes.
Projected product water quality, assuming a 75% recovery is shown in Table 3.
The RO rejected significant concentrations of all BATEA parameters to the
extent that the product water could be discharged directly or reused in the
mill. Brine treatment by either alkaline chlorination or ozonation would
be required prior to discharge. Information on treatment of this sidestream
was also obtained during the study.
TABLE 3. BLAST FURNACE PROJECTED REVERSE OSMOSIS
PRODUCT QUALITY
Product
Parameter quality range
PH
SS, mg/A
VSS, mg/A
Phenol, mg/£
NH3 as N, rag/5.
N03 as N, mg/A
TKN as N, mg/A
S, mg/A
CN, mg/A
CNA, mg/A
F, mg/A
TOC, mg/A
BOD5, tag/ A
COD, mg/A
TP as P, mg/A
TDS, mg/A
5.35-5.75
3-4
1-3
<0.054
9-10
<0.02
10-12
<0.2
<0.01
<0.01
0.40-0.45
<2
<2
<5
<0.01
60-70
Alkaline chlorination of the reverse osmosis brine consistently reduced the
ammonia concentration to less than BATEA levels provided the chlorine to
ammonia ratio was in excess of 7.6:1. Ammonia concentrations continued to
drop as the Cl2:NH3 ratio was increased above 7.6:1 until breakpoint was
reached at a ratio of about 10:1. Total cyanide and cyanide-A values were
typically less than 0.01 mg/A for the effluent wastewater.
During ozonation of the reverse osmosis brine, the influent pH was elevated
to 11.5-12.0 prior to applying ozone dosages ranging from 0-3,065 mg/A.
Hydraulic detention times of 120-400 minutes were studied. To meet discharge
requirements, a dosage of approximately 9 mg of ozone were required for each
milligram of ammonia in the effluent.
337
-------
The pilot scale data were evaluated and submitted to vendors for scale-up
to full size equipment. Costs were then summarized for the treatment trains
that met BATEA requirements. These cost summaries are shown in Table 4.
The data indicate that for a 5,678 m3/day (1.5 MGD) design, the alkaline
chlorination treatment train would have a capital cost of $1,032,700. The
corresponding operating costs including amortization of capital would be
$2.856/3,785Jl ($2.856/1,000 gal.). This compares to an operating cost of
$5.404/3,785*. ($5.404/1,000 gal.) for ozonation and $4.547/3,785£ ($4.547/
1,000 gal.) for reverse osmosis assuming ozonation of the brine. For
reverse osmosis treatment with alkaline chlorination of the brine, the
expected operating cost is $6.779/3,785£ ($6.779/1,000 gal.). The above
costs are for treatment only and do not include the cost of handling and
disposing of the sludge residues generated.
Space requirements are also given in Table for a 5,678 m-Vday (1-5 MGD)
system. The ozonation treatment train has the lowest area requirement
[754 m2 (7,980 ft2)] of all of the systems evaluated.
SUMMARY AND CONCLUSIONS
1. The results of the pilot program have indicated that alkaline
chlorination, ozonation and reverse osmosis (RO) were effective
in reducing influent contaminants to below Best Available
Technology Economically Achievable (BATEA) levels in the treat-
ment of blast furnace scrubber blowdown.
a. Pretreatment requirements include:
1. For alkaline chlorination: none
2. For reverse osmosis: chemical clarification and filtration
3. For ozonation: filtration
b. Post-treatment requirements:
1. For alkaline chlorination: chemical clarification and
activated carbon
2. For reverse osmosis: brine treatment by alkaline chlorina-
tion or ozonation
(a) Following alkaline chlorination of the RO brine, the
wastewater would require clarification with polymer and
dechlorination by activated carbon.
(b) After RO brine treatment by ozonation clarification
with polymer is required.
2. Alkaline chlorination was the least cost alternative treatment train
investigated. Expected capital investment for a 5,678 m^/day
(1.5 MGD) is $1,032,700. The corresponding operating costs including
amortization of capital is estimated at $2.86/3,7852, ($2.86/1,000 gal.)
3. Ozonation has the lowest system area requirement of 754 m2 (7,980
ft2). This compared to 1,098 m2 (11,500 ft2) for the alkaline
chlorination treatment train.
338
-------
TABLE 4. TREATMENT TRAIN COST AND
SPACE REQUIREMENTS
5,678 m3/day (1.5 MGD) DESIGN
co
Co
Process
TRAIN NO. 1
Alkaline Chlorination
Chemical Treatment
Activated Carbon
Total
TRAIN NO. 2
Filtration
Ozonation
Clarification
Total
TRAIN NO. 3
Clarification
Filtration
Reverse Osmosis
Ozonat ion-Brine
Clarification-Brine
Total
TRAIN NO. 4
Clarification
Filtration
Reverse Osmosis
Alk. Chlor. -Brine
Clarification-Brine
Act. Carbon- Brine
Total
Capital
Costs Operating Costs
$ $/3,785£ ($/l,000 gal.]
120
103
808
1,032
35
11,250
133
11,418
93
36
473
3,225
78
3,906
93
36
473
93
78
326
1,101
,900
,800
,000
,700
,100
,000
,700
,800
,000
,400
,500
,000
,500
,400
,000
,400
,500
,600
,500
,000
,000
1.
0.
1.
2.
0.
5.
0.
5.
0.
0.
1.
2.
0.
4.
0.
0.
1.
2.
0.
2.
6.
19
19
476
856
064
20
14
404
17
067
16
90
25
547
17
067
16
75
25
382
779
Space Requirements
> n.2 (ft2)
426
300
372
1,098
251
203
300
754
300
251
333
93
70
1,047
300
251
333
325
70
265
1,544
(4
(3
(A
(11
(2
(2
(3
(7
(3
(2
(3
(1
(11
(3
(2
(3
(3
(2
(16
,400)
,100)
jOOO)
,500)
,700)
,180)
,100)
,980)
,100)
,700)
,590)
,000)
(780)
,170)
,100)
,700)
,590)
,500)
(780)
,850)
,520)
-------
ACKNOWLEDGEMENTS
Rexnord acknowledges the cooperation and support of the U.S. Environmental
Protection Agency. The assistance given by Robert Hendriks, Project Officer
was received with much appreciation.
The information contained in this paper is part of a draft final report
being prepared for the U.S. Environmental Protection Agency. Modifications
to the enclosed material prior to publication of the final report are
probable.
340
-------
STUDY OF NQN-U.S. WASTEHATER TREATMENT TECHNOLOGY
AT BLAST FURNACES AND COKE PLANTS
Harold Hofstein - Manager Engineer
Harold J. Kohlmann - Sr. Vice President
Hydrotechnic Corporation
1250 Broadway
New York, NY 10001
ABSTRACT
An engineering study is currently being performed to de-
termine the applicability of wastewater treatment technology
being used at blast furnaces and coke plants outside of the
United States to the iron and steel industry in the United
States. This study is being performed under Contract No.
68-02-3123 issued by the U.S.E.P.A., Metallurgical Processes
Branch, Industrial Processes Division, IERL, RTF North Carolina.
Steel plants in Western Europe, Australia, Taiwan, Japan,
North and South America and Africa were visited and discussions
held with appropriate plant and corporation personnel. Cooper-
ation by the plants has been very good. Where permitted, the
actual treatment facilities were visited. Where available,
influent and effluent quantities and qualities were provided as
well as treatment systems operating parameters.
Regulatory agencies, an iron and steel industry association
and a centralized treatment facility were also visited to deter-
mine effluent regulations, pretreatment constraints and the ra-
tionale for the required effluent quality.
Preliminary findings will be presented in this paper. How-
ever, as of this presentation there is still ongoing correspon-
dence between Hydrotechnic and the steel companies visited to
continue to develop bases for adapting non-US technology to the
United States blast furnace and coke plant waste treatment fa-
cilities.
341
-------
STUDY OF NON-U.S. WASTEWATER TREATMENT TECHNOLOGY
AT BLAST FURNACES AND COKE PLANTS
Harold Hofstein - Manager Engineer
Harold J. Kohlmann - Sr. Vice President
Hydrotechnic Corporation
1250 Broadway
New York, NY 10001
INTRODUCTION
On September 29, 1978 Hydrotechnic Corporation received an
assignment from the U.S.E.P.A. to study non-US blast furnace and
coke plant waste treatment facilities with the view towards ap-
plying any superior technologies found at similar production fa-
cilities in the United States. This project is to determine if
such technologies do in fact exist, the efficiencies of these
technologies and their applicability to US steel plants. As of
the presentation of this paper the technologies are still being
evaluated and any conclusions drawn herein are tentative.
METHODOLOGY
The obvious first step in meeting the goals of the assign-
ment was to obtain permission to visit as many of the steel
plants outside of the United States which, based on prior know-
ledge, were apt to have facilities installed that might be con-
sidered as being superior to those presently in operation in the
United States. Using the in-house information at Hydrotechnic,
a literature search and contacts with steel people throughout
the world, the plants were selected. In addition, the assist-
ance of the International Iron and Steel Institute in Brussels
was requested but they suggested the steel companies be con-
tacted directly. Additional candidate plants were obtained
from steel plant equipment manufacturers and from other steel
342
-------
plants. As a result, twelve steel plants in seven West European
countries, seven in Australia and the Far East, two in Mexico,
three in South Africa and two in South America were visited to
observe and discuss their treatment systems. Prior to the vis-
its, questionnaires were prepared and mailed to each of the
plants with a request that they be completed and given to us
when we arrived at the plant. These questionnaires were fair-
ly detailed and it was not expected that all data requested
would be available at every plant. Most questionnaires were
completed to some degree and they were ready when we arrived
to hold more detailed discussions. After the discussions, most
plants permitted visits to the treatment facilities.
Upon our return to the United States, trip reports includ-
ing our understanding of the information provided were prepared
and sent to the respective plants visited for verification of
the process and data contained therein.
Meetings were also held with government regulatory agencies
in Sweden, Holland, the United Kingdom, Australia, Taiwan,
Japan, Mexico, Brazil and South Africa to determine what regula-
tions have been established for steel plant effluents. The West
German government referred us to the Verein Deutcher Eisenhut-
tenleute, an. organization similar to the AISI in North America.
A centralized wastewater treatment plant was observed where an
entire river polluted with municipal and industrial wastes (in-
cluding those from steel plants) is diverted, passed through the
treatment plant and discharged.
Since the primary objective of this project is to determine
foreign technologies that offer advantages over domestic prac-
tices, this study also includes the researching of existing methods
utilized by steel plants in this country for treatment of coke
plant and blast furnace wastes and near term research for
343
-------
treatment of these wastes.
As originally conceived, a second phase of this project
was to consist of the sampling of foreign was,te treatment faci-
lities influents, effluents and selected intermediate streams
for parameters regulated under the existing effluent limita-
tions guidelines and also for priority pollutants. Not all
plants visited were receptive to this sampling program but many
expressed a willingness to cooperate.
However, based on the analytic data provided, only por-
tions of the observed wastewater treatment practices appeared
to be applicable for transfer to domestic practice. This sam-
pling phase was deleted from the project in favor of more de-
tailed engineering study to determine the adaptability of pro-
cesses observed to the United States practice.
Major differences exist between the establishment and en-
forcement processes of regulations in the United States and
other nations. Foreign regulations are most often established
on a case by case basis with respect to receiving water quality,
feasibility and economics of treating the discharges to a given
quality, benefits 'to be gained and quantities of water used. In
addition, in many cases local agencies rather than the national
governments exert the major authority. In Great Britain, it is
the local water basin authority; in Japan, it is the prefecture;
in Australia and the Republic of China, it is the state al-
though the national government establishes the national goals.
Most government agencies do not set standards which are
then passed on to the industry for comment. The industry takes
an active part of the standard setting procedure. In two of
the countries visited, additional government agencies are also
included in the standard setting procedure. In the Netherlands,
344
-------
the Ministry of Economic Affairs is consulted and in Japan the
Ministry of International Trade and Industry takes an active
part in the procedure. In some countries, the government owns
or has a major investment in the industry.
In countries that are members of the European Economic
Community, a directive of the Council of the European Communi-
ties with respect to "pollution caused by certain dangerous
substances discharged into the aquatic environment of the Com-
munity" is used as the general guideline for all discharges.
This includes surface and ground water within the member coun-
tries and also ocean discharges.
Although intervener groups are permitted to contest regu-
lations, the technical merits of their arguments are determined
by hearing procedures rather than courtroom adversary proceed-
ings. There is a minimum of media comments.
«
Some countries, notably the Netherlands and West Germany,
establish a fee for discharges based on quantity and quality.
For example, one plant in Europe has to pay a tax calculated on
the basis of:
Annual Tax=$10 (m3/day (COD + 4.57 N)) COD & N in mg/1
180
If the plant does not meet minimum concentration standards
for specified parameters there are additional fines levied.
The objective of this paper, however, is to discuss the
technology observed. Some aspects of regulations have been dis-
cussed only to provide an understanding of the possible incen-
tives there are for the steel plants to treat their wastes.
Other incentives exist, however, in addition to regulatory
345
-------
constraints. Water withdrawn from public water supplies, be it
a stream or aquifer, in many cases, must be paid for. One plant
visited obtains the highest quality water that it requires from
wells located on its own property and must pay the equivalent of
14* per cubic meter withdrawn (53* per 1000 gallons) to local
water authorities. It is therefore to the plants' economic ad-
vantage to treat water and maximize reuse. The plant mentioned
above reports that 97 per cent of all water used in the plant is
water that has been used, and recirculated either with or with-
out treatment.
BLAST FURNACE WASTE TREATMENT
Blast furnace gas cleaning wastes are generally treated by
sedimentation - in some instances with polyelectrolyte addition,
cooling and recirculation. One plant visited discharged their
flue dust laden water without settling although they are operat-
ing under a directive to install settling facilities by 1982.
Other plants - generally those using salt water - do not recir-
culate after settling.
Only one of the plants visited practiced any treatment of
the settled gas cleaning wastewaters for removal of any of the
U.S.E.P.A. guideline parameters. Many of the plants do not
blow down their recirculation systems per se. They blow down
via sludge discharge or they use the blow downs to quench slag.
The one plant that does treat for removal of cyanide and phenol
uses Caro's Acid (permonosulfuric acid - H2S05) to reduce the
levels to the required 0.2 mg/1 CN and 0.5 mg/1 phenol. The
normal influent level of phenol at that plant is 2 mg/1.
One plant reported that the addition of polyphosphate in
the cooling tower aids in CN removal but at CN levels over
10 rag/1 it does not work. Another plant reported that they had
346
-------
tried to remove CN using polyphosphate but successful removals
could not be obtained. That same plant reported that a form of
CN has been produced in their blast furnaces that is resistant
to alkaline chlorination.
At two of the plants visited CN reductions were noted be-
tween gas cleaning discharges and the plant effluent without
any intentional CN treatment installations. One of these
plants operates its lone blast furnace on a five day per week
basis and shuts down on week-ends. Since this plant is in a cold
climate, steam was injected into the gas cleaning water collec-
tion pond to prevent freezing and CN was observed to be reduced
from 30 mg/1 to 2.4 mg/1. Stripping appears to be the reduction
phenomenon.
The other plant that showed incidental reduction of CN was
one that utilized the sludge withdrawal from the clarifiers as
the total gas cleaning system blowdown. The sludge is deposited
in sludge lagoons which is maintained at an alkaline pH, by
either adding caustic or by combining cold rolling mills waste
with the sludge. The water seeps through the ten meter depth of
sludge and is collected at the underside by open-jointed pipes
and discharged to a river. The CN is reduced from a level of
0.2 mg/1 to 0.1 mg/1. The plant has theorized that the removal
mechanism is the formation of metallo cyanide complexes which
are adsorbed in the sludge. Zinc reductions have also been no-
ticed and have been accounted for as the precipitation of zinc
as a hydroxide which remains in the sludge due to filtering ac-
tion.
At two plants, one in West Germany and one in Japan, aera-
tion prior to discharge to the clarifiers was an integral part
of the gas cleaning water recirculation system. The purpose is
to strip CO and C02 from the water and to precipitate CaC03. A
347
-------
portion of the clarifier sludge is recycled to act as a seed and
enhance precipitation and sedimentation.
At each of the plants visited inquiries were made as to the
sources and compositions of their ores and coals. The objective
was to attempt to correlate wastewater characteristics with the
type of coal and/or ore used. Unfortunately virtually every
plant has a multitude of sources and the ores and coals are
blended to meet the plants' production objectives. Therefore,
the correlation hoped for was impossible.
Data for several of the gas cleaning waste recirculation
water systems observed are shown in Table 1.
COKE PLANT WASTE TREATMENT
Many modes of operation and distribution of responsibility
were reported at the coke plants visited. One plant has a phe-
nol recovery system owned by an outside agency but operated by
the plant; some plants have the coke plant owned and operated
by chemical companies although located on the steel plant site
and one coke plant was not an integral part of any steel plant
proper. Most coke plants, however, are owned and operated by
the respective steel plants. Of the sixteen coke plants vi-
sited, four did not treat the final coke plant effluent other
than passing the excess ammonia liquor through free ammonia
stills; one of these plants also operated a dephenolizer. Two
of these plants utilized the excess ammonia liquor diluted with
other plant wastes for coke quenching and the other two dis-
charged the wastes with other plant wastes to a river. This
river receives wastes from other industries and, at the mouth,
the entire river is diverted from an improved natural course,
treated biologically and discharged to its natural course.
348
-------
All of the biological treatment plants-observed were single
stage, activated sludge. One plant operated three banks of
three basins each in series for a total of nine basins. Each
series of three basins has its own final settling facility. In
effect, three separate cultures are maintained.
Table 2 presents data from five plants that operate biolo-
gical treatment systems. It can be seen that there are wide va-
riations in concentrations of the various parameters entering
the plants, and leaving the plants, although the detention times
are relatively uniform. There are also operational variations
i.e., some do and some do not utilize dilution water. Generally
it can be seen that the single stage biological systems follow-
ing free ammonia stills and with nutrient addition "in the form
of phosphoric acid produces good reductions of phenols and cya-
nide, poor reductions, if any, of ammonia and erratic COD re-
ductions.
One plant visited had the complete treatment train as sug-
gested in the development document guidelines, i.e. biological
treatment, settling, filtration and activated carbon adsorption.
The data for this plant is not presented here because the li-
mited amount of data provided us has not been verified by the
plant.
Two of the plants visited remove tar from their coke plant
waste stream in coke filters prior to discharge to the aeration
basins. The tar laden coke is then blended with coal and re-
charged to the coke ovens. One plant combines sanitary wastes
with coke plant wastes for biological treatment.
Some of the plants visited controlled charging and pushing
emissions by scrubbing systems. The water from all of the sys-
tems observed was discharged as makeup to the quench systems.
349
-------
GENERAL
Good housekeeping was evident at all of the plants visited.
Very few leaks and drips of equipment were observed, no treat-
ment facilities appeared to be hydraulically overloaded and no
water was observed to be running for no apparent purpose.
Valves are opened only when needed and closed when the need has
been satisfied.
Green areas are prevalent in all of the post-war plants and
in fact, at one plant many rabbits were observed within the
plant grounds.
There is a great deal of concern with respect to noise.
Most equipment is sound-proofed arid green buffer zones are
created between the plants and the surrounding populated areas.
CONCLUSIONS
Based on the data obtained during the plant visits, it was
concluded that sampling of the waste treatment facilities of the
plants would not serve any useful purpose. None of the plants
visited as of this writing have shown a complete system that
would be considered as exemplary and the limited data available
bears this out. However, portions of systems appear to show
promise for further research and do merit further investigation.
Extremely preliminary conclusions are that coke filters at bio-
logical plants may be useful to enhance the biological treatment
even though electrostatic precipitators are used for tar remo-
val. Aeration of blast furnace gas washer water may allow in-
creases in the cycles of concentrations, thereby reducing the
blowdown that would require treatment. Utilizing flue dust as
a filter medium at an optimum pH level may be able to reduce
350
-------
cyanide and metals concentrations. This could be used as a unit
operation in the treatment of blast furnace gas washer blowdown.
As stated earlier, only limited data was available for the
guidelines parameters and no data was available for priority
pollutants. All of the data received as of this writing is from
steel plants in Western Europe and Australia and the Far East.
These data are still being analyzed; therefore, further conclu-
sions may still be drawn. Visits to the additional plants in
Mexico, South America and South Africa may provide additional
treatment concepts that may be applicable to the U.S. iron and
steel industry.
351
-------
TABLE I
Ln
ro
Plant
E-l
FE-I
E-2
E-3
FE-2
E-4
E-5
Total Plant
BF
Capacity
No. Cap.
Mq/day
5 50,000
1 8,000
7 62,000
2 4,500'
.(Typ) 800
2 1,400
1 1,200
BLAST FURNACE
BF Cleaning
Water
Appl ied
m3/hr
7600
1 125-1500
4445
600-900
1000
470
155
GAS CLEANING WATER
BF Cleaning
Water
Slowdown
mVhr
480
83-125
68
50
40-60
150
20
1.7-4.2
3
Percent
Recycled
93.7
92.1
97.3
93.3
83
>99
98
Disposition of BD
To Bay
To si ag quench
Sludge and
to river
To river
To Bay and
sludge
To river
To river via pig-
E-6
2 9,000
2400
21
99
caster
Sludge, f iItrate
to river
-------
TABLE 2
Plant
E-7
E-l
E-2
E-8
W LO
s a
E-9
Coke
Ovens
100
430
440
500
54
Capacity
(Mg/day)
4730
3920
6500
2000
COKE
Dilution
WAL Added
(m3/hr) (m3/hr)
25 up to 75
850(2>(3>
227(3)
70 0
40 0
PLANT
Det.
Time
(hrs)
24
22
21
24
WASTE TREATMENT
Phenol CN
In Out In
100(1) 1.02 (1) 3.72*1
510 2.0 98
900 18 13
400- 5-35 1-10
700
500- 0.3 0.2-
600 0.3
) o
1
0
1
0
Out
.62
1010
115
1000-
1800
900-
1000
Out
340(1)
1010
470
1000-
1800
1350
COD
i In
662™
1380
3300
2000-
3000
NR
Out
180 (
68
411
900-
1000
NR
(1) Better of 2 values reported,
(2) Design capacity.
(3) Individual flows not reported.
All analyses reported as mg/1
-------
Symposium on Iron and Steel
Pollution Abatement Technology
October 30 to November 1
Chicago, Illinois
"Formation and Structure of
Water-Formed Scales"
Dr. George R. St. Pierre and Mrs. Rhonda L. McKimpson
Department of Metallurgical Engineering
The Ohio State University
Columbus, Ohio 43210
ABSTRACT
After a brief review of the general problems associated with
water-formed scales in steelplant recycle/reuse systems, the results
of a detailed investigation on the initiation of calcium carbonate
scales are presented. In particular, the structure of thermally-
induced calcite deposits under a variety of experimental conditions
are described with the help of scanning electron micrographs. The
range of supersaturation for scale deposition on metallic substrates
is defined and the transition from equiaxed to acicular crystal
formation is characterized. The significance of the results in
connection with the avoidance of deleterious scales are summarized
briefly and a few suggestions for on-line sensor and-control devices
for steelplant water systems are made.
354
-------
"FORMATION AND STRUCTURE OF WATER-FORMED SCALES"
INTRODUCTION
The technological developments in the treatment, reuse, and
recycling of water in integrated steelplants depend in part upon
the development of adequate control procedures for the avoidance
of scale formation in a variety of steel operations. Several
papers in this Symposium have described particular problems
associated with deleterious scales. For the minimization of blow-
downs and optimization of recycle/reuse systems, it is necessary
to understand the tolerable levels of dissolved salts and, in
some cases, suspended solids in different streams. In the present
communication, some particular information on the formation of
calcium carbonate scales under laboratory conditions is presented.
-*
The work is part of a more comprehensive investigation on the con-
trol of scale formation in steelplant water systems.
EXPERIMENTAL
Several types of water recirculation systems have been used in
the laboratory to study the initiation and rate of formation of
calcite scales on metal substrates. Figure 1 shows one of the
assemblies used for studying thermally-induced scales. The appear-
ance of a stainless steel substrate exposed to unstable circulated
water is shown in Figure 2.
Instability has been established by a variety of methods includ-
ing the addition of Na-CO,, pH adjustment, agitation with CO- gas of
varying partial pressure, and temperature. The heating coil shown
in Figure 1 enables the establishment of a substrate temperature in
excess of 100°C. Five thermocouples are attached to the back of the
substrate sample in order to record the temperature profile.
RESULTS
Figure 3 shows some of the results obtained with the described
test unit. At excessive levels of supersaturation, general precipi-
355
-------
tation of CaCO- occurred throughout the water system. However, a
region of intermediate non-equilibrium conditions was established
wherein scale formation occurred in the absence of general precipi-
tation. The reproducibility of these results is shown in Figure 4.
Normalized scale densities after a 10 hour exposure to water
at a degree of supersaturation equal to twenty are shown in Figure
5.
In the present communication, the initiation of scales is of
principal interest. Scanning Electron Microscopy (JEOL JXA-35 with
EDAX) was used to characterize the initial crystallization. Repre-
sentative micrographs are shown in Figures 6 and 7. Debeye-Scherrer
patterns showed no evidence of aragonite and clear evidence for
calcite as the dominant structure.
For short incubation periods and relatively low sample tempera-
tures, equiaxed crystallization was favored. At higher temperatures,
acicular crystals dominated.
DISCUSSION AND CONCLUDING REMARKS
Substrate condition influences the initiation, or nucleation,
of calcite on metallic surfaces. Minute polishing scratches can
markedly enhance crystallization even at very low levels of super-
saturation. The experimental test cell described is a useful
technique for evaluating the instability of circulated waters and
for evaluating the effectiveness of surface coatings and inhibitors.
The results presented in this brief communication represent a
small part of the entire experimental program. The test cell shows
promise for use as an on-line sensor for scale formation control.
Proposed modifications include the incorporation of several additional
thermocouples which would enable the detection of scale formation
without visual observation. The thermocouple outputs can yield
information on both the initiation and rate of scale development.
However, the suitability of such a sensor in a complex water stream
containing dissolved salts in various combinations and a variety of
suspended solids must be demonstrated.
356
-------
ACKNOWLEDGEMENTS
Yu-Sue Won, Adil Khan, and John Newman have made significant
contributions to the experimental program. The support of the
U.S.E.P.A. and technical direction of Mr. Norman Plaks and Mr. John
Ruppersberger are appreciated greatly. While no references are
cited specifically, the contributions of many earlier and contempo-
rary researchers in the field have been of immense value.
357
-------
Cu Wire.
to Cold Junction
Ceramic
Insulation
32mm
28mm
Corks
Glass Pyrex
41 mm
Insulation
Nichrome Wire
Heating Element
Chromel-Alumel
Thermocouple (1-5)
Metal Sample Sealed on Glass
Figure 1.
Schematic Diagram for Studying Thermally-Induced
Scales.
358
-------
Figure 2. A Stainless Steel Substrate After Exposure
to Unstable Water. (Note scale deposit in
region on the right side.)
359
-------
o
I"
*»
c
o
'•£ -2
c
0)
o
c
o
o
CD
o
c
o
JD
i_
O
o
o
_J
-4
-6
-8
-10
-12
General
Precipitation
Scale
_ No Scale or Precipitation
• scale only
(§) scale + ppt
o ppt only
-4 -3 -2
Log (Calcium Concentration, g-moles/ liter)
Figure 3. Regions for Stability, Scale Formation, and
General Precipitation.
360
-------
= 96.6 ppm CaCO,
O No ppt present
• ppt present
-14
0 100 200 1200 1300 1400 1500 1600 1700 1800 1900 2000
Time (Minutes)
Figure 4. Repeated CaCC>3 Precipitation and Dissolution
Induced by pH Adjustment and Control of
C02 Pressure. (Run No. 23.)
361
-------
2.0
1.8
1.6
JE
£ ''4
E
2: 1.2
1.0
c
0)
Q
.9?
o
o
CO
-o 0.8
0.6
0.4
0.2
100 200 300 400
Calcium Concentration, ppm CaCO,
500
Figure 5
Normalized Scale Densities after Ten Hours
Exposure at a Degree of Supersaturation
of Twenty.
362
-------
Figure 6. The Appearance of Both Equiax and Acicular
Calcite on Stainless Steel.
363
-------
Figure 7. Typical Appearance of Acicular Calcite Formed
on Stainless Steel.
364
-------
Sections: SOLID WASTE POLLUTION ABATEMENT
Chairman: Eugene F. Meyer, Chemist
Hazardous Waste Management Branch
Region V, EPA
Chicago, IL
365
-------
ABSTRACT
FEDERAL REQUIREMENTS FOR CHEMICAL WASTE DISPOSAL
By
Eugene F. Meyer
Hazardous Waste Management Branch
Region V, U.S. Environmental Protection Agency
Chicago, Illinois
Subtitle C of the Resource Conservation and Recovery Act (RCRA)
of 1976 gives the U.S. Environmental Protection Agency the mandate
to provide a regulatory program in the area of hazardous waste
management and other areas of solid waste management. In response
to this mandate, the U.S. EPA published on December 18, 1978, its
proposed guidelines and regulations for the identification and
listing of hazardous waste. This document, published in the
Federal Register, is expected to be promulgated in 1980. Its
various sections will be discussed, specifically with regard to the
definition of a hazardous waste (Section 3001), its generation
(Section 3002), transportation (Section 3003), and treatment, storage,
and disposal (Section 3004).
366
-------
RTI/1603/24-03S September 1979
ENVIRONMENTAL AND RESOURCE CONSERVATION
CONSIDERATIONS OF STEEL INDUSTRY SOLID WASTE
M. R. Branscome, V. H. Baldwin, C. C. Allen, B. H. Carpenter
Research Triangle Institute
Research Triangle Park, North Carolina
ABSTRACT
The United States steel industry produces roughly 137 million metric tons of
solid waste annually (including in-plant mill scrap) and currently reuses or
recycles about 80 percent. The balance, about 30 million metric tons, is
characterized here as to origin, nature, and quantity. Current disposal meth-
ods and leachate characteristics of each type of waste are also described. The
impact of Section 4004 of the Resource Conservation and Recovery Act is dis-
cussed with respect to solid wastes not classified as hazardous. This im-
pact is to require the solid waste disposer to prevent groundwater endanger-
ment from the migration of leachate. Assuming this is accomplished through
the use of lined landfills, the incremental cost of Section 4004 to the
industry is estimated as 21 million dollars per year. This represents a 40
percent increase in current disposal costs, but it is less than 2 percent of
the current environmental expenditures.
367
-------
ENVIRONMENTAL AND RESOURCE CONSERVATION
CONSIDERATIONS OF STEEL INDUSTRY SOLID WASTE
1.0 INTRODUCTION
The iron and steel industry produces an estimated 137 million metric tons
of waste (including metallic scrap) annually in the production of 125 million
metric tons of steel. Approximately 80 percent of this waste is currently re-
used or recycled. This study examines the remaining 20 percent with respect
to the impact of Section 4004 of the Resource Conservation and Recovery Act of
1976 (RCRA), based upon proposed criteria published in the February 6, 1978
issue of the Federal Register.
The approach taken to determine this impact was to:
(1) characterize the solid wastes as to origin, nature, quantity, and
potential for groundwater endangerment,
(2) examine current disposal practices,
(3) outline the potential requirements of proposed Section 4004 criteria,
and
(4) estimate the cost of meeting these potential requirements.
2.0 WASTE CHARACTERIZATION
2.1 Coke Plant Wastes
The solid wastes generated in the coke plant, including by-product recov-
ery, are coke breeze, sludge from the tar decanter, lime sludge from the
ammonia still, sludge from the biological treatment plant, and waste sludge
from the light oil wash and neutralization. Breeze generally has several
uses, although it can become an occasional disposal or stockpiling problem.
The tar decanter sludge, lime sludge, and oleum wash sludge are classified haz-
ardous, so they are subject to the criteria for disposal of hazardous wastes.
An estimated 1.8 million tons of breeze and 140,000 tons of sludges are gener-
ated annually from the production of 48 million metric tons of coke.
368
-------
2.2 Slags
There are two types of slag wastes: ironmaking and steelmaking. A typical
analysis of blast furnace slag is 41 percent CaO, 35 percent Si09, 13 percent
i *•
Al-O,, and 8 percent MgO with a basicity ratio of 1.0. There is a wide varia-
tion in the range of compositions for steelmaking slag. The nature of the slag
varies depending upon the metallurgy of the process involved, upon the impuri-
ties in the feed materials, principally sulfur, and upon the end product. EOF
slag samples from two different steel companies contained 32-42 percent CaO,
15-30 percent FeO, 9-23 percent SiO_, 5-10 percent MgO, and 0.1-0.2 percent
2
A1-0-. The slag basicity ratio ranged from 2 to 4.
Blast furnace slag is processed at 66 major slag plants and is sold as
three general types: air-cooled, granulated, and expanded. Approximately 10
percent of the blast furnace slag that is produced is landfilled (Table 1);
however, even in these cases it may serve a constructive purpose. For example,
one major plant is using its slag as on-site fill material for future plant
expansion. Other plants pile the slag in mounds for future sale or use it to
dike a landfill area. Some old slag dump sites are being mined to recover
the slag to meet the increased demand.
Steelmaking slag is processed at 35 major plants but in much smaller
quantities than ironmaking slag. Steelmaking slag is sometimes recycled to the
blast furnace to recover iron, manganese, and lime values, and finds some use
in construction for unconfined bases, fill, and highway shoulders. Its utility
is much more limited than ironmaking slag because it can undergo uncontrolled
3
expansion due to hydration of free lime. An estimated 45 percent of the
steelmaking slag is used or recycled; the remaining 55 percent is landfilled.
The landfilled slag often is used for dikes, landfill bases, and for layering
or mixing with dust and sludge.
The amount of "stocked" slag was not estimated due to the difficulty in
determining the difference in landfilling (or dumping) and stockpiling. Many
disposal sites described as stockpiles have accumulated large quantities of
slag over a period of years. A report prepared in 1976 for the Federal Highway
Administration to examine the availability of wastes for use as highway mater-
ials estimated the quantities available at a few selected slag dump sites.
There were six locations in Pennsylvania in 1976 that had 93.5 million tonnes
(103 million tons) in slag piles.
369
-------
Table 1. SLAG DISPOSITION FROM 125,000,000 TONNES OF STEEL PER YEAR
(TONNES PER YEAR)
Source Generated Landfilled % Recycled, Used %
Ironmaking 28,300,000 2,800,000 10 25,500,000 90
Steelmaking 19,360,000 10,560.000 55_ 8.800,000 45
TOTAL 47,660,000 13,350,000 28 34,300,000 72
370
-------
2.3 Iron Oxide Wastes
The iron oxide wastes include dusts, sludges, and scale generated by a
variety of processes. Within each of these waste types are variations in
composition, particle size, water content, and contaminants. Each presents its
own problems, or lack of them, with respect to the potential for recycle or
reuse. Steelmaking dusts and sludges are particularly troublesome due to many
problems associated with recycle. , A few of the key ones are:
(1) Zinc and lead in the dust are carried into the sinter and from
there to the blast furnace, where they interfere with flue
operations of the blast furnace and cause premature destruction
of the furnace lining.
(2) The very fine particulates cause handling problems' and interfere
with smooth operations of the sintering process.
(3) The iron content of Steelmaking fines is usually small although
often highly variable.
(4) The tonnage of waste iron oxide generated in a single steel-
making facility may be too small to economically support a
sophisticated and technically correct process for recovering
the waste and converting it to a useful form.
2.3.1 Dust Treatment and Disposal. Dust is collected by dry air pollution
control equipment used in the sinter plant, blast furnace, and Steelmaking
furnaces (Table 2). Sinter and blast furnace dusts are generally recycled, but
Steelmaking dust is mostly landfilled and accounts for 73 percent of the 1.2
million tonnes of dust which are not recycled.
Dust disposition data, available from 17 major plants, revealed that 16
practiced recycle, 6 had stockpiles on-site, and 7 landfilled a portion of
their dust.
Some specific dust handling techniques include:
(1) mixing with scale and stockpiling,
(2) mixing with water to prevent wind transportation and placing in
a holding pond,
(3) recycling EOF dust by using select scrap in the EOF to keep zinc
content down,
371
-------
Table 2. DUST DISPOSITION FROM 125,000,000 TONNES OF STEEL PER YEAR
(TONNES PER YEAR)
Source
Sinter
Ironmaking
Steelmaking
TOTAL
Generated
740,000
1,290,000
1,050,000
3,080,000
Land filled
40,000
170,000
690,000
900,000
%
6
13
66
29
Stocked
___
120,000
190,000
310,000
%
—
9
18
10
Recycled
700,000
1,000,000
170,000
1,870,000
%
94
78
16
61
372
-------
(4) "storing" dust in the ground by covering with a layer of dirt,
and covering with EOF slag.
2.3.2 Sludge Treatment and Disposal. Sludge is generated by water treat-
ment facilities in which solids are removed from process wastewater and from
the water used in wet pollution control equipment. The wastewater goes through
a series of treatments that may include settlers, thickners, oil skimmers,
scale pits, polymer addition to aid settling and dewatering, clarifiers, fil-
ters and biological treatment. The type of treatment is plant specific and may
involve almost any combination of the above for treating water from various
processes individually or in central treatment plants. The resulting sludge is
recycled, landfilled, stocked, or put into a lagoon for additional dewatering
before disposal. The use of lagoons and holding ponds is widespread with each
major plant having at least one such facility. A total of 16 lagoons and ponds
were identified in 13 major plants, and each plant generated some sludge that
was landfilled.
Complete sludge disposition data was available from 17 plants. These data
indicated that 13 plants practiced recycling, 10 had stockpiles on-site for
potential reuse, and all 17 landfilled at least a portion of their sludge
(Table 3). Sludge from the rolling mills and steelmaking furnaces accounts
for 1.3 million tonnes of the estimated 1.6 million tonnes of sludge landfilled
yearly.
Some of the disposal techniques used by individual plants include:
(1) mixing with dust and slag in landfill,
(2) spreading over slag pile,
(3) mixing with dust and scale on site,
(4) placing in pits in the landfill area, then covering with slag,
and
(5) placing in lined landfill with leachate collection.
2.3.3 Scale Treatment and Disposal. Scale is generated in the rolling
operations and is usually collected in scale pits or settling basins. These
settlers serve as a preliminary treatment of direct contact process water that
is used for cooling, scale removal, and flushing. The heavy coarse pieces
settle out and the very fine scale is removed in subsequent water treatment as
a sludge.
373
-------
Table 3. SLUDGE DISPOSITION FROM 125,000,000 TONNES OF STEEL PER YEAR
(TONNES PER YEAR)
Source
Ironmaking
Steelmaking
Rolling Mills
TOTAL
Generated
2
1
3
,030
,170
758
,958
,000
,000
,000
,000
Land filled %
270
617
730
1,617
,000
,000
,000
,000
13
53
96
41
Stocked
190,000
286,000
___
476,000
% Recycled %
9
24
—
12
1,570
267
28
1,865
,000
,000
,000
,000
78
23
4
47
374
-------
Most of the scale generated in the rolling mills is recycled or stocked
for potential recycling. Some of the stockpiled scale is not recycled immedi-
ately due to a high oil content that causes problems of hydrocarbon emissions
and fouling of fabric filters in the sinter plant. Approximately 70 percent
of the mill scale (3.9 million tonnes) is recycled, 30 percent stocked, and
a small quantity is dumped. That portion disposed of in a landfill is gener-
ated by the cold rolling operation and has a high oil content, but it is only
0.04 percent of the mill scale produced.
Soaking pit scale, also called soaking pit slag, is iron oxide scale
fused with the coke breeze or dolomite that has been placed in the bottom of
the soaking pit. This scale may be contaminated with refractory or other
material. An estimated 1.3 million tonnes of soaking pit scale are landfilled
annually.
2.4 Miscellaneous Waste Treatment and Disposal
Plant debris, trash, rubble, and refractory from relining of furnaces are
landfilled. AISI estimated that these wastes are generated at a rate of 10
percent of the steel produced (200 pounds per ton), so that a national pro-
duction of 125 million tonnes of steel would give 12.5 million tonnes of this
waste. Eight plants reported to state agencies regarding the disposition of
miscellaneous debris and the quantities totaled approximately 5 percent of
the steel produced. In three cases the waste was disposed of by means of
contract disposal, in another three at an off-site landfill, and in two at an
on-site landfill.
Fly ash and bottom ash (or clinker) are solid wastes generated in coal-
fired boilers. An estimated 380,000 tonnes are generated annually by the iron
and steel industry. Information on these wastes was obtained from state
agencies for six plants and their rate of generation was approximately 13 kg
per tonne of steel. Two of these plants landfilled the ash on-site and the
other four off-site.
The sludge from neutralization of spent pickle liquor was estimated as
350,000 (dry) tonnes per year. An EPA survey of 16 plants revealed that 60.8
percent of the waste pickle liquor was disposed of untreated by deep-well
injection, dumping on a slag pile, or direct discharge; 20.5 percent was
375
-------
neutralized on-site; 11.3 percent handled by contract hauler; and 7.4 percent
was recycled, regenerated, or reused. The disposal problem is complicated by
the fact that hydrated metal oxides from the neutralization process usually
will not dewater to more than 10-20 percent solids, so this sludge is not
really a solid but a gelatinous fluid. The pickling of one millions tons of
steel could result in 200,000 tons of wet sludge, which would require 150
acre-feet of permanent fill volume. A change from deep-well disposal to
neutralization would, therefore, cause a significant increase in the amount of
sludge that must be disposed of in landfills.
In Pennsylvania, most pickle liquor is handled by two contract haulers
Q
who use the following disposal technique:
1. Pickle liquor is placed in a lagoon and neutralized;
2. The liquid is floated off and the sludge is left in place in the
lagoon, and
3. When the lagoon is full, it is covered with a sloping top of soil
and revegetated.
2.5 Summary of Waste Generation
The nature of a specific waste generated may vary widely from plant to
fi 9
plant in both composition and quantity. ' Some typical variations are shown
in Table 4; the quantities are based upon a plant producing 2.5 million
metric tons of steel per year. Compositions vary as widely as quantity. For
example, blast furnace dust may range from 5.9 to 54 percent iron, and EOF
sludge has varied from 32 to 66 percent iron, or 0.3 to 13 percent zinc.
A summary of the waste generation (Table 5) shows that 78-percent of the
138 million tons of solid waste is recycled or reused. Slag is the predominant
landfilled waste at 13.4 million tons of the 30.5 million tonne total. The
miscellaneous wastes include metallic scrap, rubble and debris, pickle liquor
sludge, fly ash, and bottom ash.
3.0 CURRENT DISPOSAL FACILITIES
3.1 Prevalence of Types of Disposal Practices
Published data were reviewed and supplemented with data from state agencies
to obtain estimates of the number of disposal sites and percentages of wastes
disposed of on-site, off-site, and by contract disposal. The data base for the
376
-------
Table 4. VARIATIONS IN WASTE GENERATION QUANTITIES6'7 (BASIS = 2,500,000
TONNES STEEL/YR)
Quantity of Waste, tonnes/yr
Waste
Coke Breeze
Still Lime Sludge
EAF Slag
EAF Dust
EOF Slag
Blast Furnace Slag
Blast Furnace Dust
Blast Furnace Sludge
Minimum
17,300
315
25,000
2,925
230,000
345,600
11,200
3,200
RTI Estimate
32,400
540
60,000
6,500
290,000
556,800
25,360
40,000
Maximum
45,000
540
164,500
8,000
400,000
820,800
54,500
44,800
377
-------
Table 5. SUMMARY OF WASTE GENERATION FOR 125 MILLION TONNES OF STEEL PER
YEAR ("MILLIONS OF TONNES PER YEAR)
Waste
Coke Plant
Slag
Dust
Sludge
Scale
Misc ellaneous
TOTAL
Generated
1.9
47.7
3.1
4.0
6.9
24
137.6
Land filled
.1
13.4
.9
1.6
1.3
13.2
30.5
%
( 7)
(28)
(29)
(41)
(19)
(18)
(22)
Recycled, Used, or
1.8
34.3
2.2
2.4
5.6
60.8
107.1
Stockpiled %
(93)
(72)
(71)
(59)
(81)
(82)
(78)
378
-------
prevalence of different types of sites consisted of the 13 plants visited by
6 9
Dravo and Calspan and 20 plants for which information was provided by state
agencies in Pennsylvania, Indiana, Maryland, Michigan, and Ohio. The various
disposal facilities for the 33 plants included 28 on-site, 11 off-site, and 10
contract disposal sites. The total for contract disposal does not include slag
processors or those contractors handling spent pickle liquor only.
The use of on-site landfills appears to be a function of plant location
and land availability. For example, a company in the Pittsburgh area has one
large off-site landfill serving four plants, while in eastern Pennsylvania
a large plant has five landfills on its own property.
Contract disposal is used routinely in combination with on- or off-site
disposal. Based upon a sample of 10 contract haulers, the types of wastes
eliminated via contract disposal (excluding slag, oil, pickle liquor) were
plant rubble, debris, miscellaneous wastes (4), sludges (4), and soaking pit
slag (2).
Complete data on the quantities of waste disposed of by each method were
-*
available for 17 plants. These quantities were summed and the percentage of
total nonhazardous waste eliminated via each of the three disposal categories
was estimated as 65 percent on-site, 29 percent off-site, and 6 percent by
contract disposal.
3.2 Estimate of the Number of Major Landfills
To estimate the number of major landfill sites, it was necessary to
establish the number of major iron and steelmaking plants. A review of the
industry revealed that there were approximately 50 plants using blast furnaces,
basic oxygen furnaces, or open hearths (often in combination with electric arc
furnaces). In addition, 13 of the 103 plants using only EAF's have capacities
exceeding 500,000 tonnes of steel per year and were arbitrarily included as
major plants. The total of 63 major plants to be used as the basis for esti-
mating the number of landfills account for more than 90 percent of steel
production (Table 6). The estimate of major landfill sites for these plants
included 53 on-site, 21 off-site, and 19 off-site landfills belonging to
contract haulers.
379
-------
Table 6. ESTIMATE OF MAJOR LANDFILLS
No. of Major Plants
On-Site
Off-Site
TOTAL
Contract Disposal*
Data Base
33
28
11
39
10
Estimate for Total
63
53
21
74
19
*Excludes slag, pickle liquor, and waste oil processors
380
-------
4.0 IMPACT OF SECTION 4004 OF RCRA
Section 4004 of the Resource Conservation and Recovery Act of 1976 (RCRA)
provides for the promulgation of regulations containing criteria for determin-
ing whether solid waste disposal facilities should be classified as sanitary
landfills or open dumps. The establishment of open dumps is prohibited, so
future sites must meet the criteria that define a sanitary landfill. At a
minimum, a facility may be classified as a sanitary landfill only if there is
no reasonable probability of adverse effects on health or the environment from
the disposal of solid waste there.
Proposed classification criteria for solid waste disposal facilities were
12
published in the Federal Register on February 6, 1978. The major impact on
current disposal practices is to require that the quality of the groundwater
beyond the disposal facility's property boundary is not endangered. Endanger-
ment is defined as the introduction of any substance into the groundwater in
such a concentration that additional treatment is necessary for a current or
future user of the water, or otherwise makes the water unfit for human consump-
et foi
13,14
12
tion. Maximum contaminant levels are set forth in promulgated National
Interim Primary Drinking Water Standards.'
Table 7 lists various permissible criteria of selected leachate components
in drinking water. Contamination beyond these limits makes the water undesir-
able for human consumption. Organic leachate components are also of concern
because certain coke plant wastes are known to contain polycyclic aromatic
hydrocarbons. National standards for suspected carcinogens such as polycyclic
aromatic hydrocarbons have not been promulgated due to a, lack of information
about health effects. Specific organic compounds which are currently monitored
have been selected on the basis of the likelihood of occurrence in treated
water, the toxicity data, and availability of practical analytical methods.
EPA is actively investigating suspected carcinogens and future water standards
may reflect this activity. The World Health Organization drinking water
standards permit only 0.0002 mg/£ of polynuclear aromatic hydrocarbons.
4.1 Water Extraction of Solid Waste Materials
Water extraction tests were reported by six plants to the Pennsylvania
Ifi 1"
Department of Environmental Research as x*ell as from an EPA survey and ASTM. '
These tests differ from the proposed EPA Extraction Procedure in that distilled
381
-------
Table 7. PERMISSIBLE CRITERIA FOR SELECTED COMPONENTS FOR PUBLIC WATER
SUPPLIES
Constituent
Permissible Criteria (mg/£)
pH
Arsenic
Barium
Cadmium
Chromium
Fluoride
Iron (filterable)
Lead
Manganese (filterable)
Selenium
Silver
Total dissolved solids
Zinc -
Carbon chloroform extract
Cyanide
Oil and grease
Phenols
Mercury
6.0-8.5'
0.05a'b
,a,b
1.0'
,a,b
0.010
0.05a'b
1.2 (63.9-70.6°F)b
a,b
a,b
a,b
0.3
0.05
0.051
0.01
0.05
500.Ob
5.0b
0.151
0.2b
Virtually absent
o.ooib
0.002a
National Interim Primary Drinking Water regulations
Water Quality Criteria, Department of Interior, FWPCA
382
-------
water was used, whereas the proposed EPA procedure uses a limited amount of
acetic acid for pH control. The ASTM leachate values were reported by Enviro
Control with additional ASTM testing provided by AISI. 'Although ASTM tested
the wastes with several different types o.f water, only the 48 hour extraction
with carbon dioxide saturated rea-gent water is included in this study.
Water extraction testing provides a common basis for comparing the po-
tential of various wastes to endanger the groundwater. It also provides a
basis for predicting which components may appear in leachate. It does not
predict natural leachate composition, since this is affected by many site
specific factors. Even within a single site, natural leachate composition
varies and at times may exceed the levels of a simple water extract.
4.1.1 Coke Plant Wastes. Coke plant wastes include coke breeze, tar
sludges, and pitches from various tar storage and processing operations,
ammonia still lime sludge, and biological treatment sludge. Due to the widely
diverse processes which can be used to treat the coke by-product gases, the
number of wastes, the amounts generated, and even the composition are expected
to vary from plant to plant. In general, coke plant wastes are expected to be
hazardous with the possible exception of coke breeze.
With the exception of pH, the water extract results are best expressed as-
the ratio of the amount of material in the extract divided by the permissible
criteria (i.e., number of times drinking water standards). The permissible
criterion, used was the largest concentration presented in the Drinking Water
Standards. This criterion, which may differ from some legal requirements, is
used to provide a uniform method for assessing potential aesthetic and health
impacts from leachate, and is'not used for the classification of a waste as
hazardous.
The tar decanter sludge extract contains relatively large amounts of oil
and grease as well as phenols (Table 8). Ammonia still lime sludge extracts
contains cyanides, phenols, and may contain polycyclic aromatic hydrocarbons in
concentrations high enough to be of concern. The water extract from cooler
sludge contained relatively large amounts of oil and phenols. Some tar is also
expected in the oil from the extract. In general, coke plant wastes should be
given special consideration because of the carcinogenic nature of the coke oven
gas from which they originate and the potential of phenols and cyanides to
endanger the groundwater. Most coke plant solid wastes are hazardous and
should be segregated from nonhazardous wastes.
383
-------
Table 8. RESULTS OF AQUEOUS EXTRACTION TESTS OF COKE PLANT .WASTES. (Results are expressed in the
Tar Decanter
Sludge
Ammonia Still
Lime Sludge
Cooler Sludge
Coke Breeze, Mine
Refuse
pH (units) Oil
8.9 1320.0
7.8 60.0
11.5 X
6.7 60 . 0
10.4 33.0
Phenols Cyanides Cd
5 x 105 3.0 X
1.3 x 105 <0.04 <3.2
2 x 104 990.0 X
12 x 105 0.2 <4.0
0.0 0.0 0.0
Cr Pb
<0.2 < 4.0
<3.7 9.6
0.4 10.0
<2.2 <10.2
0.0 0.0
co
oo
.P-
-------
4.1.2 Slags. Slags are the major solid waste generated by the iron and
steel industry. The results of aqueous extraction tests for various iron and
steelmaking slags indicate that steelmaking slags are generally of more en-
vironmental concern than the blast furnace slag (Table 9). For example, the
pH of leachate from steelmaking slags is much higher than that from blast
furnace slag and may be as high as 12.5. Other leachate components of possible
concern are chromium, lead, and phenols.
4.1.3 Dusts and Sludges. Water extraction tests performed on sludges
(Table 10) are somewhat incomplete, but show the presence of cadmium, chromium,
and lead. Oil, phenols, and cyanides were found in blast furnace sludge.
Results from dusts (Table 11) similarly show these three metals and oil,
phenol, and cyanides.
4.1.4 Miscellaneous Wastes. Many miscellaneous wastes were tested and
showed varying levels of cadmium, chromium, lead, oil and phenols. These
wastes include melt shop rubble, mill scale, soaking pit slag, wastewater
sludge, lagoon sludge, and acid rinse sludge.
4.2 General Information on Soil Attenuation and Leachate Movement
Water extraction results of steel wastes that are used to estimate leach-
ate composition were just discussed. However, for the purpose of assessing
the impact of leachate on the environment, it is important to understand the
mechanisms that may alter the leachate and the factors that affect the
accurate measurement of this impact on groundwater.
As leachate moves through subsurface soils, several mechanisms can affect'
the nature and, consequently, the environmental impact of the leachate. These
include ion exchange and adsorption by clay and organic soils, metal fixation,
18
and reactions of metal cations to yield a precipitate of low solubility.
Heavy metals in their metallic state are generally insoluble, but the heavy
metal salts (as from electroplating or pickling), may be quite soluble.
Ammonia that is present in leachate is oxidized to nitrate under aerobic con-
ditions by certain bacteria and may be nitrate by the time it reaches ground-
water.
The fate of organic leachate constituents is not well documented since
few have been identified and their toxicity is unknown. Organics may come
385
-------
Table 9,
RESULTS OF AQUEOUS EXTRACTION TESTS OF BOF SLAG. (Results are expressed in amount
Material
BOF Slag
Source
A
E
F
C
D
D
D
Solids
2.2
0.3
0.7
X
X
1.4
1.3
Oil
27
X
30
X
X
X
X
Cd
<3.2
<2.0
0.0
X
<1.0
<1.0
X
Cr
<3.7
<1.0
4.2
3.0
<1.0
<0.2
X
pH (units)
12.2
12.5
9.4
12.5
9-11
9.0
12.4
Pb
< 4.4
7.0
0.0
4.0
<0.2-1.6
1.2
X
Phenol
<23
<26
0.0
X
X
X
X
(jO
00
-------
Table 10. RESULTS OF AQUEOUS SOLUBILITY TESTS OF IRON AND STEEL SLUDGES. (Results are expressed
Material
Blast Furnace
Sludge
BOF Sludge
Open Hearth
Source
A
C
G
C
D
pH (units)
9.5
9.5
9.6
10.4
5.4-6.9
Oil
67.0
X
X
X
X
Phenols
14.0
400.0
X
X
X
Cyanides
25.0
X
X
X
X
Cd
3.2
X
X
X
1.0
Cr
3.34
0.4
3.6
1.8
1.0
Pb
4.0
4.0
X
4.0
1.0-2.0
Sludge
EAF Sludge
•C
11.5
X
X
X 1880.0 40.0
Co
oo
-------
Table 11. RESULTS OF AQUEOUS SOLUBILITY TESTS OF IRON AND STEEL DUSTS. (Results are expressed in
00
oo
Material
Blast Furnace
Dust
Open Hearth
Dust
EAF Dust
BOF Dust
Precipitator
Baghouse
>
>
Source
A
D
D
C
G
A
A
E
Solids
X
X
19.0
X
15.0
8.0
10.4
0.8
PH
(units)
11.7
6.3-7.2
6.8
12.6
7.0
12.4
8.2
12.5
Oil
X
X
X
X
13.0
53.0
20.0
X
Phenols Cyanides
250 <1.5
X 0.02-0.4
X ( X
X X
0 4.2
28.0 0.4
40.0 0.03
X X
Cd
X
63-360
330
X
353
<3.2
<3.2
X
Cr
0.6
0-1.0
0.0
6.8
25,000
<37.4
9.52
2.0
Pb
5.0
12-30
66.0
3000
6.0
<4.4
8.2
142
-------
directly from the solid waste or from decomposition products and are probably
19
subjected to adsorption and microbial degradation.
These mechanisms are described to Show the fate of some leachate constitu-
ents and not as a means of groundwater protection. 'They are often unpredict-
able in their effect, and once the soil capacity for a particular mechanism has
been exceeded, a constituent may have an unobstructed path to the groundwater.
Groundwater travels at low velocities, typically 1.5 meters/year to 1.5
meters/day. This low velocity results in laminar flow which exhibits mixing
characteristics different from the turbulent flow-usually associated with sur-
face water. When water with a composition different from the groundwater is
injected or percolated into the groundwater, it tends to maintain its integrity
and is not diluted with the entire body of groundwater. Instead, it moves with
19
the groundwater flow as a plume undergoing minimal mixing.
Some other factors that affect leachate movement and consequently affect
monitoring and sampling requirements for environmental assessment:
1. Geohydrologic conditions: Under some circumstances leachate will
percolate rapidly, as through coastal plains sand, or through channels
that may have developed in limestone. In other cases, it may move
only a few feet per year through .soils of low permeability.
2. Climatic conditions; Leachate will move differently depending on
whether or not the soil is frozen, the amount of annual precipitation,
and frequency of brief periods of intense rainfall in a dry climate.
3. Disposal methods: The type of disposal method, whether lagoon, pit,
dump, or landfill and the site preparation affect the rate of
leaching.
4. Type of wastes: Some important waste types are (a) solid, sludge, or
liquid (as in a lagoon with a continuous leachate plume), (b) organic
or inorganic, and (c) water soluble or insoluble.
5. Age of site; This is relevant in that leachate percolation may take
several months to reach the groundwater.
6. Miscellaneous: The influence of nearby wells, changes in aquifer
depth, and groundwater velocity also affect leachate migration.
4.3 Groundwater Analysis From Iron and Steel Landfills
Groundwater analyses were provided to the Pennsylvania Department of
Environmental Resources by several iron and steel companies in Pennsylvania
389
-------
(Table 12). When these results are compared with the leachate from individual
wastes, both the ground water and leachate extract contain large quantities of
oil and grease. The problems with the water in meeting drinking water stan-
dards include alkalinity (high pH), excessive dissolved solids, and significant
amounts of chromium and manganese. The overall quality of the groundwater was
difficult to assess because of the lack of testing for heavy metals such as
cadmium, and for specific organic materials in the extracts.
Five groundwater samples were provided by two steel companies for addi-
tional testing at RTI. The samples showed high levels of manganese with res-
pect to drinking water standards. Other elements for which maximum levels have
been set were generally within drinking water standards. Analysis for polynu-
clear aromatic hydrocarbons (PNA) showed levels of 3 to 30 parts per billion
(ppb). Although no national standards for PNA's have been established, this
level is 15 to 150 times the International Standards for Drinking Water of
0.2 ppb.
4.4 Cost of Proposed Section 4004 RCRA Criteria
The Resource Conservation and Recovery Act provides for the promulgation
of regulations and criteria for determining which facilities shall be clas-
sified as sanitary landfills and which shall be classified as open dumps. The
general current practice in the iron and steel industry is the dumping of
wastes in unlined sites. The major impact of the proposed criteria is to re-
quire the disposer to control the leachate migrating toward the groundwater.
All steel plant waste, with the possible exception of bricks, rubble, and
certain trash items are anticipated to have leachate which could make ground-
water unfit for human consumption. Contaminants such as oil and grease, dis-
solved solids, fluorine, chromium, manganese, lead, iron, phenol, cyanide,
cadmium, zinc, and mercury have been identified in the water extracts of some
of the various iron and steel wastes at concentrations that may endanger
groundwater.
Although most steel plant wastes are not currently classified as hazard-
ous, available leachate and/or water extraction test data have shown the poten-
tial for the extract to make groundwater unfit for human consumption. In view
of these facts and in evaluation of environmental endangerment, a lined land-
fill may be required for these wastes. A major economic impact may result if
390
-------
Table 12. SELECTED LEACHATE COMPONENTS IN THE GROUNDWATER OF VARIOUS IRON AND STELL WASTE
LANDFILLS. (Results expressed in amount measured divided by permissible
Site, Sample
Position
A.I
A, 6
B,3
C.l
D,l
E,2
T »-
Solids
4.5
6.3
5.4
1.6
X
X
Oil
206.0
120.0
120.0
0.53
81.0
22.5
pll
(units)
7.5
7.5
12.1
11.4
12.2
X
Ammo n ia
0.1
0.1
4.5
X
1.8
X
Cr
0.8
2.6
0.8
X
0.4
0.6
Mn
54.0
117.0
2.2
X
0.0
10.0
Phenols
<12.0
<13.0
<10.0
4.9
X
X
Cd
X
X
X
<2000
X
0.0
GO
VD
-------
contaminants must be removed from the collected leachate. However, since the
wastes are not classified hazardous, the leachate disposal method assumed for
these wastes is controlled discharge to waterways or recycle back through the
landfill.
Discarded steelmaking slag could need liners because of the high pH of
the water extract, the dissolved solids in the extract, and inorganic elements.
However, the slag does not require lined landfilling if it is used as a sal-
able product, for resource recovery, or if the state has exempted the disposal
area from groundwater requirements under Case 2 of the proposed rules. Since
steelmaking slags are a major landfilled waste, two calculations were performed
on the economic impact of the proposed criteria with and without the required
lined landfilling of steel slag.
The impact of the proposed Section 4004 criteria on the iron and steel
industry was calculated assuming the following: the criteria requires the
lined landfilling of certain wastes, the removal of the leachate resulting from
rainfall on these wastes, and the controlled discharge of the water which is
-»
collected. Therefore, the cost of the criteria would be the cost of converting
an existing landfill into an area for the collection and removal of leachate
and would require a substantial capital investment. The criteria do not spec-
ifically require changes in current solid waste disposal practices such as the
transportation of wastes, employment of landfill personnel, or purchase of land
for waste disposal. It should be pointed out that the costs of these criteria
do not include those costs incurred as a result of hazardous waste disposal,
which may be more expensive than for nonhazardous wastes. The elements con-
sidered in determining the additional cost include a hydrogeologic survey,
excavation, lining the landfill, leachate collection, and groundwater monitor-
ing.
The estimated annual capital cost of lining nonhazardous waste landfills
is $6.9 million (Table 13). The cost of a lined landfill for steel slag dis-
posal is approximately twice that of nonslag nonhazardous waste disposal.
Although some economies of scale are achieved with increasing waste disposal
volume, when steel slag is placed in a lined landfill, the overall cost is
still three times as high. The estimated cost is relatively low for two major
reasons, primarily because only the cost of converting a potential landfill
site to a lined landfill was considered, and secondarily because the majority
392
-------
Table 13. SUMMARY OF ESTIMATED 4004 CRITERIA COSTS
Annual % Current % Future % of
Capital Environ- Environ- Current
Cost mental mental % of Disposal
Enforcement ($ Millions) Costs Costs Sales Costs
A-Steel Slags 6.9 0.6 0.2 0-01 12
Excluded
B-Steel Slags 21.1 1.9 0.6 0.04 37
Included
393
-------
of iron and steel wastes are currently either recycled, sold, or used in a
manner consistent with the objectives of RCRA.
Current disposal costs are estimated for 30 million tonnes of nonhazard-
ous waste at an average cost of $1.90 per tonne or $57 million. This repre-
sents a small fraction of current and future environmental costs. Current
annual environmental operating costs were estimated as $8 per ton of steel,
20
including the cost of air and water pollution control. The long term total
environmental costs, including disposal of nonhazardous solid waste, are es-
21
timated as $3,620 million per year.
The proposed criteria for Section 4004 will have a major impact on the
disposal practices used by the industry and substantially increase the cost of
present land disposal systems. It is estimated that the capital costs for de-
veloping leachate collection facilities alone will double the disposal costs of
those wastes placed in lined landfills.
4.5 Conclusions
Although most iron and steel wastes are not listed as hazardous, the
available leachate testing data indicate that leachate control is needed to
protect groundwater. Most of these wastes are currently deposited in facil-
ities which do not provide for leachate collection. The major impact of the
proposed RCRA criteria is to require the disposer to prevent groundwater endap-
germent from leachate migration. Assuming that this is accomplished by using a
lined landfill with leachate collection, this will increase the industry's cur-
rent disposal costs by 40 percent. However, this 40 percent increase repre-
sents less than 2 percent of the current environmental costs.
394
-------
5.0 ACKNOWLEDGEMENTS
This study was jointly funded by the U.S. Environmental Protection
Agency's Industrial Environmental Research Laboratory and the Office of Solid
Waste. The guidance and direction of project officers John Ruppersberger,
Jan Auerbach, and William Kline are gratefully acknowledged.
395
-------
6.0 REFERENCES
1. McGannon, H. E., The Making, Shaping, and Treating of Steel, 9th Ed.,
Pittsburgh, U.. S. Steel, 1971.
2. Slag Analyses from Wheeling-Pittsburgh and Allegheny-Ludlum, Pittsburgh
Office, Pennsylvania Department of Environmental Resources, October 1978.
3- Evans, J. R. "Slag-Iron and Steel," Bureau of Mines Minerals Yearbook,
U. S. Department of the Interior, 1976.
4. Collins, R. J. "Availability of Mining Wastes and Their Potential For Use
as Highway Material," Federal Highway Administration, FHWA-RD-76-107,
May 1976.
5. Personal Communication with E. Young, Solid Waste Committee, American
Iron and Steel Institute, November 1978.
6. Pasztor, L. and S. B. Floyd, Jr., "Managing and Disposing of Residues from
Environmental Control Facilities in the Steel Industry," U.S.
Environmental Protection Agency, EPA 600/2-76-267, October 1976.
7. U. S. Environmental Protection Agency, "Development Document for Interim
Final Effluent Limitations Guidelines and Proposed New Source Performance
Standards for the Forming, Finishing, and Specialty Steel," EPA 440/1-
76-048-b, March 1976.
8. Personal Communication with Mr. C. A. Duritsa, Pittsburgh Office,
Pennsylvania Department of Environmental Resources, October 1978.
9. Leonard, R. P., "Assessment of Industrial Hazardous Waste Practices in
the Smelting and Refining Industry," Vol. Ill, Ferrous Smelting and
Refining, U. S. Environmental Protection Agency, EPA-SW 145 c.3, 1977.
10. Personal Communication with Edward C. Levy Co., Detroit, MI, October 1978.
11. Public Law 94-580, October 21, 1976. The Resource Conservation and
Recovery Act of 1976.
12. Solid Waste Disposal Facilities, Proposed Classification Criteria, EPA,
Federal Register, February 6, 1978, Part II.
13. National Interim Primary Drinking Water Regulations, EPA Federal
Register, Vol./ 40, No. 248-Wednesday, December 24, 1975.
14. Report of the National Technical Advisory Committee to the Secretary of
the Interior, Water Quality Criteria, Federal Water Pollution Control
Administration, Washington, D- C., April 1968.
15. International Standards for Drinking Water, 3rd Ed., WHO, Geneva, 1971.
396
-------
16. Leonard, R. P., "Assessment of Industrial Hazardous Waste Practices in
the Metal Smelting and Refining Industry," Volume IV, U.S. Environmental
Protection, EPA-SW 145 c.4, 1977.
17. Berney, B. W., "Hazardous Waste Listings: 'Fully Integrated Steel Mills."
U. S. Environmental Protection Agency, May 1978.
18. Weant, George E. and M. R. Overcash, "Environmental Assessment of Steel-
making Furnace Dust Disposal Methods," U. S. Environmental Protection
Agency, EPA 600/2-77-044, February 1977.
19. "Procedures Manual for Groundwater Monitoring at Solid Waste Disposal
Facilities," U. S. Environmental Protection Agency, EPA 530/SW-611,
August 1977.
20. "Prices and Costs in the United States Steel Industry," The Council on
Wage and Price Stability, Washington, D.C., October 1977.
21. A. D- Little, Inc., "Steel and the Environment: A Cost Impact Analysis—
A Report to the American Iron and Steel Institute." A.D.L., Inc.,
C-80527, May, 1978.
397
-------
DEOILING AND UTILIZATION OF MILL SCALE
S. R. Balajee
Senior Research Engineer
Raw Materials Processes Research
Inland Steel Company
East Chicago, Indiana
ABSTRACT
Screening and deoiling of mill scale are necessary in order to
recycle mill scale through a sinter plant which incorporates a
baghouse without causing environmental and operational
problems. The oil and moisture contents of the mill scale vary
with its size and affect the screening operation. Without
deoiling, oil on the mill scale volatilizes during sintering.
As a result, the non-condensable hydrocarbons increase the
opacity of the stack gases and the condensable hydrocarbons
adversely affect the baghouse operation.
In view of the above, various mill scale deoiling methods were
investigated which include both thermal and water washing
techniques. Inland mill scale is currently being deoiled on a
commercial basis in a direct-fired kiln. Finally, a comparison
between oily and deoiled mill scale properties is made to
determine the effect of deoiling on mill scale utilization in
sintering and blast furnace operations.
398
-------
DEOILING AND UTILIZATION OF MILL SCALE
INTRODUCTION
Inland Steel Company's No. 3 Sinter Plant commenced
operation in 1959. The Dwight-Lloyd sinter strand is 2.44 m
wide and 51.20 m long with a corresponding strand area of about
125 m2. The strand has a design capacity of 3628 Mg/d of
iron ore acid sinter. Over the past several years,
super-fluxed sinter of ~3 basicity [(CaO + MgO)/(Al203 +
Si02)] has been produced from a mix containing iron ore,
sinter flux, steelmaking slag, blast furnace flue dust, mill
scale and burnt lime fines. In 1978, the sinter plant
productivity averaged 3300 Mg/d of basic sinter. The sinter is
consumed in the blast furnaces and accounts for about 15-25% of
the iron-bearing burden? the balance is made up mostly of
iron-oxide acid pellets with minor quantities of steelmaking
slag.
In the sinter plant, the multiclones of the settling
chamber, as well as the dry electrostatic precipitator, remove
solid particulates from the strand off-gas. In addition, in
order to comply with the stringent environmental regulations
regarding solid dust emissions, a baghouse was installed in
October, 1975, to treat the strand off-gas. A second baghouse
treats the off-gas from the sinter breaker, hot screening and
initial cooler zone at the sinter discharge end of the strand.
Due to the start-up of the mainstack baghouse, an earlier
practice of utilizing approximately 15-30% of -4.76 mm oily
mill scale fines containing ~ 0.4% oil and~4.5% moisture in the
sinter plant mix was stopped in late 1975. The oil on the mill
scale volatilizes during sintering, the resulting
non-condensable hydrocarbons increase the opacity of the stack
gas, and the condensable hydrocarbons in the strand off-gas
adversely affect the operation of the main stack baghouse
because the pores in the filter bags are blinded by the
entrapment of fine -0.044 mm oil-bearing baghouse dust which
contains significant quantities of lime, alkali and iron
oxide. Under these conditions, deoiling of the mill scale
fines is required in order to sinter this material at
relatively high concentrations in the sinter mix. For this
reason, various mill scale deoiling technologies were
evaluated, including water washing and thermal incineration
methods. Mill scale fines are currently being deoiled by an
outside contractor in a direct fired kiln for subsequent
utilization in the sintering operation.
399
-------
In this paper, existing deoiling technologies and the
current processing of mill scale at Inland, including screening
and deoiling, for consumption in the ironmaking and steelmaking
operations are presented. The oily and deoiled mill scale are
characterized on the basis of size, oil concentration, and
chemical composition. Changes in the oxidation state of iron
in the mill scale as a result of deoiling and the subsequent
effect on the heat generated by exothermic oxidation of the
mill scale during sintering are examined.
GENERATION AND CONSUMPTION OF MILL SCALE
The annual mill scale generation at Inland Steel has ranged
from 3.7 to 4.7% of the annual raw steel production, which is
similar to published mill scale generation rates.(!»2) In
1978, approximately 283,000 Mg of mill scale were generated,
which amounted to 3.7% of the annual raw steel production.
Mill scale is generated at the various slabbing, blooming, and
rolling mill operations. The oil concentration on mill scale
varies depending on the location of its generation (Table 1).
The variation in the oil content and size of the mill scale
necessitates blending prior to its consumption. Mill scale
contains a high concentration of iron (72-75% Fe) and <4%
gangue consisting of SiC>2/ Al2C>3, CaO and MgO (Table 2) .
The coarse sized mill scale is charged directly to the blast
furnaces and the mill scale fines are sintered.
SCREENING OF MILL SCALE
Screening of freshly generated mill scale is done in order
to produce three mill scale fractions for recycling in the
ironmaking and steelmaking operations. As seen in Figure 1,
the +150 mm mill scale contains mostly scrap and capping plates
and is consumed in steelmaking (<*-0.5-2%), the -50 + 3.35 mm
coarse mill scale is charged directly to the blast furnaces
(—12-18%), and the -4.76 mm mill scale fines are utilized in
sintering (*»80-87%). Magnetic separation of the coarse
-150+50 mm mill scale upgrades this material by discarding the
non-magnetic refractory (~0.5%). Thus, it can be seen that
mill scale fines comprised most of the mill scale generated in
the steel plant.
It was found that moisture and oil adversely affect the
screening of mill scale at finer sizes, such as 4.75 mm. The
screening efficiency is improved as the moisture on the mill
scale decreases. On the average, mill scale contains ~ 0.4%
oil (range of 0.1-1.7%) and ~ 4.5% moisture (range of
1.4-10.2%). The oil and moisture concentrations decrease to
some extent after blending and allowing the oil and moisture to
drain by gravity (Table 1). This results in some improvement
400
-------
in the efficiency of screening mill scale. However, the
-50+4.76 mm mill scale still contains 0.19% oil because of the
presence of high oil bearing piggy-back fines (Tables 3 and
4). The removal of these fines from the surface of the coarse
mill scale particles by drying the mill scale prior to
screening improves the screening operation and produces an
oversize scale with a lower oil content for direct blast
furnace charging. This can be seen by comparing the oil
content of the mill scale as received and after drying at
100°C for 24 hours (Table 4).
The oil concentration of mill scale is a function of the
particle size. The oil and moisture contents increase with a
decrease in particle size because of increased surface area,
thereby resulting in increased surface absorption of the oil
(Tables 1, 3, and 4).
DEOILING OF MILL SCALE FINES
The following methods were investigated for deoiling mill
scale fines which were produced by screening mill scale at 4.76
mm:
1. Washing methods
2. Thermal incineration methods
The oil concentration of mill scale after deoiling should be as
low as possible for utilization in the sinter plant equipped
with a baghouse.
Was_hijig_ Me;thod£
Commercial deoiling of mill scale is done by water washing
in a cyclone and a series of spiral rake classifiers at
Stelco's (Steel Company of Canada's) Hilton Works, Hamilton,
Ontario, and is operated by Uramco International.(3,4) Fine
mill scale containing 3-6% oil is deoiled by this process to
0.2% oil. It is also claimed(4) that the oil content could
be lowered further by the addition of oil breaking and
emulsifying chemical compounds in the secondary screw
classifiers. Approximately 10-15% of the mill scale generated
at Stelco is deoiled by water washing and subsequently sintered
together with mill scale fines which are relatively low in oil.
As noted in Table 5, the results of laboratory deoiling of
-0.6 mra Inland mill scale fines by water washing with agitation
indicate that as much as 60% and 70% of -the oil was removed by
401
-------
using cold and hot water, respectively. In contrast, about 90%
of the oil was removed by Colerapa Industries, Ravenna, Ohio,
when the mill scale was deoiled on a laboratory scale in a 75
mm diameter cyclone by cold water washing. The mill scale in
the cyclone underflow was recycled up to six times through the
cyclone in a slurry form. No chemical additives were used in
the test. Under these conditions, the mill scale was deoiled
to a range of 0.06-0.18% oil which reflects the initial oil
concentration.
The use of hot alkaline solution increased mill scale
deoiling efficiency to 90% and achieved deoiling to 0.08% oil
from an initial 2% oil concentration (Table 5). The use of hot
alkaline water solution for deoiling fine high oil bearing mill
scale sludge was demonstrated on a pilot plant scale in West
Germany.($)
Water washing deoils mill scale to ~0.1-0.2% oil. In order
to deoil mill scale to lower oil levels, the solvent washing of
mill scale by recyclable chlorinated hydrocarbons is currently
being investigated by Colerapa Industries, Ravenna, Ohio. It
is anticipated that an oil free mill scale product and oil will
be obtained. The solvent washing process consists of oil
extraction by a solvent, drying of the deoiled mill scale, and
separation of oil and water from the solvent by distillation.
In terms of oil recovery, this process may be more applicable
to high oil bearing steel plant sludge materials containing
6-20% oil. With the increasing cost of fuel oil and natural
gas used in various thermal deoiling methods, the solvent
washing process-^6) may become economically attractive;
however, its technological viability remains to be proven.
Thermal^ Met_hods_
Laboratory static (1.1 m3/h air flow) and dynamic thermal
deoiling tests indicate that mill scale is deoiled to<0.01%
oil after 30 minutes at 315°C (Table 6 and Figure 2). The
deoiling time could be reduced by increased temperature and,
perhaps, gas flow rate.
The following pilot or plant scale thermal mill scale
deoiling methods were investigated in which oil is volatilized
from the surface of mill scale:
a) By utilizing the sensible heat of molten steelmaking
slag,
b) By flash burning of oil from mill scale-lime
mini-pellets,
c) In an external indirect-fired kiln, and
d) In an internal direct-fired kiln.
402
-------
Mo l_t en_S t ee lma]<; i n g_S l_ac[
In this method, the sensible heat of molten steelmaking
slag, a waste heat source in the steel plant, is utilized to
deoil mill scale fines by contact in a slag pit. Successive
layers of oily mill scale fines and molten slag result in the
devolatilization and ignition of oil on the mill scale. (7)
Approximately 200 Mg of -4.76 mm mill scale fines
containing *»• 0.3% oil was deoiled by about 400 Mg of No. 4 EOF
steelmaking slag. The slag temperature was 1437°C. On
contact, oil from a 100-150 mm thick layer of mill scale is
partially devolatized and subsequently forms a big fire ball.
The remaining oil continues to burn in the form of small flames
(less than 250 mm high) at several broken slag/air interfaces.
At these interfaces, bluish grey hydrocarbon fumes were noticed
when the slag-mill scale mixture was water quenched, which
indicates that hydrocarbon emissions are a problem with this
process .
After cooling, the mill scale particles beneath the slag
layer retained their size structure and did not appear to
dissolve in the slag matrix. This observation indicates the
possibility of concentrating the deoiled mill scale fines in
the screen undersize by screening the slag-scale mixture at
6.35 mm. The separation of the mill scale fines from the
screened slag-scale mixture by magnetic separation was not
satisfactory because mill scale contains significant quantities
of antif erromagnetic wustite and hematite, besides
ferromagnetic magnetite and metallic Fe.
The mill scale was deoiled to <0.07% oil. The
aforementioned hydrocarbon emission problem coupled with
excessive fine dust generation during the handling and
processing of the scale-slag mixture precludes the deoiling of
mill scale by this method, unless the process is modified in
order to render it environmentally acceptable.
Flash_ Bur_nin£ of_Mi^ll^ i3cale-Lime_Min:L-Pe;Llets_
Deoiling of mill scale fines by the Inland flash burning
process (8) involves the following:
1. Mixing and balling of-* 90% by weight of -4.76 mm mill
scale fines with ~ 10% by weight of -3.35 mm burnt lime
fines at~8% moisture to produce mini-pellets.
2. Air drying the mill scale-lime mini-pellets to less
than 6% moisture for better strength, less spelling
and ease in handling prior to deoiling.
403
-------
3. Deoiling of the partially dried mini-pellets on a pan
conveyor by an impinging burner flame at a pellet
surface temperature of about 315°C.
Deoiling by flash burning is the result of ignition of oil
at the pellet surface; the oil appears to concentrate on the
surface of the mini-pellet due to capillary action associated
with the burnt lime. It was found that the heat source for
flash burning should have a flame temperature of at least
650°C for complete combustion of the hydrocarbons in the
waste gas. In a pilot test run, natural gas burners were used
which yielded a flame temperature of about 1000°C. A 25-40
mm thick layer of the mini-pellets was deoiled by a vertical
impingement of the flame. The retention time for flash burning
was about 1.5 minutes when the pan conveyor speed was 1.8
m/min. The natural gas consumption and the throughput rate
were a function of the moisture content of the air dried
mini-pellets. Specifically, as the air-dried mini-pellet
moisture varied from 2.9 to 5.2%, the energy consumption and
throughput rate changed from 663 to 882 kJ/kg and 7.2 to 4.9
Mg/h, respectively. Under these conditions, the oil content
was reduced from 0.42% to 0.07%.
In d i r_ec t - Fi r e d_K il n
In a pilot scale test run, about 180 Mg of -4.76 mm mill
scale fines were deoiled in an unlined indirect fired kiln
utilizing natural gas as a fuel. The kiln residence time was
estimated to be about 23 minutes and the shell temperature was
427°C. The initial and deoiled products averaged 0.58 and
0.08% oil, respectively.
Deoiling mill scale in an indirect-fired kiln was done in
an attempt to recover oil by collecting condensable
hydrocarbons for reuse. The recovered oil from mill scale
could lower the fuel consumption, and its quantity would
reflect the initial oil concentration and oil recovery
efficiency. The hydrocarbon recovery system consisted of a
primary condenser, a spray washer and a skimmer tank. A
centrifugal blower with a capacity of 0.472 Nm3/s was used to
inject the kiln off-gas into the hydrocarbon recovery units.
The hydrocarbons produced by mill scale deoiling consisted
of two components: a light oil and a heavy greasy sludge. The
light oil can be readily recovered; however, the recovery of
the heavy oil from the greasy sludge requires a distillation
step. No quantitative data was obtained regarding the recovery
of the oils. A detailed evaluation of the properties of the
oils will be needed to develop a hydrocarbon recovery flow
sheet.
404
-------
In the pilot scale test, inadequate condensation resulted
in excessive generation of a white mist in the final exhaust
gas. The mist may contain uncondensed and non-condensable
hydrocarbons; if this is the case, an environmental control
device may be needed to combust the hydrocarbons prior to
discharging the kiln off-gas in the atmosphere. Another
alternative is to recycle the mist-bearing gas to the
combustion unit used for generating the external kiln heating
gas. In addition, the possibility of eliminating or minimizing
mist generation in the kiln off-gas should be investigated by
changing the kiln operating conditions such as the oxygen
potential, residence time, and temperature.
Pilot scale deoiling tests were conducted in a direct-fired
kiln at Miller Compressing Company, Milwaukee, and Walker
Corporation, Gary, respectively. The results showed that it is
possible to consistently produce a deoiled mill scale product
analyzing <0 . 01% oil. In the Miller kiln (1.4 m diameter x
6.1 m long), approximately 40 Mg of -4.76 mm Inland mill scale
fines containing 1.0% oil was deoiled at a throughput rate of
5-6 Mg/h. In the Walker kiln (3 m diameter x 20 m long), -4.76
mm Inland mill scale fines containing 0.5% oil were deoiled at
a throughput rate of 25 Mg/h. The kiln off-gas temperature for
both kilns was in the range of 500°C. The hydrocarbons in
the kiln off-gas were combusted by after-burners prior to final
discharge from the stack.
On the basis of the consistent low concentration of oil
obtained from a direct-fired kiln, it was decided that Inland
mill scale should be deoiled on a commercial basis in a
direct-fired, refractory lined kiln. Therefore, since October,
1978, about 27,300 Mg/month oily mill scale fines have been
deoiled by this method for utilization in the sintering
operation.
The direct-fired kiln used for deoiling mill scale is a
counter-current reactor in which air is drawn from the mill
scale exit end. The kiln (3 m diameter x 20 m long) operates
at 7 rpm and uses an average of 23.6 dm3/Mg of No. 2 fuel oil
to deoil the mill scale in the kiln and to combust off-gas
hydrocarbons in after-burners (Table 7) . It was found
necessary to maintain a minimum kiln off-gas temperature of
~ 315°C in order to deoil mill scale to <0.01% oil. For
complete combustion of the hydrocarbons, the two after-burners
have to be operated at 650°C; approximately 40% of the total
fuel oil consumed is utilized in the after-burners. The kiln
off -gas is cleaned of fine solid particulates by a wet scrubber
prior to discharge to the atmosphere. The deoiled wet scrubber
405
-------
sludge is added to the deoiled kiln product in minor quantities
and then sintered. The breech material, which is collected by
gravity from the kiln off-gas at the kiln feed end, is recycled
(Figure 3).
Compari_son_o:f Vari.pus_M_ill. Scale Deoi_ling_Methods
Water washing resulted in the deoiling of mill scale fines
to 0.06-0.18% oil depending upon the initial oil concentration
(0.3-2.0%). Organic solvent washing may achieve deoiling to a
lower level, however, this remains to be proven.
Of the thermal mill scale deoiling methods investigated,
the direct-fired kiln gave the best result by consistently
producing a deoiled mill scale product containing<0.01% oil
(Table 7). The other thermal deoiling methods deoiled mill
scale to about 0.08% oil, and the fuel consumption would
increase if deoiling to lower levels is required. The
direct-fired kiln deoiling of mill scale uses about 40% of the
total fuel in combusting hydrocarbons in the waste gas, which
may also be needed in an external indirect-fired kiln to the
extent that the remaining uncondensed and non-condensable
hydrocarbons in the waste gas from the hydrocarbon recovery
system cause environmental problems.
The fuel consumption in the thermal deoiling methods, as
given in'Table 7, cannot be compared realistically because of
the variability in the oil concentration of the final deoiled
product, the presence and absence of pollution control devices
to combust hydrocarbons in the waste gas, and the sizes of the
plants. Furthermore, in the case of indirect-kiln deoiling,
the kiln was not lined with refractory and the recovery of
hydrocarbons could not be quantitatively established.
The cost of fuel oil makes the thermal deoiling of mill
scale very expensive. Less energy intensive mill scale
deoiling processes, which include washing by water(3) and/or
organic solvents(6), and flash-burning(7'8) may become
attractive. In some of the thermal deoiling methods,
environmental control devices may be needed to control
hydrocarbon emissions. In addition, process changes may, in
some instances, be required to obtain a deoiled mill scale
product with a lower oil content.
UTILIZATION OF MILL SCALE
It was stated earlier that the screening of mill scale is
necessary for recycling in ironmaking and steelmaking
406
-------
operations. At Inland Steel, the +50 nun mill scale contains
mostly capping plates and scrap and is recycled in steelmaking
operations (Figure 1). The -50+3.35 ram coarse low oil-bearing
mill scale is recycled through the blast furnace as a direct
metallic charge. The -4.76 mm mill scale fines are sintered,
but must be deoiled in order to be used at high concentrations
in the mix. In this section, the utilization of mill scale
fines in sintering is discussed - at low concentrations of oily
mill scale fines and high concentrations of deoiled mill scale
fines. In addition, both oily and deoiled mill scale fines are
characterized with respect to iron oxidation state and
mineralogy, followed by thermodynamic calculations regarding
the exothermic oxidation of mill scale during sintering.
Based on the sinter plant experience, it is possible at
Inland to recycle oily mill scale fines up to a concentration
of 7% of the raw sinter mix (or 5% of the sinter burden, which
includes the return fines and coke breeze) without causing
baghouse and other operational problems.
The deoiling of mill scale is required in order to utilize
mill scale at concentration greater than 7% in the sinter
plant. Therefore, since September, 1978, about 240,000 Mg of
deoiled mill scale fines were produced in a direct-fired kiln.
The entire deoiled mill scale product was sintered in
concentrations ranging from 20-30% of the raw sinter mix.
Mill scale contains exothermic iron and iron oxide
constitutents which affect the heat balance of the sintering
operation. Therefore, the sinter mix composition becomes
important in maintaining proper heat input to the mix by the
usage of endothermic sinter flux stone and by control of
exothermic carbon- and iron-bearing materials, such as blast
furnace flue dust, steelmaking slag which contains metallic
iron, and mill scale. In this regard, the oxidation states of
the iron were determined for oily and direct-fired kiln deoiled
mill scale fines. Based on this data, calculations of the heat
generated during mill scale oxidation were made for a sintering
temperature of 1327°C.
As noted in Table 8, the metallic iron concentration varied
from 3.0 to 10.8% for three samples of oily mill scale fines
produced by screening at 16, 12.5, and 4.76 mm, which may be
the result of sample inhomogeneity, inaccurate chemical
analysis, or variability in the oxidation states of iron in
mill scale due to the varying processing conditions employed at
the various mill operations. The total Fe, Fe++, and Fe+++
concentrations in the mill scale did not change significantly
during the deoiling operation. However, in the case of the
direct-fired kiln wet scrubber sludge (Figure 3), the Fe++
407
-------
and metallic Fe concentrations were relatively low and the
Fe+++ content was relatively high because the particles in
the scrubber sludge are finer and subjected to higher
temperatures in order to combust hydrocarbons in the kiln
off-gas.
On the basis of the concentrations of the various iron
forms, and assuming that one percent of the magnetic Fe, as
determined by Satmagan, is equivalent to 2.49% metallic
Fe(9,10) an<3 that the antiferromagnetic wustite and hematite
do not contribute towards the magnetic susceptibility
measurement by the Satmagan unit,(1°) the iron mineralogy of
the oily and deoiled mill scale fines was determined. The data
presented in Table 9 shows that, as a result of deoiling, the
magnetite and metallic Fe concentrations increased while the
wustite and hematite concentrations decreased.
The heat generated by the oxidation of wustite, magnetite,
and metallic Fe to hematite at a sintering temperature of
1327°C is presented in Table 10. On a constant mass basis,
metallic Fe and wustite generate respectively about 14.4 and
3.9 times more heat than magnetite. Therefore, any change in
the initial metallic iron concentration of mill scale would
have a significant effect on the heating value of mill scale.
Comparing the respective iron mineralogy and heat generation
data given in Tables 9 and 11, the heating value of mill scale
appears to be more sensitive to the initial metallic Fe content
than to changes in the iron mineralogy due to deoiling. A
metallic Fe concentration change from 3 to 5.6% and 10.8%
resulted in a respective 24% and 42% increase in the total heat
generated by mill scale (Table 11). However, only up to 10%
excess heat can be generated by changes in the oxidation states
of iron in mill scale as a result of deoiling. The wet
scrubber sludge from the direct-fired kiln deoiling operation
is oxidized due to the high temperature combustion of
hydrocarbons at 650°C, and, as a result, contains less
oxidative heat than mill scale.
The excess heat of deoiled. mill scale may be compensated to
some extent by the combustion of oil on oily mill scale which
is a function of the type of oil, initial oil concentration and
degree of volatilization and combustion during a single or
multiple pass(es) of air or strand off-gas (waste gas) through
the sinter bed.(H) Calculations show that 0.3 to 0.4% oil
on mill scale will generate 0.13 x 10"^ to 0.17 x 10~6
kJ/Mg of heat on complete combustion. This amounts to about 7
to 13% of the heat generated from the oxidation of oily mill
scale (Table 11). The combustion of oil on mill scale occurs
to a limited extent without waste gas recirculation through the
408
-------
sinter bed. In view of the above, more heat may be generated
by the deoiled mill scale fines when compared with the oily
mill scale fines while sintering a mix containing identical
mill scale concentrations.
The effect of mill scale deoiling on the permeability of
the sinter mix containing deoiled mill scale could be another
potential problem affecting sinter production. The size
distributions of the oily and deoiled mill scale fines, as well
as that of the wet scrubber sludge, are presented in Table 12.
About 4% more of the -0.15 mm fines was generated in the kiln.
The deoiled kiln product mixed with minor amounts of the wet
scrubber sludge is sintered; the mix contains 21.5% -0.15 mm
fines, as opposed to 8.7% -0.15 mm fines in the case of the
oily mill scale fines. An increase in the concentration of
-0.15 mm fines in the mill scale decreases the mix
permeability. Furthermore, the wetting characteristics of the
deoiled mill scale by water will be different than that of the
oily mill scale because of the hydrophobic nature of the oily
mill scale. The mix permeability may be increased by balling
the sinter mix containing mill scale prior to sintering because
of an increase in the apparent mean size of the mix and a
subsequent decrease in the effective concentration of the
-0.15 mm fines contained in the mix. Balling of the deoiled
mill scale fines may have a beneficial effect on the sinter mix
permeability and sinter plant productivity.
The sinter bed permeability can be improved by reducing the
bed height for a specific suction level. The downward movement
of the flame front during sintering coupled with the flow of
preheated air and waste gas increases the maximum bed
temperature as the height increases, especially in the lower
bed region near the grate. Under these conditions, the heat
generated by mill scale oxidation in the bed could cause
slagging of the bed, particularly in the lower bed region,
because of relatively higher bed temperatures. This effect can
be seen from Table 13, where reduction in bed height at an
increased mill scale concentration of 30% resulted in the
attainment of generally similar bed permeability, as evidenced
by a similar value of suction in the last windbox (No. 21), as
when 22% deoiled mill scale was used in the sinter plant.
However, the sinter plant productivity dropped. This may be
partially due to increased cold return fines generation, which
reflects lower sinter quality, as well as lower sintering
temperatures, as evidenced by a lower strand off-gas
temperature. The adverse effects of deoiled mill scale on
sinter bed permeability, productivity, and sinter quality can
probably be reduced by proper sinter mix preparation and
process conditions.
409
-------
SUMMARY
Utilization of mill scale in sintering requires screening
of mill scale. In the case of a sinter plant equipped with a
baghouse, deoiling of mill scale fines in necessary.
Therefore, various mill scale deoiling methods were
investigated. The water washing of mill scale removes up to
90% of the oil, provided a hot alkaline solution with stirring
or intense agitation with a high volume of water in a unit such
as a cyclone is used. The deoiling of mill scale by thermal
methods was investigated by utilizing the sensible heat of
molten steelmaking slag, by flash burning of oil on mill
scale-lime mini-pellets by a burner flame on a pan conveyor,
and in a direct- or indirect-fired kiln. The direct-fired kiln
consistently gave the best deoiled product, which analyzed
<0.01% oil. Hence, the Inland mill scale fines are currently
deoiled in a direct-fired kiln. All of the deoiled mill scale
fines are utilized in the sinter plant in concentrations of 20
to 30% of the raw sinter mix.
Both the oily and deoiled mill scale fines were
characterized with respect to size, chemistry,-magnetic iron
content and iron mineralogy. Mill scale is high in iron and
low in impurities, making it a prime material for recycling.
The oil concentration increased with decreased particle size of
oily mill scale. Calculations of the heating value of mill
scale resulting from the oxidation at 1327°C of iron and iron
oxides in the mill scale indicated that direct-fired kiln
deoiling increased the heating value.
The high cost of fuel oil makes thermal direct-fired kiln
deoiling of mill scale rather expensive, especially when about
40% of the total fuel consumed is needed for environmental
control. Under these conditions, other mill scale deoiling
methods, such as the use of organic solvents for washing, or
low energy thermal methods, may become attractive.
REFERENCES
1. Pastzov, L., and Floyd, S. B., "Managing and Disposing of
Residues from Environmental Control Facilities in the Steel
Industry," Dravo Research, EPA Contract No. HR-803619,
June, 1976, p. 118.
2. Recycling of Steel Plant Waste Materials, British Steel
Corporation, Steelresearch 74, 1974, p. 19.
410
-------
3. "Steel Industry Sludge is Being Reused," Environmental
Science and Technology, Vol. 9, July, 1975, p. 624, and
"Recycling Mill Sludge-Profitably," Industry'week, Vol.
181, No. 11, June 10, 1974, p. 44.
4. Duval, L. A., "Method and Apparatus for Processing Waste
Water Slimes of Steel Mill Water Treatment Systems," U. S.
Patent No. 3,844,943, October 4, 1974.
5. Supp, A., and Zimmermann, K., "Untersuchungen Zn Entolung
Von Walzzunder, Techn. mlH. Krunp. Forsch. Ear, Vol. 33,
1975, #3, p. 89; ibid, Stahl U Eisen, Vol. 96, November,
1976, No. 23, p. 1177.
6. Bahrke, L.f Method for Degreasing Roll Mill Scale," German
Patent No. 2532689, July 25, 1975.
7. Pack, P. R., "Method for Removing Oil from Mill Scale and
Recovering Metallic Values Therein," U. S. Patent
Application No. 908894 Filed on May 23, 1978, U.S. Patent
document No. 4,177,062, dated December 4, 1979, assigned to
Harsco Corporation.
8. Schwarz, A. M., "Method of Deoiling and Agglomerating Mill
Scale," U. S. Patent Application No. 06/056,331 Filed on
July 10, 1979.
9. Gaudin, A. M., "Principles of Mineral Dressing,"
McGraw-Hill, 1939, p. 436.
10. Cullity, B. D., "Introduction to Magnetic Materials,"
Addison-Wesley, 1972, pp. 12, 157, 190.
11. Ban, T. E., "Ore-Sintering Process Reduces Air Pollutants,"
Process Technology, Chemical Engineering, Vol. 85, No. 14,
June 19, 1978, p."81.
SRB/jrb
411
-------
TABLE 1
Oil and Moisture Concentration of Mill Scale
Mills
80" Hot Strip Mill
No. 1 Pit
No. 2 Pit
No. 4 Slabber Mill
North Pit
South Pit
No. 3 Blooming Mill
No. 2 Blooming Mill
28" Structural Mill
14" Bar Mill
12" Bar Mill
10" Bar Mill
Mill Scale Fines
-12.5 mm Freshly Generated
-4.76 mm Stored Pile*
Raw Mill Scale
Oil, %
0.43
2.00
0.26
0.59
0.21
0.16
0.10
0.30
0.70
1.10
0.38
0.27
-0.40
Moisture, %
7.4
11. 0
2.1
10.2
1.4
1.5
2.5
3.2
6.4
5.6
4.4
4.0
~4.3
-3.35 mm, %
85
100
32
96
67
37
37
87
89
93
80
88
~70
For at least a six-month period.
412
-------
TABLE 2
Chemistry of -4.76 ram Oily Mill Scale Fines
Weight % (Dry Basis)
Total Fe 72-75
Fe** 45-50
Fe*** 19-23
Metallic Fe 3-11
Magnetic Fe 26-28
Gangue* < 4
Oil ~ 0.4
Includes SiC>2, Al^, CaO, MgO
413
-------
TABLE 3
Oil Concentration of Stored Mill Scale
Mill Scale
Coarse (85% -50+4.76 mm)
Fines (96% -4.76 mm)
Total
Mill Scaled Fines (96% -4.76 mm)
Screen Size (mm)
+3.35
-3.35 4-2.00
-2.00 +1.00
-1.00 +0.50
-0.50 +0.25
-0.25 +0.15
-0.15
Total
wt., %
18.4
81.6
100. 0
Air Dried
wt., %
12.3
14.8
13.9
23.1
13.0
11.1
11.8
100.0
oil, %
0.19
0.31
ND
to 0.6%
oil, %
0.16
0.21
0.28
0.30
0.35
0.45
0.61
0.32
Oil Units
0.035
0.253
0.288
.Moisture*
Oil Units
0.027
0.031
0.039
0.069
0.046
0.050
0.072
0.327
Oil Distribution, %
12.2
87.8
100.0
Oil Distribution, %
6.1
9.5
11.9
21.1
14.1
15.3
22.0
100.0
*In order to facilitate screening at finer sizes.
ND: Not Determined
414
-------
TABLE 4
Oil Concentration of Stored -4.76 mm Mill Scale Fines
as a Function of Moisture
As-Received Oven Dried (100°C, 24 h)
MDisture, % 3.2 0
Oil, % 0.35 0.26
Screen Size (mm) wt., % oil, % wt., % oil, %
+6.35
-6.35 +4.76
-4.76 +3.35
-3.35 +2.00
-2.00 +1.00
-1.00
Total 100.0 0.36 100.0 0.36
2.0
4.3
8.9
18.3
24.2
42.3
0.15
0.13
0.16
0.24
0.26
0.55
1.6
3.6
7.5
15.5
17.2
54.6
0.06
0.10
0.12
0.17
0.19
0.41
415
-------
TABLE 5
Water Washing of Inland Mill Scale
Test Done by Inland
80" Hot Strip Mill, No. 2 Pit
(-0.60 mm)
Oil on Mill Scale, %
Oil
As-Received
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
After Washing
1.00
0.90
0.80
1.00
0.90
0.70
0.60
0.46
0.08
50
55
60
50
55
65
70
77
97
Time,
10
10
5
10
Washing Condition
Cold water at 25°C,
stirred
Hot water at 80°C,
stirred
Hot alkaline solution
(PAFCO 338) at 80°C,
stirred
Test Done by Colerapa Industries
14" Bar Mill (-3.35 mm)
No. 4 Slabber, North Pit (-16.00 mm)
No. 4 Slabber, South Pit (-0.60 mm)
80" Hot Strip Mill, No. 1 Pit (-3.35 mm)
80" Hot Strip Mill, No. 2 Pit (-0.60 mn)
0.32
0.24
0.59
0.43
1.81
0.09
0.06
0.12
0.09
0.18
72
69
80
79
90
Cold water
washing in a 75 mm
diameter cyclone
-------
TABLE 6
Laboratory Thermal Deoiling of Mill Scale
Static Deoiling in a
Muffle Furnace
Screened Mill Scale Fines (-4.76 mm)
No. 3 Bloomer Mill (-12.5 mm)
No. 4 Slabber Mill (-6 mm)
80" Hot Strip Mill, #2 Pit (-0.6 mm)
Dynamic Deoiling in a 0.48 m
Diameter x 0.61 m kiln
rotating at 10 rpm
Deoiling Conditions
Mill Scale
Temperature ,
°C Time, min.
315 30
315 30
315 30
315 30
Oil
As-Received
0.6
0.3
0.6
2.0
Deo i led
Product
<0.01
<0.01
<0.01
<0.01
Screened Mill Scale Fines (-4.76 mm)
~315
~30
1.3
<0.01
-------
TABLE 7
Comparison of Various Thermal Mill Scale Depilinq Methods
00
Mater ial
Commercial Direct-
Fired Kiln
-4.76 mm mill scale
fines, 4% moisture
Unit Size
Feed Rate, Mg/h
Deoiling Temperature, °C
3 m diameter x 20 m
long kiln
55
N.D.
Exhaust Gas Temperature, °C 315-330
Volume, Nm^/Mg
Fuel Used for Deoiling
Fuel Consumption, kJ/kg
Mill Scale Oil Content, %
Initial
Final
700
No. 2 Fuel Oil
942**
0.40
<0.01
Pilot Scale Indirect-
Fired Kiln
-4.76 mm mill scale
fines, 2.5% moisture
1.5 m diameter x 6.1 m
long kiln
8.8
427*
N.D.
190
Natural Gas
1030
0.58
0.08
Pilot Scale Flash Burning
on Pan Conveyor
Mini-pellets containing
•90% -4.76 mm mill scale fines, and
•10% -3.18 mm burnt lime fines and
-8% moisture, as produced, and
2.9-5.2% moisture, after air drying
1.5 m wide x 4.9 m long pan conveyor
7.2
315
315-330
250
Natural Gas
663-882***
0.42
0.07
* Kiln shell temperature of 427°C.
** Includes fuel oil consumed for combusting hydrocarbons in the kiln off-gas.
*** Fuel consumption varied as a function of mini-pellets moisture.
ND: Not Determined
-------
TABLE 8
Analysis of Oily and Deoiled Mill Scale Fines
Total Fe (%)
Fe++
Fe+++
Metallic Fe
tMagnetic Fe
+0il, %
+Moisture, %
-16
Oily*
72.5
46.3
23.2
3.0
26.4
0.60
3.70
mm Mill Scale
Wet Scrubber
Deoiled* Sludge
72.9
45.3
24.3
3.3
37.5
<0.01
0.02
70.3
35.4
34.3
0.6
35.6
<0.01
7.60
-12.5 mm Mill Scale
Oily Deoiled
74.9 74.6
50.6 49.1
18.7 17.0
5.6 8.5
28.0 38.0
0.44 <0.01
3.70 0.00
-4.76 mm Mill Scale
Oily**
75.3
45.0
19.5
5.6
5.2***
10.8
26.6
0.34
3.70
Deoiled**
76.2
45.9
18.7
5.4
6.2***
11.6
35.3
<0.01
0.00
The various Fe analyses were made on the following ground mill scale samples:
* -0.18 mm ground fraction (~85% by weight of total sample)
** -0.21 mm ground fraction except for total Fe which was based on the weighted average of +0.21 mm
and -0.21 mm ground fractions, and for metallic Fe, marked***, the H-0.21 mm ground fraction was
used.
The various Fe analyses for -12.5 mm mill scale were based on the weighted average of +0.21 mm and
-0.21 mm ground fractions.
+ The oil and moisture were analyzed on the unground mill scale sample.
# The magnetic iron is determined by Satmagan, which is calibrated to measure the magnetic
susceptibility for magnetite.
-------
TABLE 9
Calculated Iron Mineralogy of Oily and Deoiled Mill Scale Fines
Iron Mineralogy, %
Wustite (FeO)
Magnetite (FeO-Fe^)
Hematite (F62P3)
Metallic Iron (Fe)
-16 mm Mill Scale
Oily Deoiled
51.6 ' 45.8
26.2 40.5
15.2 6.4
3.0 3.3
Wet
Scrubber
Sludge
31.0
47.1
16.6
0.6
-12.5 mm
Qiiy
59.2
19.4
13.3
5.6
Mill Scale
Deoiled
56.1
23.2
8.3
8.5
-4.76
Oily
52.6
17.6
15.7
10.8
mm Mill Scale
Deoiled
49.8
30.1
6.0
11.6
The iron mineralogy was calculated on the basis of the chemical and magnetic iron analyses as given in
Table 8.
-------
TABLE 10
Heat Generated During Oxidation to Hematite (Fe;£>3)
at 1327°C
Heat Generated Relative
Iron Mineralogy AH, kJ/Mg x 10" 6 to Magnetite
Magnetite (FeO-Fe^) -0.498 1.000
^ Wustite (FeO) -1.926 3.867
ro
Metallic Iron (Fe) -7.163 14.384
-------
TABLE 11
Heat Generated by the Oxidation to Hematite of
fO
AH, kJ/Mg x 1(T6
Relative to Oily
Mill Scale
Metallic Fe, %
AH, kJ/Mg x 1(T6
Relative
Oily
-16 ran
and
Deoiled
Mill
Mill Scale
Oily Deoiled
~" J_«
1.
-16
-1.
1.
320
000
Oily
Sludge*
-1.340 -0.880
1.015 0.667
Mill Scale
mm -12.
3.0
320
000
-1.
1.
5 mm
5.6
638
241
-4
-1
1
.76 mm
10.8
.874
.420
Scale Fines at 1327^
-12.5
mm
Oily
-1.638
1.000
Dec
-16
-1.
1.
mm
3.3
340
000
Mill Scale
Deoiled
-1.805
1.102
>iled Mill Scale
-12.5 mm -4.
8.5
-1.805 -1
1.347 1
-4
.76 mm
Oily
-1.874
1.000
Mill Scale
Deoiled
-1.940
1.035
76 mm
11.6
.940
.448
From the wet scrubber.
-------
TABLE 12
Screen Analysis of -4.76 mm Oily and Deoiled
Mill Scale Fines
Oily Mill Scale
Deoiled Mill Scale
N>
00
Screen Size
(mm)
+6.35
-6.35 +4.76
-4.76 +3.35
-3.35 +2.00
-2.00 +2.00
-1.00 +0.50
-0.50 +0.25
-0.25 +0.15
-0.15
Moisture, %
Oil, %
Feed
wt. %
1.6
3.6
7.5
15.5
17.2
22.3
13.2
10.4
8.7
3
0
to Kiln
Cum. wt. %
1.6
5.2
12.7
28.2
45.4
67.7
80.9
91.3
100
.2
.35
Kiln Product +
Kiln
wt. %
0.4
1.0
3.0
8.8
13.5
28.4
18.8
13.8
12.3
0.
0.
Product
Cum. wt.
0.4
1.4
4.4
13.2
26.7
55.1
73.9
87.7
100
00
01
Wet Scrubber Sludge*
% wt. %
0.6
1.8
3.3
8.1
16.4
24.7
13.3
11.3
21.5
Cum. wt. %
0.6
2.4
5.7
13.8
30.2
54.9
68.2
79.5
100
3.2
0.01
Wet Scrubber
Sludge
wt. % Cum. wt. %
0.0
0.2
0.1
0.2
0.2
0.2
5.3
15.8
78.0
7.6
0.01
0.0
0.2
0.3
0.5
0.7
0.9
6.2
22.0
100
Used in the sinter plant.
-------
TABLE 13
Sinter Plant Operating Data for Mixes as a
Function of Deoiled Mill Scale Concentration
Deoiled Mill Scale in the Mix
Process Variables
Bed Height, mm
Last windbox (No. 21) suction, kPa
Last Windbox (No. 21) Off-Gas Temperature, °C
22%
434
4.7
147
30%
394
5.2
114
Cold Return Fines (-6.35 mm), %
19.7
24.3
Sinter Quality
Bases (CaO + MgO), %
Basicity (CaO + MgO)/(Al 203 + SiO2)
Strength Index-Impact, +3.35 mm, %
24.3
3.3
72.0
22.2
3.1
71.5
Sinter Strand Availability, %
87.2
85.2
Sinter Production, Mg/m2/d
Actual
Normalized to 100% Sinter Strand Availability
25.3
28.9
23.1
26.9
-------
INLAND RAW MILL SCALE
(100 WT. %,~O.4% OIL)
150mmx900mm SCREEN IN
A 3.65mx3.65m
STATIONARY GRIZZLY
•HSOmm
125mmxl25mm
DUST CLEAN OUT
STATIONARY GRIZZLY
f — 150mm
1.82mx4.88m
VIBRATORY
SCREEN
DUST TO-4.76mm MILL SCALE FINES
-150mm + 50mm |CROSS BELT
"SEPARATION
-.-- +50mm AND+150mm
* CAPPING PLATES
AND SCRAP
FOR STEELMAKING
(~0.5-2 WT.%, % OIL-N.D.)
1.52mx3.65m
VIBRATORY
SCREEN
NON-MAGNETIC
REFRACTORY RUBBLE
TO LANDFILL
( »0.5WT.%)
*. -50 + 3.35mm
COARSE MILL SCALE
FOR DIRECT CHARGE
TO BLAST FURNACES
(~12-18 WT.%,~0.1% OIL)
-4.76mm MILL SCALE FINES
FOR SINTERING
(~80-87 WT. %,~0.4% OIL)
N.D.: NOT DETERMINED
FIGURE 1 SCHEMATIC DIAGRAM OF THE SCREENING OF INLAND
MILL SCALE
425
-------
2.000 i—
100°C
204°C
260°C
ANALYTICAL
DETECTION
LIMIT
I
10 20 30 40
DEOILING TIME, MINUTES
50
60
FIGURE 2 STATIC THERMAL DEOILING OF-4.76mm
MILL SCALE FINES AS A FUNCTION OF DEOILING
TIME AND TEMPERATURE
426
-------
-4.76mm INLAND MILL SCALE FINES
r
VIBRATING SCALPING
SCREEN 16mmx50mm
I
I +16mm
OVERSIZE STORED
(100 WT.%~0.4% OIL,~4.3% H2O)
103 WT.%
DIRECT FIRED DEOILING
KILN 3mx20m
— 2mm
-3 WT.%
— BREECH
MATERIAL
<0.01% OIL
0.00% H2O
I
AND
KILN
AS STEAM
HYDROCARBONS IN
OFF-GAS (~315°C)
55
— 4.76mm
-93 WT.%
KILN PRODUCT <0.01% OIL
Mg/h 0.00% H2O
-0.6mm ~3 WT. %
.WET SCRUBBER^,,
* SLUDGE **
<0.01% OIL
7.60%
I
AFTER-BURNERS (2)
TO COMBUST
HYDROCARBONS AT 650°C
WET SCRUBBER
TO REMOVE
PARTICULATES
FINAL DEOILED
MILL SCALE TO
SINTER PLANT
"96 WT.%
<0.01% OIL
~4.00% H2O<
H2O
STACK DISCHARGE
~80°C
'MOISTURE ADDED FOR
DUST CONTROL DURING
HANDLING
FIGURE 3 SCHEMATIC DIAGRAM OF COMMERCIAL DIRECT-FIRED
KILN DEOILING OF INLAND MILL SCALE FINES
427
-------
CHARACTERIZATION AND UTILIZATION OF STEEL PLANT FINES
Donald R. Fosnacht
Research Engineer
Inland Steel Company
East Chicago, Indiana
ABSTRACT
In the production of finished steel products, various waste oxide materials
are generated (e.g., blast furnace flue dust, steelmaking dusts, and mill
scale). The materials present serious environmental and resource recovery
problems. Future environmental, space, and economic considerations will
necessitate new handling and treatment methods for these materials. The
quantity of waste oxide material generated each year by American steel
producers is enormous. The contained iron content of the material is high and
if recycled would reduce the quantity of virgin materials needed by industry.
At present, however, problems associated with the dust materials make
recycling to blast furnace or steelmaking operations quite difficult.
Research is being conducted to characterize the dust materials and to develop
processing methods which will allow greater utilization of these materials in
our primary production processes. The physical and chemical properties of
mill scale, blast furnace flue dust, and steelmaking dusts have been
investigated. The various properties of the dust materials are given in this
paper. The results of this preliminary work point to methods which may allow
greater utilization of the materials in iron and steelmaking operations.
428
-------
INTRODUCTION
In the production of finished steel products, various waste oxide
materials are generated (e.g. blast furnace flue dust, steelmaking dusts, and
mill scale). These materials present serious environmental and resource
recovery problems. The amount of waste oxide material generated is dependent
on both the quality of the materials used and the operating conditions
employed in a given steel mill operation. The rates of waste oxide generation
from various Inland operations are shown in Table 1. These values are based
on long term plant experience. At its current raw steel capacity, Inland
generates over 650,000 tonnes of waste oxide material each year. Since
Inland's share of total raw steel production in the United States is
approximately 6 percent, the total amount of waste oxide materials generated
by American industry is enormous.W
Future environmental, space, and economic considerations will necessitate
new handling and treatment methods for these materials. At present, however,
problems associated with the dust materials make recycling to blast furnace or
steelmaking operations quite difficult. Research is being conducted to
characterize the dust materials and to develop processing methods which will
allow greater utilization of these materials in our primary metal production
processes.
The physical and chemical properties of Inland's mill scale, blast furnace
flue dust, and steelmaking dusts have been investigated. The various
properties of these materials are given in this paper. The results of this
preliminary work point to methods which may allow greater utilization of the
materials in iron- and steelmaking operations.
EXPERIMENTAL
Samples of mill scale, blast furnace flue dusts, and steelmaking dusts
were investigated by various techniques. Chemical analyses were made using
standard methods; the analytical results are reported on an elemental basis.
X-ray diffraction analyses were performed using a Debye-Sherrer camera and
chromium radiation. These techniques were supplemented by scanning electron
microscopy, electron microprobe analysis, and optical microscopy.
The bulk densities of the dry-collected flue dusts and mill scale were
determined by weighing the amount of material necessary to fill a 0.028 m^
container; hand tapping was used to simulate settling. "Theoretical"
densities were determined using a water pycnometer.
Particle size distribution analyses were made using various sizing
techniques. Standard sieving techniques were employed for mill scale and
dry-collected blast furnace flue dust. A Cyclosizer Sub-Sieve Analyzer was
used to examine the wet-collected blast furnace flue dust and the -0.044 mm
dry-collected blast furnace flue dust. The steelmaking dusts were
investigated using a Model TAII Coulter Counter.
Relative magnetic susceptibility measurements were made by means of a
Satmagan saturation magnetization analyzer. "Percentage magnetic material"
values were obtained as direct instrumental readings and are equivalent to
that which would be obtained for samples containing various amounts of iron as
magnetite dispersed in a magnetically inert material.
Other testing procedures such as magnetic separation, hydrocycloning, and
froth flotation were also used.
429
-------
RESULTS AND DISCUSSION
Mill Scale
This material accounts for approximately one third of the waste oxide
material generated each year by Inland. It results from surface oxidation of
ingots, slabs, and other steel materials in rolling and finishing operations.
The surface oxide products fracture off the steel materials and are collected
in scale pits. The scale is periodically cleaned from these pits and some is
recycled to iron- and steelmaking operations. Usually the larger sizes
(greater than 6.35 mm) are used directly in blast furnace operations. The
smaller materials are often recycled through sintering operations. Any unused
material is usually discarded into landfills.
The mill scale is a relatively coarse, dense, high iron content (e.g. 72.2
wt. % Fe) waste oxide material which is low in tramp impurities (see Tables
2-6). The material consists of metallic iron, various iron oxides, and gangue
material. The predominant iron oxide phases are wustite and magnetite (see
Table 2). Minor amounts of hematite are also present. The various phases
found in the scale are shown in Figure 1.
The mill scale has a high degree of magnetic character, as shown in Table
3. This arises from the large quantities of metallic iron and magnetite
present in the scale material.
Since the mill scale consists largely of metallic iron and partially
oxidized iron oxides, a significant quantity of heat is released when the
material is processed in sintering operations due to oxidation of these phases
to hematite. The amount of mill scale used in sintering must be closely
controlled to ensure efficient thermal operation of the sinter strand. If
other materials containing significant amounts of fuel substances are also
used in sintering operations (e.g. blast furnace flue dust), the amount of
mill scale that can be used may be limited.
If a baghouse is used to control sinter plant emissions, the oil content
of the mill scale must be reduced to prevent operating difficulties and to
reduce possible air pollution problems. At Inland, the scale is de-oiled in a
rotary kiln. The deoiling process effectively reduces the oil content to
acceptable levels.
Since processing of the scale in sintering operations results in oxidation
of the iron and iron oxide phases to hematite, some inefficiency is inherent
in processing scale in this fashion. It may, therefore, be beneficial to
consider other processing techniques that would maintain the original degree
of oxidation of the scale. For example, if a suitable mill scale agglomerate
could be produced with acceptable chemical and physical properties for direct
use in ironmaking operations, the reduced energy requirement of the partially
reduced iron oxides would result in lower coke rates and increased 'furnace
productivity. Alternatively, if the scale could 'be treated in a direct
reduction kiln, the low residual sponge iron produced would be useful in
either steelmaking operations as a scrap substitute or in ironmaking
operations where the coke rate would be reduced and furnace productivity
increased.(2-4) Furthermore, the scale may be useful as a cooling agent in
steelmaking operations. These possible alternatives need further technical
and economic analyses to determine if they are indeed viable alternatives to
current processing methods.
430
-------
Blast Furnace Flue Dusts
At Inland, these dusts are the second largest contributor to the total
amount of waste oxide material generated each year. There are two types of
dust materials collected in blast furnace operations. Dry-collected dusts are
obtained from cyclone dust collectors. The finer dust that escapes the
cyclone collectors is caught by wet scrubbers. This wet-collected dust
material is further processed in thickeners and filtering devices to reduce
the water content. Currently, some of this material is utilized in sintering
operations, but the bulk of the material is used as landfill.
Dry-Collected Dust. This material is a multi-component physical mixture
of degraded blast furnace burden materials (e.g. coke, pellets, mill scale,
BOF slag, ore, sinter, and limestone) with a variable chemistry. The
complexity of the dust material is well illustrated in Figure 2 which shows
typical examples of the various components. Coke, pellet, and BOF slag
fragments are the predominant components of the dust material. The diversity
of the mixture is reflected by the x-ray diffraction analysis of the dust
material which shows that the significant phases present are hematite,
magnetite, graphite, calcium carbonate, wustite, and silica (see Table 2).
The bulk density of the material is about half that of mill scale (see
Table 5). This is largely due to the large quantity of coke fragments in the
dust material. Quantitative microscopy shows that on a volume basis, coke
fragments are the single greatest component of the mixture.
The chemical composition of the dust material depends on the strength of
the burden materials used in the blast furnace operations. If coke quality is
low, the dust material will be richer in carbon components. Furnace operation
with low strength iron burden materials will result in an increased iron
content for the dust material. The material averages 30.6 wt. % Fe and 31.3
wt. % C (see Table 4), but these values can vary by + 10 wt. %. in daily
operation. The dust material contains undesirable concentrations of alkali,
zinc, and lead impurities (see Table 4). The use of agglomerates produced
from dusts containing high levels of these impurities may lead to blast
furnace operating difficulties.(5-8)
The size distribution of particles in the dust material is broad and
ranges from a top size of approximately 2.4 mm to less than 0.014 mm. Over
86% of the material is greater than 0.149 mm. All the components of the dust
material can be found over the whole particle size range, but some segregation
according to size can be seen (see Table 7). The carbon components
concentrate in the coarser size fractions (over 70% of the total carbon in the
dust is found in particle fractions greater than 0.149 mm). The iron
components are more evenly distributed throughout the particle size range and
only 50% of the iron components are greater than 0.149 mm. Most of the zinc
material contained in the dust is found in the smaller size ranges (over 62%
of the total contained zinc is found in the size fractions less than 0.149 mm).
Since the dust material is a physical mixture of diverse components,
physical beneficiation of the material may be possible. This may allow for
increased usage of the material in iron- and steelmaking operations. At the
current time, the usage of flue dust in sintering operations is limited due to
both the variability in dust composition and the high carbon levels in the
431
-------
dust. If these problems can be overcame through physical beneficiation, more
of the material might be consumed in sintering. Preliminary studies using
various physical separation techniques look promising in this regard.
Wet-Collected Dust. This material is similar to the dry-collected dust.
Similar iron oxide, carbon, and gangue components are found in each (see
Tables 2 and 4). The material contains some magnetic iron oxides, as
indicated by x-ray diffraction analysis and Satmagan testing. The dust is
generally smaller in size consist and contains more carbon than the dry dust
(see Tables 4, 6, and 8). Over 68% of the material is less than 60.5 fim and
22% is less than 17.5 ^m.
The wet-collected dust is concentrated in Dorr thickeners to an average
water content of 58%. This water content is quite variable and may range from
as low as 45% to greater than 70%. Consequently, the material must be
dewatered prior to its use in sintering operations.
This material, like the dry-collected dust, has a variable chemistry. The
material averages 23.8 wt. % Fe and 44.8 wt. % C, but these values can vary by
+ 10 wt. % in daily operation. The high and variable carbon content of this
material makes efficient use of the material in sintering operations quite
difficult. In addition, the zinc content is higher than that of the
dry-collected dust. This factor also limits the amount of wet-collected dust
that can be recycled in sintering operations, because the amount of zinc
entering the blast furnace must be limited to ensure efficient blast furnace
operation, and the zinc content is not reduced significantly in sintering.
Like the dry-collected dust, segregation of some chemical species
according to particle size was found for this dust material. Figure 3 shows
that the zinc material concentrates in the lower particle size range; alkalis
show the opposite trend. The Japanese have found a similar trend for their
wet-collected flue dust.(9) They use this property to lower the amount of
zinc in the dust materials processed for recycling by separating the smaller
size materials from the bulk of the dust using a wet cyclone. They claim to
eliminate between 70-80% of the original zinc in the dust material using this
technique.(9) Some segregation of the carbon components was also found. As
was found for the dry-collected dust, the carbon components concentrated in
the larger particle size range.
Since the wet-collected dust is similar to the dry-collected material,
some beneficiation of the material should be possible using physical
beneficiation techniques. Some preliminary tests show that froth flotation
may be useful in upgrading the dust to produce an iron concentrate more
readily amenable to sinter plant treatment. Other uses for beneficiated dust
materials are also being investigated.
Steelmaking Dusts
Steelmaking dusts account for the lowest proportion of waste oxide
materials generated each year, but they may be the most difficult to recycle
through iron- and Steelmaking operations due to their fine size consistency
and undesirable chemistry. The materials are formed during the melting and
refining stages of steel production. They are evolved as a fine fume which is
collected using electrostatic precipitators, baghouses, and scrubbers. The
collected materials may be dry or wet depending on the collection system used.
432
-------
Dry-Collected Dusts. At Inland, electric furnace dust and open hearth
dust are collected in the dry condition using a baghouse and electrostatic
precipitator, respectively. These dusts are similar in mineralogy, magnetic
properties, and size consistency, but have somewhat different overall
chemistries (see Table 2, 3, 4 and 9). X-ray diffraction analysis shows that
the predominant phases in these materials are magnetite, hematite, and
zincite. The materials are very fine (over 97% of the dust is less than
64 fim) and this causes difficulties in handling and transportation. The open
hearth dust is richer in iron than the electric furnace dust (52.0 wt. % Fe
versus 32.8 wt. % Fe) . The zinc concentrations for these materials depend on
the amount of zinc-bearing scrap charged to the furnace. When high levels are
used, zinc concentrations for the dust may exceed 20 wt. %. Agglomerates
produced from these dusts are unsuitable for use in ironmaking operations
because of their high levels of zinc, lead, and alkali.
In order to increase the potential for recycling these materials, the
amount of zinc contained in these dusts must be greatly reduced. Leaching and
other tests indicate that roughly half the zinc exists as zincite. The
remaining zinc material exists as zinc ferrites, silicates, and aluminates.
The zincite form is easily leachable using acid or basic solutions. The other
zinc forms will require more rigorous processing. Thermal treatment may be
necessary to fully remove the zinc contained in these dust materials.
Wet-Collected Dusts. In both of Inland's EOF operations, venturi
scrubbers are used to collect the fine dust material generated during steel
production. Two different hood systems are used in the EOF operations. The
older shop has the traditional open hood (OH) system which allows some air to
mix with the off -gas. The newer shop has the closed hood off-gas (OG) system
which prevents air from mixing with the off-gas stream. The off-gas from the
OG system has a higher reduction potential than that from the OH system.
Consequently, the iron oxide forms found in the BOF-OG dust are in a more
reduced state (see Table 2) . The predominant iron oxide phase for the BOF-OG
material is wustite; whereas, for the BOF-OH material a mixed ion spinel
(probably a mixture of magnetite and zinc ferrite) is obtained.
The chemistry of the dusts are similar (see Table 4) . Both dusts contain
over 52 wt. % Fe and 3-4 wt. % zinc. The zinc concentration in the dust
material depends on the amount of scrap charged, as mentioned previously. The
BOF materials have lower average zinc concentrations than the electric or open
hearth dusts, because the amount of scrap used in BOF operations is
substantially less and the scrap used is lower in zinc.
The two BOF materials have similar size consistencies (see Table 9) . The
size distributions of these dusts are comparable to those obtained for the
electric furnace and open hearth dusts (over 97% of the material is less than
64
Several problems exist which prevent wide-spread recycling of this
material. First, the zinc level, although greatly lower than that of the
dry-collected dusts, is still too high for direct recycling to ironmaking
processes. Second, the material must be dewatered prior to use in iron- or
steelmaking processes. And, third, the material must be agglomerated (because
of its fine particle size) prior to recycling into primary metal operations.
433
-------
The first problem is the most difficult. The zinc is in a highly leachable
state, in the case of the BOF-OG dust. Preliminary acid leaching tests have
shown that the zinc exists largely as zincite for this material (almost 73% of
the total zinc was soluble in acid solution). However, for the BOF-OH system,
zinc ferrite phases predominate and only 22% of the total zinc was soluble in
acid solution (pH = 1). Thermal processing may be necessary to fully remove
the zinc in these dusts. After zinc removal, acceptable agglomerates can be
made for use in iron- and steelmaking processes.(10~12/
SUMMARY
Various chemical and physical tests were conducted on blast furnace flue
dust, mill scale, and various steelmaking dusts. The mill scale was found to
be a relatively coarse, dense, high iron content waste oxide material which is
low in tramp impurities. The scale largely consists of metallic iron and
partially oxidized iron oxides.
The blast furnace flue dust materials were found to be a multi-component
mixture of degraded blast furnace burden materials. The chemistry of these
dusts is quite variable and this variability and the high carbon content of
these dusts limits the use of the materials in current sintering operations.
Various physical processing techniques are being investigated which should
allow effective beneficiation of these materials into iron and carbon
concentrates. These beneficiated materials may then be better utilized in
iron- and steelmaking operations.
The steel plant dusts were found to consist of very fine particles and to
have undesirable levels of zinc, lead, and alkali. The mineralogy of these
dusts depends on the steelmaking operation employed. Removal of zinc from
these dusts using hydrometallurgical methods will be difficult as significant
amounts of zinc ferrites, silicates, and aluminates have been detected.
Thermal treatment for zinc removal is feasible and will allow increased usage
of these materials in ironmaking operations.
Test work is continuing on various beneficiation techniques which should
allow further recycling of the various waste oxide materials in primary metal
production processes.
REFERENCES
"EAnon, "1978 Steel Industry Financial Analysis," Iron Age, 222, (17), 1979,
p. 36B.
2. Rollinger, B., "Steel via Direct Reduction," I & SM, January, 1975, pp.
10-15.
3. Bleimann, Karl R., and Ahmed, Aziz, "An Assessment of the Value of Direct
Reduced Iron to the Steelmaker," Proceedings of the 3rd International Iron
and Steel Congress, April 16-20, 1978, Chicago, Illinois, pp. 435-438.
4. Maschlanka, Walter, et al, "Utilization of Direct Reduced Iron in Iron and
Steel Production Processes," Proceedings of the 3rd International Iron and
Steel Congress, April 16-20, 1978, Chicago, Illinois, pp. 422-434.
5. Schroth, P., and Robinson, G. C., "The Effects of Alkali Attack on Various
Carbon Refractories," Ironmaking Proceedings, 32, 1973, pp. 60-73.
434
-------
6. El Kasabgy, T. and Lu, W-K., "Conbinative Effect of Gangue and Alkalies on
the Behavior of Iron Ore Pellets During Reduction," Ironmaking
Proceedings, 36, 1977, pp. 2-8.
7. Sasaki, Minoru, and Nakazawa, Takao, "On the Mineral Composition and
Formation of the Blast Furnace Scaffold," Trans. ISIJ, 9, 1969, pp.
413-422.
8. Chow, C. K., and Lu, W. K., "Degradation of Coke in the Blast Furnace Due
to Alkali Vapors," Paper presented at 62nd National Open Hearth and Basic
Oxygen Steelmaking Conference and 38th Ironmaking Conference, March, 1979,
AIME, Detroit, Michigan.
9. Toda, Hideo, et al, "Blast Furnace Wet-Dust Zinc Removing Installation
Using Wet Cyclone and Its Operation," Tetsu-to-Hagane, 64, (8), 1978, pp.
78-A91-78-A94.
10. Kanda, Y., et al, "Production of Pre-Reduced Pellets from Iron and
Steelmaking Dust," Tetsu-to-Hagane, 62, Lectures 9 and 10, S9 and S10.
11. Saito, Y., "Direct Reduction Process for Recycling Steel Plant Waste
Fines," Ironmaking Proceedings, 34, 1975, pp. 464-481.
12. Sugasawa, K., et al, "Direct Reduction of Metallurgical Dusts," Stahl und
Eisen, 96_, (24), 1976, pp. 1239-1245.
DRF/jrb
435
-------
TABLE 1 APPROXIMATE GENERATION RATES OF
INLAND WASTE OXIDE MATERIALS
Material Generation Rate
Blast Furnace Flue Dust 40 kg/tonne HM
Mill Scale 36 kg/tonne RS
EOF Dust 14 kg/tonne RS
Electric Furnace Dust 18 kg/tonne RS
Open Hearth Dust 11 kg/tonne RS
HM = Hot Metal; RS = Raw Steel
436
-------
TABLE 2 X-RAY ANALYSIS OF VARIOUS WASTE OXIDE MATERIALS
Material
Mill Scale
Dry-Collected
Blast Furnace
Flue Dust
Wet-Collected
Blast Furnace
Flue Dust
Electric Furnace
Dust
Open Hearth Dust
BOF-OG Dust**
BOF-OH Dust**
Major Phases
Minor Phases
Fe
C (graphite)
Ca(X>3
C (graphite)
CaC03
ZnO
ZnO
Fel-x°
Fe
(Fe!_yXy)
CaCO3
CaC03
'Mixed Ion Spinel — Some Zinc Ferrite may be present.
** BOF-OG = (BOF operating with Closed Hood Off-Gas System.)
BOH-OH = (BOF Operating with Open Hood System.)
437
-------
TABLE 3 RELATIVE MAGNETIC CONTENT OF VARIOUS
WASTE OXIDE MATERIALS*
Magnetic Pe Content
Material for Sample Tested
Material (wt. %) (wt. %)
Mill Scale 48.0 76.8
Dry-Collected 3.2 23.6
BF Flue Dust**
Wet-Collected 4.6 21.6
BF Flue Dust**
Electric Furnace 14.6 35.8
Dust
Open Hearth Dust 29.9 58.2
BOF-OG Dust*** 3.0 58.6
BOF-OH Dust*** 44.4 52.1
* Magnetic Determinations by Satmagan Analyzer
** BF = Blast Furnace
*** BOF-OG = (BOF Operating with Closed Hood Off-Gas System.) BOF-OH =
(BOF operating with Open Hood System.)
438
-------
TABLE 4 AVERAGE CHEMICAL ANALYSIS OF VARIOUS WASTE OXIDE MATERIALS (Wt. %)
OJ
Species
Fe
Fe (met)
Fe++
Fe+++
C
Si
Al
Ca
Mg
Mn
S
Zn
Pb
Na
K
P
VM
Oil
ND = Not
*
**
Mill BF Flue*
Scale Dust (Dry)
72.2
1.2
48.5
22.1
ND
0.4
0.7
1.4
0.1
0.5
0.05
ND
ND
0.1
ND
0.03
ND
0.5
Determined
BF = Blast
BOF-OG = (B
30.6
ND
12.8
ND
31.3
3.0
0.9
5.1
1.2
0.6
0.5
0.1
0.1
0.1
0.3
0.1
14.9
0.02
Furnace;
OF operal
BF Flue*
Dust (Wet)t OH Dust*
23.8
ND
5.5
ND
44.8
3.0
1.1
3.8
1.1
0.3
0.4
0.4
0.1
0.1
0.2
0.1
ND
ND
52.0
ND
1.7
ND
ND
0.2
0.1
1.0
0.4
0.5
1.4
10.6
1.2
0.5
0.7
0.1
ND
ND
OH = Open Hearth; EF =
bing with Closed
Off-Gas
EF Dust*
32.8
ND
1.1
ND
0.3
1.1
0.3
6.6
1.6
3.5
0.5
10.3
2.1
0.9
1.0
0.1
ND
ND
BOF-OG**
Dust
57.4
ND
36.2
ND
ND
0.9
0.1
4.2
0.7
1.3
0.2
3.2
1.0
0.1
0.1
0.1
ND
ND
BOF-OH**
Dust
52.8
ND
16.5
35.8
ND
0.8
0.2
6.4
1.6
1.1
0.3
4.1
0.4
0.1
0.1
0.1
ND
ND
Electric Furnace
Hood System)
BOF-OH = (BOF operating with Open Hood System)
-------
TABLE 5 "THEORETICAL DENSITY" AND BULK DENSITY MEASUREMENTS FDR
VARIOUS WASTE OXIDE MATERIALS
"Theoretical" Bulk
Material Density Density
(kg/m3) (kg/m3)
Mill Scale 4300 2130
Dry-Collected 2700 980
BF Flue Dust*
Wet-Collected 2100
BF Flue Dust*
OH Dust* 4400 1130
EF Dust* 3600 900
BOF-OH Dust** 3500
BOF-OG Dust** 3000
* BF = Blast Furnace; OH = Open Hearth; EF = Electric Furnace.
** BOF-OG = (BOF operating with Closed Hood Off-Gas System)
BOF-OH = (BOF operating with Open Hood System)
440
-------
TABLE 6 SIZE DISTRIBUTION FOR MILL SCALE AND DRY-COLLECTED
Screen
Aperture
Size
(mm)
2.38
1.68
1.41
1.19
1.00
0.84
0.59
0.42
0.297
0.210
0.149
0.074
0.044
0.036
0.026
0.018
0.014
BLAST FURNACE FLUE DUST
Percentage Less Than Size Indicated
Mill Scale* BF Flue Dust
92.2 96.6
95.0
83.8
92.9
72.0
89.3
67.4 81.6
62.2
31.2 49.8
17.0 30.2
7.8 13.7
6.3 13.5
0.5 6.2
4.3
3.1
2.2
1.3
This size analysis is for material processed through
sintering operations.
BF = Blast Furnace
441
-------
Screen
Aperture
Size
(mm)
+0.59
0.297
0.210
0.149
0.074
0.044
-0.044
ACCORDING TO SIZE FOR DRY-COLI
Chemical Analysis
of Incremental Size
Fractions (wt.%)
Fe C Zn
11.0 63.8 0.05
16.6 50.0 0.08
27.2 33.7 0.09
28.8 31.5 0.10
32.8 28.5 0.11
27.0 21.4 0.22
57.8 10.8 0.35
^BCTKU BLAST FURNACE FLUE DUST
Cumulative Percentage of Total
Species Contained in Particles
Greater than Size Indicated
Fe C Zn
0.5 2.6 0.4
8.0 25.3 8.2
25.0 46.3 19.9
48.3 71.1 37.1
63.4 84.5 47.3
77.1 95.3 70.7
100.0 100.0 100.0
442
-------
TABLE 8 SIZE DISTRIBUTION FOR WET-COLLECTED BLAST FURNACE
FLUE DUST*
Nominal Particle
Diameter ( jim)
60.5
46.7
33.0
22.7
17.5
Weight Percentage Less
Than Size Indicated
68.8
38.9
30.4
24.7
22.0
Cyclesizer Sub-Sieve Analyzer Used for This Analysis.
443
-------
TABLE 9 PARTICLE SIZE DISTRIBUTION FOR STEELMAKING FLUE DUSTS*
Flue Dust Source
Open Hearth Dust
Electric Furnace Dust
BOF-OG Dust**
BOF-OH Dust**
Weight Percent of Total Dust
3 fim
100.0
77.8
80.3
82.6
4 ptm
78.8
66.3
68.7
70.1
5 tim
65.0
59.1
60.0
60.5
6 nm
54.0
53.1
52.1
53.5
Greater Than Particle
8 urn
44.4 '
47.8
45.3
48.4
16 urn
21.0
34.5
28.8
33.2
Diameter
32 urn
9.9
19.8
13.4
16.5
64 fxm
1.6
2.2
2.5
2.9
* Coulter Counter Model T^u Used for this Analysis.
** BOF-OG = (EOF Operating with Closed Hood Off-Gas System)
BOF-OH = (BOF Operating with Open Hood System)
-------
200
FIGURE 1 SINTER PLANT MILL SCALE-100X
445
-------
FIGURE 2 DRY-COLLECTED BLAST FURNACE
FLUE DUST-100X
446
-------
100
»
•
\
•
\
ALKALI (K + Na)
ZINC
I
1
I
1
15
25 35 45 55
PARTICLE SIZE (/Ltm)
65
FIGURE 3 CUMULATIVE PERCENTAGE OF TOTAL SPECIES
CONTAINED IN PARTICLES GREATER THAN SIZE
INDICATED FOR WET-COLLECTED BLAST FURNACE
FLUE DUST
-------
INTERNATIONAL MINERAL RECOVERY, LTD., DEZINCING PROCESS
by
John E. Allen
President
International Mineral Recovery, Ltd.
ABSTRACT
The "basic intent of the Process is primarily for process-
ing steelmaking dusts containing low amounts of zinc and lead
in combination with higher iron values.
Because of well founded problems in recycling the steel-
making dusts through normal steel plant dust agglomeration
facilities and the blast furnace a new method has been developed.
The Process follows the lines of a previously developed
process for the binding together of ferrogenaceous and carbon-
aceous substances with a heavy hydrocarbon binder. Further
treating of the agglomerates produces a strong product for
further processing in high temperature facilities to effect
the separation. The Process is covered by U. S. Patent No.
3,850,613.
Specific handling steps begin with conventional receipt
of materials in bins and proportioning by adequate feeding
arrangements. All materials are well blended and dried in
conventional equipment.
Binder addition and briqueting are done at +200 °F using
a conventional roll type briquet machine. Dehydrogenation of
the binder is accomplished in a stream of hot air.
Final processing is done in conventional iron making
cupola or electric arc furnace with the amount of carbon in
the raw dusts being the determining factor.
Conventional high efficiency dust collectors capture the
zinc and lead oxides, while the molten iron is pigged in con-
ventional pig machines. And, slag is air cooled for shipment
to the usual processors.
448
-------
INTERNATIONAL MINERAL RECOVERY, LTD., DEZINCING PROCESS
By John E. Allen, President
The primary intent of this Process is for the removal of
zinc from steelmaking dusts. And, it applies to other zinc and
iron bearing substances . Lead to a small percentage is also
associated with zinc as both are found in scrap which is used
as a portion of every steelmaking process. The analysis of
the major tonnages of steelmaking dusts is as follows i Fe 35 -
6<#j Zn 1.2 - 20#; Pb 0.2 - Z%\ S 0.03 - 1.7#j Si02 1 - 2.57&J
and, CaO & MgO 2 -
The first characteristic limiting recycle of these steel-
making dusts through normal channels is the zinc content, which
causes disintegration of the blast furnace refractory linings.
And, sintering plants, the normal agglomeration facility used
in steel plants, experiences operating difficulties with these
primarily submicron sized particles.
The problems presented by the steelmaking dusts inside the
steel plants and the surrounding communities became urgent with
the advent of oxygen steelmaking. With the old basic open hearth
process with its white to yellowish emissions were considered
just a nuisance and the emissions were carried away by the air
currents and spread over a very wide area. Oxygen steelmaking,
however, brought about the dense red clouds of dust in spectac-
ular amounts over a short period of time. Thus they required
installation of gas cleaning equipment. The dusts containing
the oxides of heavy metals had to be disposed of with other
forms of "in-plant" debris. While the generation of unsightly
steel plant waste materials has been common practice for many
years, it wasn't until these new steelmaking dusts were added
that another potential problem was detected. An investigation
by the United States Enviromental Protection Agency showed that
the zinc and lead contents of these dusts could be leeched out
into the natural waterways under specific conditions. These
findings were published in February, 1977 under EPA 600-2-77-044.
The extraction of zinc and lead is by no means new. Lee-
ching is a well known technique and numerous patents have been
granted employing this technology. The leeching technology
when applied to steelmaking dusts, costwise, seems to 'break-
even' around a zinc content of 12 - 15# • However, only a very
minor tonnage of steelmaking dust ever reaches zinc contents
this high. The bulk of the tonnage is in a zinc content range
of 3 - 7JC, with no guarantee of any degree of consistancy, and
of course iron is the major metallic substance.
The thermal extraction of zinc from its ores has been
practiced for centuries by blending in an appropriate amount
of carbon and heating same in an enclosed vessel. One early
process of 'our time' has been the horizontal retort which col-
lected metallic zinc in a condenser, while any iron remained in
449
-------
the retort as a residue along with excess carbon, small per-
centages of lead and zinc, and gangue materials, all of which
was wasted in a dump. These dump areas still dot communities
where these plants once operated. The process was discontinued
because emissions from these plants killed plant life in the
communities where these plant operated so that large tracts of
land were devastated. Utalizing the same technology, the ver-
tical retort was devised, but without the devastation problem.
Hence, at least two thermal methods are well proven for the
extraction of zinc.
Our analysis, based upon the study of various processes
and their economics, led us to believe that all components must
be salable products, in order that the Process be economically
viable. Zinc being the smaller percentage must be considered
a coproduct. Iron, in order to bring the maximum return, must
be in a form capable of bringing the highest price. Steel, we
analysed required a major investment for a very minor tonnage.
Pig iron on the other hand is a top dollar product which can be
produced easily. And, a third product, slag, is salable as an
aggregate. Hence, we could see that our Process was one pro-
ducing products all of which were salable.
Work done on earlier dates for the agglomeration of waste
fines from steel plants had indicated that strong agglomerates
could be made which retained all the metallic and carbon values,
while providing good strengths at high temperatures. The process
utalizes a heavy hydrocarbon binder and the processing of same
in hot air. Test agglomerates were made of selected blends of
zinc and iron dusts plus coke dust and heated same in a muffle
furnace. Zinc vapors came off in a very short time and it was
evident that we had a process which would work. Further work
lead to the filing for a patent in late 1973» which was granted
on November 24, 1974, as U. S. Patent No. 3,850,613.
I believe that at this point it might be well to stop and
look at the modern day trends in a changing society.
It seems apparent that the accumulation of solid wastes are
becoming an ever increasing cost problem. As communities grow
the land available in close proximity to the industrial sector
for dumping waste materials is disappearing, hence, disposal
costs are increasing. Furthermore, the toxic materials which
have been highlighted in numerous news releases causes adverse
public reactions to any dump situation so that permits for new
dump sites are increasingly difficult to obtain. There is no
doubt that dumps over the years have been unsightly. Pressure
seems to be toward the cleanup of many such dumps and certainly
to gain insurance that any future dumps be 'beautified1 will
simply mean that extra costs can be expected.
Another factor which is just beginning to reach the 'head-
lines' is the growing scarcity of metals in the United States.
An article appearing in the July 2, 1979 issue of BUSINESS WEEK
45'0
-------
states that this shortage will have an even greater impact on
costs than the Energy Crisis. Recycle of metallics in the steel-
making dusts will help.
The combination of the increasing dump costs and the
scarcity of metals bring out a new need for reacessing our wast-
age practices. It is our belief that into process costs must
be taken the full concept of processing all segments. The old
concept of processing that protion which represented the major
profit and dumping the rest on the fback ^0' is coming to a
close. And, since it is a well established fact that "matter
is never destroyed, but only changed", we have sought by this
Process to bring about change without increasing existing costs.
We believe that the Steel Industry have and will continue
to seek out corporations who have special skills in handling
waste substances. There operates today numerous corporations
which have built up reputations of service in handling large
amounts of waste substances such as slag and scrap metallics.
We believe that more sophisticated processes will be required as
work toward total processing becomes a 'must1. It is into this
category that our Dezincing Process falls.
We fit where dusts containing zinc and lead are inter-
mingled with iron bearing substances and where mechanical
separation is impossible. With the original intent based upon
the solution of the current production of steelmaking dusts,
we have found that it may be profitable and necessary to also
consider dusts previously segregated and stocked or dumped
with refuse, plus substances from other industries that contain
zinc, lead, iron and carbon values.
Each situation is generally sufficiently different that the
economics require individual evaluation in order to determing
the profitability.
The Process integrated into a complete plant consists of the
followingi
Materials received consist of dusts of various size and
moisture content. The method of receipt may be either by rail,
truck, barge or belt. Steelmaking dusts maybe bone dry as
received from precipitators, bag filters or cyclones. Or, as
a filter cake or sludge from vacuum filters, centrifuges, or
settling ponds. Hence, the moisture may vary from 0 to 50#.
In actual practice there will generally be a variety of dusts
with as many varied moisture contents. Also, of importance is
the particle size of the various dusts. As a result we have
found it to be most desirable to proportion these individual
dusts with special feeders so that a controlled blend is achieved.
It is further most important that the blend be thoroughly mixed
so that it is homogeneous.
The homogeneous blend of .proportioned dusts must be dried
first to remove the moisture and secondly to preheat these
451
-------
materials to 200 - 225 °F so that the binder may be mixed into
the dry dusts. After drying the materials can be easily screened.
The plus 1/8" materials should be crushed. The minus 1/8"
materials are now stored as a process surge point.
Binder should have a ring and ball softening pint below
212 °F. The binder may be heated to as much as 400 °F in order
to secure the best mixing. Generally the dry dust mixed with
the binder still appears dry and only slightly darker in color.
Should excess binder be added such is uneconomical hence good
control is needed. Moisture content is a further serious factor
because of drying costs.
While we have indicated binder and moisture content, carbon
content is indeed a major cost factor. Since this is a reducing
process, carbon to reduce the zinc, lead and iron oxides is
necessary as a first consideration. Secondly, this is a smelting
process, hence heat must be provided for melting iron and slag
for a quality pig iron and a slag capable of containing the sulfur
Hence, we are looking for an iron temperature of 2650 - 2750 °F
in the melting unit. Because of this heat requirement the carbon
content must first be the minimum to remove the oxygen from the
metallic oxides. And, the heat required for smelting may be
furnished by combustion of carbon contained in the brieuet, or
added by combustion of fuel added along with the briquets, or
electrical energy. Since carbon fines is mixed with many iron-
zinc bearing dusts such is the least expensive method of obtain-
ing the heat required. Should additional heat be required such
can be in the form of coke dust added with the dust mix.
Briqueting is by a standard roll type briquet machine.
The shape of the briquet is of importance as +40$ voids is
required. We have found that a l£ X l£ X 1 inch briquet is a
good compromise in size vs cost consideration.
"Green" briquets are weak, hence a minimum of handling is
essential. Again the presence of fines reduces the voids,
hence, screening must be done. Again extreme care is required
in charging the briquets on to the hot air processing conveyor
line. Once placed on the conveyor line for processing with hot
air no movement of the briquets takes place until the binder
has been dehydrogenated.
This denydrogenation process consists of converting the
hydrocarbon binder to char by oxidizing off the hydrogen by the
stream of hot air. The process generates heat so that the
stream of air passing through the briquets both heats the
briquets to the temperature that the reaction takes place and
cools the briquets to a point that char does not ignite. Hence,
the importance of a good void pattern and temperature control
is a key factor in the process.
Cured briquets possess good strength, 600 - 800 Ibs crush-
ing strength. Thus, they may be screened thoroughly at £" and
all fines recycled back through the process. The +1" sized
452
-------
cured "briquets are stored in bins for further processing at this
logical surge point in the total process.
At this stage in the process it seems appropriate to discuss
the pollutants from the process that may occur and must be
handled.
In the transporting of the dusts 'drip1 of the wet dusts
from cars and trucks must be given careful attention. Bone dry
dusts must be handled in enclosed containers. And, as all of us
know who have handled these substances over the years, house-
keeping in the unloading area is difficult and a constant chore.
At the processing plant, for those bins receiving the bone
dry dusts adequate suct5.on around the dumping area is necessary
to capture such dusts that become airborne and collection of
same in a bag filter. As previously discussed the proportioning
and mixing of all dusts provides us with a material that is not
difficult to handle. Drying produces a moisture vapor cloud
which exits through the main stack, but no other substances are
carried out when using a dryer consisting of the hollow screw
design with high temperature oil circulating in the screws, as
there is not sufficient agitation to have the particles become
airborne. Each transfer point in the belt conveyor system must
be hooded and connected to a bag filter. The same coverage is
required for the dry dust screen, oversize crusher, and surge bin,
as well as the heat treated briquet sizing screen.
The hot air processing of the 'green1 briquets with its
high volume of hot air generates some hydrocarbon smoke from
the light distillates which exist in the binder and lubricating
oils present in some dusts. The exhaust gases are used as
combustion air for the hot oil heater, with any excess consumed
by a thermal incinerator. The products of combustion from the
hot oil heater exit with the moisture cloud from the dryer to
improve the dew point.
Briquets are drawn from several bins by a program to provide
a uniform analysis for the dezincing and melting unit. As
previously discussed the type of melting unit depends upon the
amount of carbon available from the various dusts, and determined
by a study as to the best economic picture. In a high percentage
of the studys made a cupola is the most practical melting unit.
A cupola melting unit consists of« skip charging, gas
cleaning, hot blast, and runout systems. In the cupola the
zinc comes off in the top region as zinc oxide which is swept
out with the gas stream. This zinc rich dust is partially
collected by a hot cyclone, with the balance collected by a
final high efficiency collector. This zinc oxide rich dust
is one of the salable products. A typical analysis is as
followsi Zinc - 63.3^1 Lead - 6.9#l Iron - 1.2£i Acid solubles -
3.6^i Tellurium - 0.28#j Tin - 0.05#j Cadmium - 0.1#? and,
other substances 0.01J6 or less. However, it must be remembered
453
-------
that the analysis of each plant differs and though our infor-
mation does not have a long history may vary over a long period
as the materials charged into the steelmaking furnaces differs.
From the tap hole and into the skimming box flows the molten
iron and slag where the slag is skimmed off and flows into a
pit for cooling. The molten iron flows into a fore-hearth
where any needed ferroalloys are added to adjust for silicon,
manganese or phosphorous requirements. From the fore-hearth
the molten iron flows into the pouring box of the pig machine.
Pig iron is the largest volume product. Pig iron is handled
by magnets for shipment or stocking for future shipment. The
cooled slag from the pit is dug and loaded into trucks for
shipment to aggregate merchandisers for sale.
This melting process of course produces some pollution
which must be taken care of. The first being the top gases
as has been covered in the above discussion. Where water
cleaning is employed in the high efficiency scrubber the water
from same will be processed through a settling pond and recycled.
And, the smoke and graphite from the slag and iron runout will
be hooded and the particles captured in a bag filter.
While it may not be essential it is our plan that at least
for the first installation the whole operation will be contained
inside a building so that the entire building may be evacuated
through a bag filter system in order that if any of the air
cleaning units fail or experience surges beyond the design
capacity the individual systems will be 'back stopped* by the
building evacuation system.
In conclusion we feel we have by the Process and the plant
design strived to accomplish the beliefs stated in the beginning,
namely that a waste substance be converted to all salable productts.
454
-------
Closing Remarks - Norman Plaks
Chief, Metallurgical Processes Branch
IERL-RTP-EPA
During the past two and one-half days twenty-six speakers have discussed
a wide range of pollution abatement technology topics relating to the
iron and steel industry. We have heard discussions of advances in
technology that have been made, of problems that are remaining and of
new problems emerging.
We were told by the AISI that almost every new, major environmental
control installation made by the industry is actually a prototype and
becomes, in fact, a full scale development program. Several of the
papers presented here tend to support that theory.
A presentation was made by EPA about it's role and attempts to operate a
R&D program benefiting both the Agency and industry in the face of the
industries' high capital intensiveness and the relatively low funding
levels available. Joint-funding arrangements involving both EPA and the
industry were present in a number of projects.
Some of the speakers recounted the environmental conditions of the
industry in the past and the advances that have been made. Credit must
be given to the industry for these advances.
455
-------
However, it is apparent that we still have a long way to go, especially
with respect to hazardous pollutants.
We have heard both here and elsewhere that economic conditions within
the industry make it difficult for the industry to undertake research
and development on the broad front that will be necessary to solve these
problems.
Engineering and consulting firms and equipment suppliers have played a
major role in the development of the pollution abatement technology.
However they can not be expected to take on the burden by themselves.
One area that has not been discussed to any significant degree at this
symposium is the role of innovative process technology. What I am
talking about is new production process technology which minimizes the
discharge of pollutants while simultaneously increasing the efficiency
and lowering the cost of the iron and steelmaking process.
In summary there have been major and significant advances made in abating
pollution from the iron and steel industry. There are still both remaining
and newly emerging problems requiring solutions. Neither the industry,
equipment supplies, engineering and consulting firms, or the Government
can individually fund and solve these remaining problems especially if
innovative technology is considered.
456
-------
As a closing thought I would like to issue a challenge to the industry,
to the equipment suppliers and consultants, and to the Government to
continue discussing, planning, and then in unison implementing programs
that will make it possible for the iron and steel industry to meet the
environmental control needs in the most cost-effective manner.
457
-------
UNPRESENTED PAPERS
Air Pollution Emissions Characterization of a Coal Preheater
Anthony J. Buonicore,
B. Drummond,
Carl Rechsteiner, and
Julie Rudolf
458
-------
AIR POLLUTION EMISSIONS
CHARACTERIZATION OF A COAL PREHEATER
by
A.J. Buonicore, P.E,
B. Drummond
YORK RESEARCH CORPORATION
One Research Drive
Stamford, CT 06906
and
Carl Rechsteiner
Julie Rudolph
Arthur D. Little, Inc.
Cambridge, Massachusetts 02140
ABSTRACT
The body of information presented in this paper is directed
to those individuals concerned with current research efforts into
quantifying gaseous and particulate emissions from coal preheat
systems at coke plants.
Emission test results on a Coaltek Pipeline Charging System
using a Cerchar coal preheater at a steel plant in Pennsylvania
are presented. The primary objective of this EPA-sponsored research
project was to investigate the inlet and outlet of the system's
venturi scrubber for particulate and polycyclic aromatic hydrocarbon
emissions.
459
-------
Non-chloroform soluble particulate emission rates (front and
back half) at the venturi scrubber outlet ranged between 0.256 -
0.281 pounds per ton of coal feed. Chloroform soluble particulate
emission rates at the venturi scrubber outlet (front half, back
half, silica gel) ranged between 0.215 - 0.549 pounds per ton of
coal feed.
Gas Chromatography/Mass Spectrometry analysis indicated that
scrubber removal efficiency of selected POM organic species was
highly variable, ranging from 41.2% - 79.1% depending upon process
operating conditions. On comparing coal preheater POM emission
levels to Discharge Multimedia Environmental Goal (DMEG) values,
the DMEG levels were exceeded for phenanthrene, benz(a) anthra-
cene, benzo(a) pyrene, 7,12-dimethyl benz(a)anthracene , and 3-
methyl cholanthrene.
EPA Level I organic analysis indicated that aliphatic hydro-
carbons, fused aromatic hydrocarbons, phenols and esters were
the major components in the stack air samples and aliphatic hydro-
carbons, carbazoles and phenols in the water samples.
SUMMARY
An EPA-sponsored research project to characterize particulate
and polycyclic organic material (POM) emissions in the effluent
from a coal preheat system on a coke battery was conducted in
August, 1978 at a steel plant in Pennsylvania. Effluents were
sampled upstream and downstream of a venturi scrubber during various
preheat system operational modes, including varying coal feed rate
and preheater outlet temperature. Evaluation of chloroform soluble
and non-chloroform soluble particulate emission rates was performed
using a conventional EPA Method 5 train. A POM train was used for
POM collection. The POM sampling train was essentially a Method 5
train with an adsorbent sampler located downstream of the filter
and condenser and upstream of the impingers. Analysis was performed
using Gas Chromatography/Mass Spectrometry.
Non-chloroform soluble particulate emission rates (front and
back half) at the venturi scrubber outlet ranged between 0.256 -
0.281 pounds per ton of coal feed. Although the data base was
not sufficiently large to statistically validate conclusions a
number of trends were evident. Increasing the coal feed rate appeared
to increase the wash and filter catches. Increasing the pre-heater
outlet temperature appeared to increase the impinger catch. Overall
scrubber efficiency (at 20 in. W.C. pressure drop and an L/G of 8)
was approximately 90%, with relatively high removal efficiency in
the front half of the train and relatively low removal efficiency
in the back half of the train.
460
-------
Analysis was also made for chloroform soluble particulate.
Emission rates at the venturi scrubber outlet ranged between
0.215 - 0.549 pounds per ton of coal feed. Total particulate
emissions (including both chloroform soluble and non-chloroform
soluble particulate) from analysis of the complete sample train
were 0.471 - 0.813 pounds per ton of coal feed. Front half train
total particulate emissions (EPA Methodology) ranged from 0.191 -
0.280 pounds per ton of coal feed. Front half and back half
(less silica gel) train total particulate emissions (DER Methodo-
logy) ranged from 0.425 - 0.752 pounds per ton of coal feed.
The GC/MS analysis showed that the scrubber removal efficiency
on selected POM organic species was highly variable, ranging
from 41.2% - 79.1% depending upon process operating conditions.
In general, POM emissions increased with increasing coal feed
rate. Among the compounds analyzed, the most prevalent species
found were naphthalene, anthracene and phenanthrene. Lesser
amounts of fluorene, pyrene, fluoranthene, benzanthracene,
chrysene and benzopyrenes were also found. Discharge Multimedia
Environmental Goal (DMEG) levels were exceeded for phenanthrene,
benz(a)anthracene, benzo(a)pyrene , 7,12-dimethyl benz(a)anthra-
cene, and 3-methyl cholanthrene. None of the species found in
the scrubber inlet/outlet water samples exceeded their DMEG values.
The GC/MS analyses showed that, in general, the levels of the
selected organic species were higher in the inlet stack samples.
The selected POMs and their isomers represented about 2 to
7% (by weight) of the total organics collected in the sample.
EPA Level I organic analysis indicated that aliphatic hydro-
carbons, fused aromatic hydrocarbons, phenols and esters were
the major components in the stack air samples and aliphatic
hydrocarbons, carbazoles and phenols in the water samples.
INTRODUCTION
The concept of utilizing preheated coal is currently an
attractive solution to improve throughput quality and economy
in the coke-making operation. One such scheme is the Coaltek
pipeline charging system with Cerchar preheater (refer to Figure 1).
In this particular system, wet coal is withdrawn by screw conveyor
from existing coal storage lines, sized and fed to the preheater
by a variable speed screw. The wet coal is fed into a flash
drying entrainment section where it comes into contact with a
stream of hot oxygen-free gas. The gas carries the partly dried,
entrained coal up through the preheater. In the combustion
chamber, low-sulfur coke oven gas is burned stoichiometrically
with air for complete combustion. The hot gases leaving the
combustion chamber are a mixture of freshly burned gas and recycled
products of combustion. The mixed gases leave the combustion
chamber with a temperature range from 725°F to 1200°F, and pass
461
-------
through the venturi section of the preheater. The wet coal
is pushed into this high volume gas stream by the feed screw.
Temperature at the upper end of the flash-drying section (pre-
heater outlet) is in the 500°F range. All preheated coal goes
overhead and is recovered in primary and secondary conventional
cyclone separators. Hot coal from the bottom of the cyclones is
distributed to charge bins and fed by a specified charging sequence
through pipelines to the ovens. Gas from the outlet of the
secondary cyclones is split into two streams. Gas volumes equiva-
lent to the combustion gases and moisture driven off the coal go
to a venturi scrubber for cleaning before being exhausted to the
atmosphere. The remaining gas volume is boosted in pressure by
means of a recycle blower and then returned to the combustion
chamber where it is used to temper and add to the flow of combus-
tion gases passing up through the preheater. Automatic controls
adjust the pressure differential across the secondary cyclones
to maintain the desired flow of gases through the preheater.
PROGRAM OBJECTIVE
In an effort to better understand the nature of the emissions
from the coal preheat system, the Environmental Protection Agency
(EPA) sponsored a research program on the Coaltek pipeline charging/
preheat system at a steel plant in Pennsylvania. The test program
was formulated by York Research Corporation (YRC) and directed
toward inlet/outlet emission characterization around the venturi
scrubber. Particular emphasis was to be placed upon the specific
polycyclic organic materials (POMs) listed in Table 1.
SAMPLING METHODOLOGY
The test program was conducted during the months of July and
August, 1978. Velocity, temperature, gas analysis and other para-
meters required for particulate and polycyclic organic materials
(POM) sampling were recorded with isokinetic sampling maintained
as specified by the following E.P.A. Methods:
• Method #1 - Sampling and velocity traverses
for stationary sources
• Method #2 - Determination of stack gas velocity
and volumetric flow rate (type S pitot
tube)
• Method #3 - Gas analysis for CO?, O2 and dry
molecular weight using an Orsat unit
• Method #4 - Determination of moisture content
in stack gas, derived from actual
sampling train
462
-------
• Method #5 - Determination of particulate in the
inlet and outlet to the coal pre-
heater scrubber.
In addition to the above methods, extensive sampling was
performed for POM determination using a POM sampling train with
an Arthur D. Little XAD-2 adsorber.
Particulate Train
The particulate sampling apparatus consisted of a probe,
cyclone bypass, filter, four impingers, dry gas meter, vacuum
pump and flow meter (see Figure 2). The probe was 5 feet in
length and glass lined. The stainless steel button-hook type
probe tip was connected by a stainless steel coupling with
Teflon packing to the probe. The probe consisted of a 5/8 inch
outside diameter tube with a ground balljoint on one end. The
probe was logarithmically wound from the entrance end with 26-
gauge nickel-chromium wire. During sampling, the wire was con-
nected to a variable transformer to maintain a gas temperature
of 250°F in the probe. The wire wound tube was wrapped with
fiberglass tape and encased in a 1-inch-OD stainless steel casing
for protection. The nozzle was attached to the end of the probe
casing. The probe was connected to a cyclone bypass and a very
coarse fritted glass filter holder containing a tared glass fiber
filter. The filter was contained in an electrically heated en-
closed box thermostatically maintained at a temperature of 250°F
to prevent condensation. Attached to the heated box was an ice
bath containing four Greenburg-Smith design impingers connected
in series with glass balljoints. The first impinger was modi-
fied by replacing the tip with a h inch ID glass tube extending
to 0.5 inches from the bottom of the flask, and, with the second
impinger, filled with 100 milliliters of distilled water. The
third impinger was left dry. The fourth was modified as the first
and charged with silica gel.
The effluent stream from the fourth impinger flowed through
a check valve, flexible rubber vacuum tubing, vacuum gauge, a
needle valve, a leakless vacuum pump (rated at 4 cubic feet per
minute at 0 inches of mercury gauge pressure and 0 cubic feet per
minute at 26 inches of mercury gauge pressure) connected in
parallel with a bypass valve, and a dry gas meter rated at 0.1
cubic foot per revolution. A calibrated orifice completed the
train and was used to measure instantaneous flow rates. The dual
manometer across the calibrated orifice was an inclined-vertical
type graduated in hundreths of an inch of water from 0 to 0.1
inch and in tenths from 1 to 10 inches.
POM Sampling Train
The POM sampling train, as shown in Figure 3, consisted of
463
-------
a Method 5 train with an adsorbent sampler located downstream
of the filter and condenser and upstream of the impingers. Utili-
zing this arrangement, POM emissions could be determined by analysis
of the probe wash, filter catch, condensate, and adsorbent sampler
catch. The impingers were only used to cool and dry the stack
gases before they entered the dry gas meter.
The POM train used in the field consisted of a heated glass-
lined probe with a stainless steel nozzle at the probe head, a
heated filter assembly, one Greenburg-Smith type condenser impinger,
the adsorbent sampler and four additional impingers.
In order to insure adequate POM collection efficiency in the
adsorbent sampler, the flue gas temperature had to be kept as low
as possible without condensing large quantities of water vapor.
For this reason, a condenser (Greenburg-Smith impinger) was used
between the filter and the adsorbent sampler. Connected to the
impinger assembly was an umbilical, vacuum pump, dry gas meter
and an orifice. All connections in the filter, adsorber and impin-
ger assemblies were glass. Thermal control at the probe and the
filter assembly was maintained in a heated mode at 325°F (versus
250°F when using the EPA Method 5 sampling train). Maintaining
the probe and filter at the higher temperature prevented conden-
sation and/or adsorption of 803 and POM (followed by the destruc-
tive reaction of 303 with POM) in these components. Thermocouple
connections at the probe head, the inlet to the filter assembly,
the inlet to the adsorber, the fourth impinger outlet and the
inlet/outlet of the dry gas meter, allowed for monitoring sampled
flue gas temperatures throughout the sampling train.
Once completely assembled, the sampling train was leak-
checked to insure collection of a representative flue gas sample.
To perform the leak-check, the vacuum pump was started and the
nozzle orifice covered to insure an air-tight seal. After bring-
ing the vacuum pressure up to 15 psi, the dry gas meter was checked
for any air leaks. Once the required leak check was performed,
the probe was inserted into the duct at the specific sampling
point. Velocity and temperature measurements of the flue gas
at the pitot head were recorded and a sampling rate determined
for isokinetic sampling.
SAMPLE RECOVERY
Recovery procedures for particulate analysis were essentially
those published in the Federal Register (Vol. 36, No. 247, December
23, 1971) for the front half (acetone wash), the filter and the
back half (water).
Sample recovery from the POM sampling train for POM analysis
involved washing of four separate portions of the sampling train:
464
-------
1) probe and glassware up to the filter,
2) filter,
3) condensate,
4) adsorbent sampler.
The extract from each component was sealed and kept in darkness
while awaiting analysis.
ANALYTICAL METHODOLOGY
Particulate emission results were analyzed and characterized
as either chloroform soluble or non-chloroform soluble using the
methodology outlined in Table (2).
Two sets of organic analyses were performed on the samples
obtained from eleven tests utilizing the POM sampling train.
•• Gas Chromatography/Mass Spectrometry (GC/MS) Analysis
All stack samples and six water samples (Tests 1, 2
and 3, both inlet and outlet) were carried through
the procedures in Figure 4. Total particulate load-
ing (probe methylene chloride wash, front half acetone
wash, back half filter and condenser acetone wash) was
determined and analyses carried out for twenty-five
organic species (POM).
• EPA Level 1 Organic Analysis
Six samples (inlet and outlet air samples from Test
No. 2, inlet and outlet water samples from Test No. 2,
stack blank, and water blank) were carried through the
EPA level 1 organic analysis illustrated in Figure 5.
The individual procedures used in the analyses are described
briefly as follows:
A. Particulate Weights
The solids from the front half methylene chloride
probe wash, acetone wash, and filters were dried
in a desiccator at room temperature to constant
weight.
B. Soxhlet Extraction
All Soxhlet extractions were carried out for a 24
hour period using high purity methylene chloride
(Burdick and Jackson, distilled-in-glass). The
XAD-2 resin samples were extracted with about 500 mL
of methylene chloride and particulate samples were
extracted with about 300 mL of methylene chloride.
465
-------
C. Liquid Extraction
Water samples were extracted with methylene chloride
in separatory funnels fitted with Teflon stopcocks.
The pH of the aqueous sample was adjusted first to
2.0 with hydrochloric acid and subsequently to 12.0
with sodium hydroxide. Two extractions were done
at each pH, using a sample/methylene chloride volume
ratio of about 20/1.
D. Total Chromatographable Qrganics Analysis (TCP)
The quantity of the total organic material with boiling
points in the range of 100-300°C was determined by gas
chromatography using a flame ionization detector. The
concentration of each sample was calculated from the
ratio of the peak areas of the sample to that of the
known standards.
E. Gravimetric Analysis (GRAV)
The amounts of organic material with boiling points
higher than 300°C were determined by the gravimetric
analysis method (GRAV); one or five mL samples were
pipetted into precleaned, dried and weighed aluminum
dishes, and were dried at room temperature in a desic-
cator to constant weight.
F. Liquid Chromatography (LC)
The methylene chloride extracts from each test were
combined and concentrated to 10-25 mL using a Kuderna
Danish apparatus. From each concentrated extract 0.5
- 8 mL aliquots were subjected to three consecutive
solvent exchanges with cyclopentane. The resultant
cyclopentane solutions were chromatographed on a silica
gel column, collecting seven fractions by elution with
solvent mixtures (pentante-methylene chloride-methanol)
of increasing polarity. A portion of fractions 2, 3
and 4 were combined for GC/MS analysis.
G. Infrared Spectroscopy (IR)
The IR spectra of all samples as potassium bromide
micro-pellets were obtained on a Perkin-Elmer 521
grating spectrometer.
H. Low Resolution Mass Spectroscopy
LRMS analysis was carried out on a Dupont 21-110B
spectrometer. Sample sizes varied from 20 yL to
50 yL. Typically, a sample was run at 15 ev and
70 ev ionization potentials over a temperature range
of 70-350°C.
466
-------
I- Gas Chromatography/Mass Spectrometry (GC/MS) Analysis
The combined aromatic fractions from each extract were
analyzed for 25 POM species by the GC/MS technique.
An OV-17 glass capillary GC column and a Finnigan
Model 4023 Mass Spectrometer/Data System were used
throughout the entire analysis.
The detection limits for the coal preheater stack air
samples ranged from 0.8 yg/m to 18 yg/m^ for species
having a molecular weight (MW) less than 252 and from
6 yg/nH to 120 yg/m^ for species having a MW greater
than 252, depending on the volume of concentrated ex-
tract and the volume of air sampled. The detection
limits for the water samples were: 1 yg/L for MW
<252 species and 8 yg/L for MW >252 species.
TEST RESULTS
Particulate emission results are summarized in Table 3 with
the process conditions indicated in Table 4. On the basis of
one test, scrubber inlet total particulate (chloroform and non-
chloroform soluble particulate in front and back half catches,
not including silica gel) concentration was 3.94 gr/DSCF (3.44
Ib/ton coal). Outlet conentrations ranged from 0.416 gr/DSCF -
0.715 gr/DSCF (0.424 - 0.752 Ib/ton coal) with the scrubber
operating at 20 in. W.C. pressure drop and a liquid-to-gas ratio
of approximately 8 gallons per 1000 ACFM.
Selected POM emission results with the specific scrubber
water flow arrangement at the plant are summarized in Tables 5
and 6 with the process conditions indicated in Table 7.
Scrubber removal efficiency for selected POM species varied
over a wide range (see Table 8) and was not particularly selective
for the high molecular weight species. Table 9 indicates that,
in general, POM emissions increase with increasing coal feed rate.
Over the preheater outlet temperature range existing during the
test program (500-550°F) no correlation between POM emission rate
and preheater outlet temperature was identified.
Table 10 summarizes POM emission factor variability for
selected individual species in light of the appropriate NAS carci-
nogenicity level. Tables 11, 12 and 13 indicate scrubber inlet/
outlet particulate loadings from the POM train, scrubber inlet
particulate loading by train component and scrubber outlet parti-
culate loading by train component, respectively. Scrubber liquor
inlet/outlet concentrations from selected POM species are presented
in Table 14.
Among the compounds analyzed, the most abundant species
found in all samples were naphthalene, anthracene, and phenanthrene
Lesser amounts of fluorene, pyrene # fluoranthene, benzofluor-
467
-------
anthene, benzanthracene, chrysene, and benzopyrenes were also
found in the air samples. Only the low molecular weight species
such as naphthalene and anthracene were found in the water samples,
with concentrations below the PPM level.
Although the analysis has been restricted to a narrow specific
list of POM species, the GC/MS data system also made it possible
to estimate the relative weight of selected POMs compared to the
total aromatic fraction of the extracted sample. Original mass
spectral data were acquired over a mass scan range of MW 125-310.
Integration of this entire mass range provided a measure of the
total aromatic content. It was then possible to extract the
abundance of only those species with 3 or more rings (MW>178)
from this total as a measure of total POM. Finally, a summation
of the selected mass POM profiles (178, 202, 228, etc.) gives
a measure of total selected POM.
For those samples investigated, the specific POMs of interest
represented 3-10% of the sample, with the outlet sample having
reduced amounts of the lighter species (i.e., 3-ring and higher-
POMs represent ^50% of the inlet samples compared to ^ 90% of
the outlet samples).
The percentage of the aromatic fraction of each sample
represented by the specific POMs analyzed by GC/MS can also be
estimated from the gravimetric analysis data. Table 15 shows
that these values range from 3 to 14% for all the stack samples
and less than 1% for the water samples from Test 3. Since the
aromatic fraction represents about half of the total extract
for most of these stack samples (Table 16), the specific POMs
analyzed by GC/MS represent 2-7% of the total sample.
The precision of the POM data reported can be estimated
from the scatter in the calibration curves and from results of
other on going programs at Arthur D. Little, Inc. In studies
whose scope has allowed for extensive quality control analysis
of replicate and spiked samples, Arthur D. Little, Inc. has found
that intralaboratory precision of ±15% relative standard deviation
is regularly attainable by the GC/MS procedures described here.
In interlaboratory comparisons, relative standard deviations
of £30% have been commonly observed. This 30% value probably
represents an upper limit on the precision of the POM data
in this report. The accuracy of the data is somewhat more diffi-
cult to estimate because recoveries of POMs from samples can be
matrix dependent and because the sample extracts were stored
(dark, refrigerated) for some time prior to analysis. In another
Arthur D. Little, Inc. program, however, recoveries of POMs
from water were generally found to be better than 75%.
468
-------
DISCUSSION
One basis for evaluating the coal preheater emissions
is to compare the concentrations of the various POM species
found in the effluent stream to previously specified "levels
of concern". The appropriate level of concern might be an
NSPS or other emission guideline for some species but such
values have not presently been defined for POMs. A set of
Multimedia Environmental Goals (MEG's) for a wide range of
chemical pollutants, including POM's, have been developed by
and for EPA-IERL to provide consistent criteria for evaluating
and comparing a variety of emission types. The coal preheater
POM emission levels can be compared to the DMEG (Discharge MEG)
values for those species to obtain a numerical discharge severity
(DS) rating (3). The DMEG values for the POMs of interest are
presented in Table 17. The ratio of estimated preheater emission
levels to DMEG values are summarized in Tables 18, 19 and 20.
In cases where two or more compounds of the same MW are not
resolved in the GC/MS analysis, the assumption is made for the
worst case, i.e., assuming all of the concentration is due to
the species having the lowest DMEG value. For example, to get
the discharge severity for benz(a) anthracene/benz(c) phenanthrene,
the concentration found is divided by the DMEG of benz(a) anthra-
cene (45 yg/m^ in the air). Data thus obtained show that in most
of the air samples the following POMs' exceed the DMEG level:
phenanthrene, benz(a) anthracene, and 3-methylcholanthrene when
present. These are all species for which the DMEG levels of
concern are less than 50 yg/m^ (ppb) in air. None of the POM
species found in the water samples exceed their DMEG values for
that medium.
The Level 1 Organic Analysis results for the two stack samples
and two water samples are presented in Table 21. The concentration
of each category was estimated using the method described in the
EPA Level 1 procedure manual. Aliphatic hydrocarbons, fused
aromatics, phenols, and esters were found to be the major com-
ponents for both inlet and outlet stack samples. Aliphatic hydro-
carbons, carbazoles, and phenols were found in the water samples.
RE COMMENDATION S
The initial results generated from this research project
suggest further study of:
1. The effect on particulate and POM emissions at varying
scrubber liquid-to-gas ratios.
2. The effect on particulate and POM emissions at varying
scrubber pressure drops.
3. The effect of scrubber recycle water quality on parti-
culate and POM emissions.
4. The effect of different coal mixes (and sizing) on POM
and particulate emissions.
5. The effect of other gas cleaning systems on POM emissions.
469
-------
REFERENCES
(1) "Particulate Polycyclic Organics Material" Committee on
Biologic Effects of Atmospheric Pollutants, Division of
Medical Sciences, National Research Council, National
Academy of Sciences, Washington, D.C., 1972
(2) Cleland, J.G., and G.L. Kingsbury, "Multimedia Environ-
mental Goals for Environmental Assessment," Volumes 1 and
2f EPA-600/7-77-136a and b.
(3) Schalit, L.M., and K.J. Wolfe, "SAM/IA: A Rapid Screening
Method for Environmental Assessment of Fossil Fuel Energy
Process Effluents," Aerotherm Report TR-76-50, August
1977, EPA Contract 68-02-2160, T.D. No. 4.
(4) Rudolph, J.L., and Rechsteiner, C.E. "Analysis of Samples
from Coal Preheater Effluent", A.D.L. report to York
Research Corporation, ADL-C-82450, May 1979.
470
-------
TABLE 1
POLYCYCLIC ORGANIC MATERIAL INVESTIGATED IN
COAL PREHEATER SYSTEM EMISSION TEST PROGRAM
Species
MW
Carcinogenicity
Rating*
Naphthalene 128
Fluorene 166
Anthracene/Phenanthrene 178
Fluoranthene 202
Pyrene 202
Benz (a) anthracene/ (c) phenanthrene 228
Chrysene/Triphenylene 228
Benzo (b or k) fluoranthene 252
Benzo (j) fluoranthene 252
Benzo (e) pyrene 252
Benzo (a) pyrene 252
Cholanthrene 254
Dimethyl benz anthracene isomers** 256
(7,12-dimethyl benz (a) anthracene)
Dibenzo (c,g) carbazole 267
3-Methylcholanthrene 268
Indeno (1,2,3-cd) pyrene 276
Benzo (ghi) perylene 276
Dibenz (ah or a j ) anthracenes 278
Dibenzacridines 279
Coronene 300
Dibenzo (a,h) pyrene 302
Dibenzo (a,i) pyrene 302
*Carcinogenicity Code
-not carcinogenic
+uncertain or weak carcinogenic
+carcinogenic
++,+++ , strongly carcinogenic
+
+
++
++
+++
++
++++
+++
++++
+++
++
+++
+++
**Includes dimethyl- and ethyl-chrysenes , benzophenanthrenes,
and benzanthracenes.
471
-------
TABLE 2A
EPA-5 SAMPLING TRAIN ANALYTICAL METHODOLOGY FOR
EVALUATION OF CHLOROFORM-SOLUBLE AND NON-CHLOROFORM SOLUBLE PARTICULATE
Nozzle,
Probe,
Front Half
Filter Holder
(Acetone Wash)
Filter
Sealed/Labeled
Sample Bottles
Back Half Filter
Holder, Impinger
Lines, Impingers
(Acetone Wash)
Sealed/Labeled
Petri Dishes
Evaporate to
Dryness at Room
Temperature
Impinger
Solutions
(Water)
Silica
Gel
Sealed/Labeled
Sample Bottles
Sealed/Labeled
Sample Bottles
Evaporate to
Dryness at Room
Temoerature
Ichloroform-T
Soluble
[["articulate!
Non-
Chloroform-
Soluble
Particulate
Chloroform-
Soluble
Particulate
r Chloroform
Extraction
Non-
Chloroform-
Soluble
Particulate
Sealed/Labeled
Sample Bottles
Chloroform
Extraction
Chloroform-1
Soluble
Particulate
Non-
Chloroform
Soluble
Particulatei
r
Chloroform
Extraction
Chloroform-
Soluble
Particulate
Non-
Chloroform
Soluble
Particulate
Chloroform
Soluble
Particulate
"Front Half Catch"—
"Back Half Catch1
K
k
EPA Method 5 "Particulate" - »-|
State of Pennsylvania DER "Particulate1
-------
TABLE 2B
POM SAMPLING TRAIN PARTICULATE
ANALYSIS METHODOLOGY
Nozzle,
Probe
Front Half Filter Holder
Filter
Methylene
Chloride
Wash
Sealed/Labeled
Sample Bottles
Sealed/Labeled
Petri Dishes
Dry to Constant
Weight at Room
Temperature
Dry to Constant
Weight at
Room Temperature
473
-------
TABLE 3
CHLOROFORM SOLUBLE AND NON-CHLOROFORM SOLUBLE FARTICULATE EMISSION RATES
A. NON-CHLOFORORM SOLUBLE PARTICULATE EMISSIONS
Particulate Emission Rate
Nozzle, Probe, Front
Half Filter Holder
Test Scrubber
No. Location
1 Outlet
2 Inlet
Outlet
3 Outlet
B. CHLOROFORM SOLUBLE
1 Outlet
2 Inlet
Outlet
3 Outlet
A.
B.
gr/DSCF
0.0396
0.403
0.026
0.033
Ib/ton
0.0396
0.352
0.01B
0.034
% Wt
14.2
11.2
7.2
13.1
gr/DSCF
0.100
2.956
0.25
0.171
Filter
Ib/ton % Wt
0.1025 36.4
2.581 82.0
0.178 69.4
0.18 68.1
Back half filter holder
Impinger lines, Impingers,
Impincjer Solutions
gr/DSCF Ib/ton
0.136 0.1385
0.246 0.215
0.084 0.06
0.047 0.05
% Wt
49.4
6.8
23.4
18.8
PARTICULATE EMISSIONS
0.00437
0.0408
0.00327
0.0011
0.00446
0.0356
0.00233
0.0012
1.3
9.2
1.1
0.2
0.0437
0.1022
0.1757
0.0616
0.0446 12.5
0.159 41.0
0.125 58.2
0.065 11.8
0.0931 0.0951
0.1165 0.102
0.0862 0.0614
0.4014 0.4216
26.7
26.2
28.6
76.9
NON-CHLOFORORM SOLUBLE PARTICULATE EMISSIONS
Test
No.
1
2
3
Scrubber
Location
Outlet
Inlet
Outlet
Outlet
CHLOROFORM SOLUBLE
1
2
3
Outlet
Inlet
Outlet
Outlet
gr/DSCF
0.275
3.605
0.36
0.251
Total
Ib/ton
0.2806
3.148
0.256
0.264
% Wt
100.0
100.0
100.0
100.0
PARTICULATE EMISSIONS
qr/DSCF
OT2078~
0.1047
0.0365
0.05B1
Silica Gel
Ib/ton
5.2122
0.0914
0.026
0.061
I Wt gr/DSCF
517B" 077*9
23.6 0.4442
12.1 0.3017
11.1 0.5222
Total
Ib/ton % Wt
0.356 ToTTTOO
0.388 100.00
0.215 100.00
0.549 100.00
-------
TABLE 4
COAL PREHEATER PROCESS DATA
DURING PARTICULATE EMISSION RATE EVALUATION
Scrubber
Flowrate
-Coal Preheater Scrubber
Test Feed Rate Outlet Temp. Pressure Drop Moisture
No. (.tph) (°F) '__ (in W.C.) ACFM DSCFM (%)
1-outlet 84 532 20 26,823 10,011 53.2
2-inlet 105 520 20 43,186 10,700 53.5
2-outlet 105 520 20 20,791 8,728 52.0
3-outlet 90 520 20 32,192 11,030 58.0
475
-------
TABLE 5
POM EMISSION FACTORS, LB/TON OF COAL
Test No.
Anthracene
Location Naphthalene Fluorene Phenanthrene Fluoranthene Pyrene
1
(60/530)
2
(90/520)
3
(92/520)
4
(120/500)
5
(119.5/550)
6
(119/550)
7
(90/550)
8
(120/520)
10
(90/520)
11
(90/500)
Inlet
Outlet
Inlet
Outlet*
Inlet
Outlet
Outlet
Outlet
Outlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
0.0012
0.0063
0.0062
0.0091
0.013
0.0085
0.016
0.0031
0.0053
0.0074
0.049
0.012
0.0096
0.008
0.0047
0.0042
0.00031
0.00058
0.0007
0.00049
0.0018
0.00039
0.0013
o.ooia
0.0007
0.0019
0.0056
0.0016
0.0013
0.0013
0.0006
0.001
0.0022
0.0023
0.0039
0.002
0.0071
0.0015
0.0032
0.009
0.0033
0.0064
0.018
0.006
0.0091
0.003
0.0031
0.0036
0.00021
0.00016
0.00036
0.00018
0.00052
0.0001
0.00034
0.00077
0.00033
0.00033
0.0025
0.00068
0.00064
0.00015
0.00029
0.00027
0.00022
0.00017
0.00032
0.0001
0.00055
0.00011
0.00031
0.00067
0.00037
0.00037
0.0022
0.0006
0.00064
0.00016
0.00025
0.00025
0.00015
0.00005
0.00024
0.00038
0.00009
0.00017
0.00054
0.00026
0.00037
0.0014
0.00044
0.0004
0.00007
0.00013
0.00014
Benz(a)anthracene/ Chrysene/ Benzo (b or k)
(c)phenanthrene Triphenylene Fluoranthene
0.00026
0.00008
0.00048
0.00083
0.00029
0.00034
0.001
0.0007
0.00074
0.0022
0.00068
0.0009
0.00016
0.00031
0.00028
0.00092
0.00003
0.00026
0.00052
0.00016
0.0005
0.0004
0.00055
0.0015
0.00064
0.00038
0.00009
0.00011
0.00013
* Not Simultaneous
-------
TABLE: 6
POM EMISSION FACTORS
(Ib/ton of Coal)
Test No.
1
(60/530)
2
(90/520)
3
(92/520)
Location
Inlet
Outlet
Inlet
Outlet*
Inlet
Outlet
Benzol j) Benzo(e)
Fluoranthene Pyrene
0.00016
0.00004
0.00022
0.00008
0.00044
0.00004
Benzo(a)
Pyrene Cholanthrene
0.00011
0.00003
0.00009
0.00005
0.00027
Dimethyl
Den 2 anthracene
Isomers
0.00035
0.00005
0.00061
0.00091
Dibenzo 3- Methyl-
(c,g) Cholanth-
Carbazole rene
0.00011
0.00002
Indeno
(1,2,3-cd
Pyrene
0.00006
0.00001 *
0.00003
Benzo
(ghi )
Perylene
0.00006
0.00002
0.0001
Dibenz
(all or aj)
Anthracenes
0.00006
(120/520)
Outlet
0.00015
0.00006
0.00091
(119.5/550) Outlet
0.0002
0.0012
(1J9/550)
Outlet
0.00036
0.00015
0.0012
0.00004
(90/550)
8
(120/520)
10
( 90/520)
11
(90/500)
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
0.00003
0.00017
0.00004
0.00002
0.00001
0.000008
0.00044
0.0011
0.00035
0.00034
0.00002
0.00009
0.00011
0.00023
0.00067
0.00025
0.00013
0.00003
0.00005
0.00004
0.0019
0.00)5
0.00091
0.00014
0.00025
0.00034
0.00006
0.00007
0.00007
0.00004
0.00003
o.ooooi
0.00002
0.00007
0.00023
0.00007
0.00003
* Not Simultaneous
-------
TABLE 7
COAL PREHEATER PROCESS DATA
DURING POM EMISSION RATE EVALUATION
Scrubber
Flowrate
Test
No.
1
2
3
4
5
6
7
8
9
10
11
Scrubber
Location
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Inlet
Outlet
Coal
Feed Rate
(tph)
60
60
90
90
92
92
120
120
119.5
119.5
119
119
90
90
120
120
32
90
90
90
90
Preheater
Outlet
Temp. (°F)
530
530
520
520
520
520
500
500
550
550
550
550
550
550
520
520
500
520
520
500
500
Scrubber
Pressure
(in W.C. )
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Drop
ACFM
30,330
18,921
41,649
32,230
33,984
28,394
45,348
30,431
57,046
35,461
58,907
35,304
33,030
32,584
54,431
34,809
19,012
43,779
22,610
24,553
22,376
DSCFM
9,864
7,096
14,693
16,826
9,726
14,213
12,239
12,088
17,225
14,352
17,378
13,032
9,241
11,905
16,409
12,755
6,414
12,815
8,330
6,290
9,213
Moisture
(%)
40.6
53.2
36.4
34.5
49.6
34.5
53.7
51.9
45.8
50.4
44.4
53.0
46.6
53.5
44.3
53.0
56.7
46.6
53.0
53.9
53.0
Note: Test No.2 inlet/outlet not simultaneous.
Test No.4 inlet, Test No.5 inlet, Test No.6 inlet, Test No.7 inlet and
Test No.9 outlet sampling above acceptable isokinetic range (due to
moisture variance)
478
-------
8
POM SCSC33ZS EFFICIENCY AICALiSIS
^^ "cc*^^ ^ A /•• ^ *^ "^ ^,'-''**i*i^e-^a'1^-*" C^ •»•*• * X v> ^ •»•
—**1—ai» .w*i . i« ww _ \ woi*5 ——., — wC, „ -hww€.
Process 3a.~* ~r---=- ^n-'ar
r Me.
L
2
3
4
6
3
_G
---/'
SO/
90/
OT
530
520
92/520
120/
500
119/550
12 O/
90/
"2^
520
^ -1.W /'
J
0
0
0
0
0
0
ton
.015
of coal)
5
.0134
.025
.055
.039
.086
.024
5
2
2
7
4
' Ib /
3
0
0
0
G
0
^
.00
.00
.01
. 02
a
of coal)
-
23
1
3
C
2
.0131
.32
* J V
5
3
3
o
r--- -
A ^
79
58
53
56
70
46
ier.cv %
f 2
• -
. 5
.0
. 5
. S
.7
£ 9
POM EMISSION RATZ A3 A FUNCTION Cr COAL TZZi: SATS
Coal Feed
\_—-4++)
50
90
90
92
119
120
120
Eniss
0
0
0
0
0
0
0
ion Fac.or
Inlec
.0155
.0244
.0134
.0265
.0392
.0552
.0867
( IbX-on
Ou
,-i
vrf •
0 .
0 .
• o.
n
•J m
*\
•J .
/->
•J -
cf coal)
clet
0097
0130
0028
0110
0121
C222
0253
479
-------
TABLE 10
POM EMISSION FACTOR VARIABILITY
Species
Emission Factor Range*
(Ib/ton coal feed)
Carcinoqenicitv
(1)
0.0017
0.00016 -
0.00062 -
0.00004 -
0.00004 -
0.00003 -
0.00006 -
0.00003 -
0.00001 -
0.00002 -
0.00001 -
0.039
0.0044
0.016
0.0007
0.0007
0.0005
0.001
0.0006
0.00004
0.00044
0.00025
0.00005 -
0.0015
0.00005 -
0,00002 -
0.00001 -
0.00003 -
0.00006
0.00004
0.00007
Naphthalene
Fluorene
Anthracene/Phenanthrene
Fluoranthene
Pyrene
Benz(a)anthracene/(c)
Phenanthrene
Chrysene/Triphenylene
Benzo (b or k) fluoranthene
Benzo(j)fluoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Cholanthrene
Dimethyl benzanthracene
isomers (7,12-dimethyl
benz(a)anthracene)
Dibenzo (c,g) carbazole
3-Methylcholanthrene
Indeno (1,2,3-cd) pyrene
Benzo(ghi)perylene
Dibenz(ah or ah)anthracenes
Dibenzacridines
Coronene
Dibenzo(a,h)pyrene
Dibenzo(a,i)pyrene
*Venturi Scrubber outlet (20 in W.C. pressure drop, L/G=8), 90-120 tph coal
feed rate, 500-550 F preheater outlet temperature range.
++
480
-------
TABLE 11
SCRUBBER PARTICULATE REMOVAL EFFICIENCY
DATA FROM POM TRAIN
/Preheater
Coal Feed / Outlet Inlet Outlet Scrubber
Test No. (tph) / Temperature ( F) (gr/DSCF) (gr/DSCF) Efficiency%
1 60/530 5.623 0.176 96.9
2 90/520 4.556 0.041 99.1
3 92/520 1.889 0.155 91.8
4 120/500 0.330
5 119.5/550 0.743
6 119/550 0.729
7 90/550 0.477
8 120/520 5.708 0.557 90.2
10 90/520 9.172 0.165 98.2
11 90/500 6.544 0.549 91.6
481
-------
TABLE 12
SCRUBBER INLET PARTICULATE LOADING
DATA FROM POM TRAIN
Test No.
1
2
3*
Nozzle, Probe and Front Half
Filter Holder Wash (Methylene
Chloride) , gr/DSCF (%wt)
1.353 (24.1)
1.306 (28.7)
1.609 (85.2)
Filter
gr/DSCF
(%wt)
4.27 (75.9)
3.25 (71.3)
0.28 (14.8)
1.478 (25.9)
4.23 (74.1)
10
11
3.632 (39.6)
2.01 (30.7)
5.54 (60.4)
4.534 (69.3)
* Data being re-evaluated.
482
-------
TABLE 13
SCRUBBER OUTLET PARTICULATE LOADING
DATA FROM POM TRAIN
Test No.
1
2
3
4
5
6
7
8
Nozzle, Probe, Front Half
Filter Holder Wash(Methylene
Chloride), gr/DSCF (%wt
0.047 (26.7)
0.026 (63.4)
0.016 (10.3)
0.009 ( 2.7)
0.02 ( 2.7)
0.022 ( 3.0)
0.028 ( 5.9)
0.02 ( 3.6)
Filter
gr/DSCF
(%wt)
0.129
0.015
0.139
0.321
0.723
0.707
0.449
0.537
(73.3)
(36.6)
(89.7)
(97.3)
(97.3)
(97.0)
(94.1)
(96.4)
10
11
0.031 (18.8)
0.031 ( 5.6)
0.134 (81.2)
0.518 (94.4)
483
-------
TABLE 14
SCRUBBER WATER IKLET-OUTLET
ANALYSIS
CONCENTRATION
(mg/L)
Component Test No. 1 Test No. 2
Inlet Outlet Inlet Outlet
Naphthalene 0.27 0.05 0.28 0.01
Fluorene 0.02 0.07 0.01
Anthracene/
Phenanthrene 0.06 0.04 0.17 0.02
Fluoranthene 0.01 • 0.01
Pyrene 0.02 0.02
Benz (a) anthra-
cene/ (c) phenan-
threne
Chrysene Tripheny-
lene -—
Benzo (e) pyrene — —
Dimethyl Benzan-
thracene isomers - —
Test
Inlet
0.03
0.01
0.16
0.02
0.02
0.01
0.02
0.01
0.02
No. 3
Outlet
0.05
0.01
0.04
0.01
^ ^^
Totals
0.38
0.09
0.55
0.04
0.30
0.11
484
-------
TABLE 15
Analyses of Aromatic Fractions (mg/a3)
Sample No.
1, Inlet
1, Outlet
2, Inlet
2, Outlet
2B, Outlet
3, Inlet
3, Outlet
4, Inlet
4, Outlet
5, Inlet
5, Outlet
6, Inlet
6, Outlet
7, Inlet
7, Outlet
8, Inlet
3, Outlet
9, Cutlet
10, 'inlet
10, Outlet
11, Inlet
11, Outlet
H20, 3 Inlet
H20, 3 Outlet
^
G8AV
394
154
359
93
146
1450
323
1230
549
977
1410
1610
1190
1310
1130
1370
1040
1961
770
371
593
269
38
22
**
POM
27
22
22
3.2
18
67
19
150
61
69
104
72
32
53
40
170
64
81
45
38
37
27
.3
.12
Z POM/GRAV
6.9
14
6.1
3.4
12
4.6
5.9
12
'11
7.1
7.4
4.5
2.7
4.1
3.5
12
6.1
4,1
5.8
10
6.2
10
.9
.5
* GKAV value for combined aromatic LC fractions (LC 2, 3,
and 4).
**
Sum of concentrations of specific POM determined by GC/MS.
485
-------
TABLE 1.6
Gravimetric Analvsis of Coal Preheater Samples
Sample No.
1, Inlet
1, Outlet
2A, Inlet
2A, Outlet
2B, Outlet
3, Inlet
3, Outlet
4, Inlet
4, Outlet
5, Inlet
5, Outlet
6, Inlet
6, Outlet
7, Inlet
7, Outlet
8, Inlet
8, Outlet
9, Outlet
10, Inlet
10, Outlet
11, Inlet
11, Outlet
H20, 3 Inlet
H20, 3 Outlet
Aromatic Fractions
(ms/m3)
394
154
359
93
146
1450
323
1230
549
977
1410
1610
1190
1310
1130
1370
1040
196
770
371
593
269
38
22
Total
(mg/m3)
668
603
807
175
439
2580
807
2680
1520
2080
2740
3610
2730
1950
2040
2960
2340
843
2140
1130
1290
840
59
92
Aromatic Total
(*)
59
26
44
53
33
56
40
46
36
47
51
45
44
67
55
46
44
23
36
33
46
32
64
24
* GRAV value for combined aromatic LC fractions (LC 2, 3,
and 4).
** GKAV value for total sample extract prior to LC separation.
486
-------
TABLE ]7
POM DriEG Values Based on Health Effects
Species
Naphthalene
Fluorene
Anthracene
Phenanthrene
Fluoranthene
Pyrene
Benz (a) anthracene
Benzo (c) phenanthrene
Chrysene
Triphenylene
Benzo (b) f luoranthene
Benzo (k) f luoranthene
Benzo (j ) f luoranthene
Benzo (e)pyrene
Benzo (a) pyrane
Cholanthrene
12-Dimethyl benz (a) anthracene
Dibenzo (c , g) carbazole
3-Methylcholanthrene
Indeno (1,2, 3-cd) pyrene
Benzo (ghi) perylene
Dibenz (ah) anthracene
Dibenz (a , h)acridine
Dibenz (a , j ) acridine
Coronene
Dibenzo (a , h) pyrene
Dibenzo (a, i) pyrene
MW
128
166
178
178
202
202
228
228
228
228
252
252
252
252
252
254
256
267
268
276
276
278
279
279
300
302
302
Air, ug/m3
5 x 10*
*
5.6 x 101*
1.6 x 103
9 x 101*
2.3 x 10s
45
2.7 x 10^
2.2 x 103
900
1.6 x 103
6.5 x 103
3 x 103
0.02
0.26
3.8
1.6 x 103
0.093
220
250
3.7 x 103.
43
Water, yg/L
7.5 x 105
8.4 x 105
2.4 x 104*
1.4 x 10s
3.5 x 106
670
4.1 x 105
3.3 x 104
1.3 2. 101*
2.5 x 101*
9.S x 104
4.6 x 101*
0.3
3.9
.56
2.4 x 10"
1.4
3.4 x 103
3.7 x 103
5.6 x 101*
650
* All blanks are data not available
487
-------
TABLE 18
DISCHARGE SEVERITY CALCULATED FOR POM IN
COAL PREHEATER SAMPLES, OUTLET
Discharge Severity (DS)
Spaclaa
Naphthalene
Fluorene
Anehracene/Pheaanthrena
Fluoranthaaa
Pyreaa
3anz(a)anthracaaa/(c)phaaaathreae .
Chryaeae/Trlphanylaaa
B«n£o
0.04C
0.03d
0.008
500
300e
-
3
0.3
-
2a
0.002
0.0008
4b
0.3C
0.02
4 I
0.9
-
5a
0.01
0.004
IQb
0.5C
0.5d
0.1
9000
9000«
5
1
-
10a
0.02
0.006
30b
lc
Id
0.2
10000e
6
0.3
-
5a
0.009
0.004
IQb
0.9C
Id
0.3
13500
100006
—
* Includes dimethyl- and ethyl-chrysenes,benzo-
phenanthrenes, benzathracenes.
** All blanks- are items below detection limit.
- DMEG values are not available.
a. Based on DMEG of Phenanthrene.
b. Based on DMEG of Benz(a)anthracene.
c. Based on DMEG of Chrysene.
d. Based on DMEG of Benzo(b) fluoranthrene.
e. Based on DMEG of 7,12-Dimethyl benzCa). anthracene,
488
-------
TABLE 19
DISCHARGE SEVERITY CALCULATED FOR POM IN
COAL PREHEATER SAMPLES, OUTLET
Discharge Severity (DS)
Species
Naphthalene
Fluorene
Anthracene/Phenanchreae
Fluor an thene
Pyrene
Senz (a) anthracene/ (c)phenanthreae
Chrysene/Triphenylene
Benzo (b or It) f luoranthene
Benzo{ J ) f luoranthene
Benzo(e)pyrene
Senzo(a)pyrene
Cholanthrene
Dimethyl benzanthracene iaomerB*
Dlbenzo (c , g) carbazole
3-Methylcholanthrene
Indeno (1,2, 3-cd) py rene
Benzo (ghi) perylene
Dibeaz
-------
TABLE 20
DISCHARGE SEVERITY CALCULATED FOR POM IN
COAL PREHEATER SAMPLES, OUTLET
Discharge Severity
Species
Naphthalene
Fluor ene
Anthracene/Phenanthrene
Fluoranthene
Pyrene
Benz (a) anthracene/ (c)phenanthrene
Chrysene/Triphenylene
Benzo(b or k)f luoranthene
Benzo ( j ) f luoranthene
Benzo(e)pyrene
Benzo (a) pyr ene
Cholanthrene
Dimethyl benzanthracene isomers*
Dlbenzo (c , g) carbazole
3-Methylcholanthrene
Indeno (1,2, 3-cd) pyrene
Benzo (ghi) pery lene
Dibenz(ah or aj) anthracenes
Dibenzacridines
Coronene
Dibenzo ( a, h) pyrene
Dibenzo(a,i)pyrene
m/e
128
166
178
202
202
228
228
252
252
252
252
254
256
267
268
276
276
278
279
300
302
302
Test Number
1, H70
7 x 10-5
2 x 10-3
1 x 10~6
2, H?0
2 x 1CT5
-
9 x 10-^
3, H20
7 x 10-5
-
2 x 10-3
3 x 10~5
2 x 10-6
* Includes dimethy- and ethyl-chrysenes, benzophenanthrenes, and
benzanthracenes.
** All blanks are items below detection limit.
DMEG values are not available.
490
-------
Table 21
Total Organics for Stack and Water Samples
Compound Categories
Aliphatic Hydrocarbons 190
Aromatic Hydrocarbons
Fused Aromatics, MW 216 280
Heterocyclic N Compounds 20
Aldehydes & Ketones
Alcohols, Phenols 120
Esters 150
Carboxylic Acids
Stack Samples (mq/Nra ) Water Samples (mg/L)
2A Inlet 2A Outlet 2 Inlet 2 Outlet
74
8
48
33
31
69
13
13
16
14
14
1
54
28
15
14
15
48
3
491
-------
TO
FLOTATION
0 0««0 0 aas^r 4
Y V -,.JC T Y INDIVIDUAL—^.J
y^3'"15 I l^ i OVSNS |w==3
?c£D HGP°£3
Q8K-. L
TWIN '
?==oes
C3KH-OV6N GAS
»—£F
ffl3
> (
V6NTUBI
»!»!!.: N£S TCCVSNS I ^»
V
"V
V6NTUSI
> ^
: = E2 -
-------
FIGURE 2
PARTICULATE SAMPLING TRAIN
(Without Cyclone)
-------
Figure 3
POM SAMPMNG TRAIN
Jampl my
noz'/.le
Stack
thermocouple
Pyrometer
Inclined manometer
(AP)
.Thermocouple Connection
\
Filter
holder
Coarse
control
valve
The rinome te r s
By pass'
valve
/
t'ondonser
Irapi.nyer
train
Vacuum gauge
Ice
bath
Orfice
Inclined
inanonu'.t or
(A h)
-------
^s. Procedure
Sample \v
Component \v
Probe Wash, Front
Half (CH2C12)
Acetone Wash, Front
Half, Back Half Filter
Holder and Condenser
Filter
Condensate, Back
Half (CH2C1?)
XAD-2 Adsorber
Water Sample
Filter
^^
sol
fr
Q
ids
•a
•H
5
,
%
/
\
Combine 1
1
filtrat
\
/
Liquid
Extraction
3
^
•
0 „ . .
Soxhlet
Extraction
\
^^
/
1
Combine and
Concentrate
\
\
^
-7
/
u
c
o
u
M-l U
0 a
> n
« 4J
PM X
o w
LC (GRAV of
Aromatic
Fraction)
CO
5
u
o
Figure 4 GC/MS Analysis
-------
Procedure
Sample No.
Ul
g
•H
O
i-l
O
O r-
H ^
(0
g
•rl
*J
O
at
M
O
O
CO
g
•rl
4J
O
cd
n
u
c
O
•H
•M
U
at
M
2, Inlet
2, Outlet
H20, 2, Inlet
}12Q, 2, Outlet
Stack Blank
H20 Blank
Figure 5 EPA Level 1 Organic Analysis
-------
APPENDIX A
Attendees
497
-------
AT'ITJIPI'.F.S
1 ROW AMI! STF.LT. POLLUTION
AHAIT.MKHT TF.GIIHOLOCY SYMPOSIUM
OclnlnM- 3(1, II, and November 1, 197*>
Pi'-k-<'on|j.ros.". llnlel Chie.ipo, II,
Ackermann
Akacem
AH-Klwn
Al laman
A 1 1 eu
Allen
Al ton
Amendol a
Anderson
Anl oine
Arent
Armour
Aust
Ayer
Ral a je.e
Rail a
Ban
Hanister
Ba 1 1 a
*- Eau.irv
VD
oo Heck
Men?.er
Hergman
Ik'i-kehi le
Ilhargava
Bliattacharyya
Bha I. tacha ryya
Bi ] J my re
Blair
Blaszak
Bloom
Bodnaruk
Boros
Boil f Turd
Boyer
Branscoiiie
Bridle
Broc^kowski
Bronian
Brown
Brown
Browne
Bnoni core
Bnrchard
Cai rns
Campbp 1 I
Capel I ini
Case
CnKhman
Centi
Kurt J.
Amed
M.
G. L.
C. Clark
John R.
n. F. .
C.ary A.
Peter
J. A.
David
F. K.
Walter I).
Franklin A .
Shank R.
Paul A.
Tom
Ilirk R.
Ma resh
Robert
Lee
W. C.
Kathleen
David G.
panka j
A.
S.
Richard D.
Thomas R.
Thomas
Berna rd
B. .1.
Joseph A.
K.rnest J.
Ilowa rd A .
Ma rv in
Trevor
Kdmund
Carl
Cli fford M.
II. M.
Wil liam R.
Anthony J.
John K.
D. F.
J. M.
Albert
Paula J.
Peter L.
Thomas .1 .
3426 E. 8'Hh St.
161 E. 42nd SI .
900 East Chi'-ago Avenue
lf>2 Floral Avorine
P. 0. Box 12194
f . 0. Dox R43
315 E. Wacker Dr.
25089 Center lUdge Knnd
Burlington Road
BP 36
6th and Walnut SI. reels
150 W. ]37th Street.
3 Parkway Ceul'-r, Suite 3 1 ri
V 0. Box 12l'l/i
3001 F,. ColiiinhiiK Drive
1465 Mart in Tower
ID St. Clai r
2401 West 26th SI .
200 North 71 h St reel
One Oliver 1'l.iza
MD- 1 3
1000 16lh Street , N. V.
4.30 W. Randolph
8th Street. M. F. . , V. 0. Uox 700
1800 FMC Drive
P 0. Box A - South P.-irli St. it ion
10 West 35th Streel
5 I") So. Ha ••mini
P. 0. I'.ox Q94H
465 I1'"] lei 'on Avenue
401 M Street , S. Iv .
710 S. Dearborn, Ail I'.nlnri '•im-nl
l<)10 Cochran koad
One Penn Pla^a
Box 1013R
P. 0. Box I21-I4
P. 0. Box "iO'iO, Waslfwaler lech.
North Poinl Uunlevai'l
I u:'»l>
2.301 Adams
One Research llriv.-
1910 Cochr.iri Ki-a.l
Clr.
Chi cago
New York
KasL Chicago
Murray II j 1 1
Hesearch Triangle Park
Lake Forest.
Chicago
Wesllake
IV-dforri
Mai 7. ierPN-I.es-llel 7.
Phi ladelpliia
Ch i cago
Pit tKbnrgll
l\r:pearih Triangle Park
1-i.isL Chicago
llelhlehelil
C 1 evel and
\: i i e
Lebanon
Pitt sbiirgll
Research Triangle Park
V.';i<-,h i ngt on
Ch i cago
Canton
1 1 a K c a
IliiTr.-ilo
Ch i cago
1 nd i atiapo 1 i s
Aunt in
Klmlinrsl
Washi ngt on
('h i cago
Pill !:bnrgh
Mew York
Mori i:.l ovii
Research Triangle Park
Hiir 1 i ngl on , tint .1 r i o
Si1., r rows Po ; n t
r..ir.l (.hi. -ago
Cleveland
Cleveland
Pi 1.1 !;hui gh
St an lord
Hesoarch Triangle Park
C.I ani te Ci 1 v
Niildleshrmlj-.h. Cleveland
1 herrv Mill
Cr.inile Cily
Sl aniltird
I1. 1 1 .;lill!Kb
II,
NV
IN
N.I
NC
! 1,
II,
oil
HA
I'rance
PA
11,
PA
NI;
1 N
PA
(lil
PA
PA
PA
NC,
DC
11.
1)11
II.
NY
ir,
IN
TX
11
lie
11,
PA
NY
N.I
NC
Canada
MD
IN
(III
nil
PA
CT
NC
1 L
KIIR land
NJ
II.
CT
PA
606 1 7
1 00 1 '!
463 1 2
07") T*
2770')
fiOO/i")
606d 1
kk l'(5
07130
57210
]<)10(>
60627
If) 2 '2 0
27709
/.6312
1B016
Ml 21
16r> Of.
170/i2
15222
27711
200.16
60606
W01
601/.3
1/.220
606 1 6
/i 6 2 2 5
7R766
60126
20/i60
f.060/1
LSI 22
10001
O7')o0
2770'!
I.7R /.Al
2121')
/.6:u2
4/iioi
'4/1 122
T.237
(K.'JOli
27711
62(l'.l)
TS8 <)Kl
OH03/4
f.2()'.()
06'IOft
ir.^.'o
United States Steel Corp. - South Works
Koch Engineering
City of E. ChU.-ijM* Air Quality Dept.
Wilputtr Corporation
Research Triangle Ins:? itnie
J. E. Al len and As son -lies , Inr .
Kaiser Knginoers, Inr.
U.S. EPA, Region V Kiistrin Hist. Office
GCA/TechnoJogy Division
L. E. C. E. .S.
II. K. F.PA
1 nl^rl akn , Inc.
AMI lul rrnnl i on.i I , Inr .
Resoa rch T» i a ngl f Tnsti tnLc
hi I and Steel Comp.iny
Bethlehem Steel Corp.
MeDowe11 -We IIm.in Co -
Kri e 't>s Li ng I,n horn tor i ea
Bnel I Env i rot rrh Corpo rat i on
Dravo (Corpora Li on
U.S. F.PA, RTF
American Iron and SLce I Inst i tute
Rooks, Pills, FuJlagnr & Poust
Re.publi < SIrel Corporali on
1-TIC Corporal ion
Donne r-Hnnna Coke Joint Von In re
1 FT Research I nst i tule
Crown Knvi ronntenla1 Contio1 Systems,Inc.
Rn<\ i nn I'.t-i p .
1'ari f i c- Knv i rmmirnl a I Sr rvi f^s
U.S. F.PA
U.S. l,PA
NUS Corpor.it ion
Cheini co Air Po 1 1 n t i on Ro,i rd
Alii ed Chnn i c.i 1 Co rp .
HP sea rrh 'I i i ,mj> I r Institute
Knv i ronmrnta I i'rot ert i on Service
l^tlilchrm Si PP| Corj>.
Jones fsi Laugh 1 i n SI rr 1 (>> i p .
Rrpnb 1 i r Si f-c I Corporal i i-n
I.urja Hi-others & Co., Inr
R i :-rhnl f F.nv i rnninrnl a I Sy,t ems
York Re sea rrh Corjioral ion
U.S. KPA, RTP
Na I iorwi 1 St re I Cor porn I. i on
British Sire 1 Corp., Teess i de f,nh .
Stone ft Wrhs I f-r KIIR i neer i IIR Corporal ion
Gran i te City Air PoI Iut i on Cont. Agrv -
York Resea rrh (-orpor.iL ion
NUS Corporation
-------
Chabala
Chung
Clark
C 1 Ulll
Cole
Cost atit ino
Cowherd
Cox
Coy
Craig
Craig
Crawford
Current
Dalil gren , Jr .
Dale
Davison
Desa i
deSante
Dick
Dincher
Draper
DuBrof f
Dulaney
Duvall
Egan
Elphins tone
Eri ckson
Erskine
Fa rnsworth
Fosnacht
Franco
Friedman
Fill lerton
Gage
Garzella
Geddis
Geinopolos
Goodfellow
Goodman
Goodwin
Go cma n
Gravenstreter
Green
Greene
Greene
Greenfield
Gronberg
Hackbarth
Haggin
David M.
Nevi lie K.
Wi 1 liam U.
James A.
W. T.
Michael S.
Chattel!
Terry E.
David W.
A. B.
Richard
David
Gene P.
Allan G.
Larry
James W.
Sudhir
Richard K.
Marshall
Tom
Glenn fi.
William
Edward I..
I.. A.
Richard P.
David D.
Boyd C.
George
Douglas
Donald R.
Nicholas I!.
Max
R. W.
Stephen J.
Jack
Robert
Anthony
llowa rd
Irvin G.
Don R.
Edmond
James P.
Lois
Kevin
Robert A.
Murray
Stephen
Manfred
Joseph H. S
lln i ve rsa I -Cyc I ops Div., Ma-'er SI .
50 Stan!ford St .
One Oliver Pl.iza
1509 University Avenue
100 King St,, W.
628 Wr-sl Parklauo Tow IT.
425 Volker Boulevard
.10') W. W.ishiiiKtrin
P. 0. Box 12I'»4
MD-62
26 Federal Plaza, Room 802
1567 Old Abt-rs Creek Uoad
Wetrfon Steel Division
Rox 401
867R Ridge Field Ro.id
One Research Itrivr
1701 S. Firsi Avenue
401 M Streel, S. W.
BOO E. Northwest Highway, file. 3'IC
1201 Kim Street
3001 F.. Columbus Drive
401 M Street, S. W.
4017 Nanwiiy Blvd.
P. 0. Box 642S
600 Grant St.
P. 0. Box 510
1820 Del ley Madison Boulevard
EN-341, 401 M Street, S. W.
3001 E. ColumliiiK Drive
Homer Research Labs
One Penn Plaza
401 M Street, S. W.
11236 So. Torrence Ave.
345 Conrtlnnd St.
5103 W. He 1 nit Uoad
21 St. Clair Avenue F.nst
309 V;. Washington
OAQPS, MD-13
EN-341, 401 M Street, S. W.
600 Grant St.
215 Fremont St., Enforcement Div.
59 E. Van Bnrcn
P. 0. Box 226
Box 460
Burlington Road
3 Parkway Center, Ste. 315
176 W. Adams St., Ste. 1433
Hr i dgev iI Ie
Host .nil
I'i I tsburgh
riadison
ll.imi I ton, Ontario
Dearborn
Kansas City
Ch i cago
ller.earrh Triangle Park
Kese:n rh Triangle Park
New York
llnnrnevi I le
We I rt on
(leaver Fa 11 n
Crystal Lake
SI amfonl
Maywood
Wi 1m iugton
Wash i np,l on
P.-ilal (lie
Dal las
Kast ('hicago
Washington
liavenna
Fort Myers
Pi 11 sbiirgh
Provo
Mi l..ean
Wash ington
F-ast Chicago
Bethlehem
New York
MonroeviIle
Washi ngton
Cli i en go
At Ianta
Mi Iwaukee
Toronto, Ontario
Chicago
Research Trtangle Park
Washington
Pi ttsburgh
San Francisco
Chicago
Midland
Hamilton, Ontario
Bedford
Pi ttsburgh
Chicago
PA
MA
PA
WI
Canadii
Ml
I ;*'
PA
v-V
PA
11.
i-T
11,
liK
TX
IN
DC:
on
KI.
PA
(IT
VA
DC
IN
PA
NY
PA
DC
II,
CiA
W|
<^.inada
II,
NC
DC
PA
C:A
K,
PA
Ca nada
MA
PA
II,
15017 Cyclops Corporal ion
021 14 Metcal f J» Kddy, Inc.
\'-,'l>'t Dravo Coi porat i on
S:t706 llniversilv "1 Wisconsin - Madison
I.RN 3T1 The Steel Co. ol Can.).la, Ltd.
4RI.'.6 Ford Mo I ->r Co.
64110 Midwesl Research Inslitule
60606 State of 111., Pollution Control Board
27/09 Research Triangle Institute
27/11 U.S. KI'A, RTP
IOOH7 U.S. KI'A, Region 11
I5l'i6 Materials Consul t nuts and Laboratories
26n'i2 National Sleel Corporation
15010 llabeock f, Wi Icox Co.
601)10 Baxter S Woodman, fnr.
06')06 York Research Corporal ion
6DK.3 Illinois F.PA, Div. of APC
1981B E. I. du Ponl de Nemours & Co., Inc.
204M) U.S. EPA
600(i7 Dincber Associalcr, , Inc.
75270 U.S. F.PA, l!egi<.|i VI
46312 Inland Steel Company
20460 U.S. EPA
44266 Colerapa Industries, Inc.
3390] The Hunters Corporation
)5l):i2 U.S. Steel Corporal ion
R4601 U.S. Steel Corporation - Geneva Works
22102 The MTTRE Corporation
20460 U.S. EPA
46312 Inland Steel Company
1R016 Bethlehem Steel Corp.
10001 Chemico Air Pollution Control Co.
15146 U.S. Steel Corporation
20460 U.S. EPA
60617 Interlnke, Inc.
30341 U.S. EPA
53214 Rexnord, Inc.
M4T 1L9 Hatch .Associates, Ltd.
60606 State of 111., Pollution Control Board
27711 U.S. EPA, RTP
20460 U.S. EPA
15230 U.S. Steel Corporation
94105 U.S. EPA, Region IX
60605 Citizens for a Better Environment
15059 Crucible, Inc.
L8N 3J5 Dominion Foundries & Steel Co.
01730 CCA/Technology Division
15220 Hartung, Kung & Co.
60603 Chemical & Engineering News
-------
Hall
•Hanson
Harms
Heeney
Hendrickson
Ilcndriks
Henz
Her 1 i hy
Hesse
llickman
Hi iovsky
Hoffman
Hof s t r. i n
Ho Iowa ty
Hurt, Jr.
1 nna ce
Jahlin
Jacen ty
Jackson
-Jaeger
Jeffrey
Johnson
Johnson, Jr.
Jones
Kamme rmayer
Kane
Kassi m
Keith
Keller
Komnor
Kerecz, Jr.
Ki essl ing
Ki rider
Klngt-
Knech t ges
Ko tnsberg
Kopt a
Kri kan
Krzymowsk i
Knriz
Lace, Sr.
Landreth
Larson
Lawrence
Lawson
Lee
Lorn ing, Jr.
Levine
Lisk
Lower
Scott
Michael A.
Russell F. .
John
Tcdford M.
Robert V.
Dona Id J .
Jim
Carolyn S.
Constance
Robert
AT.
Ha ro 1 d
Michael 0.
Georp.e
Joseph J.
Richard
Jeffrey W.
Walter F. .
Rudy
John 11.
Byron f I .
John 11.
N. Stuart
R.
John F
An to i ne 1 1 e
W. II.
Thomas G.
Wil ] j am
Be la J.
Frank J.
Thom.ir, J.
Charier;
Richard C.
Henry J.
David
F. G.
Cpza r y
Joseph W.
Robert II.
Ronald R.
Gregory ]..
James R.
Allan T.
John R.
John W.
Sy
fan
C,e.orv," W.
4800 E. 63rd Street
208 S. La Sal le St..
6200 Oak Tree Boulevard
2001 Sprjnj. R.I.
900 Agnow Road
MD-62
11499 Chester Road
401 M Street , S. W.
309 W. Wnsliingl on
20th k State SI reets
22850 Coolev Drive
505 King Avenue
1250 Broadway
3001 E. Columbus Drive
230 S. Dcaihi-i-n St .
1221 Avp. ol Americas
Stnt ion 1 , Kox P-J
2845 Clearview Dr.
600 Grant. Pt .
1701 S. First Avenue
Hurl inj-ton Ro.-id
1500 Front age Road
345 Court land Street.
f. 0. IV) x 12194
P. O. Box 460
One Penn PI axa
Office of Technology A'i
401 H Street. S. W.
Ma r t i n Towe r
11499 Chi-si cr Koad
Homer Rpr-.earrh I.-ibs
Onp (11 ivet Pla,-a
2735 Hroadv.iv
9542 llardpali Road
6801 Brri-ks"i i le Ro.nl
125 Si Inf. IK-ni.- Highway
P. 0. Box .'.'ill'!
150 W. n?fh SI reel
1701 S. First Avenue
6th and Walnut SI reel ;;
f . 0. Box (•'>:'!'.
3001 F. . Col bus Drive
One Ol i vei Pl.i^.i
S-3556 Lake Sb'.ro Road
P. 0. Box 1237
11236 So. Torn-ni e Ave .
1636 Mart in To-.-er
1701 S. First Avenue
1301 Soul h Hi ovo
Department of Mela Mini;'
Kansas City
Ch i capo
Independence
Oakhrook
Pittsburgh
Research Triangle P. irk
Ci nci nnat i
Wash i n^ton
Chicago
Granite Ci ly
Co It on
Co I nnibns
New fork
F.ast Chicago
Ch i caRO
New York
WrightHvil le (leach
Atlanta
Pi tt sl>nrgh
Maywood
lied ford
Northbrook
At l.iiit.-i
Research Triangle !'-ii'k
llaini 1 1 on , Out a r i.o
N'-w York
Wash i m;ton
Washi nj'.r on
Helhlclie.m
Ci nci nna t i
Bel hlehem
Pi t t shnrgh
Cleveland
Anfto 1 ;j
1 Mfleppiidenre
WcUlPrr.f icld
Sal l I,. ike City
Ch i t. a£->
n.i V^'ood
Plu ladrlphia
Fort Mvers
K.I si Chiraj.o
Pill slmrp.h
Illlffaln
North Svrlnev
Ch i c-af;;i
Ret hlehem
Maywood
Ha rr i nj> ton
llnnKlit on
nn
II.
('II
II.
I'A
NT
(M!
PC
1 1,
1 1,
CA
Oil
N'I
IN
1 1.
r.~.
NI.
(iA
PA
II,
HA
1 1
C.A
!'C
Cnnad.i
N\
!C
It
•A
III
'A
•A
III
!V
HI
CI
i !
1 L
1 1.
I'A
H
IN
PA
NN
An*! r.i 1 i.i
II.
PA
1 1.
II.
ni
64130
60604
4U31
60521
15227
27711
45246
20460
60606
62040
92324
43201
10001
46312
60604
10020
2R480
30362
1S230
(.0153
01730
(,(1062
'I030R
27709
LBN 3J5
10001
205 1 0
20406
1 flO 1 7
45246
IR016
15222
44114
14006
44131
06109
84 1 1 0
60627
00153
19106
,)3'J01
46312
15222
14219
2060
60617
18016
60153
(".0010
49931
Burns and McDonnell
U.F. Steel Corporation
Davy-McKee , Inc .
Brown & Root, Inc.
Jones f< Lanphlin Steel Corp.
U.S. F.PA, RTP
PEDCo F.nvi ronmenl al , Inc.
U.S. EPA
State of III., Pollution Control Board
National Sleet ('<• rp"ra ti on
So. Coast Air Qn:, lily Management I)ist.
Uatte 1 1 e-Co 1 nmbii:; L.ihora tori es
Hyd rot echo i ( Co rpora L i on
Inland Steel Company
U.S. F.PA, Region V
MrCrnw Hill, 33 Mela] Producing Mag.
Richard -lahlin S Associates
The CADKF, Corpora I ion
I). S. Steel Corporation
11 1 inois ETA, Div. ol APC
CiCA/Technol i.Ry Di vi r. ion
llav.leton F.nv ii oninen 1 a 1 Sciences Corp.
U.S. KPA
F
-------
Manda
Margo I i n
Marlzolf
Maslany
Ha I a
McAdams
McClusKey
McCrill is
McDermott
McMeans
Mears
Meffert
Merri tt
Meyer
Mi celi
Mi Ihan
Miller
Miller
Miner
Molnar-
Mueller
Murphy
Nagano
Nakamnra
Nel son , Jr .
Nenfeld
Nicola
Nicoll
Nogava
Notar
Oclisenfeld
Oda
Onsgard
ParJkh
Pasx.tor
Pan 1
Pekron
Perl
Peterson
Pile 1 ps
Pike
Piper
Plaks
Polglase
Pnrdy
Rad i gan
Raman i.
Ramchandani
Reilly
Riddle
J. G.
Stanley V.
J. A.
Thomas J.
Cora
I.elia M.
Ernest J .
Robert C.
James E.
Wi 1 1 i am I..
Conna 1 1 y E .
Donald P.
John 1-,.
Eugene F.
Joseph V .
Alain
A . l,es J ie
Bruce
Robert P.
James A .
Patrick
Samuel M.
Kenichi
Mas.itoshi
Theodore W.
Ronald D.
Art
Harry F. .
Hi rotoshi
John
Joe
Terry
Henry
I.ax
Laszl n
Ralph I.
Phillip G.
Conrad
Joseph C.
Richard G.
Daniel F, .
Steve
Norman
Wil li.im L.
Ralph W.
Pa t ri ck
R. V.
Chandrn
Robert P.
M. J.
20th fi State St reels
Acorn Parl-.
')400 Cherry A-'enn-
6th (t Wfi linn Si reels
1701 S. First Avenue
Middletown Work.'i
3210 Watling Si reel
IKKL, Mll-62
230 S. nearhorn St. .
4600 N. Clarendon, Ste. 1306
I860 Lincoln St.
12161 l.acl.l.ind Itoad
1406 Chainher of Cfmmert '• H'iiidinf>
230 S. Dearborn St ie.-t
485 Clyde Avenue
74 Boulevard o
llami 1 ton , Out ari-i
1 nil i an.i|nt 1 is
F.ar^ t Cli i r'a^o
I'LriR 1 ewood
Heel ford
Research Triangle Parli
Research Trian^l*1 I'-irl,
Pi t tnhnrpji
Granite City
Murray Hill
Toronto, Onlario
Buffalo
Otl.iwn, Onlarin
11
;IA
<:A
PA
i i.
on
IN
NI:
u.
ir
CO
fin
PA
II,
CA
France
PA
CA
NY
III
II,
KV
II.
NY
II.
PA
PA
nil
NY
II.
II.
PA
n,
CA
PA
< .
1 1.
'.'.-in ifl.i
IK
l\
N.J
MA
N(
MC
PA
II.
N.l
Can.ida
NY
Canada
0^040
0^140
'ims
1'>'426
6(llri3
/.'>043
4h312
27711
60604
60640
80203
63141
13219
60604
94042
I'i219
:!0308
10580
48121
60104
40601
60521
10017
60606
15261
15201
44) 18
10022
60604
60153
1 ') 1 06
60604
10362
15222
43201
60604
I.«N yn
46225
46112
07631
01730
2771 1
27711
15219
62040
07')74
M/iV 1P5
14240
K1A 1C8
National Stool CDI poration
Arllnir I). Little-, Inc.
Kaiser Utcr I Corporation
U.S. F.TA, Region I I I
II lino is El'A, Hiv. nl ARC
Armco, Inc.
Inlami Ste.ol Company
U.S. FPA, R'll'
U.S. EPA
Apol I o Tech
U.S. EPA, Region VIII
Envirodyne KMK i nee rs , Inc.
Koppe rs Co. , I IK- .
U.S. KI'A, Region V
Acnrex Corporation
French Envirnnniciii.il Protection Agency
Koppors Co., I IK .
U.S. EPA, RPR ion IV
Chemiro Air PoJlnlion Control Corp.
Ford Motor CD.
Faville l.e V.il ly Corp.
Kentucky Div. of Air Pollution Control
Nippon Steel Corp.
Kawasaki Stool Corp.
Peter F. Loftns Corp. (Illinois)
University of Pittsburgh
Pennsylvania F.nginooring Corporation
J. P. Bergron f:n.
Nippon Steel USA, die.
U.S. EPA, ReRion V
Illinois EPA, Div. of APC
U.S. EPA
U.S. KPA
The CADKK Corporal inn
Dravo Corporation
Battel 1 e-Col unihiir. Laboratories
U.S. EPA
The SteeJ Co. of i:,in.irl.i , Ltd.
Crown Envi roiiwonl a J Control Systems,Inc
Inland Steel Cnmp.uiy
Neptune Ai rPo I , Inc.
CCA/Techno I,,Ry ])i vis ion
U.S. F.PA, RTP
U.S. F.PA, RTP
National Steel Corporal ion
National Slo.-l Coipuiation
Wilplltte Corporation
Ontario Ministry of the Environment
Andco Environmental Processes
Environment Canada
-------
Riley
Rodgers
Rosenlhal
Rudy
Ruggiero •
Ruppersberger
Sandona t o
Schaf Cer
Schraff
Schrcibei s
Kchuldt
Sharpe
Sharpe
Shetzler
Shoup
Siehenberger
Silvcrman
Silverm.lntz
Smith
Smi th
Smith
Soko lowski
Sonkup
St. Pierre
Staellle
Stagi as
SI ebb ins
Ste i ncr
Ste-incr
S tens land
S t il es
Stromness
Szuhay
Tarta ron _
Tejuja
Tel ford
Tennyson
Thayil
Thomas
Thompson
Throop
Tomes
Towe r
Trains
Trenholm
Truskowski
Tucker
Turk
Twork
Umene
William .1.
Peter
Steve
Bill
Dominick D.
Jolin
Henry L.
Robert B.
Christopher R
I.ee
A. A.
Robert
Susan
James F. .
Stefan P.
L. G.
Stuart
Ma rk
Dan
David B.
William M.
Jery.y Z.
Stanley
George R .
Wi 1 1 i am
N. .1.
T.loyd II.
Bruce A.
..James
John G.
N. R.
Lawrence A .
Gary
Raincsh
Anton M.
Richard P.
Boni Face S .
Jean G.
Ronald J .
William
Steve
Kevin C.
Vincent P.
Andrew
Steve
A. I,.
Diana
John V.
Ki.H
Martin Tower
Staveley Works, C
230 S. Dearborn,
147 E. Second St.
1250 Broadway
MD-62
584 Delaware Avenue
401 M Street, S. W.,
37 W. Broad St.
1635 Martin Tower
Stelco Tower, 100
2200 Churchill Road
MD-62
3210 Walling
20th & State Sit
F.N-341, 40) II SI
EN-341, 401 M Street,
2845 Cle.irvjpw I
Prudential Cent*
2ROO Grant Bnildinji
320 North Clark Si.
152 Floral Avenue
2041 N. Col lege Ri.ad
One F.. Wnrlter Drive
152 Floral Aveniif-
P. 0. Box 5200
P. 0. Box 000
485 Clyde Avenue
dm N. Court , SI e.
22850 Co I ton Oi ive
Stale Street
6B01 Brecksvil l(
327 Fifth Aveinir
1330 West Mirliign
2.200 Church il I Rd.
P. O. Box fi42R
2.30 S. Dear IK, i
Somerlon Ko.id
P. 0. Box 1.0(1
P. O. Box I Of, 7
11499 Chester Road
125 Silas Deal
2130 West Park Drive
425 Volker Ron I
9542 llardpnn !<>>
3001 Dickey Una
2970 Maria Aven
North Point Boillrv;ird
Avenue ll:urn>i r I 2/
:hesterfir>ld
5 AHAP
Hie
w. , wn-552
r
:> King St. , W.
nad
'•etS
r <>,•(., S. W.
reel , S. W.
ill!'.
Si .
f
Read
i vr
<:. 325
Ve
Rund
nil K 1. .
'1 , , Air IV nil i t !io(
Alirill-Air PI-OKI imr.
.ul
li i p.hway
r i ve
VMI'd
.1
rv;i rd
/>'.
Bethlehem
Derbyshi re
Chicago
M i nrol a
New York
Research Triangle Park
Buffalo
Washington
Col umhiis
Bethlehem
Hamilton, Ontario
Springfield .
Research Triangle Par);
Wi 1 mington
I'.ast Ch i cago
Grani te Ci ty
Wash i ngl on
Wnshi rig ton
Al Lanta
ROH t on
I'i ttsl'Utgh
Chi ca go
Murray Mill.
Co 1 timhiis
Chicago
Murray Hi 1 1
Cherry Mil 1
Middle town
Monti 1 ;i i il V i cw
Pa Lit inc
1,'oi ton
('. l.-ii rtmi
I ndcpoiidenro
I'i I I shin git
lnol is
Sjir i ng( i p 1 d
l-'orl Myers
Cli i cago
Trcvose
Middlrtnwn
Waiikesha
C i tit- i nna t i
Wel.hors 1 i r 1 '706
2/M1
1<)H9R
/i63)2
0/iOAO
?.0/i('.0
2()'idO
.)03(')0
02I(I')
ISZl-i
60610
0/
6??0/t
3.TI01
66604
l'l()47
4M)63
ri :i 1 R i
'i',2 4 6
nt.o-53
•'i'ior>:i
041 10
14006
46112
r>0062
2 1 ;M '!
Bethlehem Steel Corp.
British Steel Corp. (Chemicals) Ltd.
U.S. F.PA
The Huron Co. , fin:.
ilyrlrot echn i r Corporation
U.S. EPA, RTF
NYS Dept. Environmental Conservation
U.S. KPA
Porter, Wright, Morris & Arthur
Bethlehem Steel Corp.
Steel Company ol Canada, Limited
Illinois KPA
U.S. F.PA, RTF
E. I. (in Ponl fie Nemours & Co., Inc.
Inland Steel Company
U.S. EPA
U.S. EPA
Thi? CADRK Corporation
Ch.TS . T. Ma.i n , ( nr .
National StoeJ C" rpor.it ion
Fnerpy and F.nv i ronmrnra } Protrrti on
Wi Iput te Corporal i on
Oh i o State lln i vcrs i 1 y
1 1 lino i s~ Indi .'iiKi Ri -State Commission
Wi. I put te Corporal, i on
Stone f* Wehs ter F,ng.i nrrr ing Corporation
Armco , 1 nr .
Acurpx Corpornt ion
Envi rotrrh Ai r Qua 1 i t.y Control Group
So . Coast Air Qua 1 i ty Managrmeni Dist .
U.S. St rp } Corpo r.'i t i on
Rrpuh lie- St tjp 1 Corporat ton
Sa 1 vurri Engi nct'rs , Inc.
Iiulinna State Ro;ir
-------
Va 1 i r jora
Vil lalnhitK
Voruz
Wall, .Ir.
Wallace
Weak 1 and
Wrntz
WeLt JauffM-
Wh i Lino re
Wj nga rd
W i sc
Wi Lli row
Wo jrie.chowski
Wo 1 f e
Wooil
Yot.om
Young
YouriR, Jr.
Yulinkr
Z. -mien
Ziel inski
Zwikl
K i r-o
Mr I
Ted
William T
Anna W.
.). H.
Je r frey
T. N.
Ju 1 i c
Conni p
SLcve
Will! am
Kl .
1400 Sonl.h, K-inlc (I
1701 S. Firsl Avenue
Z.'iO S. Drvirliorn SI .
7.23 W. Jackson lilv.l.
2()r)21 Chagrin III vd .
12.') Si las Dc.ine llij.livay
Sf>ntcrlon Hoad
1000 )6th ,'Urcnl , N. W.
Suite 2610, !>9 K.-i!!t Van llni
2200 Churchi1 I Road
fit h S Walnut HlrrcLt:
200 NcvilIP Road
Ma vw '>()0lr>:j
f.OIVi
f,[)?nl
'if>-U'2
60I'I.'>
6<)lr>3
604r>;)
fi(lfil)f)
ViK'2
OMII9
l<)()'.7
200 16
60GO!i
62706
19106
II I i no is KI'A
Illinois I-.PA, Div. nl AI'C
NAI.cn Clicmicnl
Knvirox, I IK .
Kpsnarrll Trianp.lr Insliliitc
Dravo Corj>oral i on
Pacific Knvi ronmi-nl ;) I Scrvirrs
Bolhlrhr-m Slecl Corp.
UNA
In I. m<1 Stcrl nompiiity
l,o(lR'--CotLrol I
Illinois KI'A, Iliv. t,| AI'C
U.S. F.PA
Prlf-r I. l.ofliis Coip. (Illinois)
(jiiria lirotlic-rr, f> Co., I in .
TKC - Knvironmciil al CIMISII I lants , Inr.
lirtx l.nbor.il orii-s, TIL.
American
Citizens
and
<-l InsLiLulr
Knvirorimnnt
Illinois KTA, Div. of AI'C
U.S. F,PA, Kf-Rion I 1 I
Shrnango, 1 nrorporat pd
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/9- 80-012
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Proceedings: First Symposium on Iron and Steel
Pollution Abatement Technology (Chicago, IL,
10/30-11/1/79)
5. REPORT DATE
February 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Franklin A. Ayer, Compiler
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
1AB604
11. CONTRACT/GRANT NO.
68-02-2630, Task 6
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COV
Proceedings; 3/79 - 2/80
OVERED
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES j.ERL-RTP project officer is Robert C. McCrillis, Mail Drop 62,
919/541-2733.
is. ABSTRACT Tne report documents presentations at the first EPA-sponsored symposium
devoted solely to pollution abatement technology for the iron and steel industry, held
in Chicago, IL, October 30 - November 1, 1979. The symposium was organized into
air, water, and solids sessions. Air pollution topics included: emission standards,
assessment of coke quench tower and by-product recovery plant emissions, sealing
of coke-oven doors, volatilization of hydrocarbons in steel rolling operations, devel-
opment of a coke-oven air pollution control cost effectiveness model, control of sin-
ter plant emissions utilizing recirculation of windbox gases, estimating fugitive con-
tributions to ambient particulate levels near steel mills , foreign technology for EOF
fugitive emission control, and fugitive particulate emission factors for EOF oper-
ations. Water topics included emission standards, total recycle of water in integra-
ted steel mills, use of spent pickle liquor in municipal sewage treatment, physical/
chemical treatment of steel plant wastewaters using mobile pilot units, foreign tech-
nology forcontrolling coke plant and blast furnace wastewaters , and formation and
structure of water-formed scales. Solid waste topics included emission standards,
environmental and resource conservation considerations of steel industry solid
waste, and de-oiling and utilization of mill scale.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATl Field/Group
Pollution Mathematical Models
Iron and Steel Industry
Emission Sintering
Assessments Dust
Coking Waste Disposal
Hydrocarbons Chemical Cleaning
Pollution Control
Stationary Sources
Emission Standards
Fugitive Dust
13B
11F
14B
13H
07C
12A
11G
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
510
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
504
------- |