United States Off ice of Enforcement EPA-68-01-4143
Environmental Protection Office of General Enforcement October 1979
Agency Division of Stationary Source Washington, DC 20460
Enforcement
v>EPA Instructional Manual
for Clarification of
Startup in Source
Categories Affected
by New Source
Performance Standards
-------
GCA-TR-79-33-G
U.S. ENVIRONMENTAL PROTECTION AGENCY
Washington, D.C. 20460
EPA Project Officer
John Busik
Division of Stationary Source
Enforcement
401 M Street, S.W.
Washington, D.C. 20460
EPA Task Manager
Robert Myers
Division of Stationary Source
Enforcement
401 M Street, S.W.
Washington, D.C. 20460
Contract No. 68-01-4143
Technical Service Area 1
Task Order No. 62
INSTRUCTIONAL MANUAL FOR CLARIFICATION
OF STARTUP IN SOURCE CATEGORIES
AFFECTED BY NEW SOURCE
PERFORMANCE STANDARDS
Final Report
October 1979
by
Douglas R. Roeck
Peter H. Anderson
Philip S. Hincman
Robert G. Mclnnes
William F. Ostrowski
-------
DISCLAIMER
This Final Report was furnished to the Environmental Protection
Agency by CCA Corporation, GCA/Technology Division, Burlington Road, Bedford,
Massachusetts 01730, in fulfillment of Contract No. 68—01-4143, Technical
Service Area 1, Task Order No. 62. The opinions, findings, and conclusions
expressed are those of the authors and not necessarily those of the Environ—
mental Protection Agency. Mention of company or product names is not to be
considered as an endorsement by the Environmental Protection Agency.
-------
ABSTRACT
L New Source Performance Standards promulgated for 27 source categories
specify that performance testing shall be conducted within certain time
periods of startup for each affected facility. This manual discusses initial
8tartup for each new facility subject to these regulations and provides the
technical basis for uniform application of the regulations pertaining to
source testing.
iii
-------
CONTENTS
Abstract
Figures vii
1. Introduction and Summary 1
Introduction . 1
Summary 2
2. Industrial Summaries 4
Source Listing 4
Fossil Fuel—Fired Steam Generators — Subpart D
§60.40 — 60.46 6
Incinerators — Subpart E
§60.50 — 60.54 10
Portland Cement Plants — Subpart F
§60.60 — 60.64 14
Nitric Acid Plants — Subpart C
§60.70 — 60.74 20
Sulfuric Acid Plants — Subpart H
§60.80 — 60.85 25
Asphalt Contrete Plants — Subpart I
§60.90 — 60.93 29
Petroleum Refineries — Subpart J
§60.100 — 60.106 35
Storage Vessels for Petroleum Liquids — Subpart K
§60.110 — 60.113 42
Secondary Lead Smelting — Subpart L
§60.120 — 60.123
Secondary Brass and Bronze Ingot Production Plants —
Subpart M §60.130 — 60.133 48
Iron and Steel Plants — Subpart N
§60.140 — 60. 144 53
Sewage Treatment Plants — Subpart 0
§60.150 — 60.154 57
Primary Copper Smelters — Subpart P
§60.160 — 60.168 63
Primary Zinc Smelters — Subpart Q
§60.170 — 60.176 69
Primary Lead Smelters — Subpart R
§60.180 — 60.186 74
Primary Aluminum Reduction Plants — Subpart S
§60.190 — 60.195 80
Wet Process — Phosphoric Acid Plants — Subpart T
§60.200 — 60.204 85
V
-------
CONTENTS (continued)
Superphosphoric Acid Plants — Subpart U
§60.210 — 60.214 89
Diammonluin Phosphate Plants — Subpart V
§60.220 — 60.224 94
Triple Superphosphate Plants — Subpart W
§60.230 — 60.234 98
Granular Triple Superphosphate Storage Facilities —
Subpart X §60.240 — 60.244 102
Coal Preparation Plants — Subpart Y
§60.250—60.254 105
Ferroalloy Production Facilities — Subpart Z
§60.260—60.266 108
Iron and Steel Plants: Electric Arc Furnaces
Subpart AA §60.270 — 60.275 113
Kraft Pulp Mills — Subpart BB
§60.280—60.285 116
Grain Elevators — Subpart DD
§60.300 — 60.304 124
Lime Manufacturing Plants — Subpart 1114
§60.340 — 60.344 129
vi.
-------
FIGURES
Number Page
1 Simplified flow diagram of a fossil fuel—fired
steam generator
2 Schematic diagram for a continuous feed municipal
incinerator
3 Conventional dry process cement plant
4 Suspension preheater cement plant
5 Flow diagram for nitric acid production by the
pressure process
6 Schematic diagram of two processes for manufacturing
sulfuric acid
7 Schematic diagram of a typical batch asphalt concrete
plant
8 Flow diagram for catalytic cracking process
9 Flow diagram for a typical three—stage claus sulfur
recovery plant
10 Brass and bronze ingot production facility
11 BOPF Steelmaking process flow diagram
12 Sewage sludge incinerator process diagram
13 Sewage sludge wet air oxidation process diagram . .
14 Flow diagram f or copper smelting and various unit
processes
15 The pyrometallurgical process
16 Lead smelting flow diagram with typical process units
17 Process flow diagram for primary aluminum reduction
21
26
7
. 11
15
16
31
36
39
49
54
58
59
64
71
75
81
vii
-------
FIGURES (continued)
Number Page
18 Flow diagram of the vet—process phosphoric acid
production process 86
19 Superphosphoric acid production by the vacuum
evaporation process 90
20 Superphosphoric acid production by the submerged
combustion process 91
21 Process flow diagram for diammonium phosphate production 95
22 Flow diagram for triple superphosphate manufacturing
process 99
23 Granular triple superphosphate storage plant 103
24 Schematic diagram of the ferroalloy production process . 110
25 Flow diagram for Kraft pulping process operation . . . . . . . 118
26 Process flow diagram for grain terminal elevators . . . 126
27 Process flow diagram for lime calcination and
hydration 130
viii
-------
ACKNOWLEDGIIENT
The authors wish to thank Mr. Robert Myers, the EPA Task Manager, for his
guidance throughout the program. We acknowledge Mr. Norman Surprenant of
CCA/Technology Division, who provided a critical review of this document.
Finally we would like to thank Mr. Paul Exner of CCA for his contribution to
the manual.
-------
SECTION 1
INTRODUCTION AND SU}IMARY
INTRODUCTION
Since 1971, New Source Performance Standards (NSPS) have been promulgated
by the U.S. Environmental Protection Agency for twenty—seven (27) industrial
categories. Section 60.8 of Title 40 of the Code of Federal Regulations (CFR)
specifies conditions for conductance of performance tests for determining
compliance with the regulations. Paragraph(a) states in part that performance
tests are to be conducted “within 60 days after achieving the maximum production
rate at which the affected facility will be operated, but not later than 180
days after initial startup of such facility ...“ Startup Is defined under
Part 60.2, paragraph (o), as “the setting in operation of an affected facility
for any purpose.” The general nature of this definition could result in a
nonuniform interpretation of “startup” by enforcement personnel as applied to
various source categories.
The purpose of this manual, therefore, Is to provide concise, descriptive
summaries of the regulated industries Including all operations and procedures
related to the startup of a new facility. These industrIal “profiles” will
provide a common basis upon which decisions can be made with respect to the
proper time for performance testing. This is an important decision since test-
ing at less than optimum conditions can yield erroneous and unrepresentative
results as well as incurring additional costs upon tl- e source operator.
LIMI TAT IONS
This manual should not be used as a substitute for meeting NSPS. require-
meats as presented in the Federal Register or Code of Federal Regulations, since
the summaries In this manual do not contain all of the requirements a source
must meet in performance testing. Additionally, new standards are nearing
promulgation and current standards are being revised; this manual represents
standards in effect only through March, 1979.
EFFECTIVE DATE
Each source section In this manual contains an effective date. Any source
in that category constructed, reconstructed or modified after the effective
date Is subject to the applicable NSPS.
1
-------
PROJECT APPROACH
Technical information for this manual has been solicited primarily from
process engineers employed at the industries involved. Additional data have
been obtained from trade associations, equipment suppliers or contractors, and
technical journals and publications.
Ai-PLICATION OP FINDINGS
The startup definitions specified for each industry have been geared
towards normal or typical circumstances in which all equipment is delivered
and tested on schedule and no major malfunctions occur during initial startup
or subsequent process evaluation. This manual points out several source cate-
gories where problems might be encountered in meeting the 180 day deadline.
In these instances, enforcement personnel will evaluate the situation on a
case—by—case basis. It must be stressed that not obtaining maximum production
rate within 180 days (as may occur with the kraft pulp mill, nitric acid plant,
or primary aluminum industry, for example) is not sufficient reason for delay
of performance tests; tests could be required during the 180 day period and
again when maximum production is reached.
SUMMARY
Contacts with the various industries involved in each of the NSPS cate-
gories have resulted in many similar circumstances regardless of source type,
which affect new source startup. Items which generally apply to any source
category are:
• The desire to come on—stream as soon as possible so as to
minimize extensive startup periods which would result in
excessive capital expenditures with no ininediate cash flow return.
• Training of operating personnel can be very important if exper-
ienced people cannot be obtained from another plant location.
• Most plants initially undergo mechanical acceptance of process
equipment which is partially carried out by the contractor or
equipment supplier and partially carried out by the source
owner or operator.
• Following mechanical acceptance, process performance evaluation
is conducted, usually in the form of a demonstration or test
run, resulting in the acceptance of the plant from the contractor.
Some specific procedures or operations that may be carried out for par-
ticular pieces of equipment are:
• Water batching of liquid vessels for leak detection and
instrument calibration,
• Gradual firing and curing of equipment containing refractory
material,
2
-------
• Hydrostatic or pressure testing of fossil fuel—fired vessels
according to specific codes, and
• Dry operation of mechanical mixing, granulating, conveying,
and transporting systems.
Other aspects of plant startup which regulatory people need to be familiar
with pertain to the phenomenon of engineering scale. Scale—up problems are
bound to occur since most new plants tend to be larger in capacity than existing
ones and also attempt to incorporate innovative designs relative to energy—
efficiency. New facilities constructed for a known process of nominal design
production rate would likely require much less time than a new plant built for
a prototype process or a much larger plant.
The industrial surveys conducted have resulted in several common suggestions
for definition of an initial plant startup, irrespective of the industry
involved:
1. 24—hours of continuous operation
2. Shipment of on—grade product to the customer
3. Product from process is used to make a profit or is
inventoried.
4. First introduction of raw material with potential
for emission of regulated pollutant(s)
5. Mechanical acceptance of plant
6. Completion of successful demonstration run
7. Contractual acceptance of plant
The selection of startup for each industry has been based upon a composite
of three criteria; the theoretical position of an enforcement agency (No. 4
above), the viewpoint of industrial contacts (all of the above), and the ability
of a source category to achieve rated capacity within 180 days of the selected
startup point.
GOOD AIR POLLUTION CONTROL PRACTICE DURING STARTUP
After the effective date of an NSPS, an applicable source must meet the
standard except during times of startup, shutdown, or malfunction. The source
owner or operator must at all times, including startup, maintain and operate
any affected facility in a manner consistent with good air pollution control
practice (40 CFR 60.11(d)). Hence, atmospheric emissions during the startup
period must always be directed through pollution control equipment.
3
-------
SECTION 2
INDUSTRIAL S1JNM LRIES
SWRCE LISTING
The twenty—seven (27) source categories currently affected by NSPS which
are the subject of this report are listed as follows:
1. Fossil fuel—fired steam generators
2. Incinerators
3. Portland cement plants
4. Nitric acid plants
5. Sulfuric acid plants
6. Asphalt concrete plants
7. Petroleum refineries
8. Storage vessels for petroleum liquids
9. Secondary lead smelters
10. Secondary brass and bronze ingot production plants
11. Iron and steel plants
12. Sewage treatment plants
13. Primary copper smelters
14. Primary zinc smelters
15. Primary lead smelters
16. Primary aluminum reduction plants
17. Wet—process phosphoric acid plants
18. Superphosphorlc acid plants
19. Dianinonium phosphate plants
20. Triple superphosphate plants
21. Granular triple superphosphate plants
22. Coal preparation plants
23. Ferroalloy production facilities
4
-------
24. Steel plants: electric arc furnaces
25. Kraft pulp mills
26. Grain elevators
27. Lime manufacturing plants
The detailed summaries that follow are arranged in the same order as the
preceding list and as they appear in CFR Part 60. Each sui ary is self—
supporting and contains the following sections:
• Introduction — Brief description of equipment and pollutants
regulated and the effective date of the standard.
• Process Description — Discussion of process(es) associated
with each industry with a flow diagram if required for equip-
ment/process clarification.
• Prestartup Operations — Discussion of equipment shakedown
and debugging procedures, time involved, and types of prob-
lems encountered.
• Startup Operations — Definition of best startup points for
each category, time and specific procedures involved, and
duration of operation prior to achieving maximum (or design)
production rate. Also, discussion of any unusual
circumstances.
• References — Listing of industrial contacts, equipment
8uppliers or other technical literature.
5
-------
FOSSIL FUEL-FIRED STEA)1 GENERATORS - SUBPART I)
§60.40 — 60.46
Introduction
The NSPS for this category encompasses fossil fuel—fired or fossil fuel
and wood residue—fired steam generating units capable of operating at greater
than 73 *J (250 x 106 Btu/hr) heat input. Performance standards were pro-
mulgated for nitrogen oxides, particulate matter, sulfur dioxide, and opacity.
Sulfur dioxide and nitrogen oxides are regulated according to the type of
fuel fired, (i.e., gaseous, liquid, or solid fuels). Particulate matter is
limited to 43 ng/J (0.1 lb/b 6 Btu) input. Opacity is limited to 20 percent
except for one 6—minute period/hr during which the opacity cannot exceed 27
percent. Continuous monitoring is required for SO 2 , NOR, and opacity.
Sources constructed, reconstructed or modified after August 17, 1971, are
subject to the standard with one exception; the effective date for the NO
provisions for lignite—fired units is December 22, 1976; all other provisions
apply to lignite—fired units constructed, reconstructed or modified after
August 17, 1971.
Process Descripdon
Fossil—fuel is defined in 40 CFR §60.41(b) as natural gas, petroleum, coal,
and any form of solid, liquid, or gaseous fuel derived from such materials.
These fuels are combusted to create heat in the boilers for the production
of steam. The steam, in turn, is used, in the case of an Industrial facility,
to provide heat and hot water, or to run process equipment, and, in the case
of an electrical utility, to drive multi—stage turbines that produce elec-
tricity for sale to regional power networks. See Figure 1.
Pre—Startup Operations
Prior to startup, certain operations are undertaken to ready the boiler
for service. These “shakedown” procedures are necessary to protect pressure
parts against corrosion, overheating, and thermal stresses; prevent furnace
explosions; to check for leaks; and insure the on—line availability of the unit.
Some of the operations which are included in the pre—startup category
Include:
1. FIlling the boiler and boilout to test components with respect to
temperature, mechanical stresses, corrosion resistance, structural soundness,
warping, gasketing, and expansion joints;
2. Curing of refractory material in the boiler and stack and any coatings
present on heat exchanger surfaces;
6
-------
TO *TMOSPNERE
*rn
STACK
-J
Figure 1. Simplified flow diagram of a fossil fuel—fired steam generator.
-------
3. By—passing of the superheater and turbine until desired steam temper-
ature is reached followed by checking of steam turbine interfacing, controls,
sensors, monitors, load switching, and safety interlocks.
Many precautions are taken during filling of the boiler to protect pressure
parts. High quality water is used to minimize corrosion and scale deposits.
To prevent thermal stresses, the temperature gradient between metal and water
is kept less than 100°F. Higher temperature differentials would limit the life
of pressure parts and if high enough could cause distortion. Mr is completely
purged from the system through vents to limit oxygen corrosion and assure that
all tubes are filled. On drum—type boilers, the glass gauge level should be
about 1 inch of water prior to firing the boiler in order to fill all circulat-
ing tubes.
Boilout is necessary to remove all grease and other deposits from interior
boiler surfaces. It is usually effected with a caustic solution at reduced
temperatures and pressures (as compared to normal operating conditions). Boil—
out also facilitates the slow curing necessary to condition refractory material.
These operations enable the detection of defects in materials, fittings,
and welds which can then be corrected without a loss of on—line availability.
To protect the superheater from overheating, each tube must have sufficient
steam flow to operate properly. A by—pass system is used to accomplish this.
By—pass systems; (a) protect the superheater against shock from water, (b) pro-
vide a means for conditioning water during startup without delaying boiler/tur-
bine warming operations, and (c) reduce temperature and pressure of the steam
leaving the boiler during startup to conditions suitable for turbines and
condensers.
Complete checkout of the superheater and turbine components is thus effected
to ensure that they are completely operational and to detect any defects in
installation.
Estimates obtained from several utility companies indicate that these pre—
liininary operations can take from 2 to 12 months depending on such site—specific
factors as equipment delivery schedules and the extent of any encountered prob-
lems. Industrial facilities would represent the low end of this range while
utility plants would require the longer time periods.
Startup Operations
For fossil fuel—fired steam generators, startup is best defined as the
first time steam is produced by the boiler and used in the case of an indus-
trial facility, to provide heat and hot water, or to run process equipment, and,
in the case of an electrical utility, to drive turbines that produce electricity.
8
-------
REFERENCES
1. Personal Communication with Mr. Smith, Babcock and Wilcox,
February 28, 1979.
2. Personal Communication with Mr. Williams, Stone and Webster,
February 28, 1979.
3. Personal Communication with United Engineers personnel,
February 28, 1979.
4. “Steam its Generation and Use”, Babcock and Wilcox,
161 East 42nd Street, New York, NY 10017, 1975.
9
-------
INCINERATORS - SUBPART E
§60.50 — 60.54
Introduction
The NSPS for this category encompasses incinerators which burn more than
50 percent municipal type waste and are capable of charging more than 45 met-
ric tons per day (50 tons/day). Incinerators handling municipal type waste
are generally referred to as municipal incinerators, although they may be
owned and operated by either a municipality or a private firm. A performance
standard was promulgated for particulate matter which limits emissions to
0.18 g/dscm (0.08 gr/dscf) corrected to 12 percent CO 2 . Sources constructed,
reconstructed or modified after August 17, 1971, are subject to the regulation.
Process Description
Incineration is defined as the process of burning solid waste for the
purpose of reducing the volume of the waste by removing the combustible matter.
While most municipal incinerators have been historically designed solely to re-
duce the volume of the refuse, an increasing number are also recovering and
utilizing the heat generated by this process in the form of steam and electric
production.
The basic components of a municipal incinerator are a) Refuse holding and
Charging, b) Combustion Chambers, c) Mr Supply, d) Residue Handling, and e)
Air Pollution Control Equipment. Refer to Figure 2 for a schematic of a typi-
cal continuous feed installation. While most incinerators regulated by this
standard are continuous feed systems, batch systems will have essentially the
same integration of component systems. Refuse is delivered by truck to a
storage pit, from where it is charged to a feed hopper by means of an overhead
crane. Once in the furnace, the refuse undergoes combustion, in which the
moisture in the refuse Is first evaporated and then the combustible portion Is
vaporized and oxidized. Complete combustion of the refuse is aided by moving
grates, which may be traveling, reciprocating or rocking types. Combustion
Chamber designs may also vary, with rectangular, water walled and rotary kiln
being the more coim n types. Particulate emissions are due to several factors,
including undergrate air velocity, refuse ash content, burning furnace
temperature, grate agitation and combustion chamber design. Of these, under—
grate air velocity has been shown to have the greatest effect. Overfire and
secondary air ports are also provided to increase turbulence and aid in oxi-
dation of the combustible fraction. Residue is discharged from the end of the
grates in quench tanks, from where it is haulea to a sanitary landfill. Com-
bustion gases exiting from the furnace enter a quench chamber or a heat recov-
ery section prior to being vented to an air pollution control device. Wet
10
-------
1
FEED HOPPER
STACK
INDUCED
DRAFT
FAN
UNDERF IRE
AIR
FORCED
DRAFT FAN
ASH TO
$ A N I TAR Y
LANDFILL.
Figure 2. SchematIc diagram for a continuous feed municipal incinerator.
-------
acrubbers, electrostatic precipitatators and to a limited extent, fabric fil-
ters, have all been used to control incinerator particulate emissions. Gases
leaving the control device pass through an induced draft fan before exiting
to the atmosphere.
Pre—Startup Operations
A shakedown of all mechanical and electrical equipment precedes the start-
up of a municipal incineration facility. Chief mechanical components including
overhead cranes, shredders, furnace grate operation, combustion air supply fans,
dampers, induced draft fan, ash handling equipment and all pumps connected with
air pollution control devices and/or boiler feedwater (for units with heat re-
covery) are inspected, adjusted and tested at this time. In addition, all
electrical equipment including instrumentation and process control devices are
“dry” (no load) tested and the primary ranges set on instrumentation. These
checks serve to insure that the equipment is operational and has been correctly
installed. Pre—startup tests such as this are coumon to all mechanical/electri-
cal system installations. The real test of these components, necessarily comes
after facility startup. The last step prior to startup involves the curing of
the refractory in the furnaces and the boilout of the boiler tubes with caustic
for facilities with heat recovery capability. These two steps will occur
simultaneously and take 2 to 3 days. Individual component checks are typically
made immediately after component installation. This entire pre—startup check
lasts 2 to 3 months.
Startup Operations
Startup for municipal incinerators is the first day refuse is fired into
the incinerator furnace. The date may be obtained from facility and/or con-
tractor personnel, as it marks a transition from installation oriented person-
nel to those who service and fine tune the equipment. Most units are initially
run at only 25 percent load for the first several days in order to check all
equipment for proper clearances, inspect all expansion joints and insure that
there is no binding or warping of mechanical components due to thermal expan-
sion. Once the facility is assured that there are no major equipment mating
problems, the refuse feed increases on an incremental basis. During this
time, process flow variables such as furnace and grate temperatures and corn—
bustion air flow rates are monitored and adjusted to give the optimum values.
Since the successful, controlled Incineration of municipal refuse involves
many variables, this fine tuning may take several months.
12
-------
REFERENCES
1. “Systems Study of Air Pollution from Municipal Incineration,” Arthur 0.
Little, Inc. March 1970. APTD 1283, pp. V—25.
2. Personal communication with Mr. Maddigan, Refuse Energy Systems Company,
Saugus, Massachusetts. March 27, 1979.
3. Personal communication with Mr. Kirkpatrick, Nashville Thermal Transfer
Plant, Nashville, Tennessee. March 23, 1979.
4. Personal communication with R. W. Beck Associates Personnel, Denver,
Colorado. March 26, 1979.
5. Personal communication with Babcock and Wilcox Personnel, Canton, Ohio.
March 27, 1979.
6. Personal communication with Dan Schram, Wisconsin Department of Natural
Resources, Madison, Wisconsin. March 26, 1979.
13
-------
PORTLAND CEMENT PLANTS - SUBPART F
§60.60 — 60.64
Introduction
This category applies to the following facilities manufacturing portland
cement by either the wet or dry process, regardless of facility size or
material throughput: kiln, clinker cooler, raw mill system, finish mill sys—
tern, raw mill dryer, raw material storage, clinker storage, finished product
storage, conveyor transfer points, bagging and bulk loading and unloading sys—
tems. Performance standards were promulgated for particulate matter from
the kiln and clinker cooler, and opacity from any affected facility. Particu-
late matter emissions from the kiln must not exceed 0.15 kg per metric ton of
feed (dry basis) to the kiln (0.30 lb/ton). Particulate matter emissions from
the clinker cooler must not exceed 0.050 kg per metric ton of feed (dry basis)
to the kiln (0.10 lb/ton). Opacity for any affected facility is limited to
lees then 10 percent, except that gases discharged from the kiln must not ex-
ceed 20 percent opacity. Sources constructed, reconstructed or modified after
August 17, 1971, are subject to the regulation.
Process Description
Refer to Figure 3 for a flow diagram of conventional cement plants, and
Figure 4 for a diagram of suspension preheater units. There are four major
steps in the production of Portland Cement: Quarrying and crushing, grinding
and blending, clinker production, and finish grinding and packaging. The
first step, quarrying and crushing, involves mining of the raw materials
(limestone, cement rock, clay, shale and gypsum), on site primary crushing,
transport to the plant, secondary crushing, and raw mill storage. In the
second step, the raw materials are ground to a powder and blended to the
required composition. This step will also Involve raw material drying or
slurry production depending on whether the dry or wet process is used. For
suspension pre—heater plants material drying and preheating take place in
a preheater section which immediately precedes the kiln. After blending,
the ground raw material Is fed to a rotating kiln where it is heated, dried,
calcined and finally heated to a point of incipient fusion at about 1,593°C
(2,900°F), a temperature at which a new mineralogical substance called clinker
is produced. As the clinker is discharged from the kiln, it passes through a
clinker cooler which serves to reduce the temperature of the clinker before it
is stored, and to recover the sensible heat for reuse in preheating kiln air
supply. The final step involves proportioning the clinker with gypsum, grinding
this mixture to a final consistency, bulk storage of the product, and bagging
and/or bulk loading/unloading of the final product. The fineness of the raw
and finished product (90 to 100 percent minus 325 mesh) requires that all
transfer points be controlled to prevent emissions. Emissions are usually
controlled with electrostatic precipitators or fabric filters.
1.4
-------
I - ’
I ) ,
0 .1 -L I,
sTACK
Figure 3. Conventional dry process cement plant.
-------
FlED
CRU 4E! j $TO ML j ii1II1 LI1 _Li
I I I
FUEL.
14.MNERS
C’
TO
S E INS SIL
BAGO ING,
BULk LOADINS
AND UNLOAD*NS
Figure 4. Suspension preheater cement plant.
-------
Pre—startup Operations
Prior to the passage of any raw material through the various processing
steps of the facility, all mechanical and electrical systems in each production
phase are independently operated. These checks serve to insure that all motors,
fans, screw conveyors, electrically controlled valves, material handling opera—
tiona and air pollution control devices are ready for production. In vet
process operations, all pumps and slurry feed systems are also checked at this
time. While the real test of these systems comes once the entire facility is
run under full load conditions, these pre—start up checks screen out inoperable
or poorly installed equipment and therefore minimize down time when the entire
facility goes on line.
Once all raw material feed systems are checked “dry” (without feed), raw
materials are fed into the primary crusher and follow the processing steps up
to the kiln feed. The quantity of material throughput at this time is limited
to the size of the raw material storage bins, typically one weekb supply.
The kiln undergoes pre—startup inspection and equipment checks of all
motors, bearings, and instrumentation. In addition, the refractory of the
kiln is conditioned by running the unit at low fire for several days. This
conditioning step is typically the last pre—startup check and raw feed into
the kiln follows shortly to avoid a second conditioning cycle. Finished
product handling, milling and storage systems are checked dry during this
pre—startup period, as well. Cement industry personnel indicated that approx-
imately 2 months are needed to perform all pre—startup checks, for both dry
and wet process plants.
Startup Operations
Startup for Portland Cement operations is generally considered to be the
first day raw material is fed into the preheater or the kiln, depending on
facility configuration. This date is obtainable from plant personnel, as it
is often marked by on—site ceremonies attended by plant, corporate and contractor
officials. This start—up, however, also begins an intensive shakedown period
for the facility. As the production of portland cement is a chemical process
involving many variables, (fuel and feed composition, flow rates, temperatures,
air flow rates, etc.) various adjustments to operating parameters must be made
to produce a final product meeting desired clinker specifications. This process
is a trial and error procedure that may take several months. In addition, once
the entire facility begins to operate with feed material, mechanical and elec-
trical breakdowns occur, delaying the time required to reach full capacity.
While several cement industry contractors felt that startup might better be’.
defined as the first day the plant is run for 24 continuous hours, or the
first day following 1 month’s continuous operation, most agreed that for
conventional plants, 180 days after initial kiln firing is sufficient time to
allow all but the rare cases an opportunity to debug the system, set up the
process operating parameters and produce at or near rated capacity. With
modified plants, much of this debugging has already been accomplished, and
they are online before the 180 days. For the newer suspension preheater units,
17
-------
however, this time frame may not be sufficient. Contacts with personnel at
plants with these units indicate that more than 180 days may be needed
for equipment shakedown. This is due to their complexity and the increased
number of interlocks on each phase of production that will shut the entire
plant down, should an upset occur. These personnel indicated that up to 1
year may be needed to reach normal operation.
18
-------
REFERENCES
1. Personal communication with Mr. Steuch, F.L. Smidth Company, Creskill,
New Jersey. March 21, 1979.
2. Personal communication with Mr. Knoflicek, Allis Chalmers Company,
Milwaukee, Wisconsin. March 20, 1979.
3. Personal communication with Mr. McCord, Portland Cement Association,
Chicago, Illinois. March 20, 1979.
4. Personal communication with F.L. Smidth Company personnel, Creskill,
New Jersey. March 20, 1979.
5. Personal communication with Lehigh Portland Cement Company personnel,
Allentown, Pennsylvania. March 19, 1979.
6. Personal communication with Southwestern Portland Cement Company per—
aonnel, Odessa, Texas. March 21, 1979.
7. Personal communication with Mr. Rite, South Dakota Cement Company,
Rapid City, South Dakota. March 21, 1979.
‘9
-------
NITRIC ACID PLANTS — SUBPART C
§60.70 — 60.74
Introduction
This NSPS category regulates weak nitric acid production facilities (30 to
70 percent strength) that utilize either the pressure or atmospheric pressure
process. Emissions of nitrogen oxides (N0 ), expressed as nitrogen dioxide, are
limited to 1.5 kg per metric ton (3.0 lb/ton) of 100 percent nitric acid. This
is roughly equivalent to about 200 ppm NO in the tail gas. Opacity from these
plants is limited to less than 10 percent. In addition, continuous monitoring
for NO, is required. Sources constructed, reconstructed or modified after
August 17, 1971, are subject to these standards.
Process Description
The manufacture of nearly all commercial grades of nitric acid in the U.S.
is accomplished by the single or high—pressure catalytic oxidation of ammonia,
illustrated in Figure 5. In the process, anhydrous anm onia (N H 3 ) is evaporated
continuously and uniformly in an evaporator using heat supplied by steam. Air
for the chemical reactions is supplied by a power—recovery compressor after
passing through an air filter. Anmonia is vaporized and oxidized with air to
nitric oxide (NO) at about 689 kPa (100 psi) and 900°C (1,650°F). The oxida-
tion is accomplished in a converter consisting of a 95 percent platinum and 5
percent rhodium wire gauze. Compounds leaving the oxidation chamber pass
through an air preheater, a waste heat boiler and a platinum filter (used to
catch any platinum driven out of the converter), before being cooled further
and introduced to the bottom of the absorption tower. Tail gases from the top
of the tower pass through an entrainment separator to remove acid droplets or
mist, are heated in an exchanger counter to the reaction gases, expanded
through the compressor, and exhausted to the atmosphere. An alternative con-
trol system consists of passing the gases from the entrainment separator
through a molecular sieve adsorption bed and on to the reheater and tail gas
expander prior to discharge to the atmosphere.
Pre-Startup Operations
The most important piece of equipment requiring mechanical checkout is the
drive train consisting of a steam or electric—driven turbine, an air compressor,
and a tall gas expander. Items which must be analyzed in the drive train include
couplings, low, high, and overspeed operation, alignment and balance of rotors
and pistons, synchronization between front and back ends, the lubrication sys-
tem, and all operational turbine trips. Examples of turbine trips (shutdowns)
which niist be evaluated are:
20
-------
NITRIC
AC O P OOUCT
Figure 5. Flow diagram for nitric acid production by the pressure process.
-------
• high water level in suction pot,
• radial and thrust vibration,
• lube oil supply, and
• air/ammonia ratio.
Once all of these equipment checks are performed, the complete unit is dis-
assembled, all parts and bearings are rechecked and oiled, the lubrication
system is drained and flushed, and the train is reassembled. A time—consuming
aspect of the drive train checkout involves plotting of the unit’s surge curves.
Once the unit is operational, the air compressor can be used to blow out down-
stream air and steam lines.
Other equipment debugging procedures are performed according to individual
“puiich lists” and are summarized as follows:
Licluid piping and coded vessels — pressure tested with water at maxinum
working pressure.
Gas lines — cannot be checked until plant is operating.
Relief valves — bench tested with required pressure——if serious problems
exist, they are sent out for repairs.
Heat exchangers — flushed with water or a cleaning solution.
Waste—Heat Boiler — undergoes a hydrostatic check followed by pretreatment
with chemicals to prevent corrosion due to oxygen and/or water prior to
plant startup. A final procedure before production starts consists of
filling the boiler with water and warming with steam to prevent shock to
the system.
Absorber column — shipped to the plant as a complete package and can be of
either a bubble cap or sieve tray arrangement. The column is prepared by
flushing with water to clean and check flow and level indicating instru-
ments. Sieve tray columns are more sensitive to gas versus liquid flow
and may require 1 hr to seal properly whereas a bubble cap unit may take
about 20 mm.
Instrumentation — cannot be installed until all other equipment is in place.
A critical component is the rnnmnnia/air ratio control system which must be
accurately calibrated to read concentrations of about 9—11 percent nwnonia
in air.
22
-------
Startup Operations
Once all equipment is installed and thoroughly checked for proper
mechanical operation (this may take from 2 to 6 months), the plant is ready
to undergo initiation of nitric acid production. Preliminary startup opera-
tions consist of the following steps:
1. etartup of air compressor system
2. initiation of water flow to adsorber tower and caustic flow
to scrubber (where a scrubber is used)
3. platinum gauze lit by hydrogen torch to initiate burning of
ammonia (flame is self—sustaining)
4. Ammonia flow is begun.
Within 2 to 3 weeks of this initial startup, the plant is ready for a test
or demonstration run. Test runs usually last 3, 7, or 14 days depending on
the contract. During this time, the plant must achieve its peak efficiency,
maximum design rate, and meet all applicable emission regulations. A violation
of any of these conditions or other equipment malfunctions results in a cessa—
tion of the test run. The conclusion of a successful test run results in the
“legal acceptance” of the plant from the contractor.
The best point in time to define plant startup is when the ammonia flow
to the converter is initiated. Barring no’unusual problems, the completion of
a successful test run and the achievement of maximum production rate should be
about one month or less from this startup point. An important point with re-
spect to nitric acid facilities is that the summer months are the most critical
for proper operation due to cooling requirements for the exotherinic reaction
involved. For this reason, most new plants try to come on line during the
summer when a successful test run would be most meaningful. Because of the
requirement for performance testing within 180 days of starçup, it is conceiv-
able that testing could be required during the winter months when a plant would
find It easiest to meet applicable emission limitations. In this instance,
regulatory agencies might want to conduct testing as soon after startup as
possible, consider postponement of tests until the following summer, or con-
sider winter testing and subsequent summer testing.
23
-------
REFERENCES
1. Personal communication with Mr. Bill Ryan, ICaiser Agricultural Chemicals,
Savannah, Georgia, March 30, 1979.
2. Personal coimarnicaUon with Mr. Louis Pebworth, Farmland Industries, Dodge
City, Kansas, Apri]. 18, 1979.
3. Personal communication with Mr. Lloyd Thomas, Agrico Chemical Co.,
Catoosa, Oklahoma, April 18, 1979.
4. Rosenberg, Harvey S., Molecular Sieve N0 Coatro].Process in Nitric Acid
Plants. EPA—600/2—76--015. January 1979.
24
-------
SULFURIC ACID PLANTS - SUBPART H
§60.80 — 60.85
Introduction
Facilities affected by this regulation are those plants producing sulfuric
acid by the contact process by burning elemental sulfur, alkylation acid, hydro-
gen sulfide, organic sulfides and mercaptans, or acid sludge. Facilities which
convert sulfur dioxide or other sulfur compounds to sulfuric acid primarily as
a means of preventing emissions of the former are not subject to the regulation.
Emissions of gaseous sulfur dioxide cannot exceed 2 kg per metric ton (4.0
lb/ton) of 100 percent sulfuric acid. Acid mist emissions, expressed as sul-
furic acid, cannot exceed 0.075 kg per metric ton (0.15 lb/ton) of 100 percent
acid produced. Opacity must be less than 10 percent. In addition, continuous
monitoring for sulfur dioxide is required. Sources constructed, reconstructed
or modified after August 17, 1971, are subject to the regulations.
Process Description
Three methods used to manufacture sulfuric acid (H 2 SO ) are the contact,
spent acid, and metallurgical processes, the first two of which are showa
in Figure 6. Spent acid utilization occurs at petroleum refineries where
the acid is available. The metallurgical process is employed as a sulfur
dioxide abatement method at plants roasting metallic sulfide ores and for
this reason is not subject to the regulation.
The most widely used manufacturing method is the contact process, In which
sulfur is burned to produce sulfur dioxide. Sulfur is transported to the plant
in either the molten or elemental state; if not shipped in the molten state, the
sulfur is melted and filtered prior to burning. Combustion air required for
burning Is dried with 93 to 98 percent acid in drying towers and then fed to
the sulfur burner. Before the sulfur dioxide (SO 2 ) gas can be introduced to the
converter, It must be cooled to about 427°C (800°F), the minimum temperature
at which the catalyst (usually diatomaceous earth impregnated with vanadium
pentoxidt ) will hasten the chemical reaction. Heat is usually recovered in a
waste heat boiler, where steam is produced, and a heat exchanger, in which sul-
fur trioxide (SO 3 ) from the converter passes through and SO 2 surrounds the ex-
changer tubes. The SO 2 gases are purified by filtration prior to Introduction
to the converter. SO 2 is oxidized to SO 3 In the converter usually in four
stages with most ( ‘75 percent) of the conversion taking place in the first
stage. The temperatures are critical at the point of introduction to the next
pass and hot gases are again used to provide heat to other points in the plant.
The SO 3 can then be absorbed in an oleum tower (for production of oleums or SO 3
in H 2 SOi ) and/or an absorbing tower (for production of H 2 S0 in water). The
acids are then cooled and pumped to storage and the absorbing towers. Best
25
-------
SUAf.INGJ
AIR
COMPRU$OR
U’
I SO 2 +t/2O 2 — So 3
$02
LI PENT *ci aI
COMPRESSOR
HUMfOIPYIMI
TOWER 0* 5 COOLER
us
COOLINS
FINAL
ARSORBING
TOWER
Figure 6.
Schematic diagram of two processes for manufacturing sulfuric acid 2 .
-------
demonstrated technology for control of SO 2 emissions is the dual absorption
process and for acid mist is the high efficiency mist eliminator. An alternate
control system consists of a hydrogen peroxide scrubbing system (DuPont’s
I y 5 IS system).
Pre—Startup Operations
Once equipment is delivered to the plant, certain mechanical checks can
be performed prior to final hookups. Gas and steam piping systems undergo
rigorous pressure testing at 1˝ times their design operating pressure. The
waste heat boiler undergoes hydrostatic testing according to ASME codes for
fossil fuel—fired pressure vessels. All other acid and water piping are tested
(primarily for leaks) at their normal operating pressure.
Pumps are rotated to ensure proper operation and blowers may be run for a
day or two to check for balance and vibration. Lubrication systems would also
be analyzed.
During the final stages of construction, major vessels such as gas to gas
heat exchangers, absorbers, converter, and sulfur storage tanks are manually
inspected by an experienced startup engineer prior to “closing.” Following
this inspection procedure, the plant would be ready to begin the initial
heating of the catalyst.
Preparation for catalyst bed heating begins with the initial low level
firing of the combustion chamber (sulfur burner). The combustion chamber con-
sists of multiple layers of brickwork which must be heated slowly to about
8710C (1,600°F) to eliminate any entrained moisture. This procedure can take
from 7 to 12 days.
The waste heat boiler will then undergo a boilout procedure wherein the
unit is filled with water and degreasing chemicals, warmed, flushed, and
drained, and then has all gaskets replaced and is refilled with normal boiler
water. Concurrently, the drying and absorbing towers are water-batched for
leak detection followed by acid recirculation.
Startup Operations
Prior to the initial feed of raw material, the heating of the system is
accomplished in three phases totaling about 30 hr. Phase I consists of an
18—hr period to preheat the furnace; Phase II takes about 2 hr to blow pre-
heated air through the converter; and Phase III requires 10 hr to reheat
the furnace and th en heat the catalyst bed to the ignition temperature.
Startup follows the preheating period and is best defined as the first
time raw material is fed to the system, whether it be sulfur or spent acid
(alkylation chemical nitration sludge and hydrogen sulfide). The raw material
feed is begun at about 25 percent of the design rate and is gradually brought
up to full load depending on the proper functioning of all other plant com-
ponents. If there are no difficulties, full production rate can be achieved
in 24 to 72 hr. The most common problem which results in plant shutdown
is gas leakage in some part of the system and this will usually occur in
1/3 of new plant startups.
27
-------
REFERENCES
1. Personal communication with Mr. Tony Corey, Monsanto Co., Everett,
Massachusetts, April 4, 1979.
2. Personal communication with Mr. Robert Crendel, Monsanto Enviro—them,
St. Louis, Missouri, April 9, 1979.
3. Personal communication with Mr. Edward Watts, E. I. DuPont Co., Cleveland,
Ohio, April 21, 1979.
4. Calvin, E. L. and F. D. Kodras. Inspection Manual for the Enforcement of
New Source Performance Standards as Applied to Contact Catalyst Sulfuric
Acid Plants. EPA—340/1—77—008, May 1977.
5. Drabkin, Marvin and Kathryn J. Brooks. A Review of Standards of Per-
formance for New Stationary Sources — Sulfuric Acid Plants. EPA—450/
3—79—003, January 1979.
28
-------
ASPHALT CONCRETE PLANTS - SUBPART I
§60.90 — 60.93
Introduction
The facilities affected under this subpart are those which comprise an
asphalt concrete plant, which Is used to manufacture asphalt concrete by heat-
ing and drying aggregate and mixing with asphalt cements. An asphalt concrete
plant is comprised of any combination of the following: dryers, systems for
screening, handling, storing, and weighing hot aggregate; systems for loading,
transferring, and storing mineral filler; systems for mixing asphalt concrete;
and the loading, transfer, and storage systems associated with emission con-
trol systems.
Affected facilities which commence construction, reconstruction or modif i—
cation after June 11, 1973, are subject to the requirements of this subpart.
Those requirements state that no owner or operator of an asphalt concrete
i1.ant shall, discharge or cause the discharge into the atmosphere from any
affected facility any gases which:
1. Contain particulate matter in excess of 90 ing/dscm
(0.04 gr/dscf),
2. Exhibit 20 percent opacity or greater.
Process Description
Asphalt concrete plants are generally classified as a widespread jobbing
operation. A large majority of the plants operate on a seasonal basis, nor-
mally during the warm months of the year. Within the industry there are
portable and permanent plants. Most of the plants are of the permanent type,
primarily located in or around urban areas were there is a constant market for
their product. Portable plants are generally moved to the job site of large
construction projects. Plant operations at a portable source tend to be more
sporadic than permanent plants because they depend on work at a field site.
Permanent plants tend to produce most of their asphalt concrete for Federal,
state and local highway departments.
The raw materials used in an asphalt batch plant are the aggregate,
asphalt cement, and possibly mineral, filler. Paving mixes are produced for
different uses with correspondingly different characteristics, which are
primarily determined by the aggregate size distribution. The three main
distributions are coarse aggregate (retained on No. 8 mesh sieve), fine ag-
gregate (passing through a No. 8 mesh sieve), and mineral dust (passing through
a No. 200 mesh sieve). The coarse aggregate can be as large as 6.4 cm (2 in.) in
29
-------
diameter. It consists of crushed stone, slag, gravel, and naturally fractured
aggregate. Fine aggregate is usually natural sand. Mineral dust is a filler
used In special applications. It consists of finely ground particles of crushed
rock, limestone, hydrated lime, or Portland cement. Asphalt cement is mixed
to about 7 percent by weight, depending upon the desired characteristics of
the f in. 1 product. Asphalt cement is manufactured from crude petroleum and
is semisolid at ambient conditions. Consequently, It Is heated to 135° to
163°C (275° to 325°F) to facIlitate pumping and mixing. Asphalt cement Is
graded based on an industrial classification and/or its penetration.
A schematic diagram outlining the operations of an asphalt concrete plant
is presented in Figure 7. In general, the preparation of asphalt concrete
involves hauling the aggregate from on—site storage piles and placing it in
designated surge bins. The material Is then metered onto a conveyor belt and!
or bucket elevator which feeds the aggregate into a gas or oil—fired rotary
dryer. As it leaves the dryer, the hot material drops into a bucket elevator
and is transferred to a set of vibrating screens where it is classified by
size and distributed to designated storage hoppers. Depending on the product
required, various amounts of the sized aggregate are weighed and then charged
into a mixer, usually a puginill. The dry aggregate is mixed for a short
period of time to produce a homogenous blend. Asphalt cement is then pumped
from heated storage tanks, weighed and injected Into the mixer. Once the
aggregate has been uniformly coated with asphalt, the product may either be
dropped into an awaiting truck or transferred to a storage silo and sub-
sequently loaded onto a truck. The former operation is typical of a batch
plant whereas the latter represents a continuous operation. Particulate
emissions generated during the process are typically controlled by a mechanical
collector followed in series with a baghouse or wet scrubber. Because the
physical size of a plant is relatively small, one air pollution control system
is all that is normally required to treat process emissions.
The equipment used for asphalt production differs from plant to plant.
Conveyors can be used interc hangeab1y with bucket elevators; storage bins
can be arranged differently, or replaced ;altogether with open pile storage.
The most critical piece of equipment from the standpoint of emission abatement
is the rotary dryer. The counter—current design is the most popular. It is
basically a rotating cylinder which is horizontally inclined with a burner
near the axis at the depressed end and an aggregate feeder at the elevated end.
The aggregate flow is toward the burner flame and opposed to the burner corn—
bustion gas flow. Often internal flights are used for agitation. Commonly,
the temperature in a dryer is 121 to 232 C (250 to 450°F), the gas stream
velocity is 2.3 to 4.1 rn/sec (450 to 800 ft/mm), and the volumetric flow rate
is 33,980 to 118,931 m 3 /hr (20,000 to 70,000 acfm). During normal startups,
the air in the control equipment is warmed sufficiently before production is
started; this is particularly important for baghouses. Similarly, the blower
or fan is always running before the dryer burner Is fired and after the cold
feed material or burner and dryer are turned off.
Rotary dryers employed solely for the purpose of drying aggregate have
been used extensively in the past. However, since the early 70’s drum mixers
have become more widely used. The drum mixer serves the same function as the
30
-------
EMISSIONS
-I
PRIMARY CUST COLLECTOR
Figure 7. Schematic diagram of a typical batch asphalt concrete plant. 1
-------
rotary dryer, but also functions as the aggregate—asphalt cement mixer. Graded
aggregate is metered into the front half section of the inclined drum mixer
where it is dried employing gas or oil—fired burners. As the mixer rotates,
the dried aggregate flows around a steel plate which separates the drying end
of the drum from the asphalt mixing end. As the aggregate flows into the back
half section of the drum, asphalt cement is injected beginning the production
of asphalt concrete. Flights, mounted in both ends of the drum, facilitate
drying and mixing. Once thoroughly blended, the asphalt concrete is trans-
ferred by a slat conveyor to storage silos prior to truck Loadout. Drum mixers
are associated with continuous operations.
Pre—Startup Operations
Once the physical plant structure has been erected, various pre—startup
checks are made prior to the production of asphalt concrete. The first step
is to connect and test (including calibration) all electrical systems, asso-
ciated Instrumentation and control panels. Once this has been completed,
conveyor belts, bucket elevators and vibrating screens are checked for align-
ment, rotation and free clearance. Process equipment and air pollution equip-
ment are checked for hydraulic and dust leaks.
The aggregate dryer is usually seasoned before It Is put into continuous
operation under design conditions. During this period, the burners are
operated under low fire and a limited amount of aggregate Is fed into the
dryer. The aggregate is fed into the rotating dryer to absorb some of the
heat which protects new parts and control equipment (especially a baghouse)
from exposure to high temperatures. If a baghouse is the primary control
device, it is during this phase when the initial dust cake forms on the fabric
filters, conditioning the bags for normal continuous use. Seasoning of the
dryer will last from 10 to 40 hours. After the dryer has cooled down It is
inspected for proper expansion, rotation and alignment.
Typically, plant operators will conduct a trial run with dry aggregate to
assure that the conveyor belt feed rate and distribution system are functioning
properly; the aggregate dryer is working; and bucket elevators, sizing screens,
and mixer are all operational. During this period, the asphalt cement storage
and distribution system is also checked. The pre—startup shakedown operations
will last approximately 3 weeks. There are relatively few differences in the
pre—startup operations and time required for shakedown between batch and con-
tinuous operation plants and permanent and portable plants.
Portable plants, which are moved from one job site to another either
before, during or after the production season, require about 1/3 to 1/2 the
amount of time for startup as for its Initial original startup (I.e., 1 to
2 weeks). To minimize problems associated with relocation, the plant is re—
erected in the same configuration as it was at the previous site. This as-
sures continuity and reduces the number of components which have to be re-
constructed, particularly pipe fittings and connections.
32
-------
Startup Operations
The initial plant startup is considered to be when asphalt cement is first
mixed with the dried aggregate. As indicated previously, this will occur ap-
proxImately 3 weeks after the plant has been physically erected. Various ad—
Juatments and recalibrations are made after the plant has been started up to
fine tune process operations.
Once the production of asphalt concrete has commenced, It may take any-
where from only a day or two to 2 months for a plant to achieve its maximum
production rate.
33
-------
REFERENCES
1. Background Information for New Source Performance Standards: Asphalt
Concrete Plants. February 1974. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina. EPA—450/2—74—003.
2. Inspection Manual for Enforcement of New Source Performance Standards:
Asphalt Concrete Plants. June 1975. U.S. Environmental Protection
Agency, Division of Stationary Source Enforcement, Washington, D.C.
3. Standards of Performance for Asphalt Concrete Plants, Subpart I,
Environment Reporter, April 1979, P. 121:1521.
4. IGCI Air Pollution Control Technology and Costs in Nine Selected Areas,
Septeither 1972, APTD—1555.
5. Mr Pollution Engineering Manual, May 1973. AP—40, Second Edition.
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina.
6. Personal couinunication with a representative of the National Asphalt
Pavement Association, Riverdale, Maryland. May22, 1979.
7. Personal coninunicacion with Mr. hans Coering, Barber Greene Company,
Aurora, Illinois. May 22, 1979.
8. Personal comunication with Mr. Joseph Laird, Maryland State Asphalt
Association, Baltimore, Maryland, May 30, 1979.
9. Personal coimnunication with Mr. Joseph Munsil, Cedar Rapids Equipment
Company, Iowa Manufacturing Company, Cedar Rapids, Iowa. May 30, 1979.
34
-------
PETROLEUM REFINERIES - SUBPART J
§60.100 — 60.106
Introduction
The NSPS for this category is applicable to the following facilities
within petroleum refineries:
1. fluid catalytic cracking (FCC) unit catalyst regenerators
2. fuel gas combustion devices
3. Claus sulfur recovery plants except those of 20.3 metric ton/day
(20 long ton/day) or less,
Since each of these processes and the performance standards pertaining to
them are different, the following sections discuss each one individually:
A. Fluid Catalytic Cracking (FCC) Unit Catalyst Regenerators
Performance standards were promulgated for particulate matter (not to
exceed 1.0 kg/l000 kg or 1.0 lb/1000 lb of coke burn—off), opacity (30 percent
maximum, except for one 6—minute average opacity reading in any 1 hour period),
and carbon monoxide (not to exceed 0.05 percent by volume). Where gases from
the FCC regenerator pass through an incinerator or waste heat boiler in which
auxiliary or supplemental liquid or solid fossil fuel is burned, the incre-
mental rate of particulate matter emissions cannot exceed 43.0 g/MJ (0.1C
lb/b 6 Btu) of heat input attributable to the fossil fuel used. Continuous
monitoring is required for opacity and carbon monoxide. Sources constructed,
reconstructed or modified after June 11, 1973, are subject to the regulations.
Process Description
Catalytic cracking is a high—temperature, low—pressure process that con-
verts certain heavier portions of crude oil into gases, gasoline blend stocks,
and distillate fuels. A fluid catalytic cracking unit Is composed of two
basic sections; reactor and regenerator. The cracking reactions take place
continuously in the cracking section, with the spent catalyst being continuous—
ly regenerated and returned to the reactor. Both the cracking and regeneration
sections operate on the fluidization principle, which makes possible a contin-
uous flow of catalyst as well as hydrocarbon feed. Regeneation of the catalyst
is necessary due to metal contamination and poisoning or deposits that coat the
catalyst surfaces and thus reduce the area available for contact with the re—
actants. Figure 8 presents a flow diagram of the catalytic cracking process.
35
-------
noo”cT$
FLUE Us
0 ’
SLIDE
VALVE
AIR INLET
HEATER
FEED
Figure 8.
Flow diagram for catalytic cracking process.
-------
Pre—Startup Operations
Prior to startup, certain operations are performed to ready the FCC for
service. These preliminary activities are necessary to help insure a smooth
and orderly startup of the operating processes. Some of the pre—startup
activi ties Include:
1. General mechanical checkout of equipment. This includes pumps
and other material handling equipment.
2. Instrument System Review. This includes continuity checkouts
as well as calibration of all transmitters, transducers and control
valves.
3. Pressure testing of welds and a general washout of equipment
with air, nitrogen or water.
4. Slow warmup of equipment to dry out refractories and minimize
the possibility of thermal shocks.
This curing of the refractories is done slowly and can take place over
a number of days depending on the quantity of material involved.
Startup Operations
After the refractories are dried out and conditioned, the next procedure
is to load and circulate the catalyst. Once the catalyst is circulating, it is
heated by means of an air heater. At this point, feed can be introduced to the
FCC unit. The date of this introduction of feed to the FCC unit can be defined
as the date of startup.
B. Fuel Gas Combustion Devices
Performance standards were promulgated for sulfur dioxide based on either
a limit on hydrogen sulfide (not to exceed 230 mg/dscm or 0.10 gr/dscf) In the
fuel gas or an equally effective control of sulfur dioxide emissions. In addi-
tion, there is the requirement either for continuous monitoring of sulfur diox-
ide in the gases discharged to the atmosphere from the combustion of fuel gases
or for continuous monitoring and recording of the concentration of hydrogen
sulfide in fuel gases burned in any fuel gas combustion device. Sources con-
structed, reconstructed or modified after June 11, 1973, are subject to the
regulations.
Process Description
A fuel gas combustion device Is defined as any piece of equipment, such
as process heaters, boilers and flares used to combust fuel gas, but does not
include facilities In which gases are combusted to produce sulfur or sulfuric
acid. Flare combustion of process upset gas or fuel gas which is released to
the flare as a result of leakage from a relief valve is exempt.
37
-------
Pre—Startup Operations
As discussed previously, preparation of process equipment involves
primarily a mechanical and instrument checkout.
Startup Operations
Startup of fuel gas combustion devices is initiated with a test firing.
Further information on procedures for startup of combustion devices can be
found in the discussion of boiler startup (Subpart D).
C. Claus Sulfur Recovery Plants
Emissions of sulfur dioxide (SO 2 ) are limited to 0.025 percent by volume
at 0 percent oxygen (dry basis) if emissions are controlled by an oxidation
system or a reduction control system followed by incineration; or 0.030 percent
by volume of reduced sulfur compounds and 0.0010 percent by volume of hydrogen
sulfide (H 2 S) calculated as SO 2 at 0 percent oxygen (dry basis) if emissions
are controlled by a reduction control system not followed by incineration.
Continuous monitoring and recording of SO 2 in the gases discharged to the atmos-
phere is required if an oxidation or reduction control system followed by in-
cineration is used. Continuous monitoring and recording of H 2 S and reduced
sulfur compounds in discharged gases is required If a reduction control system
not followed by incineration is used. Sources constructed, reconstructed or
modified after October 4, 1976 are subject to the regulation.
Process Description
A Claus sulfur recovery plant is a process unit which recovers elemental
sulfur from hydrogen sulfide by a vapor—phase catalytic reaction of sulfur
dioxide and hydrogen sulfide. A typical three—stage Claus unit designed to
operate on a “standard” acid gas rich in hydrogen sulfide is shown in Figure 9.
Pre—Startup Operations
As discussed previously for the FCC process unit, preparation of process
equipment involves the following activities:
1. general mechanical checkout (pumps, etc.)
2. instrument checkout
3. confirmation of “as—built” conditions with design specifications
4. pressure testing of welds
5. acidize compressor lines
6. general washout (air, nitrogen, or water).
7. commIssion utilities
38
-------
HIGH PRESSURE
STEAM
L A - ,
ACID
GAS
WATER
AIR
TAIL
Figure 9. Flow diagram for a typical three—stage claus sulfur recovery plant 3 .
-------
8. dry out refractories (slow warmup)
9. load catalyst
Startup Operations *
Startup for Claus sulfur recovery plants can be defined as the introduc-
tion ,f feed to the unit. Once feed is introduced, additional checks for leaks
are performed. If no leaks are detected, process variables are aimed at test
conditions previously determined.
40
-------
REFERENCES
1. Personal communication with Mr. Michael Volker, Ashland Oil Refining,
Ashland, Kentucky. March 7, 1979.
2. “Chemical Process Industries” fourth edition, Shreve and Briak, McGraw—
Hill Book Company, New York, 1977.
3. Chute, Andrew E. Tailor Sulfur Plants to Unusual Conditions. Hydro-
carbons Processing, April 1977, pp. 119—124.
41
-------
STORAGE VESSELS FOR PETROLEUM LIQUIDS — SIJBPART K
§60.110 — 60.113
Introduction
The provisions of Subpart K of 40 CFR 60 dictate performance standards
for storage vessels for petroleum liquids. Subject to the requirements of
Subpart K are storage vessels with capacities greater than 151 m 3 (40,000 gal),
but not exceeding 246 m 3 (65,000 gal), which have commenced construction, re-
construction or modification after March 8, 1974. In addition, those storage
vessels having a capacity greater than 246 m 3 (65,000 gal) and having commenced
construction, reconstruction or modification after June 11, 1973, are subject
to Subpart K. If the true vapor pressure of the petroleum liquid, as stored,
is equal to or greater than 10.3 kPa (1.5 psia) but not greater than 76.5 kPa
(11.1 psia), the storage vessel must be equipped with a floating roof, a vapor
recovery system or their equivalents. If the true vapor pressure is greater
than 76.5 kPa (11.1 psia), the storage vessel must be equipped with a vapor
recovery system or its equivalent.
The owner or operator of a storage vessel is required to maintain a file
of each type of petroleum liquid stored, Reid vapor pressure, dates of storage
and date on which the storage tank is emptied. In addition, the owner or
operator shall record average monthly temperature and true vapor pressure. No
performance testing is required for petroleum storage vessels.
Process Description
There are two types of petroleum storage tank emission reduction systems
presently being used: floating roof and vapor recovery.
Floating roof storage tanks are used for storing volatile material with
vapor pressure in the lower explosive ranges minimizing fires or explosive
hazards. There are three types of floating roof tanks: pan, pontoon and
double deck. The pan floating roof type has been used for the past 40 years
and is now being phased out. Many operators have had tilting and sinking
problems with this type of roof; besides, the pan roof also has a high vapor-
ization loss around the periphery of the roof. The pontoon roof is mainly
used on tanks with larger diameters and provides for better stability.
Some pontoon roofs have a center drain with hinged or flexible connections for
roof drainage problems. In addition, traps or dams are provided on the under—
side of the roof. The traps retain any vapors formed as a result of localized
solar boiling. The double deck floating roof, which is the most expensive,
reduces the effect of solar boiling and has more rigidity than the pan and
42
-------
pontoon floating roofs. Double deck roofs have compartmented dead—air spaces
over the entire liquid surface. The roof also has drains for water accumulation
and vapor traps on the underside of the roof.
A vapor recovery system is designed to handle vapors originating from
filling operations. The recovered vapors are compressed and charged to an
absorption unit for recovery of condensable hydrocarbons. The system includes
vapor lines interconnecting the vapor spaces of the tanks that the system
serves. Each tank should be capable of being Isolated from the system. Tank
isolation and operation is by manual and pressure controlled butterfly—valves,
regulators, and check valves. Knockout pots are normally used at low points
in the vapor line to remove condensate. Noncondensable vapors are piped to a
fuel gas Pystem or to a smokeless flare. When absorption of the condensable
vapors is not practical from an economic standpoint, these vapors, too, are
either sent directly to a fuel system or incinerated in a smokeless flare.
Startup Operations
Since no performance testing is required for petroleum storage vessels,
the 180 day test period does not apply in this case. However, the proper
operation of either the floating roof and/or vapor recovery system is essential
for the minimization of hydrocarbon emissions. Storage vessel systems are in-
spected and checked for structural, mechanical, electrical and hydraulic prob-
lems. The primary checks for the petroleum storage vessel are water pressure
tests for structural defects and mechanical operation and clearances of the
storage vessel’s roof and seals, compressors and vapor control valves and
actuators.
43
-------
SECONDARY LEAD SMELTING — SUBPART L
§60.120 — 60.123
Introduction
The NSPS for this category is applicable to the following facilities in
secondary lead smelters: pot furnaces of more than 250 kg (550 ib) charging
capacity, blast (cupola) furnaces, and reverberatory furnaces. Performance
standards for blast (cupola) or reverberatory furnaces were promulgated for
particulate matter and opacity. Particulate matter cannot exceed 50 mgldscm
(0.022 gr/dscf).. Opacity must be less than 20 percent. Performance standards
for pot furnaces were promulgated for opacity only, which must be less than
10 percent. Sources constructed, reconstructed or modified after June 11, 1973,
are subject to the regulations.
Process Description
The processing of secondary lead centers around the utilization of three
furnaces. Smelting operations on the scrap lead are carried out in the blast
(cupola) furnace and/or reverberatory furnace and the final purification steps
in pot furnaces.
The blast (cupola) furnace used in processing secondary lead is similar
to those in the ferrous industry; cylindrically shaped and standing vertically.
Forced air, sometimes oxygen enriched, is introduced near the bottom of the
furnace. The furnace is batch fed at the top. A typical charge is made up
of about 80 percent scrap lead (generally battery plates and including 8 per-
cent return slag) 8 percent coke, 2 percent iron, and 10 percent limestone.
Heat is produced by the combustion of the coke which also provides an atmosphere
for reducing the lead oxide feed. The lead metal collects at the bottom of
the furnace and Is customarily drawn off through a tap hole.
Reverberatory furnaces operate by radiating heat from the gas or oil fired
burners and the surrounding hot refractory lining onto the contents of the
furnace. The flame and products of combustion come in direct contact with the
charge material. The furnace is commonly rectangular in shape with a shallow
hearth and constructed of fire brick and refractory materials. The principal
use of the reverberatory furnace involves the melting and purification of
lead by removal of extraneous ingredients.
The reverberatory furnace may be charged with molten lead from the cupola
• a continuous basis. In this case, air is blown through the bath either
ntinuously or intermittently to oxidize metal impurities. The metal dross,
w *ich is formed, floats on top of the lead and is removed Intermittently by
44
-------
slagging. The lead product is tapped from the furnace into molds on an
intermittent basis. If lead oxide drosses are charged to the furnace, a
reducing agent such as granular carbon must be added to the bath to reduce
the lead oxide to metallic lead. The furnace operates at about 1260°C
(2300°F) principally to allow the reaction between metallic impurities and
the oxygen aparged into the bath. The high temperature also allows for
afterbu ning in the furnace proper. This is accomplished by maintaining a
tigit furnace, that is, excessive air leakage into the furnace is prevented
and the amount of oxygen introduced to the furnace is thus controlled. The
reverberatory furnace product is a seal—soft lead which is more pure than
that which the blast furnace produces.
Pot furnaces are used for remelting and for final alloying and refining
processes before pouring into product molds. They are open—top, ceramic—lined
kettles, hemispherically shaped and generally range in size from 0.9 to 45
metric ton (1 to 50 ton) capacity. They are normally under—fired by natural
gas burners. Refining is a batch operation that can vary from several hours
to two or more days, depending on the required final composition. Drossing
agents or alloys are generally added individually and the bath is normally
agitated or, in some cases, air is bubbled through the bath. Drosses are
normally skinned off the surface of the lead by hand.
Particulate matter is typically controlled by fabric filters in secondary
lead smelting operations.
Pre-startup Operations
Prior to startup, certain operations are undertaken to ready the furnace
for production. These preliminary activities are concerned not only with the
operation of the furnace but also with the operation of material handling
systems, and air pollution control systems.
For the furnace, some of the operations which are considered in the
pre—startup category include:
1. Delivery, assembly and hook—up of utilities
2. Installation of pollution control equipment
3. Installation of air handling system including stack
4. Check out of wiring and control systems
5. Installation of refractory
These activities will typically reqńire 1 month to complete.
Related to the operation of the furnace and crucial to achieving maximum
production are the start—up of material handling systems and the air pollution
control systems. Problems in either of these two areas can cause delays of
several months in preventing a smelting operation from achieving design
capacities.
45
-------
In efforts to reduce worker exposures to lead, some new installations
of secondary smelting may utilize completely enclosed, remote controlled
operations. In these cases, debugging the new technology of remote control
material handling may require several months.
Startup Operations
The lead blast furnace (cupola) has a startup procedure similar to the
blast furnace used in the iron and steel industry. The startup procedure first
requires the hand stacking of wood in the furnace. The time required for this
operation will depend on the size of the furnace, however, it should not take
more than a few hours. The next procedure which takes 2—3 hr to accomplish,
is the charging of the furnace with oil—soaked coke and upon completion the
furnace is ready for ignition. Combustion air is introduced at the bottom of
the furnace through tuyeres at a gauge pressure of 3.45 to 5.17 kPa (8—12 oz/
in 2 .) About 6—8 hr are allowed for the blast furnace to achieve operating
temperature. Hard lead is charged into the cupola at less than full capacity
to provide molten metal to fill the crucible. Normal charges of slag, coke,
iron and limestone are added to the furnace. As the level of molten metal
rises, the slag is tapped off at intervals while the molten lead flows from
the furnace at a more or less continuous rate. If the first few preliminary
pours are successful, larger quantities of scrap lead will be charged until
maximum capacity is achieved. Assuming no problems occur during initial
startup operations, a facility can usually achieve its maximum production rate
in 4 to 7 days. The official startup date should begin when oil—soaked coke
is ignited.
After pre—startup and heatup operations of a reverberatory furnace, the
material is slowly fed into the furnace to prevent thermal shock and to main-
tain a stable temperature while keeping a small mound of unmelted material on
top of the bath. As the mound becomes molten, ore material is charged. The
time required to reach maximum production is 3 to 5 days after material is
fi rst introduced into the furnace. The total time required from a cold start
to maximum production will range from 4—6 days. The 180—day performance test
period should begin when material is charged Into the furnace.
Following pre—startup and heatup procedures, the charge is then slowly
introduced into the furnace developing a molten pool of lead. Since this is
a batch operation, the charging continues until the vessel is filled to rated
capacity. The time required to achieve maximum production depends on the
size of the furnace. Normally, maximum production can be reached in a few
days. The official startup date should begin when the furnace is first charged
with lead.
46
-------
REFERENCES
1. Personal communication with Mr. Hess, U.S. Smelting Furnace Company,
Beilville, Illinois, March 9, 1979.
2. Personal communication with Mr. Borell, Could Inc., Industrial Battery
Division, Sun Valley, California, March 13, 1979.
3. Personal congnuncation with Mr. Pike, General Battery Corp., Reading,
Pennsylvania, March 15, 1979.
4. Personal communication with Warrick Furnace, Chicago, Illinois,
April 24, 1979.
5. Personal communication with Steve Meuller, ANAX Lead/Homestake,
Buick Mine, Missouri, April 10, 1979.
47
-------
SECONDARY BRASS AND BRONZE INGOT PRODUCTION PLANTS - SUBPART M
§60.130 — 60.133
Introduc t ion
The NSPS for this category applies to reverberatory and electric furnaces
of 1,000 kg (2,205 ib) or greater production capacity and blast (cupola) fur-
naces of 250 kg/hr (550 lb/hr) or greater production capacity. For reverbera-
tory furnaces, the standard limits particulate discharge to 50 mg/dscm (0.022
grldacf) and opacity to less than 20 percent. Blast (cupola) or electric
furnaces are governed only by an opacity standard, which must be less than
10 percent. All facilities that commenced construction, were modified or re-
constructed after June 11, 1973, are subject to the regulations.
Process Description
Brass and bronze ingots are produced from three types of furnaces —
reverberatory, rotary and crucible. In the Industry, 95 percent of the ingots
are produced by direct oil or gas fired reverberatory furnaces. Rotary fur-
naces are also classified as reverberatory units for the purpose of NSPS. Due
to the overwhelming predominance in the industry of these units, and the lower
particulate emissions associated with electric furnace use, only reverberatory
furnaces are required to meet a mass particulate emission unit. Figure 10
presents a schematic of a typical Ingot production facility.
Processing begins with raw materials which consist of copper bearing scrap
including faucets, telephone or electric cable, radiators and brass turnings.
Before blending in a furnace, the scrap is preprocessed to remove impurities
and to concentrate like alloys. Pretreatment operations take on several forms.
Mechanical methods include hand sorting of scrap into piles of like material,
stripping or shredding of wire covering or Insulation, magnetizing of iron
particles to remove them, and briquetting for size reduction. Pyrometallurgical
(heating) methods include sweating of scrap to remove low melting point metals,
burning to remove wire insulation, drying In a rotary kiln to vaporize cutting
fluids on machine shop scrap, and use of a blast furnace or cupola. Of these
operations, only the blast furnace (cupola) is governed by NSPS.
The blast furnace operation is a continuous process that accepts slag
skimmings and other metal oxides which are byproducts of the ingot produc-
tion furnaces. Coke, copper oxides and other materials are charged into the
top of the furnace and combustion air, sometimes enriched with oxygen, is
blown In through tuyeres at the bottom. The coke acts as both a fuel and a
reducing agent through the production of carbon monoxide. The reducing
48
-------
“Ac’.
‘.0
‘asilousE
FN SION POINIS
lEvEIS(Ia!oA
FUINACI
• titcilic
C*U ISLL FUANACC
Figure 10. Brass and bronze ingot production facility. 5
-------
atmosphere allows direct reduction of metallic oxides and the resultant
dense molten metal settles from the nonmetallic glass—like slag. The metal
(black copper) is drawn off for further refining in the ingot production
furnaces.
The reverberatory furnace is commonly rectangular In shape with a shallow
hearth and constructed of fire brick and refractory materials. The furnace
oper teB by radiating heat from the gas or oil—fired burners and the surround-
ing hot refractory lining onto the contents of the furnace. The flame and
products of combustion come in direct contact with the charge material. There
are five distinct steps in producing brass or bronze of the desired specifica-
tion in a furnace. First, the scrap materials are charged into the preheated
furnace (charging) and, second, oil or gas is fired directly into the charge
to melt the materials (melting). The charge is next brought to the desired
temperature and fluxes are added to remove Impurities such as carbon, metal
oxides, gases, etc. (refining or smelting). Fourth, after the Impurities are
removed, metal is added to bring the alloy to the proper metallic composition
(alloying). Finally, when analysis indicates the correct grade has been
achieved and the melt is at the proper temperature, the metal is poured Into
ingots (pouring).
The electric crucible furnace follows a similar production sequence.
Heat, however, is applied indirectly using high or low frequency induction
heaters to raise the crucible temperature. The absence of direct flame con-
tact on the metal minimizes metallic volatilization and thereby reduces
particulate generation.
Particulate emissions vary with the content of the alloy being produced
and the presence of impurities in the scrap feed. )bst of the particulate
emissions are metal oxides, predominantly zinc oxides (45 to 77 percent) and
lead oxides (1 to 13 percent). Fabric filters are extensively used to control
emissions from all three types of furnaces, although electrostatic precipitators
have recently been adopted as well.
For various reasons, including the promulgation of an NSPS for this industry,
only one new plant has been constructed in the country since 1973. Total ingot
production has remained relatively stable during this period, and producers have
consolidated operations to the point where only 30 ingot manufacturers are now
in operation.
Pre—Startup Operations
There are numerous material handling subsystems in a secondary brass and
bronze Ingot production facility that must be Individually tested before Ingots
are poured. The number and types of these systems is entirely dependent upon
specific plant practice. Virtually all facilities will have a sorting!
classification process for the Incoming scrap, a magnetizing station for re-
moval of ferrous metals and a briquetting operatiop to facilitate furnace
charging. These subsystems incorporate cranes, conveyors, pulleys, motors,
hydraulic presses, etc., all of which must be operated before startup. Less
common pretreatments such as wire stripping, shredding, sweating, burning and
use of cupolas can be Installed, although they are finding less favor in the
50
-------
industry due to economic and environmental reasons and increased reliance on
large scrap dealers who will clean and process raw scrap. Air pollution con—
trol devices, typically baghouses, must be operated to insure proper installa-
tion and structural integrity of all ductwork, housings and supports. Finally,
the furnace(s) nvst be inspected for proper refractory installation (all units),
freedom of rotation (rotary units, tilting reverberatory and electrical units)
and correct charging clearances. Pretreatment (curing) of furnace refractory
is c . .tional and left to the discretion of the plant operator. When practiced,
curing typically takes 2 to 3 days. Electric induction furnace equipment
(transformers, rectifiers, etc.) is also energized prior to startup. Complete
checkout of this equipment necessarily requires actual plant startup. Plant
construction typically takes 6-12 months, and individual pre—startup checks
will be performed immediately after the specific piece of equipment is installed.
Startup Operations
Startup of a new or modified brass and bronze ingot production facility is
the first day metal is charged into the furnace and melting begins. Application
of heat, either directly or indirectly, will cause generation of metallic fumes,
which are the pollutant of concern. This startup date can be obtained by check-
ing plant production records. Immediately prior to metal charging and light—off,
the collection equipment is activated to insure particulate emissions are con-
trolled at all times. When baghouses are used, activation may also include in-
jecting lime into the flue gas stream to neutralize acids and prolong bag life.
As the operation of conventional design furnaces is straightforward, most op-
erators will fully charge the unit with scrap and attempt to reach maximum
production as soon as possible. Under these conditions mechanical and electri—
Cal probleme will appear and be corrected immediately. Furnaces with new designs,
although infrequently installed, will require a slower break—in period to insure
all metallurgical conditions are correct and the furnace is operating as designed.
Regardless of furnace type, industry contacts indicated that the 180—day test
time frame could be met by all plants.
51
-------
REFERENCES
1. Personal communication with Mr. Drickenmiller, American Brass Company,
Dothan, Alabama. May 9, 1979.
2. Personal communication with Mr. Pinsof, Sipi Metals Corporation, Chicago,
Illinois. May 10, 1979.
3. Personal communication with Mr. Stefinitis, I. Schumann Company, Bedford,
Ohio. May 10, 1979.
4. Personal communication with Mr. Bowman, Brass and Bronze Ingot Institute,
Chicago, Illinois. May 14, 1979.
5. Inspection Manual for Enforcement of New Source Performance Standards:
Secondary BraBs and Bronze Ingot Production Plants. EPA—340/l—77—003.
January 1977.
52
-------
IRON AND STEEL PLANTS — SUBPART N
§60.140 — 60.144
Introduction
The NSPS for iron and steel plants applies to the basic oxygen process of
steel production. The facility affected under this subpart is the basic oxygen
process furnace (BOPF) which produces steel by refining a high carbon—content
charge using a large volume of an oxygen—rich gas.
The mass emission standard promulgated under this subpart states that
gases discharged from a BOPF shall not contain particulate matter in excess
of 50 mg/dscm (0.022 gr/dscf). In addition, the exhaust gas from a control
device shall not exhibit an opacity of 10 percent or greater, except that an
opacity of greater than 10 percent but less than 20 percent may occur once
per steel production cycle . Facilities constructed, reconstructed or modified
after June 11, 1973, are subject to the requirements of this subpart.
Process Description
Steelmaking by a BOPF involves the melting, mixing and subsequent refining
of scrap metal, molten Iron, and fluxes charged to the vessel by the Injection
of a high volume of an oxygen—rich gas. Oxygen may be top blown into the
vessel by a water—cooled retractable lance or bottom blown through a set of
fixed tuyeres. The former method is the more common type of BOPF, whereas
the latter is a recent process modification specific to the Quelle—Basic
Oxygen Process (Q—BOP) furnace. A simplified flow diagram outlining steel—
making in a BOPF is presented in Figure 11.
The oxygen blow commences Immediately following charging and continues
for a predetermined length of time depending on the grade of steel desired and
the charge composition. After the blow, the molten steel is checked to deter-
mine if the endpoint chemical composition and proper tap temperature have been
att iined. If the desired temperature has not been reached, oxygen is reblown
into the bath until the proper temperature is reached. When the melt meets
the required grade specifications, the furnace Is tapped. Tapping involves
rotating the vessel, pouring the finished steel into a teeming ladle. After
tapping, the vessel is rotated back through the vertical and slag (the solid
waste formed during refining) is poured from the furnace into pots located
below the furnace floor. Following slagging, the furnace refractory lining
and tap hole are checked for wear before another heat Is started.
Typical air pollution control equipment installed on BOP furnaces include
a primary hood to collect emissions during the oxygen blow and possibly a
secondary hood and furnace enclosure to capture charging and tapping emissions.
53
-------
PARTICULATES
WASTE GASES
ANOtHEAIT
ALLOY ADDITIONS
SCRAP METAL
MOLTEN IRON
Fl UXES
,J1 OXYG N — RICH GAS
A DOlT STEEL
VESSEL ALLOY IN _____ ____________
_____ SLAG
Figure 11. BOPF Steelinaking process flow diagram.
-------
Collected fumes are normally vented to an electrostatic precipitator or high
energy wet scrubber.
Pre—Startup Qperations
Pre—startup operations of a BOPF include:
• conditioning of refractory lining and tap hole
• testing of mechanical controls
• instrument checks
• checking raw material, process gas and coating water distribution lines
The refractory lining, including the tap hole, must be preheated before
the furnace is put in operation to prolong the life of the lining and to pre-
vent sudden thermal stresses when the furnace is initially charged. The con-
ditioning period, which varies with furnace size and type of refractory, typic-
ally lasts up to several weeks.
Instruments and mechanical equipment are checked to ensure proper opera-
tion and to correct any defects in installation. The usual time period re—
quf red to check equipment varies from 1 to several weeks.
Start—up Operations
The start—up period (ranging from initial heat until attainment of maximum
production) will vary in length from 2 days to a year, depending on the type
of furnace, and production demands. Even though melt cycles will remain
lengthy at first, each steel heat will be produced at the rated furnace
capacity.
Once the furnace and ancillary equipment has been thoroughly checked out
it will take approximately 8 hours to bring the vessel on—line. The cyclical
nature of the process suggests startup should be defined as the time the furnace
is put into operation for the first steel production cycle.
55
-------
REFERENCES
1. Personal communication with Mr. J. Gaibreath, Whiting Corporation,
Harvey, Illinois. 13 March 1979.
2. Personal communication with Mr. S. Hanley, Pullman-Swindell, Pittsburgh,
Pennsylvania. 13 March 1979.
3. Personal communication with Mr. F. Schick, John I4ohr and Sons, Chicago,
Illinois. 13 March 1979.
4. Personal communication with Mr. B. Moore, Penn Engineering Corporation,
New Castle, Pennsylvania. 13 March 1979.
5. Personal communication with Mr. Bonder, William B. Pollock Company,
Youngstown, Ohio. 13 March 1979.
6. Personal communication with Mr. B. Wolf, Loftus Engineering Corporation,
Pittsburgh, Pennsylvania. 13 March 1979.
56
-------
SEWAGE TREATMENT PLANTS - SUBPART 0
§60.150 — 60.154
Introduction
The standard contained in this part applies to sewage sludge incinerators
that combust wastes containing more than 10 percent sewage sludge (dry basis)
produced by municipal sewage treatment plants or that charge more than 1000 kg
(2205 ibs) per day (dry basis) of municipal sewage sludge. Particulate ezais—
alone are not to exceed 0.65 g/kg (1.3 lbs/ton), based on dry sludge input,
and opacity is limited to less than 20 percent. Sources constructed, recon-
structed or modified after June 11, 1973, are subject to the regulations.
Process Description
There are essentially two forms of ultimate municipal sewage sludge destruc-
tion of practical importance today: incineration and wet oxidation.
Incineration (see Figure 12) is a two—step process involving drying and
combustion. The drying step should not be confused with preliminary dewatering
by mechanical means. Drying and combustion may be done in separate units or
successively in the same unit. Where a two step process is employed, startup
should be applied only to the combustion step. Many drying operations are
brought on line well in advance of the combustion unit f or both checkout and
to prepare sludge for landfill while incinerator checkout is proceeding.
Two major Incineration systems employed in the United States are the
multiple hearth furnace and the fluidized bed incinerator. The key difference
between these systems, from a startup standpoint, is that multiple hearth
furnaces employ refractory material whereas fluidized bed units are basically
of all—metal construction. Both incinerator types typically employ scrubbers
for air pollution control.
The wet air oxidation process (depicted in Figure 13) is based on the
principle that any substance capable of burning can be oxidized in the presence
of liquid water at temperatures between 121°C (250°F) and 371°C (700°F). Air
pollution is minimized because the oxidation takes place in water at low tem-
peratures and no fly ash, dust, sulfur dioxide, or nitrogen oxides are formed.
Some particulate and odor pollution can result and catalytic burning is typically
used to control these pollutants.
57
-------
_________ ——
$LUO4( liii DIWATLRtNS $LUDU PREfISt*TMtNT I.
TO
ATMOSPHUt
Figure 12. Sewage sludge incinerator process diagram.
-------
ENNAUST GAS
AIR POLLUTION
TO
__J...!INAUSLGAS ATMOSPHERE
I I
SLUDGE PRETREATMENT I
l (OPTIONAL) f
I I
I WET SOLIDS TO
DISPOSAL
—
I STEAM I
_ GENERATOR WAT R —
(OPTIONAL)
I I
Ui
WET AIR
0 * IOAT ION
STERILE
SOLIDS
AND WATER -
SOLIDS
SEPARATION
WATER
CONSUSTI0N J AIR I L — — — — —
“ 1 COMPRESSOR
WATER
BOO,
Figure 13. Sewage sludge wet air oxidation process diagram.
-------
Pre—Startup Operations
1. Multiple Hearth : One to two weeks are typically allowed for
instrumentation and mechanical equipment shakedown. This in-
cludes the operation of the combustion air source and the
sludge feed system. Upon completion of this pre—startup pro-
cedure, the unit is heated to an operating temperature of about
927°C (1700°F). This must be done gradually In order to pre-
vent stressing of the refractory material. Once at the operating
temperature, the unit remains fired until it is thoroughly
“dryed out.” The heat—up and dry—out procedure consumes
approximately one week.
2. Fluidized Bed : One to two weeks is required to perform an
instrumentation and mechanical equipment shakedown similar to
that for a multiple hearth furnace. Following this, the unit
Is heated up to its operating temperature of about 871°C
(1600°F). This can be done relatively quickly since there is
no refractory construction material. A typical heat—up period
18 three days. A drying—out procedure Is not necessary.
3. Wet Air Oxidation : One to two weeks is required for instru-
mentation and mechanical equipment shakedown. The unit is
operated by feeding water only as part of the procedure. Since
the unit operates at only 93°C (200°F) to 204°C (400°F), the
heat—up period is only a matter of hours. Dry—out is not
required.
Startup Operations
1. Multiple Hearth : After the pre—startup operations of heat—up
and dry—out are completed, the unit Is ready to receive sludge.
The sludge mixture, which is mainly water, must be charged
slowly at first to prevent thermal shocks and gradients within
the refractory material. The unit must be run at partial
capacity for approximately one week while the load is gradually
increased. The day on which sludge is first introduced to the
unit should officially be considered the startup date. Up
until this point, including post heat—up, the unit could be
shut down fc:mechanical or instrumentation failure, or more
often, for a lack of sludge of proper consistency and quantity.
For a multiple hearth furnace, the maximum production rate can
be attained one week from startup.
2. Fluldized Bed : After the pre—startup operation of heat—up is
complete, the unit is ready to receive sludge. The sludge is
usually introduced at a rate within 20 percent of design capacity
or greater. The day on which sludge is first introduced to the
60
-------
unit should officially be considered the startup date. Since
sludge is introduced at or near the design rate, there is no
time elapsed between startup and the point when the maximum
production rate is reached.
3. Wet Air Oxidation : After the pre—startup operation of heat—
U is complete, the unit is capable of receiving sludge at or
near the maximum production rate. The day on which sludge is
first introduced to the unit should officially be considered
the startup date. Since sludge Is Introduced at or near the
design rate, there is no time elapsed between startup and the
point when the maximum production rate is reached.
61
-------
REFERENCES
1. Process Design Manual for Sludge Treatment and Disposal, U.S. EPA Tech-
nology Transfer, October 1974. EPA—62511—74—006.
2. Inspection Manual for the Enforc sent of New Source Performance Standards —
Sewage Sludge Incinerators, by Devitt and Kulujian, U.S. EPA, Research
Tripngle Park, N.C., February 1975.
3. Private counication: Richard Shedlow of Nichols Engineering and
Research Corporation, Belle Mead, New Jersey. March 5, 1979.
4. Private Cotanunication: Ed Sweeney of Dorr—Oliver Company, Stamford,
Connecticut. March 6, 1979.
5. Private Coisiunication: Mike Mayer of Zimpro, Rothschild, Wisconsin.
March 12, 1979.
62
-------
PRIMARY COPPER SMELTERS - SUBPART P
§60.160 — 60.168
Introduct ion
The standards of performance for primary copper smelting affect the dry-
ing, roasting, smelting and copper converting facilities. Gases which contain
particulate matter in excess of 50 mg/dscm (0.022 gr/dscf) from any dryer and
sulfur dioxide in excess of 0.065 percent by volume from any roasters, smelt-
ing furnaces or copper converters cannot be discharged into the atmosphere.
The standard for visible emissions limits opacity to 20 percent for dryers and
facilities using a sulfuric acid plant to comply with the standards. Addition-
ally, sources shall install and operate a continuous monitoring system for
opacity and sulfur dioxide. Sources constructed, reconstructed or modified
after October 16, 1974, are subject to these requirements.
Process Description
The copper bearing ores are either smelted as they come from the mine or
are subject to a preparatory process of grinding and floatation to transform
low—percentage ores into high—percentage ores. Ore coming directly from the
mine normally passes through a dryer to remove the moisture while grinding—
flotation processed ore is subsequently smelted down either directly or after
partial roasting. Roasting removes part of the sulfur from the ore. There
are two main types of roasters used in the industry: multiple hearth and
fluidized bed. In the smelting furnace, iron oxide combines with siliceous
flux to form a slag and leaving a material known as matte composed of copper,
iron and sulfur. Two commonly used smelting furnaces are reverberatory and
electric smelting. The matte In the smelting furnace is reduced to copper in
two stages of blowing with air in a unit called a converter. The first stage
eliminates sulfur and forms an iron oxide slag by adding a siliceous flux.
The slag is then removed from the melt. In the second stage, the copper sul-
fide is reduced to metal; sulfur is eliminated as SO 2 leaving a material known
as blister copper. The blister copper may be further refined to remove sulfur
and oxygen to be cast into anodes for electrolytic refining. Figure 14 illus-
trates a combined flow sheet of a variety of choices in current and developing
technology offered to smelter designers. The unit/processes chosen for startup
evalatuton appear to be typical for the copper smelting industry in the United
States.
63
-------
i. F DRYER
I
2. ROASTER
• Multiple hearth
• Fluid Bed SO 2
1 .
3. SI LTINC FURNACES
• Blast furnace
• Momada Blast furnace
• Reverberatory furnace
• Flash smelting
• Electric smelting
• Cyclone smelting
m
4. CONVERTER
• Pierce Smith
• Iloboken
• Top blown rotary
• Stationary
slag Blister Copper
to waste
• REFINING FURNACEI
Anodes
Figure 14. Flow diagram for copper smelting and various unit processes.
64
-------
Pre—Startup Operations
Dryer : A rotary direct heat or indirect heat dryer is typically used in
the industry. Four to seven days are allowed for dryer shakedown. During the
shakedown period, a manufacturer’s field engineer is on hand checking electri-
cal, mechanical, hydraulic and combustion systems. Upon completion of pre—
startup checks, the unit will be turned over to the operator for startup or a
manufacturer’s application engineer is available for complex startups. Nor-
mally, startup is straight forward and an application engineer is not needed.
Pre—startup operation should take no longer than 1 week.
Roaster:
Multiple Hearth : Two to three weeks are allowed for pre—startup operations.
If problems develop, an additional week is allowed. Curing and heatup of the
furnace takes about 5—7 days. This operation must be done very slowly to pre-
vent stress on refractory material. Total pre—startup consumes a 3—4 week
period.
Fluidized Bed : Seven to nine days are required to perform the pre—startup
operations. Following a successful pre—startup, the system is brought up to
operating temperature (3—4 days); the system is now ready for process material.
Total pre—startup operation may consume a 2—3 week period.
Smelting Furnace
Electric Smelting : Two to three weeks are required for pre—startup opera-
tions. According to manufacturers, this period of time is known as green tag.
It also allows time for replacements of minor defective parts. As each sub-
unit is Installed, they are checked mechanically, electrically and hydrauli-
cally as needed. The electrodes and rigging, a major system for the furnace,
is actuated to check alignment and operation. The electrical system is given
special attention for grounds, insulation, dead shorts, and high resistance.
Secondary voltage is then applied to the transformer conductors at a very low
current. Transformers are then energized with primary voltage (Power company
electricity). Curing the furnace is usually accomplished by heating with
electric energy or a supplementary fuel. At this time, the furnace is ready
for charging.
Copper Converters : When a converter system is being installed, it requires
the cooperation of many suppliers and contractors. Upon completion of sub—unit
installation and checking, the unit is normally turned over to the owner. The
period of time consumed during installation and checking is 3—6 months. During
this time, all electrical, mechanical and hydraulic systems are debugged. Other
units and systems checked are: compressors, fuel lines, levelings, fans, pumps,
gauges, refractory linings and the cooling, chemical feed and emergency systems.
Two to three weeks are allowed to cure the refractory material of the converter.
The final step in pre—startup operation is a dry run of the entire system to
check out and coordinate the operation of individual units into a coherent sys-
tem. A dry run normally takes 1 week. Heating the converter to operating
temperature requires 3 to 4 days. The total preliminary startup operation
takes from 5 to 8 weeks.
-------
Startup Operation
Dryer : Startup of a dryer begins when ore is first fed into the dryer.
If no problems develop, the dryer should achieve maximum production in 1˝ hours.
Special procedures sometimes require the presence of a manufacturer’s application
engineer to make adjustments such as drum rotational speed, material fall slope,
and temperature to attain the designed moisture content in the ore. No more
than 2 days are required to make final adjustments. Manufacturers indicated
that the 180—day period performance test period should begin when material is
first fed to the dryer.
Multiple Hearth : After the pre—startup operations, the unit Is slowly
charged with ore to prevent thermal shocks and gradients within the refractory
material. The load is gradually increased to maximum capacity consumIng 1—2
weeks. The official startup should be when material Is first introduced into
the furnace. Prior to charging, the furnace could be shut down very easily
for repair without major time expended.
Fluidized Bed : Following pre—startup operation and heatup, ore is slowly
introduced into the furnace taking 2—4 days to reach maximum production. The
official startup date should be when material is first introduced into the
furnace.
Smelting Furnaces
Reverberatory Furnace : Following pre—startup and heat—up, ore is slowly
introduced into the furnace developing matte. Matte is composed of copper,
Iron and some sulfur. Flux is then added to the matte to develop an Iron
oxide slag. In order to have a continuous process and to develop maximum
production, a good matte is necessary. The development of a good matte, essen-
tial for maximum production, may take as long as 3 to 4 weeks and is dependent
upon operator skills. Inexperienced operators may take as long as 2 months to
develop a good matte. The official startup date should be when material is
first introduced into the furnace.
Electric Furnace : After pre—startup operation and heat—up, the charge is
slowly introduced Into the furnace to develop a good matte and to prevent
thermal gradients and shocks to the refractory material. During this period,
flux is also added to develop a slag. Maximum production cannot be reached
unless a good matte exists and this again may take 3—4 weeks depending on
operator skills. The official startup date should be when ore is first intro-
duced into the furnace.
Converter : After pre—startup operations and heat—up, the matte from the
smelting furnace is introduced into the converter. If the first heat goes
smoothly, the process may continue while slowly increasing the quantity of
matte introduced into the converter until naaximum production is reached. It
may take as long as 6 to 8 months to develop a maximum production rate. The
official startup date should begin when material is first introduced Into the
converter.
66
-------
REFERENCES
1. Personal communication with Mr. Hess, U.S. Smelting Furnace Co.,
Belleville, Illinois. March 1979.
2. Personal communication with Colorado Smelting and Mine, Denver, Colorado.
March 15, 1979.
3. Personal communication with Treadwell Corp., New York, New York.
March 13, 1979.
4. Personal communication with Pyro Industries, Mineola, New York.
March 13, 1979.
5. Personal communication with General Precision Systems, Glendale,
California. March 1979.
6. Personal communication with Mr. Fluggie, Hevi Duty Heating Equipment,
Watertown, Wisconson. March 9, 1979. -
7. Personal communication with Mr. Gaibreath, Whiting Corporation, Harvey,
Illinois. March 9, 1979.
8. Personal communication with Mr. Bruecker, Surface Combustion, Toledo,
Ohio. March 13, 1979.
9. Personal communication with the Copper Development Association, New York,
New York. March 14, 1979.
10. Personal communication with Mr. Farkes and Mr. Nelson, Dwight Lloyd
Research Laboratories, McDowell—Wellman Engineering Co., Cleveland, Ohio.
March 14, 1979.
11. Personal communication with Mr. Nagite, Pennsylvania Engineering,
Pittsburgh, Pennsylvaniat March 14, 1979.
12. Personal communication with Mr. Douglas, Lectromelt Corporation,
Pittsburgh, Pennsylvania. March 9, 1979.
13. Personal communication with Mr. Kuzio, GE Raymond Bartlett Snow, New
York, New York. March 16, 1979.
14. Personal communication with Mr. Persons, Lectromelt Corporation,
Pittsburgh, Pennsylvania. May 8, 1979.
67
-------
15. Personal communication with Mr. Priestly, Door Oliver, Stamford,
Connecticut. March 1979.
16. Personal communication with Mr. Defull, Colorado School of Mines
Research Institute, Denver, Colorado. March 1979.
68
-------
PRIMARY ZINC SMELTERS - SUBPART Q
§60.170 — 60.176
Introduction
Affected operations within this category are roasting (mul ip1e hearth
and fluid bed) and sintering. Particulates discharged into the atmosphere
from a sintering machine are not to exceed 50 mg/dscm (0.022 gr/dscf) while sul—
fur dioxide discharged from any roaster is not to exceed 0.065 percent by
volume. A sintering machine which drives off more than 10 percent of the sul-
fur initially contained in the zinc sulfide ore concentrates is considered a
roaster. Visible emissions are limited to 20 percent opacity from any sintering
machine or from any affected facility that uses a sulfuric acid plant to
comply with the standard. The owners or operators of a zinc smelter are re-
quired to install and operate a continuous monitoring system for opacity and
sulfur dioxide. Sources constructed, reconstructed or modified after
October 16, 1974, are subject to the regulations.
Process Description
Zinc is typically found in ore called sphalerite containing impurities
of lead, cadimum and trace elements. Sphalerite is normally processed at the
mines to form concentrates containing up to 62 percent zinc and 32 percent
sulfur. The smelting of zinc sulfide into oxide or metallic zinc is carried
out by either a pyrometallurgical or a combined pyrometallurgical—electrolytic
extraction process. The three primary steps of the pyrometallurgical extrac-
tion process are:
1. roasting (multiple hearth and fluid bed) of zinc sulfide
concentrates to ren ve impurities and to form an impure
zinc oxide called calcine.
2. sintering of the calcine to eliminate remaining sulfur,
volatilization of lead and cadmium and formation of a
dense permeable furnace feed, and
3. pyrometallurgical reduction of the zinc oxide to metallic
zinc.
The smelting of zinc sulfide using electrolytic extraction requires two
principal operations:
1. roastIng, and
2. electrolytic extraction, after chemical leaching of calcine
to produce 99.9 percent pure high grade zinc.
69
-------
Figure 15 illustrates the pyrometallurgical and associated process units
typically used in each step. NSPS limit emissions from roasting and sintering
operations and therefore, a startup evaluation was not conducted for the
pyrometallurgical—electrolytic extraction process.
Pre—Startup Operations
Roasting—Multiple Hearth : Two to three weeks are allowed for pre—startup
operations. If problems develop, an additional week is allowed. Curing the
furnace and heat—up takes about 5—7 days. The off—gas stream temperature is
about 700°C (1,292°F). The heat—up must be done very slowly to prevent stress
on refractory material. The total pre—startup operation may consume a 3—4
week period.
Roasting—Fluid Bed : Seven to nine days are required for pre-startup
operations. Three to four days are needed for the heat—up period. A fluid bed
furnace does not have refractory material so the furnace does not need
curing. The total time allowed for pre—startup is 2—3 weeks.
Sintering Machine : The Installation and checkout of a sintering machine
is a very complex operation Involving many suppliers and contractors. This
period of time is called construction and test—out and may require 3—4 months
to complete. A major factor affecting the length of this period of time is
the geographical location of the plant; isolated areas may make it difficult
to obtain spare parts thus delaying startup. The pre—startup operation is
the most hazardous period during which component parts may be damaged. The
test—out is done without load material. Some of the items checked are:
instrumentation, structure, electric motors, mechanical clearances, conveyor
beltB, mixing drums, sinter breakers, and coolers. Curing the machine
is a 5 to 7 day operation. Upon successful completion of the construction and
test—out, the sintering machine is ready to receive material. The maximum
time period for construction and test—out is 6 months.
Startup Operations
Roasting—Multiple Hearth : After pre—startup operations, ore is slowly
introduced into the furnace in order to prevent thermal shocks and gradients
in refractory material. The load is gradually increased to maximum capacity
over a 1 to 2 week period. The startup date should be defined as the time
material (ore) is first introduced into the furnace.
Roasting—Fluid Bed : Folloying the pre—startup and heat—up operations,
the ore is slowly introduced Into the bed, Increasing to the maximum feed rate
in 2—4 days. The official startup date should begin when material is first
introduced into the bed.
70
-------
concttrate
ROASTING
• ROPP & Multiple Hearth
• Flash ) SO 2
• Fluid Bed
Calcine ZnO
O 1de sintering J so 2
furnace
sinter
ZnO
Reduction
.1
Metallic
zinc
Figure 15. The pyrotnetallurgical process.
71
-------
Sintering Machine : After the pre—startup operation and heat—up is com-
pleted, the sintering machine normally goes into a one shift per day operation
for 2—3 weeks. This period allows the operators to become familiar with the
equipment. Each daily startup takes about 1˝ hours to produce a product.
Before going into maximum production, the machine will go into a 24 hr/day
operation for about 2 weeks. By the end of the second week, operators and the
machine ehould be ready for maximum production. However, the inexperienced
operator will extend this time. On the average, the sintering machine and
operator should be able to reach maximum production in 3—6 months. The time
needed to reach maximum production for a sintering machine varies; new sinter
machine designs and new processes require more attention than off—the—shelf
sintering machines. The first year’s production rate normally ranges between
70 to 80 percent because of malfunctions and repair time; however, the produc-
tion rate usually increases in the succeeding year to a maximum production
rate (90 to 100 percent). The official startup date should be designated as
the day when material is first introduced into the machine. Manufacturers
indicated that 180 days is a reasonable time period in which to reach maximum
production for conducting performance testing.
72
-------
REFERENCES
1. Personal communication with Mr. Williams, Dwight Lloyd Research Labora-
tories, McDowell—Wellman Engineering Co., Cleveland, Ohio. March 1979.
2. Personal communication with Pyro Industries, Mineola, New York.
March 1979.
3. Personal communication with Mr. Bruecker, Surface Combustion, Toledo,
Ohio. March 1979.
4. Personal communication with Mr. Nelson, Dwight Lloyd Research Laboratories,
McDowell—Wellman Engineering Co., Cleveland, Ohio. March 1979.
5. Personal communication with Mr. Coulter, Dravo Corporation, Pittsburgh,
Pennsylvania. March 15, 1979.
6. Personal communication with Mr. Farkes, Dwight Lloyd Research Laboratories,
McDowell—Wellman Engineering Co., Cleveland, Ohio. March 14, 1979.
7. Personal communication with Mr. Latoweski, Kopper Company, Pittsburgh,
Pennsylvania. March 15, 1979.
8. Personal communication with Mr. Nell, the Bethlehem Corporation, New York,
New York. March 14, 1979.
9. Personal communication with Mr. Smith, Mine and Smelter Supply, Denver,
Colorado. March 1979.
10. Personal communication with Mr. Priestly, Door Oliver, Stamford,
Connecticut. March 1979.
73
-------
PRIMARY LEA]) SMELTERS - SUBPART R
§60.180 — 60.186
Introdiction
This NSPS category regulates the following facility operations for primary
lead smelting: sintering machines, sintering machine discharge end, blast
furnaces, dross reverberatory furnaces, electric smelting furnaces and con-
verters. Particulate emissions from any blast furnace, dross reverberatory
furnace or sintering machine discharge end are limited to 50 mg/dscm (0.022
gr/dacf). Sulfur dioxide is limited to 0.065 percent by volume from any
sintering machine, electric smelting furnace, or converter. Opacity is limited
to 20 percent from blast furnaces, dross reverberatory furnaces, sintering dis-
charge end, or any facility that uses a sulfuric acid plant to comply with
standards. Lead smelting facility owners and operators are required to install
a id operate continuous monitoring equipment for opacity and sulfur dioxide.
Sources constructed, reconstructed or modified after October 16, 1974, are
subject to the regulations.
Process Description
Lead is found in nature as a sulfide ore containing impurities of copper,
zinc and other trace elements. At the mines, the lead ore is normally concen-
trated to 65—70 percent. Typically, the concentrate contains between 13—19
percent by weight of sulfur. The production of lead includes the following
operations:
Sintering : Lead and sulfur are oxidized to produce lead oxide and to
reduce the sulfur dioxide concentration to 85 percent. Simultaneously, the
charge material composed of recycled sinter, sand and inert material is
agglomerated to form a dense, permeable material called sinter. The sinter
is then charged to a reduction unit such as a blast furnace.
Reduction : In the blast furnace, lead oxide is reduced to a molten lead
bullion. The lead bullion then goes to the refinement steps.
Refinement : Impurities are removed from the lead bullion.
Figure 16 illustrates a flow diagram of a lead smelting process. Smelting
manufacturers indicated that electric furnaces and converters are not used in
the U.S. because of the high cost of energy; therefore, a startup evaluation
‘as not conducted for these units.
74
-------
1.
3.
502
lead bullion
Figure 16. Lead smelting flow diagram with typical process units.
2.
REDUCTION
Blast furnace
Dross Reverberatory
ctric_Smelting
lead
75
-------
Pre—Startup Operations
Sintering Machine : The installation and checkout of a sintering machine
is a very complex operation involving many suppliers and contractors. This
period of time is called construction and test—out; it may last 3—4 months. A
major factor affecting the length of this period is the geographical location
of the plant; isolated areas may make it difficult to obtain spare parts thus
delaying startup. The pre—startup operation is the most hazardous period dur-
ing which component parts may be damaged. The test—out is done without load
material. Some of the items checked are: instrumentation, structure,
electric motors, mechanical clearances, conveyor belts, mixing drums,
sinter breakers, and coolers. Curing the machine is a 5—7 day operation. In-
experienced operators are a major problem in achieving a maximum production
rate within a reasonable time frame. Upon successful completion of the con—
truction and test—out, the sintering machine is ready to receive material.
The maximum time period for construction and test—out is 6 months.
Reduction — Blast Furnace : The blast furnace for lead smelting is unlike
that used in the Iron and Steel industry in that it has no refractory material,
thereby eliminating the curing requirement. When a lead blast furnace is
installed or reconditioned, each system is checked and reconditioned or re-
placed. A final checkout is then conducted coordinating the functions of all
systems and units operating at rated capacity. This test period takes 2—3
hours. Some of the major systems/units examined are the water packet cooling
system, flange corking, charging, weighing, fuel and all mechanical, electri-
cal, and hydraulic systems in general. The furnace is now ready to receive
a charge.
Reduction — Dross Reverberatory : Once the furnace is installed and all
connections made, pre—startup operations begin. The pre—startup operaton
takes 2-3 days. Simultaneously, the furnace is cured for 4—i days depending
on the size of the furnace. Finally, the furnace is brought up to operating
temperature and stabilized (2—3 days). The furnace is now ready to receive
material. The total pre—startup operation consumes 5—7 days.
Reduction — Electric Furnace : Electric furnaces in the lead smelting
industry are used for melting rather than smelting. The energy is too demand-
ing for a smelting operation. Two to four weeks are allowed for pre—startup
operations. Manufacturers call this period of time the green tag period
(test—out). Individual, mechanical, electrical and hydraulic components are
checked. Special attention is given to electrode rigging, electrical grounds,
shorts and high resistance. Upon successful completion of all checks, a low
secondary voltage is applied to the transformer. If no problems develop, the
primary voltage (Electric Company) energizes the entire system. Curing the
furnace is either done with electrical or fossil fuel energy. Curing consumes
a 3—4 day period and is normally done during the green tag period. The total
pre—startup operations take about 1—2 weeks.
76
-------
Startup Operations
Sintering Machine : After the pre—startup operation and heat—up is com-
pleted, the sintering machine normally goes into a one shift per day operation
for 2 to 3 weeks. This period allows the operators to become familiar with the
machine. Each daily startup takes about 1˝ hours to produce a product. Before
going into maximum production, the machine will go into a 24 hr/day operation
for about 2 weeks. By the end of the second week, operators and the machine
should be ready for maximum production. On the average, the sintering machine
should be able to reach maximum production in 3 to 6 months. The time needed
to reach maximum production for a sintering machine varies, new sinter machine
designs and new processes require more attention than off—the—shelf sintering
machines. The first years production rate normally ranges between 70 to 80
percent because of malfunction and repair time; however, the production rate
usually increases in the succeeding year to maximum production rate (90 to 100
percent). The of fical startup date should be when material is first introduced
into the machine. Manufacturers indicated that the 180 day performance test
period is a reasonable time in which to reach maximum production.
RedUction — Blast Furnaces : Following a 2—3 hour pre—startup, the furnace
I.s ready to receive a charge and will normally reach a maximum production rate
In 24 to 48 hours. The startup procedure for a lead blast furnace first re-
quires the hand stacking of wood in the furnace. This operation takes 6—12
hours depending on the size of the furnace. Next, the blast furnace is charged
with oil—soaked coke (3—5 hours). The blast furnace is ignited, brought up to
operating temperatures and sinter is slowly fed into the furnace until the
maximum charge rate is reached. During this time, other materials such as
limestone, silica, litharge and slag—forming materials are added to the furnace
to develop a high quality lead bullion pool for continuous operation. If the
charge period goes smoothly, a blast furnace can reach maximum production in
24 hours; however, minor problems will extend this time to 2 days. The life of
a blast furnace normally does not exceed 180 days, therefore, there should be
no problems in testing this unit under maximum production conditions with the
specified time (180 days). The official startup should be designated as the
time when the wood and coke are first ignited.
Reduction — Dross Reverberatory : After pre—startup and heat—up operations,
lead builtori is slowly fed into the furnace to prevent thermal shock and to
maintain a stable temperature. The time required to reach maximum production is
3—5 days after material is first introduced into the furnace. The total time
required from a cold start to maximum production will range from 5—7 days. The
official startup date should be when material is first introduced into the
furnace.
Reduction — Electrical Furnace : Following pre—startup and heat—up, lead
bullion is slowly introduced into the furnace. The transformer is then ener-
gized, heating the furnace and charge to desired temperatures; the introduction
of more charge lowers the temperature and the furnace is energized again. This
operation occurs many times until the furnace is filled to maximum capacity.
The time required to reach maximum production Is about 5 days from a cold start.
Two to three days are considered heat—up time. The official startup date should
begin when material is first introduced into the furnace.
77
-------
REFERENCES
1. Personal communication with Mr. Williams, Dwight Lloyd Research Labora-
tories, McDowell—Wellman Engineering Co., Cleveland, Ohio. March 1979.
2. Personal communication with Mr. Schick, Mohr John and Son, Chicago,
Illinois. March 1979.
3. Personal communication with Mr. Gallaway, Chicago Bridge and Iron,
Chicago, Illinois. March 1979.
4. Personal communication with Mr. Gail, Ingall Iron Works, Birmingham,
Alabama. March 1979.
5. Personal communication with Colorado Smelting and Mine, Denver, Colorado.
March 1979.
6. Personal communication with Mr. Johnston, St. Joe Lead Co., Herculaneun,
Missouri. March 29, 1979.
7. Personal communication with Mr. Boyd, the Bunker Hill Company, Kellogg,
Idaho. March 29, 1979.
8. Personal communication with Rico Argentine Mining Co., Rico, Colorado.,
March 29, 1979.
9. Personal communication with Mr. Nelson, Dwight Lloyd Research Labora-
tories, McDowell—Wellman Engineering Co., Cleveland, Ohio. March 1979.
10. Personal communication with Mr. Coulter, Dravo Corporation, Pittsburgh,
Pennsylvania. March 15, 1979.
1]. Personal communication with Mr. Farkes, Dwight Lloyd Research Laboratories,
McDowell—Wellman Engineering Co., Cleveland, Ohio. March 14, 1979.
12. Personal communication with Mr. Latoweski, Kopper Company, Pittsburgh,
Pennsylvania. March 15, 1979.
13. Personal communication with Mr. Jensen, AMAX Lead Homestake, Buick Mines,
Missouri. March 1979.
14. Personal communication with Mr. Douglas, Lectromelt Corporation, Pittsburgh,
Pennsylvania. March 9, 1979.
78
-------
15. Personal communication with Dr. Lyman and Dr. Cole, Lead Industrial
Association, New York, New York. March 1979.
16. Personal communication with Mr. Priestly, Door Oliver, Stamford,
Connecticut. March 1979.
79
-------
PRIMARY ALUMINUM REDUCTION PLANTS - SUBPART S
§60.190 — 60.195
Introduction
(Amendments to the NSPS for this category will be promulgated shortly;
the requirements detailed here incorporate these amendments.)
The NSPS for this category applies to potrooin groups and anode bake plants,
regardless of size, which are located at primary aluminum reduction plants.
Performance tests are required at least once each month during the life of the
affected facility, but the owner or operator may petition the EPA administrator
to establish an alternative requiring less frequent testing for a primary con-
trol system or an anode bake plant.
Performance standards have been established for fluorides and opacity.
Fluoride emissions are limited to: (1) 1.0 kg/Mg (2.0 lb/ton) of aluminum
produced for potroom groups at Soderberg plants; (2) 0.95 kg/Mg (1.9 lb/ton)
of aluminum produced for potroom groups at prebake plants; and (3) 0.05 kg/Mg
(0.1 lb/ton) of aluminum equivalent for anode bake plants. Potroom emissions
could exceed the applicable Soderberg or prebake plant standard, but could
never exceed 1.30 kg/Mg (2.6 lb/ton), if an owner or operator could establish
that the emission control system was properly operated and maintained at the
time the excursion above the standard occurred. Opacity must be less than
10 percent for potroom groups and less than 20 percent for anode bake plants.
Sources constructed, modified, or reconstructed after October 23, 1974 are
subject to the regulations.
Process Description
All aluminum produced in the United States is by electrolytic reduction of
alumina (A1 2 0 3 ). Alumina is itself a product produced from Bauxite ore. Fig—
ure 17 presents a flow diagram of the aluminum reduction process. The process
is carried out in shallow rectangular cells (Pots) constructed of carbon coated
steel shells (cathodes) utilizing consumable carbon blocks (anodes) which are
suspended above and extend down into the pot. A series of these cells, when
connected to a conm n electrical supply, is considered a potline. Potlines are
normally constructed as two adjacent buildings approximately 1,200 ft long.
Cryolite, a double fluoride salt of sodium and aluminum (Na 3 A1F 6 ), serves
as both an electrolyte and a solvent for alumina. The resistance of the cells
causes them to be heated and operate between 9500 and 980°C (1,740° to 1,830°F).
During the reduction process, aluminum ions are reduced to aluminum. The
oxygen released reacts with the anode to form carbon dioxide and carbon mon-
oxide. Molten aluminum is periodically tapped from beneath the cryolite bath
and moved to holding furnaces prior to being cast into ingots.
-------
PASTE
ELECTRICITY
CHLORINE
mO CHLORIDE
SALTS
Figure 17. Process flow diagram for primary aluminum reduction. 5
-------
Three different types of cells are used for the production of aluminum:
the vertical stud Soderberg (VSS), the horizontal stud Soderberg (HSS), and the
prebake (PB). These cells differ in the design for introducing the electrical
current to the cryolite bath. Soderberg cells utilize a consumable anode which
is baked in situ . A mixture of ground petroleum coke and coal tar pitch is
periodically added to the top of the anode to renew It. Heat from the process
bakes the lower boiling organics and fuses the new material to the old anode.
This system precludes the need for a separate anode baking facility.
The PB cell uses an anode that is prebaked. Since the anode is consumed
during normal operation, the old anode remnants (butts) are replaced periodically
with new anodes. These butts are cleaned, ground, mixed with new coke and
blended together with coal tar pitch in an anode plant. The mixture is weighed,
formed into bricks, then slowly baked in a furnace. This process cycle can
take up to 60 days.
Fluoride emissions from these processes are controlled by several methods:
dry scrubbers (baghouses), venturi scrubbers, wet electrostatic precipitation,
or impingement devices. Dry scrubbers are the most conmion, and all utilize
cell grade alumina to adsorb the gaseous fluoride. This alumina is then used in
the pots for electrolytic reduction. Various methods such as coated bags,
alumina Injection, and fluldized bed dry scrubbing, are employed to provide
fluoride/alumina contact.
Pre—Startup Operations
In a prebaked operation, the anode plant is necessarily constructed and put
into operation 2 to 6 months before aluminum is produced In the potilne opera-
tion. This insures that a sufficient supply of anodes is available for produc-
tion. Anode production involves crushing, shaking, screening, classifying,
mixing, pressing, and baking of the raw materials. Each of these mechanical
systems is checked without load prior to startup. These tests are made after
each system is operational and collectively will take 2 to 3 months. When the
anode plant includes a baking furnace, it must be conditioned by curing the
refractory for 2 to 3 weeks on a low fire before the first green anodes are
baked.
Major components of an aluminum reduction plant include materials handling
equipment (cranes, conveyors, pumps), electrical equipment (rectifiers, trans-
formers, busbars), steel—shell pots and associated anode support structure, and
the air pollution control equipment. These systems are all individually checked
before plant startup to insure they are operational. However, all systems,
especially the electrical and control equipment, require a load to insure they
meet design specifications.
The green cathode is “baked—out” by placing resistor coke in the pot and
setting the anode on this coke. Current appjied through the anode serves to
heat, dry, and bake the green cathode lining. This operation takes two days
per pot, and is the initial step of startup.
82
-------
Startup Operations
Frebake plants (considered likely to be the most widely used type in the
future) require that pots be started sequentially, such that startup of a pot—
line is considered to be the day the first pot produces aluminum at its designed
capability. A preliminary step in potline startup (those equipped with
dry scrubbing/baghouse systems) is the activation of the primary air pollution
control system, including alumina movement through the system. This ensures
that fluoride emissions from all pots are controlled as they come on line. De-
sign flow rates and temperatures, however, are not reached until all pots are in
service. The individual pots are placed on line, one at a time, with a maximum
of three or more pots started per day. Some plants may elect to introduce up
to 20 pots at one time. In these cases, the pots are operated for 1 week before
additional pots are energized. For typical potline sizes of 100 to 180 pots,
the entire startup procedure may last from 150 to 300 days. In light of the
fact that this time frame may exceed the 180—day testing requirement, enforce-.
ment proceedings may be required to establish performance testing schedules
at individual plants. Greenfield (new potline in a new location) pot startup
is accomplished by adding powdered cryolite to a pot and slowly heating it to
970°C (1,780°F). During this time, the anode—cathode distance is opened, the
electrical equipment is monitored for short—circuits and the pots are contin-
ually inspected to insure against breakdown of the cathode lining and anode
overheating. This phase is critical as a poorly adjusted anode can quickly
overheat and destroy the cathode lining. Since the pots are connected in
series, an electrical failure will trip the entire potline. Initiation of
metal tapping takes from 7 to 10 days after alumina is added, as a stable mol-
ten aluminum pad must be established in every pot. Testing would be required
within 60 days of the last pot being put into service.
Startup of an anode bake plant is considered to be the first baking of
green anodes. Fluoride emissions from this operation will commence (a) with
this first bake, if the anode plant is an expansion of an existing facility
and butts, containing fluoride, are recycled, or (b) with the initial recycling
of butts in a greenfield operation.
Soderberg plants, which are not likely to be chosen for new sources in the
years to come, may require from 180 to 360 days to meet maximum potline produc-
tion, as the physics of in—situ anode production is complex and there is greater
potential for problems. Since the anodes are baked in—situ , the entire anode
production process must be de—bugged before aluminum production can begin.
The initial anode for each pot must be mixed, incorporated into a support sys-
tem, connected electrically and then baked before alumina is added. An attempt
at production can be made once the individual anode is ready. Should an anode
problem occur the entire support assembly must be disassembled, and a new anode
mixed, reassembled and baked before another attempt is made. As this time con—
suming procedure may occur at each pot, the ability of an entire potline to
meet maximum production may be delayed, requiring enforcement proceedings to
establish performance testing schedules.
83
-------
REFERENCES
1. Personal communication with Mr. Alberts, Noranda Corporation, New Madrid,
Missouri. March 29, 1979.
2. Personal communication with Mr. Howarth, Alumex Corporation, Buc Keystown,
Maryland. March 29, 1979.
3. Personal communication with Mr. Jansen, Eastalco Corporation, Frederick,
Maryland. March 30, 1979.
4. Personal communication with Mr. Yeager, Anaconda Aluminum Corporation,
Louisville, Kentucky. April 23, 1979.
5. Technical Guidance for Control of Industrial Process Fugitive Particulate
Emissions. 1977. U.S. Environmental Protection Agency, EPA—450/3—77—01O.
84
-------
WET PROCESS - PHOSPHORIC ACID PLANTS - SUBPART T
§60.200 — 60.204
Introduction
For the purposes of this regulation, wet process phosphoric acid plants
are defined as any facilities manufacturing phosphoric acid by reacting phos-
phate rock and acid. Affected facilities include any combination of reactors,
filters, evaporators, and hotwells. Emissions of total fluorides are limited
to 10 g/metric ton (0.02 lb/ton) of equivalent P 2 0 5 feed. Equivalent P 2 0 5
feed is defined as the quantity of phosphorous, expressed as phosphorous
pentoxide, fed to the process. Sources constructed, reconstructed, or modi-
fied after October 22, 1974, are subject to the regulation.
Process Description
The sulfuric acid or wet process produces phosphoric acid by the reaction
of phosphate rock with sulfuric acid which also results in the precipitation of
calcium sulfate (gypsum). Figure 18 is a simplified flow diagram of the pro-
cess. Phosphate rock is ground in a special ball mill with dilute phosphoric
acid (recycled) and the resultant slurry mixture is then passed into multiple
reactor and digesting tanks for reaction with sulfuric acid. Cooling is re-
quired to remove the heat of reaction. Water vapor and gaseous impurities are
carried to an absorber where fluosilicic acid is recovered. Acid digestion of
the slurry requires 4 to 8 hours at temperatures of about 750C (1670F). Violent
agitation and close temperature control are required for the production of
uniform, easily washed and filtered gypsum (CaS0 2H 0) crystals. Without
this close control, the anhydrite would form, become subsequently hydrated, and
result in plugging of pipes.
Slurry from the digester tanks passes into horizontal, rotating, tilting—
pan—type vacuum filters where phosphoric acid (30 to 35 percent P 2 0 5 ) is removed
from the filter cake. This acid filtrate is then concentrated to 54 percent
P 2 0 5 by evaporation.
Pre—Startup Operations
Mechanical checkout is begun as soon as specific units or groups of inter-
related units are reported as mechanically complete by the construction con-
tractor. These checks include:
• alignment of all motor driven equipment
• hydrostatic testing of all liquid handling equipment
85
-------
WATER
u —S
VACUUM FLASH
PHOSPHATE ROCK
ICAL.CINED)
PLUOSILIC IC
ACID TO RECOVERY
a,
a’
WASH WATER
F ILl RATE
RECEIVERS
PHOSPHORIC ACID
(30—35%
AC ID
154% P 2 0 5 )
Figure 18.
Flow diagram of the wet—process phosphoric acid production process.
-------
• initial checkout of all instrumentation loops, with physical
or electronic calibration where possible
• integrity check of all electrical circuits and proper rotation
of electrically driven units
• proper performance of vacuum generating equipment and vacuum
handling equipment such as flash coolers and filtration units.
These steps would be carried out by both maintenance and mechanical
engineering personnel and would take at least 1 week. Any significant design
faults or failures could appreciably alter this time period.
Once mechanical acceptance has been completed, process evaluation is be-
gun by operating all units “on water” to check for leaks and to evaluate pump
capacities and instrument calibration.
Startup Operations
Once the system has been operated on water for a long enough time to
insure equipment integrity and the ability of driven equipment to perform
within design limits, staged operation can be initated. Startup is best
defined as the point when phosphate rock and sulfuric acid are first added to
the reaction vessel. The normal feed ratio is about 726 Kg (0.8 tons) of 100
percent sulfuric acid per 907 kg (ton) of rock and varies depending on the
quality of rock. These two raw materials are added at low rates initially,
until the slurry in the reaction tank is of proper concentration and consis—
tency. Under normal circumstances, this initial charging of the reactor can
take 8 to 16 hours while the proper concentration is usually achieved in 16
to 48 hours. The plant is then operated at continuous flow conditions but
only at half capacity. As confidence in and knowledge of equipment and process
is gained, rates are increased to design levels. This can involve 1 to 3 weeks
in the absence of unusual circumstances. Once design capacity is achieved,
the plant initiates a 5 to 7 day test period, during which time all operational,
mechanical, and chemical parameters must be measured to insure that all equipment
operates within design conditions and that all guaranteed process rates,
efficiencies, and recoveries are met. Upon completion of a successful rest
period, the plant is legally accepted (possibly with provisions to upgrade
equipment if necessary).
87
-------
REFERENCES
1.. Personal communication with Mr. C. N. Hebbard, New Wales Chemicals, Inc.,
Mulberry, Florida, March 23 and March 27, 1979.
2. Personal communication with Mr. Alan Martin, W. R. Grace and Company,
Bartow, Florida, April 6, 1979.
3. Personal communication with Mr. Robert Schmidt, Gardinier, Inc., Tampa,
Florida, April 18, 1979.
4. Personal communication with Mr. James C. Daniel, U.S.S. Agrichem/Division
U.S.S. Steel, Fort Meade, Florida, May 18, 1979.
88
-------
SIJPERPHOSPRORIC ACID PLANTS - SUBPART U
§60.210 — 60.214
Introduction
Superphosphoric acid plants regulated under this part are defined as any
facility which concentrates wet—process phosphoric acid to 66 percent or greater
phosphorous pentoxide (P 2 0 5 ) content by weight for eventual consumption as a
fertilizer. Affected facilities include any combination of evaporators, hot—
wells, acid sumps, and cooling tanks. Total fluoride emissions are limited to
5 g per metric ton (0.01 lb/ton) of equivalent P 2 0 5 feed, defined as the
quantity of phosphorous pentoxide fed to the process. Sources constructed,
reconstructed or modified after October 22, 1974, are subject to the
regulation.
Process Description
Superphosphoric acid is produced by concentrating 54 percent P 2 0 5 phos-
phoric acid to about 70 percent (± 2 percent) P 2 0 5 . Two coimnercial processes
used to accomplish this are vacuum evaporation and submerged combustion, al-
though the latter process is virtually outdated. In the vacuum evaporation
process, depicted in Figure 19, clarified 54 percent P 2 0 5 acid is continuously
fed to a vacuum evaporator from which hot gases containing water vapor and
fluorides are condensed in the water—cooled barometric condenser. Condenser
water flows to the hotwell prior to draining to the gypsum pond. Concentrated
acid is drawn from the evaporator to cooling tanks and then to storage.
In the submerged combustion process, Figure 20, hot gases are forced
below the surface of the 54 percent P 2 0 5 phosphoric acid in a submerged com-
bustion evaporator. Water vapor, fluorides, and acid mist are driven from
solution and concentrated acid is drawn off as product.
Pre—Startup Operations
The major piece of equipment requiring a thorough mechanical checkout is
the evaporator. In the past, falling—film evaporators have been used but have
been replaced more recently with forced—circulation units. A disadvantage of
the falling—film type was the inability to uniformly distribute the liquid
as a film inside the exchanger tubes.
The evaporator is typically a shell and tube exchanger with high pressure
steam Input, a high—volume, low—head, recirculation pump, and is operated at
about 6.75 to 20.3 kPa (2 to 6 in. Rg) absolute pressure. Initial checking of
the evaporator assembly Is carried out by the contractor while the plant will
89
-------
STEAM
54% P 2 0 5
ACID
0
TO STORAGE
Figure 19. Superphosphoric acid production by the vacuum evaporation process.
-------
TEIT&RINS
EVAPORATOR
i-I
54%
ACID
STORASE
Figure 20. Superphosphoric acid production by the submerged combustion process.
-------
perform their own debugging procedures when the equipment is delivered. The
plant would check the circulation pump for alignment, rotation, and with
couplings disengaged, and also check bearings and packing. All acid—handling
equipment Is filled with fresh water for about 2 hours during which time leaks
are detected and pumps, ammeters, flow meters, etc. are checked for capacity
and calibration. A plant engineer will also inspect the equipment for proper
welds and to verify the construction material (usually Hastelloy G — 20 to 30
percent nickel). Operating vacuum Is then pulled on the system and held for
at least 8 hours to check for vacuum leaks. Steam lines are then blown out
to remeve mill scale, welding rods, or other extraneous materials. Water and
steam are then Introduced to the plant and the water is evaporated for 5 to 6
hours to check pumps, instrument loops, control valves, and other pertinent
components.
Startup Operations
Upon completion of the process equipment checks, the system is recharged
with acid and the acid feed Is initiated; this can be considered as plant
startup. For normal circumstances involving no equipment failures or innovative
designs requiring additional process evaluations, the preliminary equipment
checks consume about 2 weeks and the maximum design rate can be achieved in
about 48 hours. Approximately 2 to 3 weeks after the startup period, a per-
formance run of about 3 days will begin, during which time all phases of opera-
tion must be as specified including emission rates and plant efficiency ex-
pressed as unit weight of product per unit weight of steam input. Under normal
conditions, 180 days is adequate time for conductance of source tests.
92
-------
REFERENCES
1. Personal counminication with Mr. Jack Smith, J. R. Simplot Co., Pocatello,
Idaho, May 15, 1979.
2. Personal comnunication with Mr. C. G. Meier, Farmland Industries, Bartow,
Florida, May 15, 1979.
3. Personal connminication with Mr. Larry Hinderager, J. R. Siinplot Co.,
Pocatello, Idaho, Nay 18, 1979.
4. Personal communication with Mr. Joseph G. Peters, New Wales Chemicals,
Inc., Mulberry, Florida, May 29, 1979.
5. Personal communication with Mr. Thomas Faulkner, Swenson Evaporators/
Division of Whiting Corporation, Harvey, Illinois, Nay 30, 1979.
6. Background Information for Standards of Performance: Phosphate Fertilizer
Industry EPA—450/2--74—O].9a, October 1974.
93
-------
DLANMONIUM PHOSPHATE PLANTS - SUBPART V
60.22O — 60.224
Introduction
Granular dianimonium phosphate plants regulated in this part are defined
as any plant manufacturing granular diauimonium phosphate by reacting phosphoric
acid with aninonia. Affected facilities within each plant include any coinbina—
tion of reactors, granulators, dryers, coolers, screens, and mills. Emissions
of total fluorides are limited to 30 g/metric ton (0.06 lb/ton) of equivalent
P 2 0 5 feed (the quantity of phosphorous, expressed as phosphorous pentoxide,
fed to the process). Sources constructed, reconstructed or modified after
October 22, 1974, are subject to the regulation.
Process Description
Two possible methods exist for the manufacture of diannnonium phosphate
(DAP); one results in run—of—pile product (large pellets), made in a TVA cone
mixer, which cures in a storage pile and is then granulated; the other results
in a slurry which is granulated in an ammoniator or granulator. Since the
latter (slurry) process is the most likely to be employed in new facilities,
the following process description, flow diagram, and startup operations will
apply to this type.
The production of DAP (high—analysis fertilizer) results in different
compositions depending on the type of phosphoric acid used in the process. If
phosphoric acid made by burning elemental phosphorous in an electric furnace is
used, the resulting fertilizer composition is 21—54—0 (21 percent total nitrogen,
54 percent available phosphate as phosphorous pentoxide, and 0 percent soluble
potash as K 2 0). If wet—process phosphoric acid is used, the resulting analysis
is 18—46—0. A flow diagram of a typical plant is illustrated in Figure 21.
Vapor or liquid anhydrous ammonia and phosphoric acid are proportioned to
an agitated atmospheric tank (preneutralizer) to maintain a ratio of 1.3:1.5
moles of ammonia per mole of phosphoric acid. The exothermic reaction that
takes place results in the evaporation of approximately one tenth of the
product as water. The slurry obtained from the preneutralizer flows into the
aimnoniator—granulator at about 121°C (250°F). As it is distributed over the
bed of solid material, it reacts with additional ammonia fed through a distribu-
tor pipe beneath the bed to complete the reaction to a mole ratio of 2.0 (diam—
monium phosphate). Material from finished product screening is recycled to
the ammoniator along with additional solid raw materials to aid in moisture
control. The moist granules exiting the ammoniator proceed through an oil—
or gas—fired concurrent rotary dryer where moisture content is lowered to about
94
-------
WATER
CYCLONE CYCLONE
U ,
bUCKET
ELEVATOR
“ Na
Figure 21. Process flow diagram for dianunonium phosphate production .
-------
1 percent. Dry product is then cooled by a countercurrent flow of air in a
rotary cooler and screened with coarse material being returned to the anunoniator.
In some plants, hot screening is employed such that cooling is achieved prior
to storage and shipping. Exhaust gases from the dryer (and cooler) pass through
cyclones or wet scrubbers for particulate removal while effluent from the ammo—
niator are usually scrubbed with incoming phosphoric acid prior to discharge
to the atmosphere.
Pre—Startup Operations .
As depicted in Figure 21, the process consists of a wet end (prior to
ammoniator) and a dry end (anunoniator and downstream). All equipment on the
wet end is water—batched to check for leaks and proper valve operation. All
pumps, electrical connections, and instrumentation are thoroughly checked and
evaluated with respect to proper operation and calibration.
The granulator, dryer, and cooler (if used) are all inclined slightly and
must be mechanically checked for proper rotation and tracking. The gear box
and driving mechanism must mesh properly and operate without excess noise or
heat. This can take from 2 hours to 2 weeks.
The dryer furnace would undergo preliminary operations similar to those
de8cribed previously for fossil fuel boilers.
Once all equipment is installed and verified for proper installation, con-
nections and mechanical operation, the plant is operated dry with recycle ma-
terial (rather than slurry) to check conveyors (recycle, belt, and elevator)
for binding prior to initial charging. All other equipment is also run dry
prior to initial charging.
Startup Operations
Once all process equipment has been operated with dry recycle material and
has been shown to operate reliably, the preneutralizer (or reactor) is charged
with ammonia, water, and phosphoric acid and the process Is begun. Initial
charging varies depending on the size of the plant but usually would require
45—91 metric tons (50—100 tons) of DAP.
The key, controlling factor with respect to proper operation of the entire
plant is the particle size of the finished material. Fines are recycled to the
granulator while oversize material is recycled to the mills and back to the
granulator. Recycling is Important to the process and too much on—size material
can also be a problem. Normal recycle to product ratio is 3:1. Throughout the
life of the plant, the operators are constantly trying to find the one limiting
piece of equipment which could affect the entire facility’s capability.
Startup is best defined as the time of initial firing of the furnace and
charging of the preneutralizer with ammonia and phosphoric acid. This would
be the first time for potential fluoride emissions and would come after all
equipment has been thoroughly debugged. The controlling factor would be the
specific gravity of the slurry. It would normally take about 2 to 6 hours to
sufficiently heat and charge the system and full production rate could be
achieved in 2 weeks. A plant acceptance run would usually be performed shortly
thereafter.
96
-------
REFERENCES
1. Personal comnunication with Mr. George Chambers, Beker Industries, Inc.,
Hahnville, Louisiana, April 19, 1979.
2. Personal communication with Mr. Schwarer, Brewster Phosphates, Luling,
Louisiana, April 6, 1979.
3. Background Information for Standards of Performance: Phosphate Fertilizer
Industry. EPA—450/2—74—Ol9a. October 1974.
4. Shreve, R. Norris, Chemical Process Industries, 3rd EditIon, McGraw Hill
Book Co., 1967.
5. Personal communication with Mr. James C. Daniel, U.S.S. Agrichem/Division
U.S.S. Steel, Fort Meade, Florida, May 18, 1979.
97
-------
TRIPLE SUPERPHOSP}IATE PLANTS - SUBPART W
§60.230 — 60.234
Introduction
A triple superphosphate plant is defined under this subpart as any facility
manufacturing triple superphosphate (TSP) by reacting phosphate rock with phos-
phoric acid. Affected facilities include any combination of mixers, curing
belts (dens), reactors, granulators, dryers, cooletç screens, mills, and
facilities which store run—of—pile TSP (any triple superphosphate that has not
been processed in a granulator and is composed of particles at least 25 percent
by weight of which (when not caked) will pass through a 16 mesh screen). Emis-
sions of total fluorides are limited to 100 g/metric ton (0.20 lb/ton) of
equivalent P 2 0 5 feed (the quantity of phosphorous, expressed as phosphorous
pentoxide, fed to the process). Sources constructed, reconstructed or modified
after October 22, 1974, are subject to the regulation.
Process Description
A typical flow sheet for TSP production is given in Figure 22. Triple or
concentrated superphosphate can contain from 44 to 51 percent available P 2 0 5 ,
as compared to 16 to 20 percent P 2 0 5 available in normal superphosphate. Ground
phosphate rock containing 75 percent BPL (bone phosphate of lime or tricalcium
phosphate), reacts with phosphoric acid in the granulator (also called a blunger
and similar to a pug mill) with steam, water and recycled fines. Recycling is
used to control moisture and temperature for proper granulation. The phosphoric
acid is preheated (only if made by the wet—process) and fed to the granulator
from beneath the bed through a perforated pipe. The granules formed then over-
flow the dam at the end of the granulator and pass into a rotary cooler (or
bucket elevator to the screening area if the cooler is installed further down-
stream). Exhaust gases pass through a cyclone where collected dust is recycled
to the blunger. The product is screened with oversize material directed to a
mill and recycled back to the granulator. In plants utilizing hot screening,
the cooler can be located downstream of the preliminary storage bin and prior
to final product storage and shipping. Exhaust gases from the granulator and
cooler are scrubbed with water to r nove silicofluorides.
Pre—Startt! p Operations
Because of the similarity between dlammonium phosphate and triple super—
pho8phate production, the preliminary equipment checking operations are basically
the same. Equipment is rotated to ensure proper mechanical operation, align-
ment, and acceptable vibration. Electrical interlocks are checked for proper
shutdown sequences such that serious downstream spills can be prevented in the
event of upstream equipment failure.
98
-------
TER
ROCK
CONDENSATE
Figure 22. Flow diagram for triple superphosphate manufacturing process .
-------
All liquid handling equipment are hydrostatically tested (prior to delivery
to the plant) at guaranteed pressures while the plant would test these systems
at design operating levels. These pieces of equipment (scrubbers or acid pre—
heaters) are also filled with water to check for leaks and to verify that
instrumentation and monitoring equipment are functioning properly.
Other procedures undertaken prior to plant startup would include:
• check plant for safety equipment, emergency lights, adequate
egress, handrails, fire extinguishers, etc.;
• review operating procedures to establish best means of
operation;
Startup Operations
Startup would be best defined as the Initial charging of the wet system;
I.e., phosphoric acid to the preheater and water to the scrubbers. Initial
operations take place at half load and, in cases where all components are
properly designed and there are no major equipment failures, rated capacity
can be achieved in 8 hours. Abnormal circumstances can extend this time
period to 6 to 8 weeks.
100
-------
REFERENCES
1. Personal Communication with Mr. Robert H. Dewey, W.R. Grace Company,
Bartow, Florida, April 9, 1979.
2. Personal Communication with Mr. Harold W. Long, Jr., Agrico Chemical
Company, Mulberry, Florida, April 27, 1979.
3. Ba’kground Information for Standards of Performance: Phosphate Fertilizer
Industry. EPA—450/2—74—019a. October, 1974.
4. Shreve, R. Norris, Chemical Process Industries, 3rd Edition. McGraw—Hill
Book Company, 1967.
101
-------
GRANULAR TRIPLE SUPERPHOSPHATE STORAGE FACILITIES - SUBPART X
§60.240 — 60.244
Introduction
This subpart regulates any facility curing or storing granular triple
auperphosphate (CTSP). Affected facilities include any combination of storage
or curing piles, conveyors, elevators, screens, and mills. Emissions of total
fluorides are limited to 0.25 g/hr/metric ton (5 x 10 lb/hr/ton) of equiva-
lent P 2 0 5 stored (the quantity of phosphorous, expressed as phosphorous pent—
oxide, being cured or stored). Sources constructed, reconstructed or modified
after October 22, 1974, are subject to the standard.
Process Description
After manufacture, GTSP is transferred to the storage building for curing
(completion of reaction). Curing serves to increase the physical strength of
the granules and the availability of P 2 0 5 as plant food by further promotion
of the reaction between the phosphoric acid and phosphate rock. The activities
within the storage plant are f.llustrated in Figure 23. Granular product is
tranaf erred to a storage pile where curing takes place and from which fluorides
evolve. Front—end loaders move the product to the screening area from which
oversize material is rejected, pulverized, and returned to the screen; under-
size material 18 returned to the production plant; and on—grade material is
delivered to shipping.
Pre—S tartup Operations
Mechanical equipment which must be checked within the storage facility
are very much similar to the dry segments of the diaimnonium phosphate and
triple superphosphate plants. Conveyors, screens, mills, and elevators are
all operated dry, prior to delivery of material, to verify proper installation
and mechanical operation and to check for excessive vibration and correct
alignment of gear—driven components.
Startup Operations
The operation of the GTSP storage facility Is obviously dependent upon the
operation of the production plant. Therefore, startup would be best defined
as the first delivery of product to the storage pile from the GTSP plant. Pro—
blems that might occur with equipment downstream of the storage pile could
result In a shutdown and subsequent re—startup of both facilities, and there-
fore, enforcement personnel would have to closely follow the operations of
both facilities.
302
-------
SH I PPING
TO
FROM
PRODUCTION
CONVEYOR
Figure 23. Granular triple superphosphate storage plant 3 .
-------
REFERENCES
1. Personal Conm unication with Mr. Robert H. Dewey, W.R. Grace Company,
Bartow, Florida, April 9, 1979.
2. Personal Coimnunication with Mr. Harold W. Long, Jr., Agrico Chemical
Co ipany, Mulberry, Florida, April 27, 1979.
3. Background Information for Standards of Performance: Phosphate Fertilizer
Industry. EPA—450/2—74-019a. October 1974.
104
-------
COAL PREPARATION PLANTS — SUBPART Y
§60.250 — 60.254
Introduction
The NSPS for this category encompasses any facility of greater than 182
metric tons (200 tons) per day process weight which prepares coal by breaking,
crushing, screening, wet or dry cleaning, or thermal drying. Affected facili—
ties include thermal dryers, pneumatic coal — cleaning equipment, coal process-
ing and conveying equipment, coal storage systems, and coal transfer and load-
ing systems. Performance standards are promulgated for particulate matter
from these facilities. Thermal dryer emissions are not to contain greater than
0.070 g/dscm (0.031 gr/dscf) or exhibit 20 percent opacity or greater. Pneu-
matic coal cleaning equipment emissions are not to exceed 0.040 g/dscm
(0.018 gr/dscf) or exhibit 10 percent opacity or greater. An emission limit
of less than 20 percent opacity is promulgated for discharges from any coal
processing and conveying equipment, coal storage systems, and coal transfer
and loading systems. Sources constructed, reconstructed or modified after
October 24, 1974, are subject to the regulation.
Process Descri ption
Coal preparation plants crush, screen, and dry run—of—mine coal. Secondary
crushing is sometimes needed at preparation plants to ensure good separation of
coal from impurities. Classifying screens separate coal by size and route it to
various coal cleaning unit operations. These unit operations include drying,
pneumatic coal conveying, and breaking or crushing. Emission control from these
processes is usually achieved with baghouses and wet suppression systems.
Pre—Startup Operations
There are a number of operations undertaken prior to startup. These opera-
tions are necessary to ensure proper operation of equipment and to prevent
damage to machinery.
For driers, a check of the bearings is made to determine if overheating
occurs. If no problems are encountered, the drier is ready for on—line opera-
tion. Other equipment is checked to determine if it is operating properly and
is then ready to be placed on—line.
Startup Operations
Startup is best defined as the day the first shipment of coal is fed through
the crusher. Conversations with equipment manufacturers and consultants indicate
that once the equipment is installed and checked process operations can be
initiated immediately. Typical times for installation of equipment to attain-
ment of full operational level are as follows:
105
-------
Equipment Time
Crushers 2 weeks,
Driers 1—4 days,
Total Plant (large) 1—3 months.
Based on the aforementioned data, commencement of process operations is
a reasonable date for startup. Testing at full operational level can be achieved
within 180 days of this date with no anticipated problems.
106
-------
REFERENCES
1. Personal communication with Mr. D. Craveman, American Pulverizer Company,
St. Louis, Missouri. 12 March 1979.
2. Personal communication with Mr. Benning, Dravo Corporation, Philadelphia,
Pennsylvania. 13 March 1979.
3. Personal communication with Mr. Carpenter, Jeffrey Manufacturing Co.,
Columbus, Ohio. 12 March 1979.
4. McCandles, L.C. and R.B. Shaver. Assessment of Coal Cleaning Technology:
First Annual Report. EPA—600/7—78—l50, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, July 1978.
107
-------
FERROALLOY PRODUCTION FACILITIES - SUBPART Z
§60.260 — 60.266
Introduc t ion
The Standards of Performance for Ferroalloy Production Facilities affect:
(1) electric submerged arc furnaces which produce silicon metal, ferrosilicon,
calcium silicon, silicomanganese zirconium, ferrochrome silicon, silvery iron,
high—carbon ferrochrome, charge chrome, standard ferromanganese, silicomanganese,
ferromanganese silicon, or calcium carbide; (2) and dust handling equipment.
The pollutants regulated are particulates and carbon monoxide.
Owners or operators are not permitted to discharge from any electric sub-
merged arc furnace any gases into the atmosphere which: (1) exit from a control
device and contain particulate matter in excess of 0.45 kg/MW—hr (0.99 lb/?44—hr)
while silicon metal, ferrosilicon, calcium silicon, or silicomanganese zirconium
is being produced, (2) exit from a control device and contain particulate mat-
ter in excess of 0.23 kg/MW—hr (0.51 lb/14W—hr) while high—carbon ferrochrome,
charge chrome, standard ferromanganese, silico—manganese, calcium carbide,
ferrochrome silicon, ferromanganese silicon or silvery iron is produced,
(3) exit from a control device and exhibit 15 percent opacity or greater,
(4) exit and escape the capture system and are visible without the aid of
instruments (applies only during periods when flow rates are being established),
(5) escape the capture system at the tapping station and are visible without
the aid of Instruments for more than 40 percent of each tapping period. There
are no limitations on visible emissions when a blowing tap occurs (requirements
apply only during periods when flow rates are being established).
Dust handling equipment emissions are limited to less than 10 percent
opacity. Carbon monoxide from any electric submerged arc furnace is limited
to less than 20 percent by volume (dry basis). A continuous monitoring sys-
tem Is required to monitor opacity of emissions discharged from the control
device.
Sources constructed, reconstructed or modified after 21 October 1974 are
subject to the regulations.
Process Description
Ferroalloy is a material consisting of iron and one or more other metals.
Ferroalloys are used En steel production as alloying elements and deoxidants.
There are three types of ferroalloys: silicon—based; manganese—based; and
chromium-based alloys. There are four major procedures used to produce ferro—
alloy: blast furnaces, electrolytic deposition, alumina silico—thermic process,
and electric smelting. However, the primary system used is electric smelting.
108
-------
Seventy—five percent of all ferroalloys are produced in electric smelting fur-
naces, of which electric submerged—arc open type is the oldest, simplest and
most widely used. The alloys are made in the electric furnaces by reduction
of suitable oxides. For example, in the production of ferrochromium, the charge
may consist of chrome ore, limestone, quartz coal, wood chips, and scrap iron.
The production of ferroalloy requires a number of steps from materials handling
to smelting to shipment. Figure 24, illustrates a schematic diagram of the
system involved.
Experts in the electric submerged arc furnace industry indicated that there
are no new plants presently under construction and that there are none planned
for the next 5 to 10 years.
Pre—Startup Operations
The pre—startup operation of a ferroalloy facility involves the coordina-
tion of numerous systems. Each unit must undergo its own pre—startup checks and
startup procedures before it can be integrated with the rest of the system.
Very often minor malfunctions and defects in units are the primary reason for
delays in meeting the startup and maximum production dates. Some of the units
which may have problems are crushers, scales, feeders, castings and pollution
control units. At this time in the pre—startup procedure, most of the units are
checked—out without a material load. The primary unit used in the production
of ferroalloy is the electric submerged arc furnace. The furnace ranges in
size from 6.1 to 12.2 meters (20 to 40 feet) in diameter with a height of
4.6 to 6.1 meters (15 to 20 feet). The furnace has a 5 cm (2 in.) thick steel
shell and is lined with brick refractory. Between the refractory brick and
the steel shell is a carbon liner. Electrical power transmitted through
electrodes bearing down on resistant coke provides the heat for smelting in
the furnace. Systems and components that are given special attention relative
to the electric furnace pre—startup operations are: water cooling, water
treatment, roof and wall structures, air pollution controls, electrode
rigging, electrode slipping and backing machanisms, tapping, lattle, crane
movement, and all electrical, mechanical and hydraulic subsystems.
The first major step in pre—startup of an electric submerged arc furnace
depends on the type of electrodes used. Internationally, there are two types
of electrodes that may be used in producing ferroalloys: prebaked and self—baked.
The pre—baked electrode, which is usually manufactured elsewhere, is a hard
rod—shaped baked carbon/silicon body ranging in size from 0.3 to 61 cm (1/8 to
24 in.) in diameter and lengths up to 244 cm (96 inches). The self—baked
electrode is a pasty mixture of carbon and silicon which is baked into a hard,
rod—shaped carbon body within the electric submerged arc furnace (this report
examines only the self—baked electric submerged arc furnace because they are
primarily used in the United States).
The first pre—startup step is the baking of the pasty carbon—silicon mix-
ture into a hard, rod—shaped body. This step is called bake—in. The carbon
material is contained in a thin gauge, hollow electrode steel cylinder 15 to
18 meters (50 to 60 feet) in height. At bake—in, a pre—baked electrode is
inserted Into the cylinder followed by a homogeneous mixture composed of coal,
metallurgical coke, lampblack, coal tar pitch, and petroleum coke. The furnace
109
-------
TO BAGHOUSE OR
SCR UBBER
WATER
COOLED
HOOD
0
SI TERIP
I DUST
4
ELECTRODE
OJ
‘QF
SE$G r(ED’NG
(LTI G
+
——I
( USHt G SCR(E .*
STO *CE
Figure 24. Schematic diagram of the ferroalloy production process. 7
-------
is then loaded with a 0.2 (55 gallon) drum of coke (the conductive material).
The electrode cylinder and content is lowered into the coke drum bearing down
with electric power. The system electrodes are energized with low voltage,
wherein heat is created by electrical resistance from the coke. This heat
bakes the carbon paste in the upper portion of the cylinder. Voltage is in-
creased and more coke is added, thus increasing the furnace temperature and
baking more of the carbon mixture in the column. The bake—in will continue
until about 4.6 meters (15 feet) of carbon mixture in the column is baked to
a hard carbon body. As the pre—baked electrode is consumed (—‘2.5 cm or 1 in.
per hr), the self—baked electrode will slip down bearing on the coke. More
carbon mixture would then be added to the column to replace the consumed por-
tion of the electrode. Once the bake—in is completed, pre—baked electrodes
are not used again. Pre-baked electrodes are only used during bake—in
(startup). If pre—baked electrodes are not used during bake—in, the entire
electrode column is filled with carbon paste and baked with an auxiliary fuel
(oil or gas). The entire bake—in takes 2 weeks time while curing the furnace
takes another 2 to 3 weeks, and heating the furnace to operating temperatures
takes about 1 to 2 weeks. During the bake—in, curing, and heatup, the furnace
will normally be shutdown 5 to 10 times for repair of malfunctions, defects
and especially to tighten the bus bar. Shutdown and startup will only take
a few hours unless there is a major malfunction. The total pre—startup opera-
tion may consume a 5 to 7 week period.
Startup Qperations
Up to this time, no metal has been introduged into the furgace. Upon
reaching operating temperatures of 2760 to 3316 C (5000 to 6000 F), metal is
slowly introduced to the furnace. The furnace is stabilized and then more
charge is added. The technique will continue until the furnace has attained
rated capacity. If everything operates smoothly, the furnace should reach
maximum production in six months. However, in ferroalloy production, there
are numerous units involved where many problems can develop. Experience has
shown that moat plants take 6 to 12 months to reach maximum production. Two
major problems are the integration of all units and the training of operators.
If any units, systems or operators are not running at rated capacity, the
maximum production rate cannot be reached. Most furnaces operate at 50 percent
maximum production the first year and in succeeding years production will
increase to about 90 percent. The first year of operation is usually plagued
with numerous malfunctions and breakdowns. The official startup date for an
electric submerged arc furnace should begin when metal is first introduced
into the furnace.
111
-------
REFERENCES
1. Personal Cotanunication with Dr. C.R. Allenback, Union Carbide, Niagara
Falls, New York, May 9, 1979.
2. Personal Communication with Dr. R. Persons, Union Carbide, Niagara Falls,
New York, May 9, 1979.
3. Communication with Foote Mineral Company, Exton, Pennsylvania, May 8, 1979.
4. Communication with Paul Blum Company, Inc., Buffalo, New York, May 8, 1979.
5. Personal Communication with Dr. Watson, Ferroalloy Association, Washington,
D.C., May 9, 1979.
6. Personal Communication wIth Mr. John Persons, Lectromelt, Pittsburgh,
Pennsylvania, May 8, 1979.
7. Dealy, James and Arthur Kuhn. Engineering and Cost Study of the
Ferroalloy Industry, EPA 450/2—74—008. U.S. Environmental Protection
Agency. 1974.
112
-------
IRON AND STEEL PLANTS: ELECTRIC ARC FURNACES — SUBPART AA
§60.270 — 60.275
Introduc t ion
Affected facilities under this subpart include the electric arc furnace
(EAF) and associated dust—handling equipment. Electric arc furnace means any
furnace that produces molten steel by melting charge material with electric
arcs from carbon electrodes. Dust—handling equipment means any equipment used
to handle particulate matter collected by an EAF control device.
The emission standard promulgated under this subpart limits particulate
matter discharge from an EAF control device to 12 mg/dscm (0.0052 gr/dscf).
In addition, control device exhaust stacks shall not exhibit 3 percent opacity
or greater. Emissions from the dust—handling equipment shall not be > 10
percent opacity.
With respect to the EAF melt shop, emissions from the shop are limited to
0 percent opacity except during:
• charging periods, when the shop opacity may be greater
than zero but less than 20 percent.
• tapping periods, when the shop opacity may be greater
than zero but less than 40 percent.
Where the EAF capture system is operated with the shop roof closed during charg-
ing and tapping, and emissions to the atmosphere are prevented until the roof is
opened after completion of the charge or tap, shop opacity standards shall apply
when the roof is opened and shall continue to apply for the time defined by the
charging and/or tapping periods. Continuous opacity monitoring of emissions
from the control device is required. Affected facilities constructed, recon-
structed or modified after October 21, 1974, are subject to the requirements of
this subpart.
Process Description
Electric arc furnaces (EAFs) for the production of steel have been in use
since 1.906. The recent demand for higher quality alloy and stainless steels
is responsible for increasing use of EAFs. It has been estimated that 20 per-
cent of the steel produced in the U.S. in’ 1976 was made in EAFs and as open
hearth units are shut down, EAF units will see increasing use.
113
-------
EAPs utilize electric current as a source of heat to melt scrap metal.
The process begins when scrap metal and alloys are charged to the furnace.
Following charging the electrodes are lowered into the charge to start the
initial “boredown” period. During boredown, a pool of molten metal is formed
on the surface of the scrap charge. From this time on, the charge is melted
by a combination of heat from the arc, heat radiated from the bottom of the
vessel, and by the resistance of the scrap to the current. Occasionally,
the boredown period is interrupted and additional scrap and/or fluxing agents
are added to the mixture. When the charge is completely melted, the steel
and slag (the solid waste formed during melting) are tapped into ladles and
pots respectively. Molten steel is then either further processed to produce
stainless steel, poured into ingot molds or sent to a continuous caster.
Pre—Startu p Operations
There are several operations involved in the pre—atartup process for elec-
tric arc furnaces. Mechanical and electrical equipment must be checked to
ensure that they are fully operational. Lastly, the furnace refractory is
preheated and checked. Once these steps are completed, the furnace is ready
for operation.
Startup Operations
Initial startup involves charging the furnace and initiating boredown.
The time required to attain maximum production depends on furnace size. Con-
versations with industry personnel indicate that for most furnaces full
production level can be achieved in 6 months. Based on this, the startup date
is best defined as the time of the first furnace heat. However, if during
startup operational problems are encountered which aignificantly alter normal
cycle times, representative performance test data might be unattainable until
the problems are straightened Out.
114
-------
REFERENCES
1. Personal communication with Mr. J. Haley, Puilman—Swindell, Pittsburgh,
P nnsylvania. 13 March 1979.
2. Personal communication with Mr. J. Gaibreath, Whiting Corp., Harvey,
Illinois. 13 March 1979.
3. U.S. Steel. The Making, Shaping and Treating of Steel. Herbick and
Held, Pittsburgh, Pennsylvania. 1971.
115
-------
KRAY PULP MILLS - SUBPART BB
§60.280 — 60.285
Introduction
The facilities of a kraft pulp mill covered under this subpart are: di-
gester system, brown stock washer system, multiple—effect evaporator system,
black liquor oxidation system, recovery furnace, smelt dissolving tank, lime
kiln, and condensate stripper system. In pulp mills where kraft pulping is
combined with neutral sulfite semi—chemical pulping, the provisions of this
subpart are applicable when any portion of the material charged to an affected
facility is produced by the kraft pulping operation. Facilities constructed,
reconstructed or modified after September 24, 1976, are covered by the
regulation.
A particulate emission limit of 0.10 g/dscm (0.044 gr/dscf) corrected to
8 percent oxygen and an opacity of less than 35 percent have been set for the
recovery furnace. Emissions from the smelt dissolving tank are limited to
0.1 g/kg black liquor solids (dry weight) or 0.2 lb/ton black liquor solids (dry
weight). For the lime kiln, the limit is 0.15 g/dscm (0.067 gr/dscf) cor-
rected to 10 percent oxygen when burning gaseous fossil fuel and 0.30 g/dscm
(0.13 gr/dscf) corrected to 10 percent oxygen, when burning liquid fossil fuel.
Total reduced sulfur (TRS) means the sum of the sulfur compounds—hydrogen
sulfide, methyl mercaptan, dimethyl sulfide, and dimethyl disulfide — which are
released during the kraft pulping operation. An emission limit of 5 ppm of
TRS by volume on a dry basis corrected to 10 percent oxygen has been established
for the digester system, brown stock washer system, multiple—effect evaporator
system, black liquor oxidation system and condensate stripper. This emission
limit remains in effect unless one of the following conditions are met: (1) the
gases are combusted in a lime kiln, (2) the gases are combusted In a recovery
furnace, (3) the gases are combusted with other waste gases in an incinerator
or other device not subject to the standard and are subjected to a minimum
temperature of 649°C (1200°F) for at least 0.5 second, (4) it has been demon-
strated to the administrator’s satisfaction that incineration of TRS from the
black liquor oxidation system or brown stock washer system is technologically
or economically not feasible, or (5) the digester, brown stock washer, conden-
sate stripper or black liquor oxidation system are controlled by a means other
than combustion whereby TRS emissions do not exceed 5 ppm by volume on a dry
basis corrected to the actual oxygen content of the untreated gas stream.
Emission limits of 5 ppm and 25 ppm TRS by volume on a dry basis, cor-
rected to 8 Percent oxygen have been established for any straight and cross
recovery furnaces respectively. For the smelt dissolving tank, the TRS ends—
slon limit has been set at 0.0084 g/kg black liquor solids (dry weight) or
116
-------
0.0168 lb/ton liquor solids (dry weight) . Total reduced sulfur emission
limits from the lime kiln have been set at 8 ppm by volume on a dry basis,
corrected to 10 percent oxygen.
Continuous monitoring requirements apply to recovery furnace opacity and
to TRS emissions from lime kilns, recovery furnaces, digester systems, brown
stock washer systems, multiple—effect evaporator systems, black liquor oxida-
tion systems and condensate stripper systems.
Process Description
The first step of kraft pulping is to prepare the wood for cooking. Logs
are debarked, chipped and screened prior to being fed to a digester. Some
mills will operate their own wood processing facilities, whereas others will
purchase wood chips from an external source. The wood chips placed in the di-
gester are cooked in a “white liquor,” a water solution of sodium sulfite
(Na 2 S) and sodium hydroxide (NaOH), at a temperature of 170 to 175°C (338 to
347°F) and pressure of 689 to 931 kPa (100 to 135 psi). The white liquor chem-
ically dissolves the lignin of the wood, freeing the cellulose fibers. The
cooking process, which can be batch or continuous, usually lasts from 2 to
3 hours. At the completion of the cook, the charge is blown into tanks to
release steam and other gases. The cellulose (pulp) is then separated from the
spent cooking liquor by filtration. At this point, the pulp is referred to as
brown stock and the spent cooking liquor is called weak black liquor. After
filtration, the brown stock is washed with water and passed through knotters,
rifflers and screens which sieve out small pieces of uncooked wood. Once
screened, the pulp is filtered and sent on to thickeners. From here, the pulp
can go to a bleach plant or to paper machines for final processing.
A major portion of the kraft pulping process is devoted to the recovery of
cooking liquor and heat. The weak black liquor is concentrated in multiple—
effect evaporators to about 65 percent solids and then burned in recovery fur-
naces. Steam generated by the furnaces is used for process operations through-
out the plant.
Basically, there have been only two types of recovery furnaces used in the
kraft pulping industry: the direct—contact evaporator system and the more re-
cent indirect—contact system. The former type requires the oxidation of the
concentrated black liquor prior to combustion to reduce TRS emissions. One
reference source contacted stated that most new mills being built today and in
the future will be using the indirect—contact evaporator system.
The residue resulting from burning the black liquor, called smelt, is dis-
solved in water to form “green liquor,” an intermediate solution used to regen-
erate white liquor. Once formed, the green liquor is transferred to a causti—
cizing tank where quicklime (CaO) is added to convert the sodium carbonate
(Na 2 CO 3 ) to NaOlI. The formation of NaOH completes the white liquor regen-
erative cycle. The calcium carbonate (CaCO 3 ) slurry (40 to 45 percent water),
generated during the causticizing process, is converted to CaO by calcina—
non in a rotary kiln. The CaO is then reused in the process.
117
-------
A condensate stripping system, employing either air or steam in a strip-
ping column, may be used to control TRS compounds emitted from the digester
systems and multiple—effect evaporators. TRS compounds emitted from other
facilities are usually controlled by process combustion, condensers, incinera-
tors, or absorption trays. A general flow diagram of the kraft pulping process
is presented in Figure 25.
Pre—Startup Operations
Before being connected to form the contiguous operation of a kraft pulp
mill, each affected and nonaffected facility is subjected to extensive shake-
down procedures. Electrical and instrumentation systems are thoroughly checked
before any piece of equipment is actuated. Process equipment is examined for
proper installation, clearance, and rotation. Transfer lines are checked for
free passage, leaks, and correct distribution. In addition, pieces of equip-
ment which are operated at elevated temperatures and pressures are tested
initially for leaks and expansion with hot water and then with steam, which is
provided by the power boilers or recovery furnaces (fired with auxiliary fuel).
The three most important facilities affected within a kraft pulp mill are:
the digester systems, recovery furnaces, and lime kiln. Initial shakedown
operations for the digesters involves passing hot water through the units (in-
cluding the blow tanks) to check seals and as a preliminary rinse. Next,
atean is passed through the digesters for further cleaning and to test expan-
sion joints. A final shakedown step prior to actual startup is to feed chips
and steam to the digester to simulate a cook. The partially cooked chips will
be passed to the blow tank to assure proper distribution and free passage of
material. This last step may last from one to eight hours.
Pre—startup operations of the recovery furnaces are similar to those of
an industrial boiler. These procedures include cleaning the inner walls of the
furnace; boilout (which removes grease and other deposits from the water and
steam tubes); and testing of components with respect to thermal expansion, mech-
anical stresses, corrosion resistance, structural soundness, warping, and leak-
age. Process controls, sensors, monitors, loading switching, safety interlocks,
and the steam distribution system are also checked prior to startup. In addi-
tion, fans and pumps are run, dampers are stroked and burners are test fired.
All furnace tests are conducted with auxiliary fuels such as oil or gas.
The startup of a recovery furnace is subject to codes set forth by the
Black Liquor Recovery Boiler Advisory Board, a self—governing body made up of
members from the kraft pulp mill industry. Some of the requirementé of the
board are that certain welds be radiographed to assure integrity, hydrostatic
pressure tests be conducted and certain safety systems be installed and
thoroughly checked before the furnace is started. The shakedown period for
recovery furnaces will normally last from two to seven days. Electrostatic
precipitators are typically used to control particulate emissions from recovery
furnaces, whereas absorption—type scrubbers are used to control reduced sulfur
compounds.
118
-------
viwi’ s*sCs
ONDENSAT(
WOOD - TAN K
rncrtus
I • s’OD c* PULP to
K NOT TEND. N
DISI$Tt* I 4PULP)
( 1 1.0$. Ns $1 I WA$$INS JI,watt* $CRIIN$
TIN
WCAK ILACK LSOUON 1r t
r 1 i NOOCONO(N$ASLC$
CONDINU
r- 1 SL*CN N
‘ \,\ _____ ___ _____
L OUO*
NECOVIRY 1 01t1DAT 10 1 1 I IvAPORA
PUNNACE
i AVv ITANII i
$V$IIN SlACK $ OPTIONAL ) .
* SLOuOD1 I
STACK I__I
LR - . v l si’ SASh
SMIL.T
( 1 1s 2 CO 3 . N• $) _______________
eVENT
I DI$SOLV$NS _______
4 sasu CONDENI
$ 1 1 111 s i ’rn u
0 TNIATMINT
POND
I TANK
1-
SAD’S
GREEN Ll0U0N
• WUITE
I LIQUOR I TANK
LINE
(NECYCLE TO
Ř SE$TtN) I CALCIUM C**NONATI
SLURRY
Figure 25. Flow diagram for Kraft pulping
process operations. 2
119
-------
The pre—startup operations of a kraft pulp mill lime kiln are similar to
those of a rotary kiln used in the manufacture of lime. Once the kiln has been
erected and before It is bolted down, various tests are conducted to check its
alignment and rotation. Once the kiln passes these preliminary tests, it is
secured and rechecked for alignment and rotation. A third examination is made
when heat is first introduced to the kiln.
The most important pre—startup operation of the kiln is the seasoning or
drying Out of the brick refractory lining. New refractory usually contains
residual amounts of water which must be driven out before the lime slurry can
be fed into the kiln. The drying out process is slow, requiring a gradual
warming of the bricks to prevent cracking or loosening. The kiln seasoning
process normally takes 3 to 5 days.
Once the refractory has been properly seasoned and the burners and fuel
supply lines have been checked out, the kiln is ready to receive the lime slurry.
Lime kilns operated at kraft pulp mills are usually fired with gas or oil.
High energy venturi scrubbers are normally used to control kiln emissions.
Startup Operations
The performance of a facility affected by this subpart is dependent upon
process operations upstream and downstream of the particular unit. Six major
operations within a kraft pulp mill are the: power boilers, digester system,
recovery furnaces, lime kiln, bleach plant, and paper machines. The startup
of a mill requires the operation of at least the first four of these processes.
Some mills will not operate bleach facilities nor paper machines. The power
boilers, not affected under this subpart, are the first operations to be brought
on line.
Startup of the facilities affected under this subpart Is considered to be
when wood chips and white liquor are first fed to the digesters with the intent
to produce pulp. Mill operators contacted stated that it may take from twelve
to twenty—four hours from the time the first cook is initiated until recongniz—
able pulp comes Out of the thickeners. The first pulp produced is not of any
commercial or production value. Due to the large capital outlay and high daily
operating costs, this situation will not last long and mill operators will con-
centrate their efforts on fine tuning process equipment in order to produce
useable pulp as soon as possible.
Once the first cook has been initiated, It will take a day or two before
enough black liquor Is produced to fire in the recovery furnace. Likewise, it
will be one to two days after black liquor is first burned before enough lime
slurry is generated from the causticizing tank to be fed into the lime kiln.
Because white liquor will not be generated from process operations until two
to four days after commencement of the initial cook, the white liquor and
caustic (lime) must be purchased or obtained from some other source in order
to start the pulping process. Once the chemical recovery cycle has been com-
pleted, it becomes almost entirely self—sustaining. In summary, it normally
takes from two to four days after Initial startup; i.e., the introduction of
wh te liquor and wood chips into the digester, before all the affected facili-
ties of a kraft pulp mill become operational.
120
-------
Because kraft pulping is a complex process involving chemical reactions
and several different and sometimes technologically new pieces of equipment,
it may and probably will take longer than 180 days after initital startup be-
fore a mill is operating at or near Its maximum production rate. Several
reference sources contacted stated that it normally takes a year or more for a
mill to reach its maximum production rate. One source stated, however, that
barring major mechanical failures or the debugging of newly developed process
equipment, a mill could achieve its maximum production rate within 180 days.
Due to the uncertainty in the amount of time required to achieve the maximum
production rate of the affected facility, multiple testing, once prior to
the 180—day limit and once after achieving maximum production capacity may be
required. It Is noted that for certain industrial categories such as kraft
pulp mills, the implementation and enforcement of Section 60.8 of CFR Part 60
may have to be carried out on a case—by—case basis.
121
-------
REFER CES
1. Environment Reporter, Bureau of National Affairs, Inc., 1978, New Source
Performance Standards Subpart BB, Standards of Performance for Kraft Pulp
Mills.
2. Standards Support and Environmental Impact Statement Volume I: Proposed
Standards of Performance for Kraft Pulp Mills, 1976. U.S. Environmental
Protection Agency Office of Air and Water Management. Research Triangle
Park, North Carolina. EPA—450/2—76—014a.
3. Personal Communication with representative of Technical Association of
Pulp and Paper Industry, Atlanta, Georgia, April 26, 1979.
4. Personal Communication with Mr. Eddinger, Industrial Studies Branch, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina,
April 27, 1979.
5. Personal Communication with Mr. Landegger, Black Clawson Company, New York,
New York, April 27, 1979.
6. Personal Communication with Mr. Dickenson, Babcock and Wilcox, North Canton,
Ohio, April 27, 1979.
7. Personal Communication with Mr. Nicholson, Weyerhauser Company, Tacoma,
Washington, April 27, 1979.
8. Personal Coninunicaiton with Mr. Elam, Black Clawson Company, Middletown,
Ohio, April 27, 1979.
9. Personal Communication with Mr. Salkowaki, Boise Cascade Corporation,
St. Helens, Oregon, April 30, 1979.
10. Personal Communication with Mr. Milow, Combustion Engineering, Boston,
Massachusetts, April 30, 1979.
11. Personal Communication with Mr. Sutter, St. Regis Paper Company, Jacksonville,
Florida, April 30, 1979.
12. Personal Communication with Mr. Derby, St. Regis Paper Company, Monticello,
Mississippi, April 30, 1979.
13. Personal Communication with Mr. Belas, Air Enforcement Branch, U.S.
Environmental Protection Agency, Region IV, Atlanta, Georgia, April 30, 1979.
122
-------
14. Personal Communication with Mr. Tretter, Georgia—Pacific Corporation,
Atlanta, Georgia, May 2, 1979.
123
-------
GRAIN ELEVATORS - SUBPART DD
§60.300 — 60.304
Introduction
The sources covered under this subpart are grain terminal elevators which
have a permanent storage capacity of over 88,100 a 3 (2.5 million U.S. bushels)
and grain storage elevators located at any wheat flour mill, vet corn mill,
dry corn mill, rice mill, or soybean oil extraction plant which have a perma-
nent storage capacity of 35,200 a 3 (1 million bushels). The facilities affected
under this category are: truck loading and unloading station, railcar loading
and unloading station, barge and shiploading and unloading station, grain
dryer, and all grain handling operations. Standards are set for captured emis-
sions and uncaptured (fugitive) emissions.
All captured emissions are passed through a stack or control device prior
to being emitted to the outside air, and must meet an opacity standard of
0 percent. All captured emissions except those from a grain dryer (column
dryer or rack dryer) must also meet a particulate standard of 0.023 g/dscm
(0.01 gr/dscf). A fugitive emission opacity limitation has been set at 5 per-
cent for any individual truck unloading station, or railcar unloading or load-
ing station; 0 percent for any grain handling operation; 10 percent for any
truck loading station; and 20 percent for any barge or shiploading station.
Sources constructed, reconstructed or modified after August 3, 1978, are subject
to the regulations.
Process acrjption
Prior to unloading, grain received is graded for quality and dust content,
and then weighed. The grain is then unloaded by direct dumping or conveying
(e.g., pneumatic, belt, bucket, or chain) into a receiving hopper and then
transferred to a surge (garner) bin. From the surge bin the grain is lifted
to the headhouse, the top of the elevator. At the headhouse, the material is
distributed to one of several storage bins. Certain incoming grains, however,
may need to be cleaned and dried before they can be stored.
Cleaning is required when incoming grain contains a high percentage of
crop soil, weeds, insects, stones and/or stalks. The first step in cleaning
(scalping) is to pass the grain through a coarse mesh screen to remove large
foreign matter. Aspiration, the next step, is carried out to remove the fine
foreign matter from the grain. The process involves directing air crosscurrently
124
-------
and countercurrently through dispersed falling grain. The third and usually
final step is to size the grain using a stack of vibrating screens. These
three grain cleaning steps can be operated separately or together as one con-
tiguous system. Grain cleaning may occur before and/or after grain storage.
To prevent spoilage, certain grains, including barley, oats, wheat, corn
and sorghum, must be dried to a specified moisture content before they can be
stored for any length of time. Drying facilities are used predominately during
the harvest season. An alternative to drying is grain blending, a practice of
mixing incoming grain which does not have an excessively high moisture content
with previously dried grain.
Because grain fermentation and heat build up may occur in long—term storage
bins, it becomes necessary to “turn” the grain, to prevent deterioration. Grain
turning involves transferring material from the bottom of one bin to the top of
another bin. Turning may be required several times a year depending on moisture
content, grain temperature, and the length of time the grain has been stored
without aeration.
Prior to grain load out, material will be weighed and cleaned if not done
so previously. The grain is then transferred from the weigh station to an a—
waiting carrier vehicle via telescopic piping. A general process flow diagram
for grain terminal elevators is presented in 1 ’igure 26. Particulate emissions
from affected facilities are typically controlled by mechancial separators and
baghouses.
Pre—Startup Operations
Because of the various affected unit operations associated with grain ter-
minal and storage elevators, many preliminary shakedown procedures must be
undertaken. Checkout procedures common to all the affected facilities are:
verification of all electrical circuits and connections, checking clearance of
moving equipment, bumping of motors and pumps, assuring correct rotation of
equipment, calibration of scales and instrumentation, and monitoring of air
pollution control devices.
Equipment checkout procedures unique to unloading, distribution and load—
out facilities are: free passage through transferring legs (e.g., pneumatic
belt, or bucket conveyors), correct movement and discharge of distributors and
trippers of the headhouse and process control systems. Equipment employed in
grain cleaning are checked out for clearance, alignment, and free flow of ma-
terial. For the dryers, equipment alignment and rotation, burners, fuel supply
lines, air leakage, and temperature and air flow controls are checked out be-
fore the equipment is actuated.
Process equipment will be tested individually at first, then as integral
parts of an affected facility. Dry runs will be conducted to make sure pro-
cess equipment is functionin8 properly before grain is intorduced into the sys—
tern. After all equipment has been checked out, trial runs will be made handling
grain. The equipment is tested under various loads to recheck alignment, rota-
tion, clearance, free passage of conveying lines, the distribution system, belt
stretching, and conveyor speec s and to break in air pollution control equipment.
125
-------
I AILC 1
INAIL c*i]
I NtCIIVIN$J
\NOPPcI’t/
4 .
OS$TRISUYON
AND TNIPPU$
0%
Figure 26. Process flow diagram for grain terminal elevators.
-------
The facilities are tested initially at a minimum load then gradually brought up
to or near design rates. The grain used for equipment shakedowns may already
be on—site f or a modified source, or for a newly constructed elevator a small
amount of grain may have to be purchased from a local farm or country elevator.
Once each affected facility has been successfully tested under actual conditions,
it is considered to be “ready for business,” and capable of receiving grain.
Pre—startup shakedown operations may last from less than 1 to 3 months depending
on the function pf the subject elevator.
Startup Operations
The startup date for each of the affected facilities of a grain terminal
elevator or grain storage elevator will vary depending on the source’s function,
type of grain received, time of year, and climatic conditions during harvesting.
Because of inherent flexibility of a grain elevator, some facilities will be
brought online while others are still under construction, being debugged or
idle. Once an elevator starts receiving grain for processing, or export, the
minimum number of facilities required to be operated concurrently are the grain
unloading stations, bin distribution system and loadout stations. These opera-
tions, however, do not have to be run simultaneously.
Grain cleaning is a variable operation which depends on the quality of the
Incoming material and its end use. Grain drying is somewhat of a seasonal
operation. Typically, grain harvested during the late fall and early winter
months of the year require some drying before storage. Grain drying, however,
may also be required at other times of the year, when an elevator receives a
shipment of material high in moisture content.
Startup is considered to be when the first shipment of grain is unloaded,
handled, cleaned, dried, or loaded after the affected facility is “ready for
business.” As stated previously, the affected facility is ready for business
when process equipment have been tested successfully with grain under trial
conditions at or near design capacity. As an example, a column dryer could be
erected and run successfully under actual operating conditions by March 1, on
March 2, It would be ready for business, but not begin drying grain until
October 1. In this example, the startup date would be October 1. This scenario
could apply to all affected facilities of a grain terminal elevator or grain
storage elevator. Startup for each operation occurs when grain, which will be
exported or processed in a mill for commercial sale, first enters the facility,
thus initiating the generation of particulate matter . Consequently, each af—
fected facility will more than likely have a different startup date.
in regards to performance testing of air pollution control devices, the
attainment of maximum production rates does not depend solely on the performance
of process equipment but in part on the incoming grain supply. It is conceivable
that an affected facility of an elevator may not reach its maximum production
rate within 180 days of initi&tl startup. A well operated elevator, however,
wouFd not fall into this category.
127
-------
B.EFERENCES
1. Environmental Reporter, Bureau of National Affairs, Inc., 1978, New Source
Performance Standards Subpart DD, Standards of Performance for Grain
Elevators.
2. Personal Communication with Mr. John I4ealy, Grain Elevator and Processing
Society. Minneapolis, Minnesota, April 23, 1979.
3. Personal Communication with Mr. Robert Hendricks, Central Soya, Fort Wayne,
Indiana, April 23, 1979.
4. Personal Communication with Mr. Sims Roy, U.S. EPA Standards Development
Branch, Raleigh, North Carolina, April 23, 1979.
5. Personal Communication with Mr. Leroy Funk, Carter—Day Company, Minneapolis,
Minnesota, April 23, 1979.
6. Personal Communication with Mr. Dick Wilbur, Cargill, Inc., Minneapolis,
Minnesota, April 24, 1979.
7. Personal Communication with Mr. Stan Gasawski, Continential Grain Company,
St. Louis, Missouri, April 24, 1979.
8. Personal Communication with Mr. Hollingsworth, Aeroglide Corporation,
Raleigh, North Carolina, April 24, 1979.
9. Personal Communication with Mr. R. Gillund, Mathews Company, Crystal Lake,
Illinois, April 24, 1979.
10. Per8onal Communication with Messrs. James Appold and Fred Wolf. The Ander—
sons, Maumee, Ohio, April 24, 1979.
11. Personal Communication with Messrs. Al Colby and Wes Weber, Archer Daniels.
Midland Processing Company, Decatur, Illinois, and St. Louis, Missouri,
April 25, 1979.
128
-------
LIME MANUFACTURING PLANTS - SUBPART HR
§60.340 — 60.344
Introduction
The facilities subject to Standards of Performance for Lime Manufacturing
Plants are: rotary lime kilns and lime hydrators. A rotary lime kiln is a
unit with an inclined rotating drum which is used to produce a lime product
from limestone by calcination. A lime hydrator is a unit used to produce a
hydrated lime product. The pollutant regulated is particulate matter. Emis—
sion limits established for rotary kilns and hydrators are 0.15 kilogram
per megagram of limestone feed (0.30 lb/ton) and 0.075 kilogram per megagram
of lime feed (0.15 lb/ton) respectively. In addition, an opacity limitation
of less than 10 percent has been established for rotary lime kiln emissions.
Continuous monitoring of rotary kiln opacity is required, except when using
a wet scrubber emission control device. Subpart HR does not apply to facili-
ties used in the manufacture of lime at kraft pulp mills. Sources constructed,
reconstructed or modified after May 3, 1977, are subject to the regulation.
Process Description
The sequence of operations in the production of lime are: 1) quarrying
the limestone, 2) crushing and sizing, 3) conveying, 4) calcination by rotary
kiln, 5) processing the quicklime (the product of the kiln) by hydration,
6) milling/sizing, and 7) packaging (bulk or bag). As mentioned above, the
lime manufacturing standard covers only the calcination and hydration operations.
These two operations are depicted in Figure 27. “Limestone” means calcitic and
dolomitic limestone, whereas “Lime Product” means the product of the calcination
process including but not limited to calcitic lime, dolomitic lime, and dead—
burned dolomite.
Calcinat ion
A rotary kiln is basically a furnace consisting of a heavy steel shell
lined with refractory brick. Kilns are normally fired with one or more of the
following fuels: natural gas, fuel oil, or pulverized coal or coke. The kilns
are normally installed at inclines of 3.50 on four to six foundation piers and
revolve on trunnions at 30—50 seconds per revolution. Rotary kilns will vary
in size ranging from 2 to 5 meters (6.5 to 16 feet) in diameter to 18 to 183
meters (59 to 600 feet) in length.
Sized limestone is fed Into the elevated end of the kiln and travels
down the length of the kiln towards the firing end. The material is heated to a
temperature of 1100°C (2000°F) which chemically breaks down the limestone (CaCO 3 )
to produce quicklime (CaO) releasing CO 2 . Particulate laden gas flows counter—
129
-------
TO
COAL S AIR
FAN
0
GAS •I’O
ATMOSPHERE
Figure 27. Process flow diagram for lime calcination and hydration. 2
-------
current to the limestone, exhausting out the feed end of the kiln. At the
lower end of the kiln, the quicklime is discharged to a cooling system.
Systems to preheat the limestone prior to introduction to the kiln
are becoming widely used to conserve fuel consumption. Kilns operated with
preheaters are shorter in length, requiring 60 to 70 percent less space than
kilna without them. Particulate matter emitted from lime rotary kilns are
usually controlled by cyclones, baghouses, electrostatic precipitators, water
scrubbers, or gravel bed filters.
Hydration
The hydration process is initiated by blending quicklime with water
in a pug mill preinixer. After mixing, the lime—water slurry passes to an agi-
tator—hydrator. An exothermic chemical reaction takes place in the hydrator
producing steam, which is vented to the atmosphere along with any air that enters
the hydrator through the charging port. The most co on air pollution control
methods used to treat hydrator emissions are water sprays in the hydrator ex-
haust stack or wet scrubbers.
Pre—Startup Operations
The rotary kiln is the heart of a lime manufacturing operation; all other
unit operations are based on the performance of the kiln. As a result, ancillary
process operations such as crushing, screening, milling, and material and fuel
storage and conveying must be thoroughly tested under operating conditions prior
to the introduction of limestone into the kiln. Electrical systems and mechani-
cal qulpment are checked out initially on an individual basis and then as
integral parts of the process operation. Instrumentation and control panels
are also debugged during this phase. Likewise, mechanical equipment are tested
to assure that they are properly installed and aligned, and that they are rota-
ting freely and in the right direction. During pre—startup, conveyor speeds
are adjusted, feeder rates are monitored and transfer lines are checked for
free passage. Equipment shakedown of unit operations may occur simultaneously
or in some logical sequence. The pre—startup shakedown period may last from
one to three months.
In regards to the rotary kiln, once it has been erected, even before It
has been bolted down, various tests are conducted to check its alignment and
rotation. Upon passing inspection, the kiln is then secured to the support
piers and again checked for alignment and rotation. A third check is made when
heat is first applied to the kiln. During this phase, the kiln may stretch
15 to 20 centimeters (6 to 8 inches) in length and 0.6 centimeters (0.25 inches)
in diameter.
One of the most important pre—startup operations is the “seasoning” or
drying out of the kiln. New refractory brick lining the inner wall of the kiln
contains residual amounts of water which must be driven out of the bricks be-
fore limestone can be fed to the kiln. The drying ou process is slow, requir-
ing a gradual warming of the bricks, up to 149 to 204 C (300 to 400°F), to
prevent cracking and loosening. Supplemental gas or oil—fired burners are
131
-------
typically used to season the refractory. At first the kiln is not rotated, then
as the temperature builds up, operators rotate the kiln so that the refractory
is heated evenly. This process normally takes 3 to 5 days. Due to equipment
failures, however, it may take two or three attempts to season the kiln pro-
perly. Limestone is not usually introduced to the kiln during the drying out
period.
Water vapor and products of combustion generated during refractory season-
ing are vented either through the air pollution control equipment or a temporary
exhaust opening in the ductwork leading to the air pollution control equipment.
The latter case occurs most often when the control device is a baghouse, con-
sisting of fabric filters which can be affected by the relatively high moisture
content of the gas stream. Once the refractory lining of the kiln has been
dried out, the kiln can 1) be shut down while ancillary process equipment are
being installed or checked out, or 2) can be brought up to its normal operating
temperature of 1100°C (2000°F) and begin to receive limestone, provided that
all ancillary equipment have been checked out and readied to co ence production.
Pre—startup of the hydrator and ancillary equipment follows the same
sequence of shakedowns as the rotary kiln and its associated process operations.
Similarly, electrical systems and mechanical equipment are initially tested
Individually and then as a total unit prior to processing of material. The only
major difference between pre-startup of the rotary kiln and the hydrator, is
that the latter does not require seasoning because it is not lined with refrac-
tory. The hydrator pre—startup shakedown can last from 2 to 4 weeks.
Startup Operations
Startup for lime manufacturing operations is generally considered to be the
first day limestone is fed into the stone preheater or the rotary kiln, depend-
ing on the plant configuration. It is at this time that particulate matter is
first generated and emitted as an air pollutant from the limestone calcining
operation. Typically, the kiln is preheated to 315 to 371°C (600 to 700°F)
before the limestone is introduced. As more material Is fed into the kiln, it
is brought up to its normal operating temperature of 1100°C (2000°F). Most of
the sources contacted indicated that coal was the principal fuel used to fire
rotary lime kiln furnaces.
Startup begins an intensive period of equipment shakedown. Process opera-
tions must be synchronized, conveyor and feed systems have to be adjusted,
Instrumentation has to be rechecked under actual operating conditions, and air
pollution control equipment must be monitored. One source stated that once
startup has occurred, the kiln could produce at 50 to 70 percent of its rated
capacity within two days, but that It could take two to three months of debug-
ging and fine tuning before the kiln would be operating close to design capacity.
Fine tuning would Include adjustment of burner positions and primary air feed
rates. Other delays would result from the malfunction or failure of auxiliary
equipment such as conveying systems, material and fuel feeding systems, and
instrumentation. Extended time delays could result if major pieces of process
iipment had to be reordered. The sources contacted stated that the 180 day
p .iod allowed for equipment shakedown prior to required performance tests pro-
vided enough time to reach maximum production rates, unless major process equip-
ment pieces had to be replaced.
132
-------
Startup for the hydrator is considered to be when quicklime and water first
come into contact in the prenixer, usually a pug mill. Material is fed Into
the system at a slow rate at first allowing process equipment to gradually heat
up to a normal operating temperature of 107 to 121°C (225 to 250°F). After
Startup, equipment is monitored, tested and adjustments are made to assure that
all operations are functioning properly. Sources contacted stated that the
hydrator system, barring mechanical failures, could achieve its design produc-
tion rate within a week of startup. Air pollution control equipment are also
monitored carefully during the initial days of startup.
133
-------
REFERENCES
1. Federal Register 40 CFR Part 60 22506—2251. Standards of Performance for
New Stationary Sources — Lime Manufacturing Plants. Vol. 42, No. 85,
Tuesday, May 3, 1977.
2. Standards Support and Environmental Impact Statement Volume 1: Proposed
Standards of Performance for Lime Manufacturing Plants. 1977. U.S.
Environmental Protection Agency, Office of Air and Water Management.
Research Triangle Park, North Carolina. EPA—450/2—77—077a.
3. Personal Communication with National Lime Association personnel, Washington,
D.C., April 18, 1979.
4. Personal Communication with Mr. Naumann, Marblehead Lime Company, Chicago,
Illinois, April 18, 1979.
5. Personal Communication with Mr. NcCandlish, Flintkote, Monterey Park,
California, April 18, 1979.
6. Personal Communication with Mr. Smithwick, Allis Chalmers, Milwaukee,
Wisconsin, April 18, 1979.
7. Personal Communication with Mr. Friend, Vulcan Iron Works, Inc., Wilkes
Barre, Pennsylvania, April 19, 1979.
8. Personal Communication with Mr. Schwarzkopf, Kennedy Vansuan Corporation,
Danville, Pennsylvania, April 19, 1979.
134 MJ.$. GOVERNMENT PRINTING OfFICE: 1980 311—132/23 1—3
------- |