v>EPA
United States Industrial Environmental Research
Environmental Protection Laboratory
Agency Research Triangle Park NC 27711
EPA-600/2-78-118a
June 1978
Research and Development
-^
Pollution Effects of
Abnormal
Operations in Iron
and Steel Making -
Volume I.
Technical Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental Protec-
tion Agency, have been grouped into nine series. These nine broad categories were
established to facilitate further development and application of environmental tech-
nology. Elimination of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY
series. This series describes research performed to develop and demonstrate instrumen-
tation, equipment, and methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the new or improved tech-
nology required for the control and treatment of pollution sources to meet environmental
quality standards.
REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-78-118a
June 1978
Pollution Effects of Abnormal Operations
in Iron and Steel Making - Volume 1.
Technical Report
by
B.H. Carpenter, D.W. VanOsdell, D.W. Coy, and R. Jablin
Research Triangle Institute
P.O. Box12194
Research Triangle Park, North Carolina 27709
Contract No. 68-02-2186
Program Element No. 1AB604
EPA Project Officer: Robert V. Hendriks
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
This study of the environmental effects of substandard, breakdown, or
abnormal operation of steelmaking processes and their controls has been made to
provide needed perspective concerning these factors and their relevance to
attainment of pollution control. The use of the term Abnormal Operating
Condition (AOC) herein, in characterizing any specific condition should not be
construed to mean that any operator is not responsible under the Clean Air Act
as amended for designing the systems to account for potential occurrence in
order to comply with applicable State Implementation Plans or New Source
Performance Standards.
-------
ACKNOWLEDGEMENT
This report presents the results of a study conducted by the Research
Triangle Institute (RTI) for the Industrial Environmental Research Laboratory
of the Environmental Protection Agency (EPA) under Contract 68-02-2186. The
EPA project officer was Mr. Robert V. Hendriks.
The project was carried out in RTI's Energy and Environmental Research
Division under the general direction of Dr. J. 0. Wortman. The work was
accomplished by members of the Process Engineering Department's Industrial
Process Studies Section, Dr. Forest 0. Mixon, Jr., Department Manager,
Mr. Ben H. Carpenter, Section Head.
The authors wish to thank the American Iron and Steel Institute for their
help in initiating contacts with the various steel companies and for their
review of this report. Members of the AISI study committee were:
Mr. William Benzer, American Iron and Steel Institute; Mr. Stephen Vajda,
Jones and Laughlin Steel Corporation; Dr. W. R. Samples, Wheeling-Pittsburgh
Steel Corporation; Mr. Tedford M. Hendrickson, Youngstown Steel; and
Mr. John R. Brough, Inland Steel Company. Acknowledgement is also given to
the steel companies who participated in this study.
iii
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES vi
LIST OF TABLES vii
INTERNATIONAL SYSTEM OF UNITS AND ALTERNATIVE (METRIC) UNITS WITH
CONVERSION FACTORS viii
1.0 SUMMARY 1
2.0 CONCLUSIONS AND RECOMMENDATIONS 3
3.0 INTRODUCTION 5
4.0 ABNORMAL OPERATING CONDITIONS 8
4.1 Sinter Plants 8
4.2 Blast Furnaces 11
4.3 Basic Oxygen Process 14
4.4 Electric Arc Furnaces (EAF) 18
4.5 Open Hearth Furnace (OH) 19
5.0 ENVIRONMENTAL EFFECTS OF ABNORMAL OPERATION 22
5.1 Sinter Plants 23
5.2 Blast Furnace 29
5.3 Basic Oxygen Process (BOP) 32
5.4 Electric Arc Furnace Steelmaking 38
5.5 Open Hearth Steelmaking 38
6.0 EVALUATION OF ENVIRONMENTAL PROBLEMS PRESENTED BY ABNORMAL
OPERATION 43
6.1 Estimated Individual Effects of AOC, Model Plant 56
6.2 Discussion 56
7.0 COST OF PREVENTING OR MINIMIZING ABNORMAL OPERATING
CONDITIONS 57
iv
-------
TABLE OF CONTENTS (cont'd)
Page
8.0 NEEDS FOR FURTHER RESEARCH AND DEVELOPMENT 60
8.1 Technology Needs 60
8.2 State-of-Art Applications and Costs 60
8.3 Additional Quantitative Data, AOC's 64
8.4 Additional Quantitative Data, Fugitive Emissions 68
8.5 Control for Major AOC's 68
Sinter Plants 68
Blast Furnaces 69
8.6 Control Development for Low R&D Investment 69
Control Equipment Delayed Start Up and Early Shut Down 69
Rotary Drum Filters 70
8.7 Cost Reduction for Presently Available But Inordinately
Expensive Controls for Major AOC's 70
9.0 REFERENCES 72
ABSTRACT 75
-------
LIST OF FIGURES
Page
1. Breakdown, shutdown or startup card 67
VI
-------
LIST OF TABLES
1. Sinter Plant Abnormal Operating Conditions. Estimated
Duration, Frequency, and Pollutant Discharge Rates 24
2. Blast Furnace Abnormal Operating Conditions. Estimated
Duration, Frequency, and Pollutant Discharge Rates 30
3, Basic Oxygen Steelmaking Process Abnormal Operating
Conditions. Estimated Duration, Frequency, and Pollutant
Discharge Rates 33
4. Electric Arc Steelmaking Process Abnormal Operating
Conditions. Estimated Duration, Frequency, and Discharge
Rates 39
5. Open Hearth Furnace Steelmaking Process Abnormal Operating
Conditions. Estimated Duration, Frequency, and Pollutant
Discharge Rates 40
6. Description of Generalized Steel Plant 44
7. Estimated Annual Increased Pollution From Abnormal
Operation, Generalized Steel Plant 45
8. Estimated Emissions and Discharges From Abnormal Operation 47
9. Estimated Emissions and Discharges From Abnormal Operation 49
10. Estimated Emissions and Discharges From Abnormal Operation 52
11. Corrective Measures for Abnormal Operating Conditions 61
-------
INTERNATIONAL SYSTEM OF UNITS AND ALTERNATIVE (METRIC) UNITS
WITH CONVERSION FACTORS
Quantity
mass
SI Unit/Modified SI Unit
volume
concentration or
rate
energy
force
area
^6
kg
Mg (megagram = 10° grams)
Mg
Gg (gigagram = 10 grams)
o
m (cubic meter)
dscm (dry standard cubic meter)
scm (standard cubic meter: 21°C, 1 atm)
i (liter = 0.001 m3)
g/m (grams/m )
3 3
mg/m (mi Hi grams/m )
g/kg
J (joule)
kJ/m3 (kilojoules/m3)
MJ (megajoules = 10 joules)
MJ/Mg
kPa (kiloPascal)
1 Pascal = 1 N/m2 (Newton/m2)
m (square meter)
Equivalent To
2.205 Ib
2205 Ib
1.1025 ton
35.32 cf
0.437 gr/ff5
0.000437 gr/fr
2 Ib/ton
0.000948 Btu
0.02684 Btu/fr
0.430 Btu/lb
859 Btu/ton
0.146 lb/in2
10.76 fr
vm
-------
1.0 SUMMARY
\
This report discusses the findings of a study of the effect of abnormal
operating conditions (AOC's) on air and water pollution in the iron and steel
industry. The purpose of the study was to draw upon available data to describe
these conditions and to determine as quantitatively as possible the resulting
pollutants attributable to them. The investigation was limited to sintering
blast furnace ironmaking, and open hearth furnace, basic oxygen process, and
electric arc furnace steelmaking operations and the processes and air and
water pollution control equipment associated with these operations. Abnormal
operating conditions included startup and shutdown difficulties and upsets of
both process and pollution control equipment, plus any unusual deviations in
processing from conditions considered for equipment design. AOC's were
identified and assessed using data and observations from plant visits, and
data from the literature. The assessment was made by estimating, for a
generalized steel plant, the increased annual emissions rates and discharges,
over and above the rates attainable under specified normal conditions. The
efforts so estimated are summarized as follows.
AOC's increased sinter plant windbox annual particulate emissions by 102
percent; and hydrocarbon emissions by 12 percent. Sinter plant product
handling AOC's increased particulate emissions by 545 percent.
AOC's increased blast furnace particulate emissions from 53 to 216 percent,
while water pollutant increases ranged: 680-24,000 percent for suspended
solids, 22,000 percent for cyanides, and 5,900 percent for phenols.
AOC's increased BOP process particulate emissions from 113 to 3,000
percent, while water suspended solids increased 0-418,000 percent.
Based on this analysis, AOC's may be expected to add an additional annual
air pollutant emissions load of from one-half to several times the load from
non-upset operation. Where water treatment is required, AOC's may add addi-
tional discharges estimated at orders of magnitude in excess of standard.
-------
While the environmental impact of AOC's is expected to vary from plant to
plant, the AOC's assessed were found to occur to some extent in all plants
visited. AOC's appear to be a common problem of importance in the attainment
of desired air quality. This initial investigation indicates that there are
two types of AOC's: those that can be remedied by applying existing (but not
necessarily already applied) technology; and those requiring research and
development of suitable controls. Since AOC's have not previously been
addressed as a routine part of application of pollution controls, cost data
for AOC control are practically non-existent.
Further research appears to be justified, directed in order of judged
importance, toward the following areas: 1) additional quantitative data,
process and control equipment, 2) blast furnace bell leaks, 3) control equip-
ment bypass at startup, 4) reliability of water recycle systems, 5) external
desulfurization of iron, 6) BOP process upsets, 7) better emission factors, 8}
sinter plant hydrocarbon emissions control, and 9) blast furnace vacuum filter
performance.
This is Volume 1 of the final report. Volumes 2-6 serve as detailed
manuals of practice for, respectively, sintering, blast furnace ironmaking,
open hearth furnace, electric arc furnace, and basic oxygen process steelmaking.
These manuals provide further details concerning the AOC's for each of the
processes studied.
-------
2.0 CONCLUSIONS AND RECOMMENDATIONS
The published literature concerning abnormal operating conditions in the
iron and steel industry is sparse. Where data and information were found they
were usually part of a discussion of operating problems of new pollution
control systems rather than being the principle topic of the article. The
bulk of the data obtained on cause, effect, frequency, duration, and remedial
action for abnormal operating conditions (AOC's) was obtained from the files
of pollution control agencies and from the records and data supplied by the
steel companies who cooperated in this study.
The total picture of AOC's presented by this report and its five companion
volumes has been assembled from a large number of sources, with no one source
presenting a complete picture of a typical plant. The model plant generated
for this document has been mated with all the major AOC's identified as occur-
ring in plants with similar equipment. Estimates have been made as necessary
to complete the evaluation of AOC contributions to increased air and water
polluting emissions. On these bases, increased emissions due to AOC's have
been shown to be significant when compared to controlled emissions from the
same sources. In general, the estimates show increased emissions from AOC's
at the minimum levels of frequency and duration to be of the same order of
magnitude as controlled emissions. At the maximum condition of frequency and
duration, though not all the worst conditions can be expected to occur simul-
taneously, the estimates of increased pollution are one-half to three orders
of magnitude higher than controlled emissions. The implications of this
conclusion reach beyond the immediate goal of meeting emission standards to
the question of being able to achieve ambient air standards and desired water
quality under these conditions.
Virtually no quantitative data on discharge or emission rates during AOC's
were found during this study. As a result, estimates were made. Some of the
estimates are very sound from an engineering point of view and others weak.
Estimates based on complete failure of pollution control equipment are strong
-------
provided frequencies and durations are accurate (because uncontrolled emission
factors are known). Partial failure estimates are less sound because emission
factors for these situations are not available. Partial failure estimates for
precipitators can be reasonably well calculated given design data and precipi-
tator theory. The same is not true for fabric filters and scrubbers at least
at present. More quantitative AOC emission data are needed.
A necessary step in mounting an effective effort at reducing AOC contri-
butions to environmental problems is to build a data base. The wide variations
in AOC record keeping found during control agency visits demonstrates the need
to establish a model or standard procedure for obtaining and retaining AOC
data, not just for pollution control equipment problems, but also for process
problems that increase emissions. Both new and existing sources must be
included.
Total or partial failures of pollution control equipment have been identi-
fied as significant contributors to increased emissions from AOC's. As opposed
to some AOC problems for which there are no ready solutions, there are some
effective ways of dealing with control equipment partial failure. Many of the
plants visited had spare control equipment capacity or backup for some primary
emission sources. Applying this concept to all primary and secondary sources
not presently included is a means of effecting AOC emission reductions. This
is an available course of action that might be accomplished through enforcement
action as opposed to the other AOC's for which additional research is needed
to find satisfactory solutions.
For those AOC problems without ready solution, three areas of further
research are recommended: development of controls for major AOC's; develop-
ment of controls with small R&D investment; and development of less costly
controls for those AOC's the control of which is now inordinately expensive.
These recommendations are discussed in Section 8.
-------
3.0 INTRODUCTION
Air and water pollution standards are generally based upon control of
discharges during normal (steady-state) operation. When upsets occur in
either the process or its pollution control equipment the abnormal operation
may lead to discharges that exceed standards. Iron and steelmaking processes
are subject to such upsets, as are other industrial operations. Among the
nearly 200 iron and steelmaking plants in the United States, there are pollutant
discharge problems under abnormal operation that are important both with
respect to environmental severity and technological difficulty. Periods of
abnormal operation are becoming recognized as an important factor in the
achievement of environmental goals.
There is a need for further information concerning upsets: their identity,
cause, resulting discharges, prevention or amelioration.
This study provides an evaluation of pollution problems attributable to
abnormal operation, startup, and shut down. It is limited to five processes:
sintering, blast furnace ironmaking, open hearth furnace, basic oxygen process,..
and electric arc furnace steelmaking. The report is issued in six volumes.
Volume 1 provides a summary and analysis of more detailed technical data found
in Volumes 2 through 6. The latter volumes serve as manuals of practice for
use by those who must deal on a day to day basis with environmental problems
attributable to abnormal conditions. It is based on available published
information, and additional information supplied by iron and steel companies.
The evaluation includes comparison of the estimated pollutant discharges under
abnormal operation with the estimated discharges under normal operation,
discussion of the state-of-the-art technology for handling upsets, and discus-
sion of needs for further research and development.
An abnormal operating condition (AOC) for purposes of this study was
considered to be that which departs from normal, characteristics, or steady-
state operation, and results in increased emissions or discharges. In addition
-------
to abnormal operation, this study includes startup and shut down difficulties
of processes and control equipment, substantial variations in operating prac-
tice and process variables, and outages for maintenance, either scheduled or
unscheduled.
AOC experiences were sought by reviewing reports of upsets made to local,
state, and regional pollution control offices. Very little data were found to
exist at Control agencies relevant to AOC's. NPDES files offer none, but do
serve to identify the plant water handling flow plans. Permit applications
provided limited data from which control equipment operating characteristics
might be inferred, but seldom if ever contained data relevant to upsets, except
for raw gas particulate loadings. Some local agencies had developed upset
reporting procedures and thus could provide data from operating experiences.
After a preliminary characterization, by engineering analysis, of the processes
and their pollution controls, and with the assistance of the American Iron and
Steel Institute, steel p-lants were selected to provide, for each of the subject
processes, a range of process and control technology reflecting such factors
as plant size and age, type of control, and raw material utilization. Visits
were made to 10 plants to observe the processes and their control systems
and to obtain available data relating to upsets, their effects, their prevention,
and the roles of system design and maintenance.
Three types of information were then developed: the frequency of occur-
rence of an upset, its duration, and its estimated intensity. The ranges of
frequency given in Tables 1-5 (Section 5.0) were developed using both the
experience of the plants visited and the data available from Control Agencies.
Frequency values thus tend to reflect the performance of more than the number
of plants visited. Nevertheless there are data gaps, the data are not always
equally representative, and some frequencies are better defined (based on more
data) than others.
The duration of an AOC was also developed as a range, sometimes reflecting
the punctuality of response to a breakdown, sometimes indicating the serious-
ness of the underlying cause.
All intensity factors (increased discharge rates) are estimates, since not
specific data were found quantifying an AOC per se. Estimates given are based
-------
upon the projected reaction of the process control system to the causes of the
upset. Ranges are given for the estimates wherever data from the literature
of from steady-state operation could be applied.
For each process, the AOC's thus characterized were further investigated
to identify causes or suspected causes, and to determine whether corrective
methodology is available or needs development.
-------
4.0 ABNORMAL OPERATING CONDITIONS
This section provides descriptions of those AOC's which appear to result
in the greatest increases in emissions and discharges. Those described here
were selected from more comprehensive lists given in Volumes 2 through 6. They
are tabulated in Section 5, where their effects are assessed in terms of
increased pollution. Tables referred to in this section are located in
Section 5.
4.1 SINTER PLANTS
The most important sinter plant related AOC's identified under this study
are shown in Table 1 (Page 24). The process has two main emissions sources,
windbox gases from the sintering strand and product handling gases from the
sinter breaker, screens, coolers, and associated conveyors. In addition,
fugitive emissions result from materials handling and transports.
Generally, operational upsets that increase the dust or hydrocarbon
loading of windbox gases will lead to increased emissions even if emissions
control equipment continues operations at the same efficiency. Such upsets
include inadequate mixing of the feed due to blending practices in the bedding
operation, mechanical problems in material transport or proportioning from
feed bins. Direct Digital Computer control of the mixing process, to convert
the analysis of each material to limestone equivalents, develop a schedule for
the bedded pile, and sense operation during the blending and conveying has
recently been implemented at one plant with reported improvement in alle-
viating this problem.
A hearth layer of returned sinter is laid down first to provide a base
for the sintering mass and protect the grate bars. When this layer is not
uniform, increased amounts of dust pass through. Some plants do not use a
hearth layer. Grate bars distort under the thermal and mechanical stress, and
the increased openings between them pass extra dusts. Frequent (once a week)
maintenance, choice of materials of construction, and protective operating
-------
practice can minimize this problem. Most of the cost is reflected in
maintenance. Poor sintering can lead to unfused dusts, everburning and
underburning at different spots. The sinter will lack uniform quality and
there will be increased dust loadings. Causes include too much fine material
or varying moisture in the feed, and difficulties in ignition. Hydrocarbon
loadings in the windbox gases become excessive when more oils and greases
(e.g., from mill scale) are fed then can be adequately burned on the strand.
The substitution of petroleum coke can also increase hydrocarbon emissions.
The control of hydrocarbons to meet state regulations is a major problem to be
overcome in plants when recycle of such materials is practiced. Primarily for
this reason, scrubbers and wet ESP's are being considered (and are under test)
for use on windbox gases. '3
Electrostatic precipitators (ESP's) are most commonly used for control of
windbox gases, although both fabric filters (FF) and scrubbers are also used.
Fabric filters are most commonly used for product handling gases, the dust's
of which are hard, sharp, and errosive. Each type of control has its charac-
teristic upsets. Both ESP's and FF's are often held inactive during startup
until the gases warm up. This is to drive off any condensed moisture within
ESP's and prevent damage to insulators and other internal parts. FF's are
bypassed to avoid bag-blinding by cold gas contaminants. Both units may be
similarly deactivated at shut down. Weekly maintenance requires weekly shut
down and startup of these systems.
Some ESP controls include heaters for insulators; these may be turned on
ahead of startup. Solid-state transformer-rectifiers may permit initial
operation at low voltage, preventing damage while accomplishing partial
collection of particulates. Design of ESP's with individually isolatable
chambers allows system start up without complete ESP shut down. The costs of
retrofitting these features would have to be determined on an individual
basis.
Startup and shut down of fans for control equipment involves changing the
position of dampers. The resulting changes in air flow usually reentrain
dusts settled within the ducts, leading to increased emissions. No complete
remedy for this upset has been identified. Electrical equipment upsets for
-------
ESP's include broken electrodes, transformer-rectifier set failure due to
overheating, insulator failure, and abnormal rapper operation. Inspection and
maintenance on a regular basis, with a suitable spare parts inventory, are
required to minimize increased emissions from these causes.
The efficiency of collection of an ESP decreases with increases in dust
resistivity. The increase is brought about by adopting a feed mix chosen to
produce a sinter of higher basicity than that for which the control unit was
designed. A basicity may be set such that resistivity will exceed the capa-
bility of an ESP to collect particulates efficiently, in which case, an alter-
native control technique will be required.
Upsets in the handling of collected dusts include conveyor belt overloads,
screw conveyor breakdowns, and dust bridging in hoppers. These have been
alleviated by replacing the conveyor system with redesigned units. The costs
would have to be determined for a particular system.
Wet ESP's are believed to have many of the same upsets as dry ones.
Although none were observed in operation, they are candidate controls for
future use and their upset characteristics were obtained from pilot plant test
reports. '
Fabric filters used to control windbox gases are bypassed at startup and
shut down to prevent bag-blinding. Other problems include torn bags, shut
downs for major repairs, fan failures, and bypasses when inlet gases show
temperature excursions (seldom for sinter plants). Fan blades with scroll
liners resist abrasion, and fan wheels with replaceable plates are easier to
maintain. Upset experiences surveyed under this study are shown in Table 1.
Scrubbers used in sinter plants need erosion resistant (ceramic) linings,
and water pipes need to be lined with rubber or other resistant material.
The equipment cost may amount to an addition 25-50 percent, which is returned
in the form of longer equipment life. Increased emissions result from excess
oil in the strand feed, low water flow, demister failure, and plugged sprays.
Maintenance of the recycle water system affects scrubber performance; pH
control is necessary but not sufficient. Often scrubber waters are combined
with those from blast furnaces for final treatment and disposition. In any
event, upsets (clarifier failure, high sump levels, low sludge densities, and
excessive overflow of recirculating water) can increase the discharge of sus-
pended solids and other water contaminants. Costs of avoiding such upsets
10
-------
depend upon the particular system and range from the costs of redundant pumps
to the costs of complete renovating of the system to provide an adequately
designed installation.
4.2 BLAST FURNACES
The most important blast furnace related AOC's identified in this study
are shown in Table 2 (Page 30). The withdrawal of hot metal can yield in-
creased emissions often related to the furnace operation. The hot iron at the
start of the cast yields emissions at rates which are relatively low and
increase with agitation and temperature. Slag is withdrawn at the end of the
cast. Limey slag leads to excessive fumes and sulfide emissions. Such a slag
may be employed at startup to coat hearth and bosh walls and protect against
breakouts. After startup, the slag volume is decreased. To remove sulfur
with a low slag volume, the slag must be made more basic, hence the temperature
must be raised to make the slag flow properly.
Furnace top bleeders are opened at the startup of the blast furnace
campaign (once a year or so) until the existing gases show no oxygen and can
be cleaned and burned in the stoves. This is done to avoid explosions and no
control technology is in use. Bleeders are opened during shut down to release
furnace gases from an ore blank or burden quench. They may open during power
failures and furnace slips. The result of bleeder opening is the emission of
raw, dust-loaded furnace gases.
Furnace slips (or kicks) occur when the furnace burden builds up in a way
which retards gas flow. This bridging of material can occur because of im-
proper heat distribution in the furnace, too small size distribution of bur-
den, or improper slag chemistry. A slip is operator-induced, that is the
operator sees the blast pressure buildup as the bridging-sealing action takes
place. He diverts the blast and the weight of the burden causes the bridge to
collapse, slipping the furnace. A kick results when the pressure buildup is
rapid, unnoticed, or will not release, and the gas eventually blows through
the bridge-seal, blowing the bleeders on the furnace. High-alkali burdens are
especially important contributors to slipping. One of the most effective ways
to decrease the alkali content of the burden is to decrease the proportion of
those materials which are the principal sources of the alkali. Factors that
11
-------
induce good performance in blast furnace operation are also conducive to
reducing AOC from slips and kicks.
Bleeder valves, once opened, must be closed manually on many furnaces.
Automatic reset controls should help reduce the open-time to a minimum.
Partial emissions control during slips might be accomplished by venting through
the equalizer, with feed-back control from pressure sensors to open the
venturi scrubber throat and, if necessary, the associated vents. Alterna-
tively, two scrubbers might be used in series, with a slip vent between them.
This system would vent partially cleaned gas and equipment life would be
longer. Installed on a new furnace, a scrubber and equalizer might cost
$900,000; the scrubbers with control vent, $1 million. Installation on existing
systems might be limited by lack of space.
Certain furnace repairs require backdrafting--furnace gases are drawn
back through the tuyeres to a stove and burned or vented through a special
backdraft stack. Emission of gas and particulate matter from the opening of
relief valves and backdrafting occur primarily during the first 15 minutes.
Visible emissions result during this time.
Power failures, if prolonged, can result in severe damage to the furnace
with accompanying increased emissions. Residual water in the tuyeres, coolers,
bosh plates, and stack plates may boil out and the metal melt. Bleeders may
be opened to release gases formed when water leaks develop in copper coolers.
Condensation of vapors in gas mains may cause infiltration of air and explo-
o
sions. Preventive measures include an auxiliary power source, steam-driven
pumps, overhead reserve water storage, and operator training for emergency
reaction.
Breakouts are caused by failure of the walls with resulting outflow of
liquid slag or molten metal and increased emissions. Slag breakouts can be
chilled with water and the hole plugged with fireclay. Iron breakouts often
stop only upon drainage of the furnace. Most iron breakouts have occurred in
furnaces lined with carbon, and have occurred mainly while learning to use
this material.
Charging dusty material involves loading, transport, and unloading of
skip cars (some 600/day). Friable, dry material yields fugitive dust under
such handling. A second screening of pellets, sinter, and coke in the blast
12
-------
furnace stockhouse may alleviate the emissions considerably. Water sprays may
help. Extra costs for such control include the screening and spraying equip-
ment and the labor involved.
Furnace stoves can become plugged with dusts, requiring bypass of the
stove until it is cleaned. Better gas cleaning can reduce the frequency of
plugging. No single methodology (or cost) applies. Dust catchers also become
plugged, and are blown clear during shutdowns. The resulting dusts can be
partially suppressed with steam and water.
Carbon black can result from the partial burning of tuyere-injected fuel
oil and can accumulate in scrubber waters, and float on the surface of the
clarifier. The cause may be excessive fuel oil rates, non-uniform distribu-
tion of oil among the tuyeres, an improperly centered oil lance in a blowpipe,
or excessive use of mill scale in the burden. Detergents added to the water
may help to settle the carbon black.
Bell leaks result from lack of complete closure of the bell, due to
Q I
limitations in machining, and the effects of usage. ' Corrective measures
may include the application of new control technology, e.g., equalizers or
alternative furnace top designs. This problem warrants further investigation.
Increased emissions into the casthouse occur under several AOC's. A hard
blow through the tap hole is employed at shut down to remove as much iron and
slag as possible from the furnace. Molten iron may be flushed to the ground
to empty the salamander. A slow cast may result from a small tap hole, or be
due to limey slag. Cold metal (relative to desirable temperatures) due to low
silica and high sulfur within the furnace caused increased emissions of kish
during the cast. Wet tap holes are rare, but do add to increased emissions.
Tap hole enlargement can increase the flow of hot molten mass. This can often
be corrected by use of different materials, e.g., anhydrous clay instead of
water-based clay. These process and operations related AOC's are perhaps more
amenable to control, generally through control, of cast house emissions, than
by adjustment of so much of the operating practice. Such general control is
not practiced commonly in this country.
Scrubbers for blast furnace gases incur plugged nozzles and worn internals,
and if back pressure results, the furnace bleeders may open to allow the large
13
-------
bell to close. This leads to increased emissions. Corrective actions include
nozzle design change, more maintenance and control of the quality of the water
supply, especially recirculated water. The costs will depend upon the par-
ticular characteristics to be altered (improved).
Water system handling components including pumps and clarifier rakes, are
subject to breakdown due to the severity of the service. The result may be
discharge of excessive suspended solids, cyanides, and phenols. Redundant
pumps, adequate sumps, pipelines equipped with cleanout plugs and blow out
connections, and proper choice of materials of construction will help minimize
water pollution associated with scrubber operation. None of the observed
systems have resolved all their water pollution problems, and further investi-
gation of system design and materials is indicated.
4.3 BASIC OXYGEN PROCESS
Table 3 (Page 33) lists the most important AOC's identified for the BOP
during the course of this study. The process related AOC's are listed first
as they affect primary emissions from the vessel. Control equipment related
AOC's follow as they affect both primary and secondary emissions from the
vessel and ancillary operations. The paragraphs that follow briefly describe
the cause, effects, and corrective actions for these AOC's. A more detailed
discussion of each AOC can be found in the BOP Process Manual also produced
under this contract.
Startup of a newly relined vessel necessitates burning-in the tar-bonded
refractory vessel lining. Combustion of the volatile matter yields hydrocarbon
emissions. These emissions can be vented through the control device attached
to the vessel, but collection of these emissions is not necessarily achieved.
When a vessel is taken out of service for relining, the vessel is turned
upside down after cooling. The refractory lining is dumped on the ground
generating a momentary puff of emissions.
Escape of emissions around the hood skirt or puffing can be caused by
rapid vessel reactions or deterioration of the hood panel junctions. A sudden
surge in emissions may overcome the available draft and escape. If cracks
develop between the hood panels, air inleakage occurs reducing available hood
draft. Continual maintenance on the hood and improved process control are
possible solutions.
14
-------
The care used in transferring hot metal from a ladle to the furnace
vessel has a big effect on the charging emissions. Use of slow pouring and a
proper furnace and ladle angle can reduce charging emissions.
Some BOP shops have relief dampers in their fume capture systems. Excur-
sions in gas temperature coming off the vessel caused by furnace reactions may
cause the damper to open to protect attached control equipment. The effect of
this AOC can be reduced by stopping the oxygen blow when it occurs and by
improved process control.
.,•
Foaming and slopping is one result of rapid furnace reactions. It may
cause relief dampers to open or at the least cause metal to spill over the
vessel side with consequent uncontrolled fume emissions. One plant has re-
duced the frequency of this problem by revising furnace operating practice.10
Lance maintenance may also be a factor in reducing these occurrences. Slop-
ping can increase during the life of the furnace lining, the cause being
build-up of the bottom resulting in loss of free board for the bath.35
Additional emissions during charging may result from improper charge
material. Scrap and hot metal quality is the cause. Concrete, water, or oil
in the scrap causes increased emissions as does high silicon hot metal.
Improved quality control of the charged material can reduce the frequency of
this problem.
Steel tapped from the furnace is transported to ingot molds or a con-
tinuous caster by ladle. Steel may be emptied from the ladle through a bottom-
hole with flow controlled by a stopper rod. When a rod does not seat properly,
steel is spilled yielding fume emissions. This AOC may be minimized by improved
practice or in some cases, by the use of slide gates to control flow.
Many BOP shops have multiple fan installations providing draft to the
vessel. When a fan fails, a spare fan is brought on-line. If the spare fails to
start, there will be insufficient draft which results in increased hood emis-
sions. The oxygen blowing rate can, however, be reduced to match draft capacity
with emission rate.
Similarly, in many shops adjacent vessels share fan facilities. If the
out-of-service vessel is not dampered off, system draft is lost thus producing
hood emissions. The dampers used sometimes leak because they are prone to
15
-------
stick open. Attention to control, maintenance, and frequent cleaning of sealing
surfaces can minimize this AOC.
Complete failure of primary and secondary emission control systems can be
caused by power failures, fan failures, and pump failures. Wastewater treat-
ment can be interrupted by clarifier rake failures. Unless the process also
is affected, as may be with power failures, or shut down intentionally, all
the emissions escape untreated. Spare capacity in the case of fans, pumps,
and clarifiers may prevent this. While spare capacity is frequently found in
the case of the primary vessel pollution control systems, the systems for
secondary emission control do not typically have spare capacity.
Precipitators and scrubbers are the common emission control devices for
the primary vessel emissions. Fabric filters and scrubbers are commonly used
to control emissions from the ancillary operations, i.e., charging, tapping,
hot metal reladling, external desulfurization.
Manufacturers of precipitators generally recommend a warmup period after
a shut down prior to energizing the electrical sets. During this warmup,
gases pass through the precipitator untreated. One or more chambers of a
multiple-chambered precipitator may be involved. The warmup period may be
shortened by use of heat insulation, hopper heaters, and insulator heaters.
Also, cold start operation may be possible at reduced voltage when solid state
controls are present.
In a multiple fan installation bringing a spare fan into service and
shutting down another may upset gas distribution in the precipitator resulting
in poor performance. Careful design of plenums and ductwork can minimize this
AOC.
Wire breakage, transformer-rectifier (TR) set failure, cracked insulators,
and dust removal system breakdown can all cause a portion of a precipitator to
be shut down with a resulting increase in particulate emissions. Improved
materials of construction and shrouds may reduce wire breakage. Air condi-
tioning the electrical control enclosure may increase the useful life of solid
state components in the TR sets.
Cleaning and preventive maintenance can reduce the frequency of problems
with insulators. Dust removal problems can be minimized by use of hopper
16
-------
heating, heat insulation, hopper level indicators, and properly used hopper
vibrators.
Spray plugging, pump failure, and insufficient conditioning all results
in increased dust resistivity and poorer precipitator performance. Chemical
control of pH, scaling, and proper material of construction selection can
reduce pump and spray problems. Use of steam injection when the gases are too
cool to quickly evaporate water improves conditioning.
In the case of scrubbers, plugged sprays or pipes and pump failures
reduce the available scrubbing water in turn reducing collection efficiency.
As with precipitators chemical control of pH and scaling improves operations.
A plugged demister reduces available draft allowing emissions to escape
the hood. Frequent flushing and preventive maintenance can minimize this AOC.
Use of a centrifugal demister may also avoid the problem.
Spills to the sewer can result from unbalanced water flow in recirculating
systems and during acid cleaning operations to remove scale. Sufficient surge
capacity and better coordination of recycle system operation solves unbalanced
flow problems. Acid spills can be avoided by careful planning to capture and
neutralize waste acid. Care must be taken to avoid leaving an obscure valve
open or forgetting where a certain pipe goes.
In closed hood systems, carbon monoxide is generated during the oxygen
blow. Most U.S. plants do not save this gas, but choose to flare it as it is
produced. When the flare igniter fails, the gas is released unburned. Im-
mediate igniter repair is necessary. Considering the present energy concerns,
a better choice would be to store and use the gas as fuel (7,500,000 joules
per standard cubic meter ~ 200 BTU per standard cubic foot). The estimated
cost of a gas recovery facility in a new plant is about $10 million for a two
227 metric ton vessel shop.
Common problems of fabric filters include bag breakage, plugging, and bag
cleaning system failures. Bag breaks result in direct discharges for a por-
tion of the waste gas stream. Plugging or bag cleaning system failure will
cause partial or total loss of system drafts. Provision of spare compartments,
frequent inspection, and preventive maintenance are the ways to minimize the
effects of these AOC's. Poor initial system design may play a large role in
these AOC's.
17
-------
Dust removal system breakdown can also cause partial or total failure.
Appropriate action includes those things suggested for precipitators as well
as those things listed in the previous paragraphs.
Open bypass dampers may result from high temperature caused by cooling
failure (potential to damage bags) or high pressure drop. The effect is
direct discharge of emissions to the atmosphere. Frequent equipment inspec-
tion..acid ..preventive maintenance are the recommended corrective actions.
4.4 ELECTRIC ARC FURNACES (EAF)
The major identified electric arc furnace related AOC's are shown in
Table 4 (Page39). Burn-in relates a new or newly lined vessel into service.
When needed, as for example with tar-bonded refractories, burn-in is accom-
plished by putting burning coke into the furnace and operating the oxygen
lance. Excessive carbonaceous emissions reach the control device, and increased
outlet loadings result. Burn-in occurs after 100-200 heats. Where control by
total building evacuation is practiced, increased emissions may be relatively
minor.
Several furnace reactions can occur during the normally turbulent condi-
tions of steel scrap melting and furnace backcharging if the scrap has exces-
sive oil, grease, water, dirt, concrete, or ice. Shops with a canopy hood
(CH) will capture a portion of the escaping fume. Corrective measures include
reducing the oxygen blowing rate, or the electrical power input, and increasing
the furnace draft. Careful selection of scrap and proper storage are pre-
ventive measures.
The general level of emissions from an EAF goes up as the quality of the
scrap goes down, even in the absence of severe furnace reactions. Poor qua-
lity scrap can lead to a one-third increase in loadings, with increased emis-
sions from the overloaded controls. The extent to which scrap quality can be
controlled varies from shop to shop.
Shops which rely on manually placed lances for oxygen injection can
experience the varying emissions rates that accompany changes in lance position.
Excessive blowing rates also increase emissions. The upset is perhaps most
pronounced where direct shell evacuation (DSE) control systems are used. This
control system requires that an elbow be attached to the furnace roof and
that this elbow be aligned with a fixed duct for fume collection. A gap is
18
-------
left to admit combustion air. When the furnace is tilted, e.g., under condi-
tions of foaming slag, misalignment of the duct occurs and substantial emissions
escape.
Running stoppers and ladle breakouts have been described under BOF's.
Pit or charging explosions can occur when molten steel or slag contacts water.
These are accidents that may lead to increased emissions a few times a year.
Fume-capture systems generally include emergency relief or bypass dampers
for pressure relief and temperature protection. Pressure and temperature
excursions do not appear to be common because of the comparatively lesser
decarburization and lower oxygen injection rates, and substantial quantities
of dilution air used for baghouse systems.
The commonly used control system is a fabric filter. Its upsets include
stack puffs and bag failures as described for sinter plants.
4.5 OPEN HEARTH FURNACE (OH)
The most important open hearth furnace related AOC's identified under
this study are shown in Table 5 (Page40 ). Either electrostatic precipitators
or scrubbers are used to capture particulate emissions. Since fossil fuels
are fired as an energy source process, sulfur and nitrogen oxides are also
emitted. Neither of these techniques has been addressed as a separate^sub-
ject. Some removal of SO and NO may occur in the particulate control
i ** "
devices. No data are available.
Control devices may be bypassed during startup when the furnace is brought
up to temperature by preheating with natural gas, the furnace being drafted
through the waste heat boiler and out the furnace stack, bypassing the control
device. After 12 hours oil may be fired for two to three hours as needed to
make the system hot enough to provide hot gas to the control device. About 24
hours of heating are required at startup before iron and scrap can be charged
to the furnace to begin steelmaking.
Control of emissions during this period now depends upon proper oil
atomization, and employing proper air to fuel ratios in the combustion of the
fuels used. Poor oil atomization may occur at any time and with a corres-
ponding increase in hydrocarbon emissions.
19
-------
Plugged checkers cause low intake air temperature of volume and thus poor
combustion. Causes are due to poor cleaning practice and/or excessive dust,
soot and slag carryover. Of course, checkers normally plug eventually. To be
considered an abnormal operating condition, the checkers should plug and require
cleaning more often than normal for the shop.
Poor combustion can be caused by poor reversing practice, excessive fugitive
air intake, or an oxygen/fuel ratio problem. These problems are corrected by
returning to correct operating practice.
Furnace emissions through the furnace doors are the result of a high
furnace pressure, which in turn may be caused by insufficient draft, plugged
checkers or active furnace conditions such as hot metal addition, some alloy
additions, and the lime boil. The condition may be improved by either reducing
the rate of fuel input or oxygen blowing or by increasing the draft.
Molten steel escaping the furnace through an improperly sealed tap hole
and spilling onto the shop floor causes a loss of steel, safety problems, and
emissions within the building. If a tap hole breakout occurs, it is likely to
be controlled within the half-hour.
The cleaning of checkers is generally accomplished by blowing air or
steam. The ensuing dust is generally routed through the control device,
slightly increasing emissions due to the high particulate loadings. An alter-
native is to hand draw the equipment, collecting the dust by hand from the
bottom of the flue.
Boil-out from an OHF is due to occasional violent furnace reactions caused
by hot metal additions, highly oxidized scrap, a violent lime boil, or high
silicon hot metal.
Ladle reactions occur due to excessive FeO in the bath, a rapid tap, or a
furnace overcharge. Both blowing oxygen at too high a rate and blowing at a
high carbon content (> 0.3) can overwhelm the furnace control system. The re-
sult is loss of steel yield due to excessive reaction products, high particulate
loadings and gas rates to the control device and furnace puffing.
Breakouts can occur in either the furnace or the ladles. Corrective
action is to contain the spill with bags; preventive action is close attention
to the conditions of the vessels and prompt repair when necessary.
20
-------
Pit explosions occur in the slag pit or in the vessel and are generally
due to the presence of water. The explosion usually shake the building suffi-
ciently to stir up settled dust resulting in some dust emissions from various
building openings. The only recommendation for reducing these occurrences is
to avoid water leaks and spills. Unfortunately, water in the vessel may enter
with the scrap.
A running stopper has previously been discussed (Section 4.3).
If the waste heat boiler fails and is not cooling the process emissions,
some shops must bypass the control device. Emissions would be from the single
furnace affected and would amount to uncontrolled emissions for the duration of
the AOC.
The ESP bypass at startup, stack puffs, unbalanced flow, and wire breakage
problems have previously been discussed (Section 4.1). One cause of primary
collection system shut down is a catastrophic utility failure. A power failure
that affects both the process and control equipment causes both to shut down
and, therefore, the immediate environmental effect is small. If, however,
power failure leads to the failure only of the control equipment, the OHF
v
operator will have the option of shutting down or continuing the heat. As the
control devices are generally retrofits, the OHF process can sometimes operate
without controls, though at a reduced rate. In addition, the plants emergency
power system may be available to the process and not to the control device.
Scrubber related upsets have been discussed under Sinter Plants and Blast
Furnaces.
21
-------
5.0 ENVIRONMENTAL EFFECTS OF ABNORMAL OPERATION
The effects of abnormal operating conditions were assessed by estimating,
within the limits of available data, the attendant increased discharge of
pollutants. Three factors are required to characterize each condition: the
pollutant discharge rate, the duration of the abnormal condition, and the
frequency with which it occurs. Pollutant discharge rates are expressed in the
following units:
sinter plant: g/Mg of strand feed*
blast furnace: g/Mg of hot metal produced, or
g/Mg of hot metal per cast
steel making: g/Mg of steel per heat, or
g/Mg of steel per blow.
The duration of each occurrence of an AOC is expressed in hours, minutes,
or days, as appropriate. Duration often varied from plant to plant, hence a
range is given. The frequency is expressed as the number of occurrences of an
AOC per week, month, or year, and a range of frequencies is given as available.
Complete specific data, taken during AOC's do not exist. Discharge rates
shown are generally estimates, based on effects of bypassing the control
systems with gases carrying measured raw dust loadings, or on effects of likely
reductions in operating efficiency of the process or control device as well as
the frequency, was based on actual operating experience whenever possible, and
estimated when no record of experience was available.
*or kg/Mg, or Mg/Mg as dictated in the magnitude of the discharge rate.
22
-------
5.1 SINTER PLANTS
Table 1 summarizes the estimates of these three factors by AOC for sinter
plants. The main process emissions sources are covered: windbox gases and
product handling. The latter includes the sinter breaker, screens, coolers,
and associate conveyors. Only those AOC's believed to result in the most
discharges are included. Process related AOC's are listed under windbox gases
because these gases receive the attendant increased emissions. Since windbox
gases are controlled by ESP's, fabric filters, or scrubbers, AOC's are listed
for all three of these controls. Estimates are shown for the wet ESP by
analogy since although none were observed in full-scale use, several are
expected to be adopted. Where these or scrubbers are used, water treatment
AOC's can be expected to occur. Almost all the product handling gases are
controlled with fabric filters, and only AOC's for this device are listed in
the table.
\
The increased emissions rates are given relative to the sinter plant
strand feed rate because several regulations employ these units.
The increased emissions rate from an improperly formed hearth layer is
based on half the EPA Region V estimate of > 50 kg/hour for five minutes, applied
to a sinter plant with a strand feed of 220 Mg/hour (240 tons/hr). The rate under
operator error is based upon bypass at shut down of a windbox control of 400
Mg/scm particulate loadings. This loading is below the 450-700 Mg/scm reported
1 o
by Steiner for this gas. The inadequate mixing of feed results in fugitive
emissions (from pug mills) estimated at 130 g/Mg of strand feed. Excess air
leakage into the system sufficient to give an increased emission rate of 130
g/Mg was assumed. The same rate was set for dust-bridging in ESP hoppers.
Grate bar distortion was estimated using a 50 percent increase over the minimum
dirty gas particulate loading of 450 g/Mg with no decrease in efficiency of the
control device. The loading applied (675 g/Mg) is less than the top of the
12
range reported for these gases, 700 g/Mg. An equal rate was used for excess
loading of gases due to poor sintering. Uncontrolled screening of sinter
results in fugitive emissions set at 955 g/Mg, equal to the high rate from
uncontrolled windbox gases. Excess hydrocarbon emissions from petroleum coke
(reported in EPA Region V) were set at twice the controlled rate of 120 g/Mg.
23
-------
TABLE 1. SINTER PLANT ABNORMAL OPERATING CONDITIONS. ESTIMATED DURATION, FREQUENCY, AND
POLLUTANT DISCHARGE RATES
Abnormal Operation
by Source and System
WINDBOX
GASES
Ref.
Discharge Rate
Particulates Unless
Otherwise Noted
g/Mg Strand Feed
Duration
Hours
Frequency
A) Process Related
1.
2.
3.
4.
5.
6.
7.
8.
9.
Improperly formed hearth layer
Shut down NO DATA
Operator error, pol. control
shut down
Inadequate mixing of feed
Excess air leakage into sys.
Grate bar distortion
Excess loading of gases by poor
sintering
Uncontrolled screening under
equip, bndn.
Substitution of petroleum coke
19
19,12
20
21
20
12
20
19
no
546
130
130
101
101
956
240 (hydrocarbons)
0.08
0.6
1-10
84
< 84
±24
^40
8-24
>_ 1 /month
2/yr
0-4/yr
0-> 3/yr
> 6/yr
Il2/yr
1 1/yr
< 2/month
B) Control Equipment Related
ESP
1.
2.
3.
4.
5.
Used
Cold start w/ESP off
Settled dust reentrainment at
startup and shut down
Dust overload, conveyor belt
Excessively high particulate
resistivity
Dust overload, conveyor belt
12
19,21
20
12,13
12
614-956
956
100
129
100
< 1
0.08
< 1
Persistent
< 1
>_ I/week
> 2/week
1 4/day
Continuous
> I/week
at shutdown
-------
TABLE 1. (cont'd)
ro
01
Abnormal Operation
by Source and System
6. ESP shut down before process,
and bypassed
7. Transformer-rectifier failure
8. Screw conveyor breakdown
9. Dust bridging in hoppers
10. Loss of chamber, e.g. insulator
fails
11. Broken electrodes
12. Abnormal rapping, heating
13. Reduced voltage
14. Excess oil in feed
Fabric Filter Used
1. Cold start w/baghouse bypassed
2. Baghouse shut down ahead of
process
3. Excess oil in feed
4. Torn bags
5. Shutdown for major repairs
6. Fan Failure
Scrubber Used
1 . Excess oil in feed
2. Low water flow
3. Demister failure
4. Plugged sprays
5. Low pH
Ref.
22
12
21
20
20
20
12
20
12
22
20
20
20
12
20
21
21
Discharge Rate
Particulates Unless
Otherwise Noted
g/Mg Strand Feed
614-956
100
614
130
122
100
100
150
44-330 hydrocarbons
614-956
614-856
44-330 hydrocarbons
410
614-956
614-956
44-330 hydrocarbons
220-440
220-440
546-682
Daily Limited Exceeded
Duration
\
Hours
<_ 1
72
1-4
1-2
4-48
3-48
168
24-48
48
< 1
71
48
24-168
5 weeks
5-97
48
3-24
24-72
1-8
Frequency
> I/week
> 1/yr
« 7/yr
< I/week
"2/yr
> 3/yr
> 1/yr
1 -2/month
1 /month
> I/week
I/week
1 /month
1 -3/yr
—
1 5/yr
1 /month
1-6/yr
1/yr
1 -4 /month
12/yr
-------
TABLE 1. (cont'd)
ro
cr>
Abnormal Operation
by Source and System
Ref.
Discharge Rate
Particulates Unless
Otherwise Noted
g/Mg Strand Feed
Duration
Hours
Frequency
Wet ESP Used
1
2
3
4
5
6
7
8
9
10
11
12
WATER
1
2
3
4
. Cold start with ESP off
. Settled dust reentrainment
. Excessively high dust resis-
tivity
. ESP shut down before process,
bypassed
. Transformer-rectifier failure
. Loss of chamber, e.g., insulator
fails
. Broken electrodes
. Reduced voltage
. Excess oil in feed
. Plugged sprays
. Low water flow
. Discharge drain water w.o.
treatment
TREATMENT
. Clarifier failure
. High sump level
. Low sludge density
. Excessive overflow recirc. water
12
19,21
12,13
22
22
20
20
12
20
21
21
20
20
20
614-956
956
129
614-956
100
122
100
150
44-330 hydrocarbons
546-682
220-440
75-116
75-116
75-116
75-116
75-116
< 1
0.08
Persistent
I/week
>_ 2/week
Continuous
>_ I/week
> 1/yr
2/yr
> 3/yr
1-2/month
1 /month
1 -4/month
1/yr
2/month
l/yr
1-6/yr
l/yr
1 /month
-------
TABLE 1. (cont'd)
Abnormal Operation
by Source and System
PRODUCT HANDLING GASES
Fabric Filter Used
1. Excessively open hoods
2. Bypassing, cold start
3. Torn bags
4. Fan failure
Ref.
20
20
20
12
Discharge Rate
Particulates Unless
Otherwise Noted
g/Mg Strand Feed
500-700
5000
410
5000
Duration
Hours
< 84
~1
24-168
5-97
~
Frequency
2/yr
> I/week
1-3/yr
1 5/yr
Ni
-M
-------
A cold startup usually is done without either the fabric filter or ESP
turned on, at least for existing controls of these types. Similarly, the
control may be shut down ahead of the process. The increased emission rates
12
were based on the range of particulate loadings reported for untreated gases.
Settled dust reentrainment was set at the 700 g/Mg dirty gas loading rate.
Conveyor belt dust overlaods yield fugitive emissions estimated at 100 g/Mg
burden. This value is set low in consideration of the larger particulates
fraction involved. High resistivity particulate rates are based on a 10 percent
loss in control efficiency with a 450 g/Mg dirty gas loading. Screw conveyor
breakdown rates are set at the same level. Rates for ESP insulator failures,
122 g/Mg, are set at a 90 percent increase over controlled emissions. For this
purpose, "control" rates were considered to be 65 g/Mg. Broken electrode (ESP)
rates are estimated at only 100 g/Mg, the same as for rapper and heater pro-
blems.
Reduced ESP voltage emission rates are set at 150 g/Mg, based on the
13
principles of ESP operation.
Hydrocarbon emission rates from excess oil in sinter strand feed is
estimated at 44 to 330 g/Mg, based on opinions expressed by plant operating
personnel.
Torn bags (fabric filter rates (410 g/Mg) are set at three-fourths the
minimum loading of dirty windbox gases. This is high, but reported incidents
indicate that torn bags are either a severe problem or a very minor problem.
The baghouse fan failure emissions rate is based on dirty gas emissions from
continued operation.
Wet ESP emission rates were assigned by analogy since none were found in
use, but this system is favored for future use.
Discharges from water treatment AOC's are based on raw wastewater load-
•t /i
ings. Only suspended solids values are shown.
28
-------
5.2 BLAST FURNACE
Table 2 shows the estimated discharge rates, duration, and frequency for
blast furnace AOC. A fundamental factor used for these estimates is the dirty
gas loading, 42 kg of dust per Mg of hot iron produced.15 Increased emissions
at startup due to excessive lime in the slag were taken at 10 to 20 percent of
the basic value. The emissions rate from open bleeders during startup is
based on a wind rate of 13 percent of normal.16 Lindau et. al. show the 0.17
kg/Mg increased emissions used for blowing through the tap hole.17 The same
source was used for emission rates during extended startup, which presents the
problem of tap hole failure at full wind rates. Excess slips during startup
were considered to result in open bleeders at an average wind rate of 13
percent; the frequency is 0-2 per day for a 3-day average startup. The same
average rate is used for open bleeder valves during shut down. The back-
drafting emission rate was estimated at 0.01 of the normal rate for dirty gas
(0.1 of normal gas flow x 0.1 of the normal loading) since the gases are in
reverse flow and do not (fluidize" the solids in the furnace. Water and power
failure rates assume an induced draft rate equal to 10 percent of the normal
wind rate which gives negligible emission rates compared to fugitive emissions
set at 0.5 kg/Mg.
The rate for charging of dusty material is based on the fugitive emissions
rate applied 0.035 to 0.07 fraction of the operation time. Rates under plugged
18
stoves are based on clean gas loadings of 0.11 to 2.3 g/scm.
The carbon monoxide (CO) release rate under loss of ignition of bleeders
for plugged gas lines is based on a 20 percent bleed of the 3890 scm/Mg wind
gas at 35 percent CO. Carbon black is taken as a fugitive emission equal to
8-12 gms/Mg of hot metal produced.
Bell leakage rates were estimated using an average bell service life of 5
years.9 Slow casts yield increased fugitive emissions from tapping and
pouring into ladles. Lindau's measured rate of 0.59 kg/Mg is used for the
additional 10-60 minutes of the cast. Cold metal cast emission rates are
estimated as the difference between high and low rates measured at the DOFASCO
Blast Furnace No. 1 in 1976 (0.364-0.15 kg/Mg).
29
-------
TABLE 2. BLAST FURNACE ABNORMAL OPERATING CONDITIONS. ESTIMATED DURATION, FREQUENCY, AND
POLLUTANT DISCHARGE RATES
CO
o
Abnormal Operation
by Source and System
Ref.
Discharge Rate
kg/Mg raw iron
Duration
Hours
Frequency
FURNACE
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Excessive lime in slag, bleeders
open
Dirty gas bleeders open
Extended startup
Excessive slips, startup
Bleeder valve open, shut down
Severe burden slips
Backdrafting
Water and power failure
Breakouts
Charging dusty material
Stove plugging
Loss of ignition on clean gas
bleeder
Formation of carbon black
Unplugging dust catcher
Leaks from bells
15
16
17
15
15
23
17
17
4-8
5.5
0.5
5.5
5.5
37-63
0.21
0.5
0.5
18-35 gms
0.21-4.600 kg
340 kg CO/Mg
8-12 gm/Mg
42 kg
0.3
12-24
12-24
1-4 for
12 days
15 sec
16-24
5-30 sec
2 hr-7
days
1-3 days
< 30 min
5-10 sec
I/day for
21 days
1-4
3 months
1-8
Continuous
1/3-4 yrs
1/3-4 yrs
1/3-4 yrs
6/3-4 yrs
1/3-4 yrs
1 -50/month
4-50/month
1/6 or more
months
1/2 yrs
600 chgs/
day
i/yr
1 /month
i/yr
1 -8/yr
Continuous
-------
TABLE 2. (cont'd)
Abnormal Operation
by Source and System Ref .
CASTHOUSE
1. Blowing through tap hole 17
2. Hard blow, tap hole, at shut 17
down
3. Iron flush to ground, shut down 17
4. Slow cast 17
5. Cold metal
6. Hot, limey slag 17
7. Wet tap hole, trough, runner
8. Tap hole enlargement
CONTROLS
1. Scrubber problems
-
Discharge Rate
kg/Mg raw iron
0.17
0.5
0.5
0.59
0.21/Mg per cast
0.5 kg/Mg per cast
0.5
0.5
42
Duration
Hours
12-48
15-45 min
0.5 hr
10-60 min
40 min
15-45 min
5-20 min
20-45 min
Bleeder
10 sec 4
times/hr
for 0.2-48
hours
Frequency
1/3-4 yrs
1/3-4 yrs
1/3-4 yrs
1-5/week
1 -2/week
2-7/week
1/3 months
8/day
3-4/yr
WATER TREATMENT
1. Loss of water pump
2. Clarifier rake failure
SS 5-39 kg/Mg
cyanide 0-32 gm/Mg
phenol 0-13 gm/Mg
SS 0.25 kg/Mg
cyanide 13 g/Mg
phenol 6.5 g/Mg
0.5-6 hrs
1-3 days
1-6/month
1-2/yr
-------
Plugged scrubbers, leading to open bleeders are assessed using a 10 second
emission from the bleeder four times per hour for from 0.2 to 48 hours.
Water treatment AOC's that increase discharges of high waste loaded water
are assessed using data from four plant tests and assuming BATEA standards are
normally met. The data are as follows:
Range of Discharge
Untreated Waste BATEA Discharge Rate
Pollutant g/Mg g/Mg
Suspended solids , 5-39 kg 0.005 kg
Cyanides 0-32 g 0.13 g
Phenols 0-13 g 0.26 g
Rake failure discharge increases assume a 50 fold increase over BATEA
rates.
5.3 BASIC OXYGEN PROCESS (BOP)
Estimates of particulate emissions from BOP steelmaking, shown in Table 3,
are based on literature reports of uncontrolled emission rates from the BOP
vessel and secondary operations. The total range of reported vessel emissions
is 6 to 20 kg/metric ton (Mg) of raw steel. For open hood vessels, the assumed
emission factor is 20 kg/Mg; for closed hood vessels it is 10 kg/Mg.
Charging emissions are reported to range from 0.15 to 0.2 kg/Mg of hot
metal charged. They were assumed to be 0.13 kg/Mg in Table 3 (assuming 70
percent hot metal charge). Reports of tapping emissions range from 0.08 to 0.1
kg/Mg of raw steel with the assumed value of 0.1 kg/Mg. Hot metal transfer
was assumed to be in the range of 0.25 to 0.35 kg/Mg of raw steel. Uncon-
trolled flux handling emissions were assumed to be 0.75 kg/Mg of raw steel.
All the particulate emission calculations in the tables are based on the above
assumed values plus other assumptions related to capture and control efficiency
of exhaust systems as noted in the following paragraphs.
For burn in the vessel (218 Mg/heat) is assumed to contain about 318 to
363 Mg of tar-bonded refractory. Loss on ignition from tar-bonded brick is
about 6 percent of which half is volatiles. The factor in Table 3 assumes
32
-------
TABLE 3. BASIC OXYGEN STEELMAKING PROCESS ABNORMAL OPERATING CONDITIONS.
FREQUENCY. AND POLLUTANT DISCHARGE RATES
ESTIMATED DURATION,
Abnormal Operation
by Source and System
PRIMARY VESSEL EMISSIONS
1. Burn in on startup
2. Vessel dump on shut down
3. Puffing at hood
4. Improper ladle to vessel transfer
5. Relief damper opening
6. Foaming and slopping
7. Pit or charging explosions
CO
f fc
8. Improper charge material
9. Running stopper
10. Insufficient draft on startup
11. Draft loss
12. Stack puff on startup
13. Dampers stuck or jammed
14. Power failure
Primary-ESP Used
1. Precipitator warmup (1 chamber)
2. Unbalanced flow among fans
3. Wire breakage
4. Sprays plugged or corroded
Ref.
25
25
25,26
25
25,26
25,26
25-27
27
25,27
25,26
25-27
25
25,26
25,26
25-26
25-26
25-27
25-27
Discharge Rate
kg/Mg'heat
(kg/Mg'blow)
1.5-1.7 kg/Mg'
occurrence
(1.0)
0.26
(2.0)
(4.0)
0.13
0.1
6.7
2.0
2.0
10-20
(2.5)
1.0
1 wire 0.07
2 wires 0.14
1.0
Duration
minutes
(fraction of
blow time)
10-120
1
(0.05-1.0)
1 heat
(0.02-1.0)
(0.05-0.25)
20 sec
1 heat
1 heat
180
24880
1-60
60-1440
0.5
(0-0.5)
720-960
480-20160
1440-4320
Frequency
1/2-2/month
1/2-2/month
2/day
I/day
1 -1 0/month
2/week-25%
of heats
Pit - 3/yr
Charge-1/yr
2/week
3/month
0-3/yr
1/month-
1/yr
1/week-l/yr
1/month-
2/yr
3/yr-l/5 yr
1/week-l/
month
1/week-l/yr
2-12/yr
3/week-l/
month
-------
TABLE 3. (cont'd)
Abnormal Operation
by Source and System
5. Insufficient conditioning
6. Pump failure, corroded, or
eroded
7. Transformer-rectifier set failure
8. Cracked insulator
9. Dust removal breakdown
Primary-Scrubber Used
1 . Rake failure
to 2. Sprays corroded or plugged
•p*
3. Plugged or corroded pipes
4. Pump failure, corroded, or eroded
5. Plugged or failed demister
6. Drum filter failure
7. Acid cleaning scrubber overflow
8. Unbalanced water level
9. Failure to flare gas (closed
Ref.
26,27
26,27
26,27
25-27
25-27
25-27
25-27
25-27
25-27
26,27
25
25
25-27
Discharge Rate
kg/Mg'heat
(kg/Mg'blow)
(1.0)
1.0
0.07
0.07
0.14
Open hood 20.0
Closed hood 10.0
Open hood 1 .0
Closed hood 0.5
Open hood 1 .0
Closed hood 0.5
Open hood 2.0
Closed hood 1 .0
Open hood 2.0
Closed hood 1.0
pH < 6.0
110 SCM CO/Mg'blow
Duration
minutes
(fraction of
blow time)
(0.05-0.4)
120-480
120-43200
60-720
60-480
1440-4320
60-420
180
120-480
4320
10-180
60
(1.0)
Frequency
I/heat
3/yr
l/yr-1/2 yr
2-4/yr
l/week-1/2
month
0-2/yr
3/week-l/2
month
6/yr
6/yr
1/yr
1/yr
6/yr
3/yr/2 vessel
hood only)
10. Loss of instrument air (closed
hood)
25-27
(10.0) (0.025-0.035)
6/yr/2 vessel
-------
TABLE 3. (cont'd)
Abnormal Operation
by Source and System
Ref.
Discharge Rate
kg/Mg'heat
(kg/Mg'blow)
Duration
mintues
(fraction of
-blow time)
Frequency
Secondary Emissions -Fabric Filter Used
System Failure
1 . Charging
2. Tapping
3. Hot Metal Transfer
4. Flux
Bag Breakage or Plugged
1 . Charging
*> 2. Tapping
3. Hot Metal
Shaker or Reverse Air Failure
1 . Charging
2. Tapping
3. Hot Metal
Open Bypass Damper
1 . Charging
2. Tapping
3. Hot Metal
Dust Removal Breakdown
25-27
25-27
25-27
26,27
25-27
25-27
25-27
25-27
25-27
25-27
25-27
25-27
25-27
25-27
0.13
0.1
0.25-0.3
0.75 kg/Mg'day
0.0002
0.0001
0.0005
0.013
0.01
0.035
0.13
0.1
0.25-0.35
10 to 100% of
30-18720
60-5760
120-180 (1 day)
780-5760
780-5760
780-5760
120-1020
480
60-480
30/yr
18-47/yr
1-8/yr
3-6/yr
3-6/yr
3-6/yr
4-6/yr
1/yr
l/week-1/2
open bypass
month
-------
1 percent volatized during burn in producing 1.5 to 1.7 kg/Mg of vessel
capacity per burn in.
The hood puffing factor assumes a 5 percent loss of hood capture efficiency
during puffing. Improper ladle to vessel transfer assumes particulates to
be emitted at twice the normal uncontrolled charging rate. Relief damper
opening assumes all the uncontrolled emissions pass directly to the atmosphere
for the period of the AOC.
Foaming and slopping assumes the hood reaches only 80 percent capture
efficiency of fumes during the AOC. Improper charge material is assumed to
release parti oil ate equivalent to the normal uncontrolled charging emission
rate. The factor for a running stopper was estimated to be equal to normal
furnace vessel tapping emissions.
The insufficient draft factor was calculated by assuming one of three
fans serving a furnace hood would fail to start. The loss of draft was
estimated at 30 percent and hood losses estimated at 30 percent (if blowing
rate were simultaneously reduced, the factor would be reduced because not as
much draft would be required to capture the emissions).
For the case of a damper being jammed open, a 10 percent flow loss to
the non-operating hood was estimated to reduce operating hood capture efficiency
by 10 percent. As in the case of a fan failing to start, reducing blow rate
would reduce the factor.
Power failure was assumed to interrupt the process shutting down the
vessel and control system. The emissions result from an estimated 30 seconds
of uncontrolled residual emissions after the process shuts down.
The precipitator warmup factor is based on a eight-chambered precipitator,
one chamber of which is being brought on-line. Therefore, one-eighth of the
process emissions would be uncontrolled for that period. Unbalanced flow
among fans was assumed to reduce precipitator efficiency by 5 percent from
a base efficiency of 98.5 percent.
For wire breakage the precipitator condiguration was assumed to be eight
chambers by four fields in the direction of gas flow giving 32 electrically
isolatable sections. One wire failure takes one section out of service.
36
-------
Base precipitator efficiency was 98.5 percent for a specific collection area of
80.9 square meters per actual cubic meter per second. The Deutsch equation was
used to calculate the reduced efficiency in the affected chamber.
Corroded and plugged sprays were assumed to cause a 5 percent reduction in
efficiency from a base of 98.5 percent. The same assumption was applied to
insufficient conditioning and pump failure. Transformer-rectifier set failure,
cracked insulator, and dust removal system breakdown factors were'calculated
the same as for wire breakage, except that dust removal system failure assumed
two sections out of service instead of one.
For clarifier rake failure, once through scrubbing water with no terminal
treatment was assumed, giving suspended solids of 20 kg/Mg of raw steel for
open hoods and 10 kg/Mg for closed hoods. Spray problems and corroded or
plugged pipes were assumed to reduce scrubber efficiency by 5 percent. Pump
failure with no spare was assumed to be 10 percent efficiency reduction. A
plugged demister reduces hood draft; in this case 10 percent reduction was
estimated.
Unflared carbon monoxide emissions were estimated on the basis of 3.1
standard cubic meter per minute of oxygen blown per metric ton of raw steel.
For a twenty minute blow and all oxygen reacting with carbon, and 10 percent
combustion of CO to C02> the emission factor is 110 scm/Mg of raw steel. This
factor would be reduced by 50 percent or more if a CO storage facility is used.
Loss of instrument air is treated the same as a power failure. The blow
wuld be interrupted, but 30 seconds of residual emissions was assumed.
For bag breakage in secondary emissions collectors a collector size was
estimated on the basis of 4300 actual cubic meters per minute hood exhaust and
an air-to-cloth ratio of 1.5 cm per second. With 644 total bags one bag break
was estimated to release 1/644 of the total particulate load.
Shaker or reverse air failure was assumed to affect only one compartment
(total of eight). Taking this compartment out of service was estimated to
reduce available hood draft by 10 percent. Dust removal system breakdown could
affect one or more compartments producing anything from 10 percent draft loss
to complete system failure.
37
-------
Fan failures in secondary control systems generally mean complete system
failure because spare fans are typically not provided.
5.4 ELECTRIC ARC FURNACE STEELMAKING
Table 4 shows the estimated increased discharge rates, duration, and
frequency for Electric Arc Furnace AOC's. Raw gases from the process were set
po on on
at 10-15 kg of particulate per Mg of steel per heat. ' ' Burn-in emissions
were considered to be hydrocarbons from the partial combustion of new tar-
bonded ceramic furnace linings. Abnormal furnace reactions were assessed using
an 85 percent capture of dirty gas loaded at 10 kg/Mg per heat. Poor scrap was
expected to result in 33 percent greater particulate loading of the gases with
0.1 of this increase passing through the control device. Improper oxygen lance
practice is expected to increase emissions only from direct shell evacuation
systems: the rate is based on escape of 4 percent of the loading 10 percent of
the time. Capture duct misalignment emissions are based on escape of 0.5 to
0.8 of the total gas loading. Fugitive emissions (e.g., running stoppers,
ladle breakout, charging explosions) are based on 0.5 kg particulates/Mg per
heat.
Relief damper openings yield emitted dirty gas at 10-15 kg/Mg x heat.
Stack puffs are based on the same rate. Bag failures are rated very conserva-
tively using 0.01 of the raw gas loading.
Bag failures are presumed to be corrected as detected, and are assessed at
0.01 of the raw gas loading.
5.5 OPEN HEARTH STEELMAKING
Table 5 provides rough estimates of increased discharges from Open Hearth
Steelmaking AOC's. Hydrocarbon emissions increase during startup, times of
poor oil atomization, plugged checkers, or poor combustion. A discharge rate
of 240 gms/Mg x heat was selected by analogy with sinter plant excess oil
charge usage. Raw untreated gases were set at 5-10 kg/Mg of product x heat.
Fugitive emissions (e.g., breakouts, running stoppers) are estimated using
0.5 kg/Mg x heat or 0.5 kg/Mg x ladle, this being the rate measured by Landau
from blast furnace cast house AOC's of similar type. Fugitive emissions
38
-------
TABLE 4. ELECTRIC ARC STEELMAKING PROCESS ABNORMAL OPERATING CONDITIONS. ESTIMATED
DURATION, FREQUENCY. AND DISCHARGE RATES
co
Abnormal Operation
by Source and System
Ref.
Discharge Rate
Duration
Frequency
PROCESS RELATED
1 . Burn in
2. Abnormal furnace
reaction
3. Poor scrap quality
4. Improper lance position
5. Capture duct misalign-
ment
6. Running stopper
7. Ladle breakout
8. Pit or charging
explosion
9. Relief damper opening
Fabric Filter
1 . Stack puff
2. Bag failure
Est.
Est.
Est.
Est.
Est.
Est.
Est.
Est.
Est.
1.5-1.7 kg/Mg/occurrence
1 .5 kg/Mg x heat
0.03 kg/Mg x heat
0.04-0.06 kg/Mg x heat
5-12 kg/Mg x heat
0.5 kg/Mg/ladle
0.5 kg/Mg/ladle
0.5 kg/Mg/ladle
10-15 kg/Mg x heat
10-15 kg/Mg x heat
0.1-0.15 kg/Mg x heat
10-120 min 1/1 or 2/mth
2-5 min
Till quality is
Till corrected
10 min
5-30 min
5 min
20 min
10-30 min
1-5 min
1-7 days
1 /month
good
I/day
0-2/day
1/yr
1-3/yr
1/yr
1 -50/yr
1/yr
-------
TABLE 5. OPEN HEARTH FURNACE STEELMAKING PROCESS ABNORMAL OPERATING CONDITIONS. ESTIMATED
DURATION. FREQUENCY, AND POLLUTANT DISCHARGE RATES
Abnormal Operation
by Source and System
Discharge Rate
Duration Frequency
-PS-
CD
PROCESS RELATED
1. Startup
2. Poor oil atomizatlon
3. Plugged checkers
4. Poor combustion
5. Furnace puffing
6. Tap hole breakout
7. Cleaning checkers
8. Boilout
9. Ladle reactions
10. Improper control 0? blowing
11. Breakouts
12. Pit or charge explosion
13. Running stopper
14. Waste heat boiler failure
CONTROLS
1. ESP bypass, startup
2. Stack puff
3. Unbalanced flow to ESP
4. Primary collection system
5. ESP wire breakage
240 g HC/Mg per hour
240 g HC/Mg per hour
240 g HC/Mg per hour
240 g HC/Mg per hour
0.25-5 kg/Mg x heat
0.15-0.5 kg/Mg x heat
0.6 kg/Mg x heat
0.15-0.5 kg/Mg x heat
0.5-1 kg/Mg x heat
0.15-5 kg/Mg x heat
0.15-5 kg/Mg x heat
0.15-5 kg/Mg x ladle
Uncontrolled emission 5-10 kg/Mg x heat
5-10 kg/Mg x heat
5-10 kg/Mg x heat
10% additional emissions*
5-10 kg/Mg x heat
Some 1.4 times controlled emissions
8-15 hr
2 days
10 min
1 hr-
persis-
tent
30 min
1-4 hrs
1-3 min
1-30 min
15 min
1-2 min
30-60 min
15 hrs
1-15 hrs
1-60 min
1/2-1/mth
1/month
1-50/yr
I/week
1/yr
1/month
1-2/month
1/month
1/month
1-4/yr
1-3/month
1/month
2/month
1-50/yr
0-15 hrs .2-3/yr
-------
TABLE 5. (cont'd)
Abnormal Operation
by Source and System
Discharge Rate
Duration Frequency
Scrubber
1. Clogged sprays
2. Plugged pipes
3. Pump failure
4. Plugged demister
5. Vacuum filter failure
6. Acid cleaning spills
Rake failure
Transformer-rectifier failure
Insulator failure
Increased if L/G <_ 0.6
Increase rate
Increase rate
Increase rate
NO DATA BUT SAID TO BE FREQUENT
Low pH discharge
5-10 kg/Mg x heat
Some 1.4 times controlled rate*
Some 1.4 times controlled rate*
1-3 hr
10mm-3
hr
1-3 days
2 hrs-
1 month
to 3.5%
of time
1/8-3/wk
1-2 month
0-1/month
1/3 month
0-2/yr
5-1/yr
*0ne affected section out of a set of 16 sections.
-------
from furnace puffs were set at 0.05 of the uncontrolled rate. Values for
scrubber AOC's, ESP AOC's, and water treatment are selected by analogy with the
previously described processes, where possible. No suitable estimates were set
for clogged sprays, pump failures, plugged demisters, and vacuum filter failure.
42
-------
6.0 EVALUATION OF ENVIRONMENTAL PROBLEMS PRESENTED BY ABNORMAL OPERATION
Tables 1-5 provide a basis for projecting the impact of AOC's at a
specific plant. The impact would vary from plant to plant due to variations
in types of processes and control equipment, to differences in production rates,
and to differences in feed components. Emissions and discharge limitations,
both daily and average, may be increased, depending on the intensity and duration
of the AOC. To provide some perspective as to the probable importance of AOC's,
the increased discharges therefrom have been projected for a generalized inte-
grated steel plant rated at 7000 Mg/day raw steel, with two sinter plants,
two blast furnaces, and two BOP's.
Table 6 shows the material balance for the plant and the production rates
by hour, day, and year. Sinter feed includes a hearth layer. The material
mass balance relates iron, scrap, sinter, and sinter feed requirements. The
sinter is assumed to contain 62 percent iron. The feed/product ratio, 1.2/0.91
is based on current experience of plants visited. The iron production is based
on a material balance, page 457 of reference 15. The ore is assumed to contain
62 percent iron: the pig iron, 93.5 percent iron. The steel production is
based on current experience at visited plants. The BOP charge is set at 30
percent scrap, 70 percent molten iron with an 85 percent yield. There are two
furnaces. One 220 Mg BOP is in use while the other is under service or on
standby. Thirty-two 45-minute heats per day are made at full production.
The sinter usage may be higher than that found at some plants. If trends
toward higher basicity sinter and increased pellet usage continue, then the
sinter plant production will decrease, relative to the rates shown in this
example.
This plant's operation is projected for one year, taken as 350 operating
days, during which the most important AOC's are also projected in terms of in-
creased emissions or discharges.
Table 7 identifies the increased pollution by process and source. Dis-
charge rates, duration, and frequencies selected from Tables 1-5 have been used
43
-------
TABLE 6. DESCRIPTION OF GENERALIZED STEEL PLANT
A. MATERIAL MASS BALANCE
1.2 sinter feed -*• 0.91 sinter
0.074 scrap + 0.38 coke + 0.006 lime + 0.23 ore + 0.91 sinter
0.08 lime + 0.36 scrap + 0.83 iron -»• 1 steel
PLANT DESCRIPTION
0.83 iron
B.
Unit
Number
Production rate
b
each unit Mg/day
all units, Mg/day
all units, Mg/hr
million Mg/yr
Sinter Strand Feed
All units, Mg/hr
Million Mg/yr
Control Equipment
Sinter Plant
2644
5287
221
1.850
291
2.444
Windbox: ESP
Product:
Fabric Filter
Blast Furnace
2905
5810
242
2.033
BOP
2 (1 spare)
7000
7000
292
2.448
Furnace Gas: Scrubber
Casthouse: None
Vessel: Open hood
w/Scrubber
Secondary: Fabric
Filter
1 Mg = 1.1025 ton
44
-------
TABLE 7. ESTIMATED ANNUAL INCREASED POLLUTION FROM ABNORMAL OPERATION. GENERALIZED STEEL PLANT
Source
SINTER PLANTS9
Windbox Gases
Process related
P/C equip, related
Subtotal
Product Handling Gases
P/C equip, related
Subtotal , Sinter Plants
BLAST FURNACES5
Furnace
Scrubber
Cast house
Water Treatment
Subtotal, Blast Furnaces
EMISSIONS
Suspended
Parti culates Hydrocarbons Solids
Mg/yr Mg/yr Mg/yr
14.1
221.8
235.9 (240)C
534.3 (96)
770.2 (336)
172-962 (464-564)
0.1-21.6 (0.1)
76-238
248.1-1221.6
(464-564)
20.1
35.2
55.3 (443)
55.3 (443)
5.6-8.4
Carbon Monoxide
987-3949
8.8-3565
5.6-8.4 8.8-3565
(0.037-522)°
DISCHARGES
Cyanide Phenol
Mg/yr Mg/yr
0.07-3.7 0.04-1.6
0.07-3.7 0.04-1.6
(0-0.017) (0-0.27)
-------
TABLE 7. (cont'd)
en
Source
BASIC OXYGEN PROCESS
Primary Vessel Emissions
Process
P/C, scrubber and
associated water
treatment
Secondary Emissions
Subtotal, BOP
TOTAL, Plant
Parti culates
Mg/yr
36.8-1531.3
(68.4)
54.8-382.3
(0.9-10.0)
4.3-1085
(15.4-16.6)
95.9-2999
(84.7-95.0)
1114-4991
EMISSIONS
Hydrocarbons
Mg/yr
2-8.9
2-8.9
62.9-72.6
Carbon Monoxide
987-3949
DISCHARGES
Suspended
Solids Cyanide Phenol
Mg/yr Mg/yr Mg/yr
0-836.9,
(0.2)d
0-836.9,
(0.2)d
8.8-4402 0.07-3.7 0.04-1.6
c
Based on a controlled emissions rate of 65 g particulates/Mg strand feed and 120 g HC/Mg strand
feed.
Based on no startup or shut down during the year.
Numbers in parentheses are the annual pollutant discharges under continued control without any
abnormal operation.
Based on one day maximum allowable discharge.
-------
TABLE 8. ESTIMATED EMISSIONS AND DISCHARGES FROM ABNORMAL OPERATION
• — — — II - • II •! •! • •! — T • I '
GENERALIZED STEEL PLANT: TWO SINTER PLANTS
AOC
Total Emissions, Mg/yr
10" [RatexMg/hrxDuration(hr)xTimes/yr)
Equivalent
Mg/yr
Net Increased
Emissions
Mg/yr
PROCESS RELATED
Windbox Gases
1.
3.
4.
5.
6.
7.
.p* 8.
"""*
Total
Hearth layer
Operator error
Inadequate feed mix
Excess air leaks
Grate bar distortion
Poor sintering
Use of petroleum coke or
oily feed material
annual increase, Particulates
HC
110x291x. 0.08x12
546x291x0.6x2
130x291x10x4
130x291x84x3
101x291x84x6
101x291x24x12
240x291x24x24
0.030
0.19
1 .5^
9.52
14.8
8.47
40.2
0.018
0.023
0.75
4.76
9.52
5.45
20.1
0.012
0.17
0.75
4.76
5.27
3S\ f\
.02
f\ f\ ^ / * i j\ \
20.1 (HC)
14.1
20.1
CONTROL EQUIPMENT RELATED
ESP
1.
2.
3.
4.
5.
6.
7.
8.
Cold start
Dust reentrained
Dust overload
High resistivity
Dust overload, shut down
ESP shut down
Transformer-rectifier
failure
Screw conveyor breakdown
956x291x1x50
956x291x0.08x104
Assumed not to occur,
129x291x24x350
100x291x1x50
956x291x1x50
100x291x72x1
614x291x4x7
13.9
2.3
this plant
314
1.5
13.9
2.1
5.0
1.0
0.1
159
1.0
1.0
1.4
0.5
12.9
2.2
155
0.5
12.9
0.7
4.5
9. Dust bridging
Assumed not to occur, this plant
-------
TABLE 8. (cont'd)
AOC 10"°
10. Loss of chamber
11. Broken electrodes
12. Abnormal rap
13. Reduced voltage
14. Excess oil, hydrocarbons
Total annual increase, Particulates
HC
Total Emissions
, Mg/yr
[RatexMg/hrxDuration(hr)xTimes/yr]
122x291x48x2
100x291x48x3
100x291x168
150x291x48x24
330x291x48x12
3.4
4.2
4.9
50.3
55.3
a
Equivalent
Controlled
Emissions
Mg/yr
1.9
2.8
3.2
21.8
20.1
Net Increased
Emissions
Mg/yr
1.5
1.4
1.7
28.5
35.2 (HC)
221.8
35.2
PRODUCT HANDLING
CONTROL EQUIPMENT RELATED
ooBaghouse
1.
2.
3.
4.
5.
Total
Excessively open hoods
Bypassing, cold start
Torn bags
Fan failure
Screen breakdown
increased particulates
c
Percent Increase
700x291x84x2
5000x291x1x50
410x291x168x3
5000x291x50x5
956x291x40x1
Windbox: Particulates 100(24.3 + 221 .8)/240
HC 100(20.1 + 35.2)/443
Product Handling: Particulates 100(534/96)
34.
72.
60.
364
11.
= 102
= 12
= 545
3
7
1
1
1
0
3
1
0
.2°
.4
.6
.9
.76
33.
72.
56.
362.
10.
534.
1
3
5
1
3
3
Mg
Based on a controlled emissions rate of 65 g participate per Mg of strand feed and 120 g hydrocarbons
per Mg of strand feed. The use of the HC standard is indicated by (HC).
3Based on a controlled emissions rate of 25 g particulate per Mg of strand feed.
^
'Percent increase based on controlled emissions for an entire year of 240 Mg for windbox particulates,
443 Mg/yr for windbox hydrocarbons, and 96 Mg/yr for product handling particulates.
-------
TABLE 9. ESTIMATED EMISSIONS AND DISCHARGES FROM ABNORMAL OPERATION
AOC
FURNACE
1. Excessive lime in slag,
bleeders open
2. Dirty gas bleeders open
3. Extended startup
4. Excess slips, startup
5. Bleeders open, shut
down
6. Severe burden slips
^
D 7. Backdrafting
8. Water and power failure
9. Breakouts
10. Charging dusty material
11 . Stove plugging
12. Loss of ignition,
clean gas bleed
13. Formation of carbon
black
14. Unplugging dust catcher
15. Leaks from bells
GENERALIZED STEEL PLANT: TWO BLAST FURNACES
Total Emissions Equivalent
Hg/yr P.rt1c«l.t« %*»™
10 xkg/MgxMg/hrxDuration(hr)xTimes/yr Mg/yr
(4-8)x242x(12-24)xl
5. 5x242x(l 2-24 )xl
0.5x242x(l-4)xl2
5.5x242x15/60x60x6
5.5x242x(16-24)
(37-63)x242x(5-30/3600)x
12-600
0.42x242x(3)x(4-50)xl2
0.5x242x(24-72)x(l-2)
0.5x242x0.5x1
(0. 018-0. 035)x242x5-10/
3600x600x350
(0.21-4.6)x242x21xl
340x242x0 -4 )xl 2
(0. 008-0. 012)x242x24xl20
42x242x0 -8)x(l -8)
0.3x242x24x350
11.6-46
16-32
1.5-5.8
0.03
21.2-32
0.15-76.2
14.6-183
3.9-17.4
0.06
1.3-4.9
0.22-23.3
987-3949
5.6-8.4
10.2-653
610
0.7-1.4
0.6-1.3
0
0
0.9-1.2
0.001-0.3
7.8-97.6
1.3-7.8
0
0
0
Carbon monox-
ide
Hydrocarbons
0.05-3.5
455
Total annual increase, particulates with startup, shut down
HC
without startup
, shut down
Net Increased
Emissions
Mg/yr
10.9-44.6*
15.4-30.7*
1.5-5.8*
0.03*
20.3-30.7*
0.15-75.9
3.5-43
1.6-9.6
0.06
1.3-4.9
0.22-23.3
987-3949
5.6-8.4
10.1-650
155
220-1074
172-962
5.6-8.4
-------
TABLE 9. (cont'd)
en
O
AOC
CASTHOUSE
1. Blow through taphole
2. Hard blow, tap hole,
shut down
3. Iron flush to ground,
shut down
4. Slow cast
5. Cold metal
6. Hot limey slag
7. Wet tap hole, trough,
runner
8. Tap hole enlargement
Total Emissions
Mg/yr Particulates
f\
Equivalent
Controlled
Emissions
10" xkg/MgxMg/hrxDuration(hr)xTimes/yr Mg/yr
0.17x242x(12-48)
0.5x242x(0.25-0.75)
0.1x242x1.5
0.59x242x(0.17-l)x(12-60)
2x0.21x533Mg/castx(50-100)
2x0.5x533x(100-350)
2x0.5x533x4
0.5x242x(0.33-0.67)x4x90
0.5-2.0
0.03-0.09
12
0.3-8.6
11.2-22.4
53.3-187
2.1
14.3-29 5.7-11.7
Total annual increase, particulates with startup, shut down
without startup
, shut down
Net Increased
Emissions
Mg/yr
0.5-2a
0.03-0.09a
»
12b
0.3-8.6
11.2-22.4
53.3-187
2.1
8.6-17.3
88-252
75.5-238
CONTROLS
1. Scrubber problems
WATER TREATMENT
1. Loss of water pump
42x242x40/3600x(0.2-48)x(3-4) 0.07-21.7 0-0.1
SS (5-39)x242(0.5-6)x(12-72) 7.3-4078 0.007-522
Cyanide 0-0.032x242x(0.5-6)x 0 -3.3 0-0.013
(12-72)
Phenol 0-0.013x242x(0.5-6)x 0-1.4 0-0.03
(12-72)
0.07-21.6
7.3-3556
0-3.3
0-1.35
-------
TABLE 9. (cont'd)
Total Emissions Equivalent
Hg/yr Parties ^'rolled
AOC lO'^xkg/MgxMg/hrxDurationthrJxTimes/yr Mg/yr
2. Clari tier rake failure SS 0.25x242x(24-72)x(l-2) 1.5-8.7 0.03-0.2
Cyanide 0.012x242x(24-72)x(l-2) 0.07-0.4 0 -0.004
Phenol 0.007x242x(24-72)x(l-2) 0.04-0.24 0 -0.009
Total annual increase water SS
Cyanides
Phenol
Net Increased
Emissions
Mg/yr
1.5-8.5
0.07-0.4
0.04-0.23
8.8-3565
0.07-3.7
0.04-1.58
PERCENT INCREASE
Particulates:
Suspended Solids:
Cyanides:
Phenols:
100(172 + 76)/464
100(8.8/0.037)
100(3.7/0.017)
100(1.6/0.027)
53; 100(962 + 21.6 + 238)/564
24000; 100(3563/522)
22000
5926
216
683
aStartup: once in 3-4 years
Shut down: once in 3-4 years
-------
TABLE 10. ESTIMATED EMISSIONS AND DISCHARGES FROM ABNORMAL OPERATION
GENERALIZED STEEL PLANT: TWO VESSEL BOP SHOP (ONE OPERATING)
OPEN HOOD USING A VENTURI SCRUBBER
FABRIC FILTER APPLIED TO ANCILLARY OPERATIONS
Total Emissions
AOC
c
10"° [Rate
Mg/yr Parti culates9
x Mg/heat x
Duration x Times/yr]
Equivalent
Controlled
Emissions
Mg/yr
Net Increased
Emissions
Mg/yr
PRIMARY VESSEL EMISSIONS
Process
1. Burn in
2. Vessel dump on shut down
3. Puffing at hood
4. Improper ladle to vessel
transfer
Relief damper opening
en
r\i
6. Foaming and slopping
8. Improper charge material
9. Running stopper
10. Insufficient draft on
startup
13. Damper stuck or jammed
14. Power failure
Total Process
Control-Venturi Scrubber
1. Rake failure
2. Sprays corroded or
plugged
1.5-1.7(218)(6-24) =
2.0-8.9 HXCX
1.0(.05-1.0)(218)(2)(350)= 7.6-152.6
0.26(1-3)(218)(1)(350) = 19.8-59.4
20(0.02-1.0)(218)(1-10)(12)=1.0-532.2
4.0(0.05-0.25)(218)(2-54)
(52) = 4.5-612.1
0.13(1)(218)(2)(52) = 2.9
0.1(1)(218)(3)(12) = 0.8
6.7(180)(lheat/45min)(218)
(0-3) = 0-17.5
2.0(60-1440)1/45(218)(2-12)= 1.2-167.4
(10-20)(0.5)(1/45)(218)
(0.2-3) = 0-0.1
2.0-8.9 HXCX
20(1440-4320)(1/45)(218)
(0-2) = 0-837.1 0.2
1(60-420)(1/45)(218)(6-156)= 1.7-317.4 0.05-8.9
0.2-4.3
0.05-0.15
0-0.73
0.03-4.3
0.015
0
0-0.07 '
0.02-2.3
0
7.4-148.3
19.8-59.3
1.0-522.5
4.5-607.8
2.9
0.8
0-17.4
1.2-165.1
0
36.8-1531.3
836.9 (SS)
1.6-308.5
-------
TABLE TO. (cont'd)
AOC
3.
4.
5.
Plugged or
corroded
pipes
Pump failure
Plugged or failed
Total Emissions
Mg/yr Parti culates
1(180)(1/45)(218)(6) =
2(120-480)(1/45)(218)(6)=
2(4320)(1/45)(218){1) =
5.2
7.0-27.9
41.9
Equivalent
Controlled
Emissions
Mg/yr
0.15
0.10-0.39
0.59
Net Increased
Emissions
Mg/yr
5.0
6.9-27.5
41.3
demister
6. Drum filter failure
8. Unbalanced water level
Total Control
SECONDARY EMISSIONS
System Failure
1-2. Charging & Tapping
3. Hot metal transfer
4. Flux handling
Bag Breakage & Plugging
1-2. Charging & Tapping
3. Hot Metal
Shaker or Reverse Air Failure
1-2. Charging & Tapping
3. Hot Metal
NO SUITABLE BASIS FOR ESTIMATE
0.23(20-18720)(1/45)(218)
(30) = 1.0-625.7
0.25-0.35(60-5760)0/45)
(218)(18-47) = T.3-459.0
0.75(l)(218)(l-8) = 0.2-1.3
0.0003(780-5760)0/45)
(218)(3-6) = 0-0.05
0.0005(780-5760)0/45)
(218)(3-6) = 0-0.8
0.023(120-1020)0/45)
(218)(4-6) = 0.1-0.7
0.035(120-1020)0/45)
(218)(4-6) = 0.1-1.0
0.01-3.1
t
0.01-2.3
0-0.01
0.01-0.19
0.01-0.29
0-0.03
0-0.05
54.8-1219.2
1.0-622.6
1.3-456.7
0.2-1.3
0
0.5
0.1-0.67
0.1-0.95
-------
TABLE 10. (cont'd)
AOC
Open Bypass Damper
1-2. Charging & Tapping
3. Hot Metal
Total Emissions
Mg/yr Parti culates
0.23(480)(1/45)(218)(1) = 0.5
0.25-0.35(480)0/45)
(218)0) = 0.6-0.8
Equivalent
Controlled
Emissions
Mg/yr
0
0-0
Net Increased
Emissions
Mg/yr
0.5
0.6-0.8
Dust Removal Breakdown
1-2. Charging & Tapping
3. Hot Metal
Total Secondary Emissions
PERCENT INCREASE
Particulates:
Suspended Solids:
10-100% of (0.23) (60-
480)(1/45)(281)(1)
10-100% of (0.23)(60-
480)(1/45)(218)(1)
= 0.01-0.53
= 0.01-0.58
0-0
0-0
100(95.9/847) = 113; 100(2999/95) = 3157
100(836.9/0.2) = 418,000
0.01-0.53
0.01-0.58
4.32-1085
Emissions for BOP Shop are appropriately calculated here on a per-heat basis using 218 Mg per heat,
which is equivalent to the 292 Mg/hour production given in Table 6.
-------
to calculate the total annual emissions during AOC's for each of the model
plant's processes. For clarity, the calculations used for each AOC are pre-
sented in Tables 8, 9, and 10. As a means of demonstrating the potential net
effects of AOC's on the environment, equivalent controlled emission levels
have also been calculated. The net potential increases in emissions due to
AOC's are presented in the last column as the difference between AOC and
controlled emissions.
The basis for calculation estimates of controlled emissions was established
by considering New Source Performance Standards, Effluent Guideline Limitations,
proposed or discussed potential standards, and outlet concentrations achievable
with high efficiency control devices as discussed in the following paragraphs.
Controlled sinter plant windbox particulate emissions were set at 65 g/Mg
of strand feed; hydrocarbons, at 120. Product handling controlled particulate
emissions were also set at 65 g/Mg.
Controlled blast furnace particulate emissions were set at 224 g/Mg,
based on 0.11 g/scm of cleaned gases given in the Effluent Guidelines. Water
discharge limits were set at 0.005 kg/Mg suspended solids, 0.13 g/Mg cyanides,
and 0.26 g/Mg phenols, based on the best available technology economically
achievable.
For the BOP, there is a New Source Performance Standard of 50 milligrams
per dry standard cubic meter. Since it is stated as a concentration, a
variation in the amount of gas produced by the process causes the allowable
maximum emission rate to vary. The model plant uses an open hood capture
system tied to a venturi scrubber. Based on an uncontrolled particulate rate
of 20 kg/Mg of raw steel and a dry gas rate of 102 standard cubic meters per
second, the controlled emission rate is 0.028 kg/Mg of raw steel. For dis-
charged wastewater suspended solids, the current Effluent Guideline is 0.0101
kg/Mg of raw steel (30 day average).
No emission standards presently exist for secondary processes, i.e.,
charging, tapping, hot metal transfer, and flux handling. For these opera-
tions a controlled emission rate was calculated on the basis of a control
55
-------
device efficiency of 99.5 percent, which is readily achievable. The uncon-
trolled emission rates were reported previously in the discussion pertaining
to Table 3.
6.1 ESTIMATED INDIVIDUAL EFFECTS OF AOC, MODEL PLANT
Tables 8, 9, and 10 show the estimated increased annual discharges for
each AOC considered in assessing the model plant.
6.2 DISCUSSION
The interpretation of the results relates, of course, to the generalized
steel plant. For the sinter plant windboxes, AOC's increased annual parti-
culate emissions by an estimated 102 percent, and hydrocarbon emissions by 12
percent. Particulates emitted from product handling showed an estimated 545
percent increase.
For the blast furnaces, two levels of AOC were considered for which the
estimated increased particulate emissions ranged from 53 to 216 percent of
controlled emissions. Estimated suspended solids increase ranged from 680 to
24000 percent; cyanides, 22000 percent; phenols, 5900 percent.
For the BOP process, two levels of AOC were also considered, so that
estimated increased particulates ranged from 113 to 3000 percent; suspended
solids, 0-418,000 percent.
Based upon this analysis, AOC's add an additional annual air pollutant
emissions load estimated at from half to several times the load from non-upset
operation. Where water treatment is required, AOC's add additional discharges
estimated at substantial levels in excess of standards.
56
-------
7.0 COST OF PREVENTING OR MINIMIZING ABNORMAL OPERATING CONDITIONS
Essentially no cost data relevant to preventing or minimizing AOC's
were acquired under this study. AOC's appear not to have been addressed
in budgets and plans developed to date for pollution control. While entire
new systems have been built partly to remedy an existing problem, the AOC
aspect of current controls have not been identifiably addressed.
Consideration of costs relating to AOC's involves a myriad of factors;
design possibilities and priorities, operating techniques, availability of
capital and regulatory practices are some of the most important. Process
equipment and control equipment should be considered separately, as the opera-
ting philosophies differ markedly between the two.
Process Related AOC's - Costs
It is fairly certain that production equipment is designed to be as
reliable as is economically feasible, at least to the extent of knowledge at
the time. It is also likely that the equipment will be operated to prevent
damage and maximize production. With economics as the primary consideration,
the decisions which must be made have a common basis.
The most important question concerning the cost of process related AOC's
is the extent to which the operator is willing to curtail production in order
to prevent or minimize emissions due to AOC. Many process related AOC's which
cause emissions are also undesirable from an operation standpoint; the AOC's
may be dangerous, cause equipment damage or loss of product, or fill the shop
with smoke. If production is to be curtailed, there is a great deal of economic
incentive for the steel company to prevent AOC's. If not, the economic incen-
tive to prevent AOC's is lacking and other factors, of uncertain impact, dominate.
Many process related AOC's are to some extent controlled by the production
rate of the process. The equipment is under more load and requires more main-
tenance, operations must be more precise, and in general the system is less
able to adjust to deviations from normal conditions.
57
-------
The effect of all these considerations on the cost of preventing AOC's is
difficult to quantify. First a starting point must be identified. This
section assumes that the AOC's associated with a process designed to be
economically reliable is the basis for comparison. If the AOC causes the
process to shut down or curtail production in addition to causing emissions,
preventing the AOC becomes a question of how much must be spent to maintain
operations. Each situation has many solutions, and the criteria for judging
vary from place to place and from time to time.
If the AOC does not directly interfere with the process, preventing the
AOC costs the shop money. Here regulatory pressures and "good citizen" con-
siderations dominate and the results cannot be predicted. Based on visits
during this project, regulatory efforts are still concentrated on achieving
primary control, and effort is spent on AOC's only if the environmental effect
is striking.
Control Equipment Related AOC's
The prevention of control equipment AOC's in most cases costs the steel-
maker money without any benefit to the balance sheet. The reliability of the
control equipment in a shop depends on several factors:
1. the quality of the design information on the job to be done
and on the design philosophy,
2. the steelmaker's willingness to spend capital (and the availability
of capital) at the time the equipment is purchased,
3. the steelmaker's willingness to properly man and operate the
equipment, and
4. the occurrence of changes in the process; i.e., fuels, raw
materials, operating conditions.
Of first importance for control equipment design is a set of good specifi-
cations. The equipment designer must know what the equipment is to collect;
that is, dust loadings, composition, variability, size range, resistivity, gas
rate, gas composition, and similar variables. This information is not always
easily obtained, particularly for an installation that hasn't yet been built.
Bad information will lead to an inadequate design and increased problems with
reliability.
58
-------
The design philosophy is also very important. Compromises must be struck
between the competing desires to achieve the lowest possible capital cost and
lowest operating cost. Generally these goals are mutually exclusive. The
eventual design can be conservative, easy to operate and maintain, and well
instrumented (and expensive) or low in capital cost or anything in between.
Obviously, AOC's are more likely (but not certain) for the stripped down design.
This study provides strong indications that conventionally designed control
equipment will be subject to a sufficient number of AOC's to double at least
the annual emissions therefrom that might be expected on a non-AOC basis.
59
-------
8.0 NEEDS FOR FURTHER RESEARCH AND DEVELOPMENT
8.1 TECHNOLOGY NEEDS
This section of the report addresses the needs for further research and
development (R&D). The need is established by classifying each of the most
serious AOC's according to the type of corrective measure: existing technology
applies or new technology needed. Table 11 provides this categorization.
AOC's are listed here in order of decreasing importance, for those processes
considered in the general example of Section 6. The order may differ from
plant to plant, however the AOC's listed should be important wherever they
occur. Table 11 identifies AOC's for Electric Arc and Open Hearth furnaces,
but does not rank the major AOC's since these processes were not considered in
the general example.
Based upon review of Table 11, further R&D needs were ranked in the
following order: 1} additional quantitative process and control equipment
data, 2) blast furnace bell leaks, 3) control equipment by-pass at start-up,
4) reliability of water recycle systems, 5) external desulfurization of iron,
6) BOP process upsets, 7) better emissions factors, 8) sinter plant HC control,
and 9) vacuum filters.
8.2 STATE-OF-ART APPLICATIONS AND COSTS
As shown in Table 11, there is considerable need to consider application
of existing technology. For this reason, a three-year program of study is
recommended to determine what can be accomplished by use of state-of-the-art
design technology and to provide adequate cost information. An estimated two
man-year per year level of effort plus testing would be required.
In addition to this program, several areas of needed research are evident.
These are discussed in detail in the following sections.
fin
-------
TABLE 11. CORRECTIVE MEASURES FOR ABNORMAL OPERATING CONDITIONS
Process
SINTER PLANT WINDBOX
Process Related
Control Eq. Related
Product Handling Control
j Eq. Related
BLAST FURNACE
Furnace
Desulfurization
Casthouse
Scrubber
% of Existing
AOC Technology
AOC Emissions Applied
Grate bar distortion
Excess air leaks
Poor Sintering
Excessive HC in feed
High resistivity sinter to ESP
Reduced voltage on ESP
Cold start— P/C off at start up
P/C shut down before process
Excessive HC, oil
Fan failure
P/C bypass, start up
Torn bags
Excessively open hoods
Bell leaks
Bleeders open (shut down)
Unplugging dust catcher
Bleeders open (starters)
External desulfurization
Cold metal
Hot limey slag
Tap hole enlargement
Operating problems
34 x
34 x
21 x
100 (HC) x
70
13
6
6
100 (HC) x
68 x
13
11 x
6 x
90
12 x
6 x
9
<.
100
15 x
70 x
11
100
Research/
Development
Needed
x
x
x
x
x
x
x
x
x
x
x
x
-------
TABLE 11. (cont'd.)
Process
Water Treatment
BASIC OXYGEN FURNACE
Process
^
} P/C, Scrubber and
Water Treatment
Secondary
% of Existing
AOC Technology
AOC Emissions Applied
Loss of water pump
Clarifies rake failure
Improper ladle-vessel transfer
Puffing at hood
Foaming and slopping
Burn-in
Plugged demister
Pump failure
Plugged pipes
Rake failure
Charge-tap
Hot metal transfer
Open bypass damper
Baq breakage plus hot metal
83 (SS) x
90 (CN)
85 (Phenol)
17 (SS) x
10 (CN)
15 (Phenol)
54 x
20
12 x
100 (HC)
75 x
13 x
9 x
100 (SS) x
23
30
26
12 x
Research/
Development
Needed
x
x
x
x
x
x
x
x
-------
TABLE 11. (cont'd.)
Process
AOC
Existing
Technology
Applied
Research/
Development
Needed
ELECTRIC ARC FURNACE
Control Equipment
OPEN HEARTH FURNACE
CJ>
Co
Control Equipment
Burn-in
Capture duct misalignment
Running stopper
Ladle breakout
Relief damper opening
Poor furnace operation
Stack puff
Bag failure
Start up
Plugged checker
Furnace puffs
Ladle reaction
Waste-heat boiler failure
Poor furnace operation
Control bypassed
ESP malfunction
Plugged demister
Rake failure
Loss of pumps
x
x
x
x
x
x
x
x
x
x
x
x
x
x
-------
8.3 ADDITIONAL QUANTITATIVE DATA, AOC'S
In the course of this study, numerous visits and calls were made to
pollution control agencies to obtain data on steel making AOC's. Most of the
jurisdictions visited or called had some regulation related to reporting of
malfunctions, breakdowns, startups, and shut downs of pollution control equip-
ment. Beyond that the similarity ends. The procedure for reporting varied
greatly as did the procedures for recording and retaining data.
Some of the variation in reporting and recording is due to differences in
the regulations. Country-wide NPDES requirements make it necessary for all
companies discharging water pollutants in excess of their effluent guideline
limitations to report these upsets within five days of the occurrence. The
nature, cause, and duration of the occurrence must be reported. This system
produces data because the companies are required to monitor their outfalls.
No counterpart of this system exists in the case of air pollutant emissions.
In the long term, it may be possible to produce this kind of data because
continuous monitoring is required for new sources. However, at present most of
these sources are existing and not subject to this requirement. The air pollu-
tant AOC data is gained primarily through self-reports by industry or chance
observations by local enforcement personnel. This means the data is less
consistent from place to place, varying according to the staff competence,
level of staffing, and conscientiousness of the companies within their juris-
diction.
It is proper to point out that many of the local pollution control agencies
have their hands full with efforts to obtain compliance from continuously
emitting sources. Those in heavily industrialized areas, typical for steel
plant locations, were especially busy in this regard. As a result little time
is left to deal with reports of abnormal operating conditions. Nevertheless,
at least one agency in this position does find or take the time to carefully
document the occurrences and retain the records in such a form as to make
review or study a relatively simple matter. Given these records it is possible
to review and identify sources where chronic AOC's are contributing signifi-
cantly to increased emissions and where remedial action is in order.
64
-------
The Air Pollution Control Agency for Erie County, New York is the agency
referred to. Their Law, Rule 10, pertaining to operation and maintenance of
pollution control equipment is as follows:
RULE 10 OPERATION AND MAINTENANCE
10.1 PERFORMANCE
Historical Note: Section added and filed March 1967 as 6.3.1; Renumbered
October 30, 1974, filed November 4, 1975.
a. Air Cleaning devices shall be selected so as to afford the highest
efficiency or the lowest discharge rate that is reasonable and
practicable. Reasonableness and practicability shall take into
account cost, the air contaminant concentration in the emission
gas stream, particle characteristics and other properties of the
contaminant and of the emission gas stream, and applicable provisions
of this Code.
Historical Note: Section added and filed March 1967 as 6.3.2; Renumbered
October 30, 1974; filed November 4, 1974.
b. All devices used to effect compliance with these Rules shall be
installed, operated and maintained so as to minimize the emission
of air contaminants.
10.2 SHUTDOWN OR BREAKDOWN OF CONTROL EQUIPMENT
Historical Note: New sections 10.2 a, b, c, d; 10.3 added October 30,
1974; filed November 4, 1974.
a. In case of an intended shutdown of any control equipment, the
operator shall notify the Commissioner at least one County work
day in advance of the shutdown.
b. In the case of a malfunction, upset, or breakdown of any operation
or control equipment which malfunction, upset or breakdown causes
or is likely to cause any applicable Air Pollution rule to be
violated, the operator shall notify the Commissioner immediately
and shall undertake immediate actions to remedy the malfunction,
upset or breakdown in as short a time as possible.
65
-------
c. The operator shall, at the time of the notification, or as soon
thereafter as possible, inform the Commissioner of the cause as
well as the estimated duration and amount of emissions associated
with the shtudown, or malfunction, upset or breakdown.
d. If the operator wishes to temporarily continue operating the
source while the control equipment is inoperative or malfunctioning,
he must obtain authorization from the Commissioner. The Comis
si oner may allow such operations for limited time periods if he
determines that it will not seriously endanger the public health,
welfare or comfort. If the Commissioner determines that the
emissions are likely to create an immediate public health hazard
or nuisance, he shall so advise the operator and the operator
shall take immediate steps to abate the emissions by shutdown
or other appropriate actions.
e. If the Commissioner determines that the frequency of breakdowns
is excessive, he may direct a source owner to take actions which
reduce the frequency of breakdowns.
10.3 COMPLIANCE
If and only if the operator completely complies with all provisions of
this Rule, emissions in excess of those allowed in the other Rules of
this article shall not be considered violations.
Passed by Board of Health: October 30, 1974
Effective December 1, 1974
The form shown in Figure 1 is used by the Agency to record the reports.
Part of the data is entered into a computer file and the hard copy is filed.
The data retained in the computer file includes the facility identification
number, facility name, source number, source name, date of the occurrence,
time, estimated percent collector efficiency loss, the type of collector, and
the length of the occurrence. If the cause of the AOC is the needed datum, it
may be found on the file copy of the report.
Obviously not all agencies can justify the use of a computer to retain
these data. Those with a large number of industrial sources, however, may find
it useful and, in the long run, time saving as more effort is put into sur-
veillance activity. The important thing is that the data are recorded and kept
on file in a readily accessible manner.
66
-------
ERIE COUNTY AIR POLLUTION CONTROL DIVISION
BREAKDOWN, SHUTDOWN OR STARTUP CARD
NAME OF COMPANY
AND DEPARTMENT
DATE
TIME
TELEPHONE NUMBER
en
NAME & TITLE OF
PERSON REPORTING
SOURCE NO. (if known)
AHD DESCRIPTION OF SOURCE
HOW LONG IS BREAKDOWN
EXPECTED TO LAST?
CAUSE AHD DESCRIPTION
OF BREAKDOWN
ESTIMATED LOSS OF EFFICIENCY?
CALL I ) y»s
BACK ( ) no TIME
TIME.BREAKDOWN
STARTED
TYPE (
1.
2.
3-
4.
5.
6.
)F EQUIPMENT
Baghousa
Scrubbar
Cycl one
Preci pit.
Aftrbrnr.
Other
.
PERSON TAKING CALL
INSPECTOR
SUPERVISOR
Sonify DATE ENTERED IN RECORD
TOTAL DURATION:
FACILITY DUMBER:
A?C-IO (9/74)
SOURCE NUMBER:
Figure 1. Breakdown, shut down or startup card.
-------
Some agencies require no report of breakdowns or malfunctions unless they
exceed some minimum duration for example four or eight hours. Other accept
telephone reports and keep no record of the particulars. Any program aimed at
reducing the emissions caused by AOC's must of necessity maintain a data file
of some sort with which to compare the frequency and duration of future occur-
rences to determine whether progress is being made.
8.4 ADDITIONAL QUANTITATIVE DATA, FUGITIVE EMISSIONS
At the beginning of the study there was a question as to whether continous
discharge of fugitive emissions should be considered an abnormal operating
condition. Eventually it was decided not to include the continuous fugitive
emissions unless there was an increase in emissions resulting from unusual
handling or processing. The difference between the two conditions, however, is
not clear because few measurements have been made in either case. There are a
few plants in the United States with control devices applied to these secondary
(previously fugitive) emission sources. The existence of plants with controls
on the secondary sources presents an opportunity to develop and substantiate
emissions factors. The areas in which better definition of emission factors is
needed include charging, tapping, hot and cold metal handling, hot metal
desulfurization, teeming, slag handling, and flux handling in the steelmaking
shops (open hearth, electric arc, and BOP). An estimated $80,000-200,000
effort would be required.
8.5 CONTROL FOR MAJOR AOC'S
Sinter Plants
The control of hydrocarbon emissions from sinter plant windbox gases
requires further investigation. New technologies for this purpose include
windbox gas recycling, the wet ESP, and scrubbers. Other devices under con-
sideration include the steam-hydro scrubber and the venturi scrubber-incinerator
combination, both high cost technologies. States have set hydrocarbon emis-
sions limitations at levels which challenge and may defeat these technologies.
68
-------
A performance-design study should be conducted, to identify the potential
capabilities of these technologies for this application. The level of effort
is estimated at $70,000-200,000 depending on the amount of experimental work.
Blast Furnaces
In the Blast Furnace Process Manual, Volume 3 of the report from this
project, the issue of leaks from the blast furnace bells was addressed. Some
limited information available suggest that leaks from blast furnace bells may
represent a substantial emission source.
These sources might be significant enough to affect the ability of some
air quality control regions to achieve ambient air standards. The task force
from AISI reviewing the work of this study question the validity of the esti-
mates of blast furnace bell leaks. In view of these questions and the fact
that the information is not substantiated, a program of research to study this
topic is appropriate. Such a study might include an attempt to produce an
accurate gas balance around the furnace. Because the estimated quantity of gas
lost at the bell represents only about 1 percent of the total gas generated for
any given furnace it would probably also be necessary to find a measuring
technique to apply directly to the leak source. Study of this problem might
proceed in two stages: (1) methodology development and (2) tests and measure-
ments. An initial effort in methodology would require some measurements and
should be considered at the $100,000-200,000 level.
8.6 CONTROL DEVELOPMENT FOR LOW R&D INVESTMENT
Control Equipment Delayed Start Up and Early Shut Down
The need to have a warmup period for precipitators and fabric filters
applied to several processes presents a control equipment oriented startup problem.
Prevention of moisture condensation in the control devices is the reason a
warmup period is recommended by equipment suppliers. The mixture of condensed
water and particulate produces a layer or cake that leads to fabric blinding or
deposits that cannot be removed from collecting plates. In addition, condensa-
tion on precipitator insulators (insulators used to steady the weighted wire
discharge electrodes) provides a conductive path across the insulator that
leads to insulator cracking or breaking.
69
-------
An experimental program is needed to develop equipment or techniques that
would allow the use of these devices under startup conditions. If not com-
pletely eliminated, perhaps some way of shortening the warmup period may be
identified. Preheating, precoating of fabric, reduced voltage energization,
and intensive rapping are all potentially valuable steps to solving the startup
problem.
Unless the startup problem can be solved, the number of high efficiency
control devices available for use at steel plants will be severely limited.
Scrubbers and wet electrostatic precipitators do not require warmup periods,
but have other drawbacks such as high energy consumption and conversion of
potential air pollution problems to potential water pollution problems. For
instance, precipitators are not now a good choice for controlling particulate
from sinter strands producing high basicity sinter; but one would not want to
eliminate the fabric filter option (considering their inherently high efficiency
capability) because of the warmup problem. A three year study of this major
AOC problem at an estimated $75,000-150,000 per year is recommended.
Rotary Drum Filters
Problems with rotary drum vacuum filters were alluded to in some of the
plants visited during this study. No data were found or provided on the
frequency and duration of AOC's related to this equipment item, yet some
comments indicated it to be a significant problem. The consequences of such
occurrences were predicted to be land disposal of water sludge (perhaps
resulting in fugitive runoff) or increased suspended solids content in the
wastewater system blowdown. Further investigation of this area is suggested.
A one year, $20,000-50,000 level of effort is estimated to be required.
8.7 COST REDUCTION FOR PRESENTLY AVAILABLE BUT INORDINATELY EXPENSIVE
CONTROLS FOR MAJOR AOC'S
Good design practice in pollution control system installations can mini-
mize the occurrence of some of the AOC's described in this study. Two areas
that might specifically benefit from the definition of good design practice
70
-------
are recycle wastewater systems and air pollution control systems for new
steelmaking technology, i.e., Q-BOP and top blown BOP with suppressed combus-
tion.
Recycle wastewater systems bring with them the need to consider how to
achieve long term reliability. Redundant or spare capacity may be provided in
the moving parts of the system. But, in addition, it is also necessary to con-
sider those things that typically cause failures such as scaling, corrosion,
and abrasion. These problems can be dealt with through improved selection of
materials of construction and/or through steps to reduce the severity of attack.
Development of guidelines for a thorough review of the system components and
their adequacy would be a useful approach. A one year effort at $65,000-120,000
would be required.
Suppressed combustion BOP fume capture systems are more susceptible to
process upsets that exceed the main fume system's capability to capture.
Secondary systems are a virtual necessity to avoid violation of air pollution
regulations. The secondary systems are relatively new and a more uniform
approach to their design is needed. The bottom blown or Q-BOP process requires
continuation of nitrogen and/or oxygen flow during the period the vessel is
turned down, thus increasing its fume generation potential by comparison to
the top blown process. Since this process is expected to gain wide acceptance
in the industry, guidelines for the design of the capture and collection systems
are essential to preventing installation of underdesigned systems. An estimated
$55,000-225,000 level of effort would be required.
71
-------
9.0 REFERENCES
1. Communication with Bethlehem Steel Company.
2. Sinter Plant Air Pollution Control Pilot Plant Study at Houston Works,
Armco Steel Co., 1972.
3. Armco Steel Company, Sinter Plant Pilot Scrubber Study at Ashland Works,
August 9, 1971.
4. Varga, J. J., Control Reclamation (Sinter) Plant Emissions Using ESP's,
EPA-600/2-76-002, January 1976.
5. Jaasund, S. A. and M. R. Mazer, Application of Wet Electrostatic
Precipitators for Control of Emissions from Three Metallurgical Processes,
EPA Symposium on Particulate Collection Problems Using Electrostatic
Precipitators in the Metallurgical Industry, June 1-3, 1977.
6. Mazer, M. R., S. T. Hernan, and S. A. Jaasund, Adaptation of Wet Electro-
static Precipitators for Control of Sinter Plant Windbox Emissions,
Bethlehem Steel Corp., Bethlehem, Pa.
7. Discussions with Air Pollution Industries, Englewood, New Jersey,
December 1, 1977.
8. Wilson, S. W. and 6. K. Jefferson, "Operating Experience of No. 13
Blast Furnace, Gary Works," Ironmaking Proceedings, Vol. 36, AIME,
1977.
9. Jablin, R., B. H. Carpenter, D. W. Coy, and D. W. VanOsdell, Pollution
Effects of Abnormal Operations in Iron and Steel making. Volume 3:
Manual of Practice—Blast Furnace Ironmaking." Final Report prepared
for the Environmental Protection Agency, Research Triangle Park, N.C.,
January 1978.
10. Trip Report, Republic Steel, Gadsden, Alabama, July 5-7, 1977.
11. Discussions with Chemico Air Pollution Control Co., New York, N. Y.
12. Steiner, B. A., Air Pollution Control in the Iron and Steel Industry,
International Metals Review, September. 1970, pp. 171-192.
13. Varga, J. J. Control Reclamation (Sinter) Plant Emissions Using ESP's,
EPA-600/2-76-002, January 1976.
14. U.S. Environmental Protection Agency, Effluent Guidelines.
15. McGannon, The Making, Shaping & Treating of Steel, 9th Edition, p. 457.
72
-------
16. Parle, R. W., "Long-Term Shutdown and Subsequent Recovery of Blast
Furnace Plant," Journal of the Institute of Fuel. June 1972.
17. Lindau, Lars and Lars Hansson, "Fugitive Dust from Steel Works," The
National Swedish Environment Protection Board, Bo Mansson, Stoft Tekniska
Laboratoriet AB, Sweden.
18. U.S. Environmental Protection Agency, Effluent Guidelines.
19. Information supplied by the U.S. Environmental Protection Agency, Region
V, Chicago, Illinois.
20. Data obtained through visits to various plants in connection with this
project.
21. Erie County Air Pollution Control Agency, Buffalo, New York.
22. Allegheny County, Pa. Health Department, Pollution Control Division.
23. Mobley, C. E., A. D. Hoffman, and H. W. Lownie, "Blast Furnace Slips
and Accompanying Emissions as an Air Pollution Source," EPA-600/2-76-
268.
24. Bradley, J. G., "Operation and Maintenance of a Modified O.G. Gas
Cleaning System," National Open Hearth and Basic Oxygen Steel Conference,
55th Proceedings. Metallurgical Society, AIME, April 10-12, 1972, pp. 305-
311.
25. Data obtained from visits to various plants with BOP shops in connection
with this project.
26. Data obtained from the literature.
27. Estimate.
28. Background Information for Standards of Performance: Electric Arc
Furnaces in the Steel Industry; Volume I: Proposed Standards, Emissions
Standards and Engineering Division, U.S. EPA, EPA-450/2-74-017a, October
1974.
29. Flux, J. H., "The Control of Fume from Electric Arc Steelmaking," Iron
and Steel International, June 1974, pp. 185-192.
30. Kaercher, L. T. and J. D. Sensenbaugh, "Air Pollution Control for an
Electric Furnace Melt Shop," Iron and Steel Engineer. May 1974, pp. 47-51.
31. See Reference 17.
32. Weber, E., "Treatment of Waste Gases from the Basic Oxygen Furnace in
West Germany," Steel Industry and the Environment, Proceedings of the
Furnas Memorial Conference, 2 nd, SUNY, Buffalo, New York, 1971. Published
by Marcel Dekker, Inc., pp. 255-247.
33. Williams, D. B., "Fume Cleaning at the BOS Plant, BSC, Port Talbot,"
Publication No. 128, Iron and Steel Institute, London, 1970, pp. 75-80.
73
-------
34. Air Pollution Aspects of the Iron and Steel Industry, U.S. Department
of Health, Education, and Welfare, Cincinnati, Ohio, 1963.
35. Stravinskas, J. and Deamer, M.R., "Influence of Operating Variables on
BOF Yield," Iron and Steelmaker, May 1978, pp. 33-37.
74
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-78-118a
2.
3. RECIPIENT'S ACCESSION NO.
4. T.TLE AND SUBTITLE Pollution Effects of Abnormal Oper-
ations in Iron and Steel Making - Volume I. Technical
Report
5. REPORT DATE
June 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
B.H.Carpenter, D.W.VanOsdell, D.W.Coy, and
R.Jablin
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.
68-02-2186
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 COVERED
Final: 10/76-1/78
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES IERL-RTP project officer is Robert V. Hendriks, Mail Drop 62,
919/541-2733.
repOr^ jg j-^g first in a six-volume series considering abnormal oper-
ating conditions (AOCs) in the primary section (sintering, blast furnace ironmaking,
open hearth, electric furnace, and basic oxygen steelmaking) of an integrated iron
and steel plant. Pollution standards, generally based on controlling discharges du-
ring normal (steady-state) operation of a process and control system, are often
exceeded during upsets in operation. Such periods of abnormal operation are becom-
ing recognized as contributing to excess air emissions and water discharges. In
general, an AOC includes process and control equipment startup and shutdown,
substantial variations in operating practice and process variables, and outages for
maintenance. The volume evaluates the magnitude of pollutants emitted during AOCs,
Compared to normal controlled emission rates from the processes, the increases
due to AOCs are estimated to be significant. The volume describes the methodology
used to gather data for the study and sources of information. Numerous pollution
control agencies and manufacturing plants were visited. Though most jurisdictions
have regulations requiring reporting of spills, malfunctions, etc. , there is a wide
variation in the procedures and records kept. Without systematic recordkeeping,
it is difficult to determine the causes of problems and identify corrective action.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Pollution Oxygen Blown Con-
Iron and Steel Industry verters
Abnormalities Blast Furnaces
Failure Openhearth Process
Starting Electric Arc Fur-
Shutdowns naces
Sintering Steel Making
Pollution Control
Stationary Sources
Abnormal Operations
Basic Oxygen Process
13B
11F
13H
13H
13A,13A
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
83
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
75
------- |