United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory 1
Cincinnati OH 45268
EPA-600/2-80-072
April 1980
Research and Development
xvEPA
Evaluation of the
Hoboken Converter at
Glogow, Poland
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution-sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-80-072
April 1980
EVALUATION OF THE HOBOKEN CONVERTER
AT GLOGOW, POLAND
by
Zbigniew Smieszek
The Institute of Nonferrous Metals
Gliwice, Poland
Contract No. 5-533-5
Project Officer
Alfred B. Craig, Jr.
Industrial Pollution Control Division
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
U.S. Environmental Protection Agency
Region V, Library
230 South Dearborn Street
Chicago, iiiinois 60604
-------
DISCLAIMER
This report has been reviewed by the Industrial Environ-
mental Research Laboratory, U.S. Environmental Protection Agency,
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
U,S. Environmental Protection Agency
-------
FOREWORD
When energy and material resources are extracted, processed,
converted, and used, the pollution related impacts on our en-
vironment and even on our health often require that new and
increasingly more efficient pollution control methods be used.
The Industrial Environmental Research Laboratory - Cincinnati
assists in developing and demonstrating new and improved method-
ologies that will meet these needs both efficiently and economi-
cally.
This report presents the results of an investigation into
the control of air pollutant emissions from the converting pro-
cess in the primary copper industry. The study was performed to
develop improved methods for operation of the copper converter to
allow more effective control of sulfur oxides and particulate
matter. The results are being used within the Agency's Office of
Research and Development as part of a larger effort to develop
improved technologies for reducing pollutant discharges in the
nonferrous metals industries. The findings will also be useful
to other Agency components and industry in dealing with environ-
mental control problems. The Metals and Inorganic Chemicals
Branch of the Industrial Pollution Control Division should be
contacted for any additional information concerning this program.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
-------
PREFACE
In 1975, the U.S. Environmental Protection Agency (EPA)
awarded a contract to the Ministry of Smelting in Poland for
research to minimize the emission of fugitive pollutants from
copper smelters and to assist in the control of all smelter
pollutants. The Ministry of Smelting assigned this work to the
Institute of Nonferrous Metals in Gliwice, which conducts both
basic and applied research and development for the nonferrous
metals industries in Poland. This work was accomplished by the
research workers at the Institute, with the assistance and close
cooperation of the staffs of the Polish copper smelters, the
design bureau BP "BIPROMET," and the Association of Nonferrous
Mining and Metallurgy.
This project was initiated with EPA's Office of Research and
Monitoring, Control Systems Laboratory, and subsequently trans-
ferred to the Industrial Environmental Research Laboratory,
Industrial Pollution Control Division, following an internal EPA
reorganization. It was supported by PL-480 funds through EPA's
Special Foreign Currency Branch.
IV
-------
ABSTRACT
This research project was initiated with three overall
objectives—to develop procedures for operating copper converters
to provide steady gas flows containing relatively high concentra-
tions of 862; to improve cleaning and treating of particulates in
converter gas streams to allow better operation of S02 removal
systems such as contact sulfuric acid plants; and to show how the
procedures and results that were developed could be applied to
the various types of copper smelters encountered in industry.
The intermittent flow of gases from the converter compli-
cates the operation of the sulfuric acid plants typically used to
further process SO2. In addition, the converter is a significant
source of fugitive emissions. The Polish copper industry employs
the siphon converter exclusively. The most important difference
between this converter and the more commonly employed Peirce-
Smith is that in the ^siphon converter the gas offtake is sepa-
rated from the charging and pouring port. This design thus has
the potential to reduce both air dilution of the gases and fugi-
tive emissions through the converter mouth.
The Institute of Nonferrous Metals conducted project re-
search in both the laboratory and in full-scale converters. It
was demonstrated that the operation of two or more converters
could be properly scheduled to ensure a steady high-strength gas
stream suitable for feed to the acid plant. The gas cleaning
system that achieves these results consists of gas coolers and
dry electrostatic precipitators, all synchronized with the con-
verter and acid plant operations by a two-stage gas pumping
system. Some problems remain in stabilizing gas removal from the
converters while simultaneously minimizing air dilution of the
gases, fugitive emissions through the mouth, and carryover of
materials from the bath into the gas offtake. During normal
operation, however, the system effectively controls sulfur emis-
sions, with the converters the only source of feed gas to the
acid plant. With appropriate modifications, this gas cleaning
system could be adapted to other smelters, including those using
Peirce-Smith converters.
This report was submitted in fulfillment of Contract No.
5-533-5 by the Institute of Nonferrous Metals under the sponsor-
ship of the U.S. Environmental Protection Agency. It covers the
period from March 1975 to February 1978, and work was completed
in December 1978.
-------
-------
CONTENTS
Foreword
Preface 1V
Abstract _ _y
Figures viii
Tables *
Acknowledgment
1. Introduction 1
2. Converter Operating Procedures 5
Background 5
The Converting Process 8
The Operating Cycle 15
A Three-Converter System 17
Dilution of Converter Off-Gases 22
Operating Problems 24
Operating Cost Evaluation 27
3. The Gas Cleaning System 28
Converter Emission Characteristics 28
The Converter Gas Cooling System 30
Electrostatic Precipitators 52
Fugitive Emissions 68
Pressure and Flow Rates 73
The Gas Pumping Control System 74
Converter Pressure Sensor 81
4. Application to U.S. Practice 85
5. Recommendations for Further Research 90
Bibliography 91
Appendices
A Other Converter Studies yj
Lime Addition to Converter Gases 93
Oxygen Enrichment 100
B Pressure and Gas Flow Calculations 101
Bibliography for Appendix B 118
C Verification Testing 11^
Summary and Discussion 120
Sampling and Analytical Methods 123
Description of the Sampling Sites 129
Modifications for Verification Tests 130
vii
-------
FIGURES
Number
Paqe
1 Shaft Furnace 7
2 Siphon Converter - Longitudinal Cross Section 9
3 Siphon Converter - Cross Section 10
4 Flowsheet - Glogow No. 1 Smelter 11
5 Flow Diagram - Converter .Gas Cleaning System 31
6- Converter Gas Temperature - Cooler and Collector
Outlets 33
7 Original Forced-Air Circulation Cooler 34
8 Flow of Converter Gases and Cooling Air Through
Original Cooler 36
9 Cooler Temperatures - Symmetrical Air Flow 37
10 Cooler Temperatures - Asymmetrical Air Flows 38
11 Incease in Cooler Resistance Over Time 41
12 Improved Forced-Air Circulation Cooler 43
13 Flow of Converter Gases and Cooling Air Through
Improved Cooler 44
14 Measurement Points for Cooler Test Program 45
15 Gas and Cooling Air Temperatures During Two
Converter Operating Cycles 47
16 Gas and Cooling Air Temperature 50
17 ESP Performance as Function of Gas Velocity 61
18 Angle of Repose of Dusts in ESP Hoppers 65
(continued)
Vlll
-------
FIGURES
Number
19
20
21
22
23
B-l
B-2
B-3
B-4
B-5
B-6
B-7
B-8
B-9
B-10
C-l
C-2
C-3
C-4
C-5
(continued)
Converter Position - Charging and Pouring
Operations
Proposed Automatic Gas Control System
Location of Converter Pressure Sensor
Strip Charts - Converter Blast and Pressure at
Mouth
Effects of Increased Converter Mouth Size
Gas Flow Through the Acid Plant System
Gas Flow Through a Two-Converter System
Gas Flow Through a Single-Converter System
Basic Fan Curve
Fan Curves - Reduced Flow
Fan Curves - Reduced Pressure
Gas Flow and Pressure Through a Single Converter
System
Gas Flow and Pressure Through a Single Converter
System
Gas Flow and Pressure Through a Two-Converter
System
Gas Flow and Pressure Through a Two-Converter
System
Diagram of the Polish Particulate Sampling
Method Train
PPSM Filter Holder
Sketch of Sampling Locations
Sketch of Sampling Locations
Probe Alignment
Page
71
77
82
84
87
106
107
108
110
111
112
114
115
116
117
126
127
131
132
133
IX
-------
TABLES
Number
Page
1 Typical Concentrate Composition 6
2 Inputs and Products of Copper Smelting - Glogow
No. 1 12
3 Converter Time Under Blast 18
4 Gas Characteristics in a Three-Converter System 20
5 Operation of Three Converters Under Blast 21
6 Fan Draft on Individual Converters in a Three-
Converter System 21
7 Converter Off-Gas Prior to Dilution "" 23
8 Converter Gas Dilution During Slag and Finish
Blows 25
9 Converter Gas Dilution at Various Fan Drafts
During Copper Blows 26
10 S02 Catalysis Found in Siphon Converter Dusts 29
11 Operating Parameters - Original Converter Gas
Cooler 35
12 Increase in Cooler Resistance During Campaign 40
13 Average Cooler Operating Parameters 49
14 Fan Specifications 52
15 Electrostatic Precipitator Data 53
16 ESP Performance as Function of Discharge Power
and Gas Velocity " 55
17 ESP Performance - Temperature >_290°C, Power >_15 KW 56
18 ESP Performance - Temperature >290°C, Power <15 KW 58
(continued) ~
-------
TABLES (continued)
19 ESP Performance - Temperature 270°-290°C 59
20 ESP Performance - Temperature <270°C 60
21 Analyses of Dusts from ESP and Cooler Hoppers 62
22 Density of Converter Dusts 64
23 Converter Dust Resistivity 66
24 Converter Dust Grain Size 67
25 characterization of Fugitive Emissions 70
26 Results of Air Sampling Near Converters 72
A-l S02/S03 Content of Converter Gases at ESP Inlet 94
A-2 CaO Content in Hopper Dusts 95
A-3 Stoichiometric Lime Requirements for S03
Neutralization 97
A-4 ESP Efficiency During S03 Neutralization Testing 98
A-5 Lime and Lead Content in Dusts From ESP Hoppers 99
A-6 Calculated Effects of Oxygen Enrichment 100
B-l Calculated Scaling Factors 109
B-2 Loops Used to Calculate Converter Pressure and
Flow Values 113
C-l Dust Concentration Obtained Using Both Trains
(g/NM3)a 121
C-2 Description of Problems Encountered in EPA Method
5 Runs 122
C-3 Expected Results for EPA Method 5 122
XI
-------
ACKNOWLEDGMENT
This report was prepared by the Institute of Nonferrous
Metals in Gliwice, Poland, under the direction of Dr. Zbigniew
Smieszek. The project manager for the Institute was Zbigniew
Dalewski, and the principal authors were W^odzimierz Babik,
Julian Bystrori, Zbigniew Dalewski, Zbigniew Greniuch, Jerzy
Kuznik, Stanis^aw Peszat, Ryszard Skrzys, Stanislaw Sobierajski,
W^odzimierz Subbotin, Robert Szeliga, Wac^aw Traczewski, and
Marian Witkos. During the development of this report some of the
authors visited several U.S. copper smelters.
.Project Officer for the Industrial Environmental Research
Laboratory of the U.S. Environmental Protection Agency was George
S. Thompson, Jr., with associated project input by Alfred B.
Craig, Jr. Project activities were coordinated by Thomas J.
LePine of EPA's Special Foreign Currency Branch.
PEDCo Environmental, Inc. provided consulting support to EPA
under the direction of Timothy W. Devitt. The PEDCo project
manager was Thomas K. Corwin. The verification testing in Poland
was conducted under the direction of Darrell L. Harris of
Monsanto Research Corporation.
-------
SECTION 1
INTRODUCTION
Copper smelters contain a variety of continuous and batch
pyrometallurgical processes that are sources of emissions of
sulfur oxides and particulate matter that are difficult to con-
trol. One such process that requires S02 control is the con-
verter, a batch-type operation in which the copper/iron/sulfur
matte formed in a smelting furnace is blown with compressed air
to oxidize the iron and copper sulfides, producing crude (blis-
ter) copper and an iron silicate slag. The intermittent flow of
gases from the converter complicates the operation of the sul-
furic acid plants typically used to further process the S02. The
converter is also frequently the most significant source of
fugitive emissions at a copper smelter.
No method has been found to eliminate the converting process
from conventional copper smelting. The most common type of
converter used in U.S. copper smelters is the Peirce-Smith, a
horizontal, refractory-lined, cylindrical furnace with an opening
in the side that serves as a mouth for charging feed materials
and pouring off molten blister copper and slag. The mouth of the
Peirce-Smith also serves as the exhaust port for the off-gases,
which are partially collected by a loose-fitting hood situated
above the converter. The hood must be retracted during charging
and pouring operations, which permits large quantities of fugi-
tive emissions to escape and causes additional air dilution of
the gas stream, thereby decreasing gas temperature and SO2
content. A variety of modifications have been made to the
design of the hood and converter mouth to minimize these emis-
sions.. Retractable gates, secondary hoods, enclosed converters,
and computerized control of the gas off-take based upon blast
intensity or gas pressure have been reported to be used at some
U.S. smelters.
The siphon converter is an alternative design to the Peirce-
Smith that is intended to eliminate the problem of excess air
infiltration into the gas removal and cleaning system. The basic
furnace design is similar to the Peirce-Smith. However, instead
of allowing the gases to exit through the converter mouth, the
siphon converter has an integral side flue located at one end.
Shaped like an inverted "U," this flue, or siphon, rotates with
-------
the converter. It is attached to a cylindrical duct, which also
rotates, that leads to a fixed vent flue. This design provides
an airtight connection between the converter and the gas removal
ductwork. It thus minimizes gas dilution by external air and
facilitates the maintenance of the high-SO2 stream desirable for
sulfur recovery in a conventional sulfuric acid plant.
Siphon converters are in only limited use in the world
copper industry. The first such units to be designed are still
being operated at the Hoboken smelter in Belgium; larger models
have been installed under license from Mechim at the Inspiration
Consolidated Copper smelter in Miami, Arizona, and are reported
to be used at several smelters in other countries. Siphon units
are used in the three modern copper smelters in Poland. At the
oldest of these, small converters with capacities of 33 Mg of
copper per cycle were constructed, using furnace drawings and the
gas take-off design purchased from Mechim. At the two newer
smelters, Glogow Nos. 1 and 2, larger converters with capacities
of 80 Mg of copper have been designed and installed. Polish
engineers have incorporated a number of modifications to the
design of both the converter and the gas cleaning system, and
these units may represent second- and third-generation improve-
ments of the original Hoboken design.
In order to furnish information to the United States En-
vironmental Protection Agency on how these converters are used to
minimize the emissions of fugitive pollutants from copper con-
verters and optimize the control of furnace off-gases, the
following objectives were developed for this study:
0 Develop procedures for operating copper converters to
provide steady gas flows containing relatively high
concentrations of S02. The off-gases from converters
are normally used for recovery of sulfuric acid by the
contact method. At some smelters, these gas streams
are combined with those from fluidized-bed roasters or
continuous smelting furnaces. In such cases, even the
leanest converter gases can be processed as they are
combined with other SO2~rich gases. In the majority of
smelters, however, sulfuric acid is recovered exclu-
sively from the converter gases. Because converting is
a^batch operation with a fluctuating gas flow, it is
difficult for the converter department alone to provide
the acid plant with a feed gas of sufficiently stable
and strong SO2 concentration (minimum 3.5-4.0 percent
volume). The operating cycles of two or more con-
verters must be synchronized to give the overall con-
verting process a relatively continuous gas output.
Sophisticated gas removal, control, and monitoring
systems are required to achieve these results.
-------
Improve cleaning and treating of particulates in con-
verter gas streams to allow better operation of S02
removal systems such as contact sulfuric acid plants.
Converter gases must be cleaned of particulate matter
prior to S02 control; this is accomplished in both dry
and wet control devices, the latter of which are an
integral part of the acid plant. Effective particulate
removal is essential because settled dusts can block
the active catalyst surface, decreasing the SO2-S03
conversion rate. The dusts fill the spaces between the
catalyst grains, increasing the rate of gas flow and
reducing contact time. The quantity of gases processed
drops because of the increase in flow resistance.
Dusts that remain in the acid circuit contaminate the
product and can form slimes that will clog nozzles and
impair heat exchange. Monitoring and control devices
can be contaminated, and there can be increased wear of
pumps, blowers, and pipelines, as well as accelerated
corrosion caused by heat imbalances. Production stops
will then be necessary to regenerate the catalyst or
repair or replace worn or damaged equipment. The
maximum dust concentration in the gases entering the
contact apparatus is 0.02 g/Nm3. In addition, no
arsenic or other catalyst poisons should be present.
If these requirements are met, the contact apparatus
should have an effective life of about 2 years, which
is equivalent to a typical acid plant using a pyrite or
elemental sulfur feed.
Show how the procedures and results that are developed
could be applied to the various types of copper smelt-
ers encountered in industry. At copper smelters where
only converter gases are processed in the acid plant,
there must be at least three converters, two of which
are under blow while the third is undergoing mainte-
nance, serving as a holding furnace, or on standby.
The working converters must operate according to a pre-
cisely synchronized schedule to ensure the proper gas
conditions. The scheduling of these operations is
facilitated with a greater number of converters.
Mathematical models of the converting process have been
developed, and automatic controls for operation of the
converter and gas removal system have been installed at
a number of smelters in addition to those already in-
stalled on the Polish smelters. These include the
Sociedad Mineria El Teniente smelter in Caletones,
Chile, and the Norddeutsche Affinerie plant in Hamburg,
Germany. Automatic control systems have thus been
installed on both Peirce-Smith and Hoboken converters
-------
of various sizes. The Polish experience with automated
process control should provide 'data potentially ap-
plicable to any modern copper smelter. However, an
automatic gas removal system based upon blast intensity
is expected to have more satisfactory results with
siphon converters because of the reduced air infiltra-
tion.
-------
SECTION 2
CONVERTER OPERATING PROCEDURES
BACKGROUND
The mining and processing of copper ores and concentrates in
Poland did not begin until after World War II. Because the
original deposits mined were small and of low grade, the indus-
try's growth was at first severely limited. The discovery of new
ore deposits in 1957 permitted the development of a modern copper
industry, whose production now more than meets Poland's domestic
needs, allowing a surplus to be exported. The mines, concen-
trators, smelters, and associated facilities are all located in
the mining districts in western Poland.
Mined ores are first upgraded by flotation to produce a
concentrate which is low in iron and sulfur content and which
contains from 20 to 25 percent copper. Table 1 presents an
example of a typical concentrate composition. After arrival at
the smelter, the concentrates are mixed with sulfite liquor, then
dried in a rotary dryer and briquetted. The briquettes, which
contain 19 to 24 percent copper, are charged to a rectangular
water-jacketed shaft smelting furnace through double bell valves.
Figure 1 is a diagram of the shaft furnace. The briquettes
comprise about 80 percent of the furnace charge, with the re-
mainder about equally divided between coke and recycled converter
slag. Shaft furnace smelting is a continuous reduction process
with heat provided by partial combustion from air blown through
nozzles near the bottom of the furnace. Liquid products drain
continuously from the furnace and into a settling tank. The
liquid matte and slag separate in the tank because of their
different specific gravities and mutual insolubility. Matte is
tapped from the tank into 21-Mg ladles and transported to the
converter aisle by overhead cranes. Exhaust gases from smelting
exit the upper part of the furnace and pass through particulate
control devices.
The copper matte obtained from smelting contains 55 to 67
percent copper, 6 to 14 percent iron, 19 to 23 percent sulfur,
1.5 to 3.0 percent lead, and 0.3 to 0.8 percent zinc. Matte
composition may fluctuate because of variations in the concen-
trates and in the operating conditions of the shaft furnaces.
-------
TABLE 1. TYPICAL CONCENTRATE COMPOSITON
Species
Cu
Fe
Pb
S
so4
CaO
MgO
A12°3
Si°2
Ag
co2
Organics
Other
Concentration
%
25.24
2.00
1.40
8.22
0.08
8.64
5.10
6.28
17.80
0.041
12.40
8.23
4.5
-------
1 - FURNACE BOTTOM 8
2 - SIDE AND FRONT WATER JACKETS 9
3 - WATER JACKET, CENTRAL TAP 10
4 - CENTRAL TAP 11
5 - SIPHON TAPPING SPOUT 12
6 - AIR NOZZLES 13
7 - WATER-COOLED SUPPORT CON-
STRUCTION
SHAFT
GAS OFFTAKE
FURNACE ROOF
DOUBLE BELL VALVES
CHARGE MIXTURE
MANHOLES
Figure 1. Shaft furnace,
-------
Siphon converters were chosen for use in the Polish copper
industry for technical, economic, and environmental reasons.
When the first smelter was constructed, it was believed that the
current technology was not sufficiently advanced to allow the
operation of Peirce-Smith converters in a manner that would
guarantee a steady supply of high-strength S02 gases to feed an
acid^plant. The Hoboken design appeared to ensure more efficient
utilization of SC>2, while simultaneously minimizing atmospheric
emissions. The design was purchased from Mechim, S.A., of
Brussels, Belgium.
•
Figures 2 and 3 are diagrams of the Polish siphon converter.
The design is essentially the same as a Peirce-Smith, with a
mouth in the middle of the furnace for charging the feed mate-
rials and skimming or pouring the liquid products. A side flue,
or siphon, is attached to one of the converter end walls. It is
connected to the exhaust tower with cylindrical ductwork and a
seal ring, both of which rotate with the converter. For weight
stability, the siphon is balanced with a counterweight. The
converter assembly is supported on concrete pillars and is
equipped with a rotating mechanism similar to that of a Peirce-
Smith. At Glogow Nos. 1 and 2, the inner volume of the furnace
permits an actual blister production of about 80 Mg per operating
cycle, depending on the copper content of the matte.
The siphon converters installed at these newer smelters
incorporate a number of improvements over the original Hoboken
design. They are more sturdily constructed, and quartzite flux
is fed continuously to them through automatic weighers. Fully
automatic tuyere punchers of Polish design, which are controlled
by blast air pressure, are also installed (See Engineering and
Mining Journal, February, 1979). A gate is provided below the
siphon to allow removal of buildups without stopping the con-
verting process or breaking the link to the gas removal system.
The seal ring connecting the ductwork to the exhaust tower is
also of Polish design, and an improved counterweight is used that
reduces the strain on the bearings and motors and provides more
stable operation.
Four siphon converters are installed at the Glogow No. 1
smelter. Except where indicated otherwise, all data in this
report refer to this smelter. The scope of the process opera-
tions is detailed in Figure 4, and the accompanying Table 2 pre-
sents the compositional range of the more significant process
streams.
THE CONVERTING PROCESS
Converting is a batch operation that consists of blowing the
liquid copper matte from the shaft furnace with compressed air to
-------
vo
CONNECTING
DUCTWORK SIPHON
REFRACTORY BURNER
LINING MOUTH HOLE
DRIVE
MOTOR
Figure 2. Siphon converter - longitudinal cross section.
-------
ROLLING
RING
LINING
MOUTH
TUYERES
MAXIMUM
60° FORWARD
INCLINATION
MAXIMUM
30° REVERSE
INCLINATION
Figure 3. Siphon converter - cross section.
10
-------
MINING
MINING, CRUSHING,
GRINDING, CLASSIFYING,
BENEFICIATING, DRYING
ll
•26
CHAR!J.E
BLENDING, GRINDING,
CLASSIFYING, MIXING
WITH HASTE SULFITE
LIQUOR, DRYING,
BRIQUETTING, AND
RECYCLING OF FINES.
AGGREGATE
MANUFACTURE
OR DUMP
SHAFT GAS
DUST CONTROL SYSTEM
CONVERTER GAS
DUST CONTROL SYSTEM
CUPRIFEROUS MATERIALS:
1. CONCENTRATE
2. BRIQUETTES
3. MATTE
4. BLISTER COPPER
5. ANODE COPPER
6. ELECGROLYTIC COPPER
7. SHAFT FURNACE DUSTS (COARSE)
8. CONVERTER SLAG
9. ANODE FURNACE SLAG
10. ANODE STUBS
11. RECLAIMED ELECTROLYTE COPPER
12. COPPER SCRAP TO CONVERTER
13. COPPER SCRAP TO ANODE FURNACE
BY-PRODUCTS AND HASTES:
NON-CUPRIFEROUS INPUTS:
14. SHAFT FURNACE SLAG 23. BLAST AIR
15. SHAFT FURNACE OFF-GASES 24. HATER
16. CLEAN SHAFT FURNACE GASES 25. SULFURIC ACID
17. SHAFT FURNACE DUSTS (FINE) 26. HASTE SULFITE LIQUOR
18. CONVERTER OFF-GASES 27. COKE
19. CLEAN CONVERTER GASES 28. QUARTZITE
20. CONVERTER DUSTS 29. HOOD AND OIL
21. HASTE ELECTROLYTE
22. ANODE SLIME
Figure 4. Flowsheet - Glogow No. 1 smelter.
11
-------
TABLE 2, INPUTS AND PRODUCTS OF COPPER SMELTING - GLOGOW NO. 1
Number referenced to Fig. 4
1 Concentrate
2 Briquettes
3 Matte
4 Blister
7 Shaft furnace dusts (coarse)
8 Converter slag
9 Anode furnace slag
10 Anode discards
11 Copper from electrolyte purification
12 Copper scrap to converter
13 Copper scrap to anode furnace
14 Shaft furnace slag
17 Shaft furnace dusts (fines)
20 Converter dusts
22 Anode slime
Cu
20-25
19-24
55-67
98.5-99.0
13-21
5-8
40-60
99.2-99.3
53-78
60-99
98-99
0.3-0.5
0.4-3..0
1.0-2.5
1.5-2.0
Analysis, %
Pb
0.7-2.0
1.0-3.0
1.5-3.0
0.1-0.3
5-15
0.5-1.0
0.5-1.0
0.1-0.2
-
-
-
0.2
35-45
45-55
20-45
S
6-11
7-12
19-23
-
6-12
-
-
-
—
-
-
-
7-14
6-14
-
Oraanics
3.5-7.0
4.0-6.5
_
_
8-18
-
_
_
_
_
-
_
6-13
-
-
-------
oxidize the sulfides and convert the remaining iron into a
slag. Three distinct stages of converter operation are used in
the Polish smelters:
1. Iron sulfide oxidation with siliceous fluxing mate-
rials, forming an iron oxide slag and sulfur oxides
(slag blow)
2FeS + 3O2 + Si02 •*• (FeO) 2« SiO2 + 2SC>2
3FeS + 502 -> Fe304 + 3S02
Copper sulfide oxidation, forming copper and sulfur
dioxide (copper blow)
Cu2S + 02"* 2Cu + S°2
Partial copper oxidation to further lower the lead
content (finish blow)
2Cu +
Cu.,0 + Pb -> PbO + 2Cu
^
Additional reactions occur in the first two stages, including the
oxidation of lead and zinc sulfides to oxides and sulfates and
partial oxidation of SO2 to SO^ .
The product of converting is blister copper, which has a
typical analysis of 98.5 to 99.0 percent copper, 0.1 to 0.3
percent lead, and 0.4 to 0.6 percent oxygen. The blister copper
is transported in ladles either directly to a fire-refining
furnace or to another converter in which it may be held in a
molten state. The converter slag contains up to 50 percent iron,
24 to 26 percent silica and 5 to 8 percent copper, and its
quantity is 21 to 23 percent of the input matte. The slag is
recycled to the shaft furnace to recover its copper content. The
high iron content of the slag also improves the quality of shaft
furnace slag, about 60 percent of which is used in the cement
industry and as trap rock in roadbeds. The high concentration of
copper in the slag occurs because the blast is not shut off
during slag pouring, but only reduced in intensity; this is done
to ensure a more continuous feed of gas to the acid plant. The
converter off -gases, containing a high SO2 and 803 content, are
withdrawn from each converter by individual fans to provide the
control necessary to minimize both excessive dilution and fugi-
tive emissions. The gases are cooled and cleaned of particulate
before they are used for acid production. Dusts from the dry
particulate control equipment (ESP's) are rich in lead and zinc
sulfates; a typical analysis is 45 to 55 percent lead, 4 to 8
percent zinc, and 6 to 14 percent sulfur. They are further
processed to recover the lead values.
13
-------
In addition to liquid matte, a converter is charged with
silica flux and cold recycled copper-bearing materials from
within the smelter. The flux is quartzite or quartzite gravel
with a grain size of 10 to 30 mm, containing 92 to 95 percent
Si02, 1.5 to 5.0 percent A1203, and 1 to 2 percent iron. The
amount of flux to be charged to each operating cycle is calcu-
lated based on the matte analysis; the intent is to convert all
of the iron in the matte to the silicate (FeO)2Si02. For ex-
ample, when the iron content in the matte is 9.72 percent and the
average silica content in the flux is 93 percent, 56.2 kg of dry
flux is needed per Mg of matte. The use of cold recycled mate-
rials prevents excessive temperature rises in the converter and
thus prolongs lining life; it also permits the recovery of this
additional copper. The recycled materials include slag from the
refining furnaces (which has an analysis of 40 to 60 percent
copper, 5 to 10 percent iron, 15 to 25 percent Si02, and 0.5 to
1.5 percent lead), and waste materials from the converter depart-
ment such as skulls, slag, and accretions.
In spite of the high copper content of the matte charged to
the converter, the process is autogenous and the heat released
within the furnace is sufficient to compensate for heat losses to
the atmosphere as well as that necessary to melt the cold re-
cycled materials. The exothermic reaction of the iron and oxida-
tion of copper create the heat, and a relatively high blast
intensity with maximum oxygen utilization is possible.
Blast air is supplied to the converter up to 30,000 Nm3/h at
a pressure of 8300 to 12,400 kg/m2; variations result from the
number of tuyeres used and their degree of obstruction. From 38
to 40 tuyeres are normally used, with 4 to 6 kept in reserve.
The tuyeres are cleaned by automatic punchers, which may also be
overridden manually.
The theoretical air requirements and time for converting one
Mg of matte, assuming total oxygen utilization and medium blast
intensity (23,000 Nm3/h), are shown below:
Slag blow 258.50 Nm3 0.674 min
Copper blow 541.00 Nm3 1.411 min
Finish blow 12.67 Nm3 0.033 min
812.17 Nm3 2.118 min
In practice, however, both air requirements and operating time
are greater than the theoretical because of air losses at the
tuyeres and incomplete oxygen use within the converter.
14
-------
In the siphon converter, the mouth is open at all times,
whereas in the Peirce-Smith, a gas collection hood covers the
converter mouth. Because the mouth is open, the efficient opera-
tion and control of the siphon off-take is essential. Improper
operation can result in either a pressure increase in the con-
verter that allows particulate and gases to escape through the
open mouth, or an excessive inflow of air which causes a carry-
over of substantial amounts of materials (slag, matte, and
blister) into -the siphon system. Solids drawn into the siphon
form deposits that impede the gas flow. Two operating factors
are of particular importance in preventing the carryover of these
materials. First, the converter should only be charged when the
surface of the melt is 0.2 to 0.3 m below the horizontal axis.
Second, only moderate blast intensities (about 12 Nm3/min) should
be applied at each tuyere. The other operating requirements for
siphon converters are similar to those for the Peirce-Smith
design. These include holding operating temperatures below
1350°C, use of high quality refractory materials, a careful
program of maintenance (especially tuyere cleaning), proper
addition of dry flux, and maximum utilization of the operating
time under blast.
Assuming total oxygen utilization, the theoretical volume
and SO- content of converter gases are as follows:
Slag blow - 240.2 Nm3/Mg matte, 14.99 percent S02
Copper blow - 539.3 Nm3/Mg matte, 20.83 percent S02
In actual practice, the gas volume is about twice and the SC>2
content about half of the theoretical because of incomplete
oxygen use and air inleakages into the furnace and ductwork.
THE OPERATING CYCLE
A typical complete operating cycle for a single converter
includes the processing of 147 Mg of liquid matte (seven 21-Mg
ladles), 9 Mg of solid recycled materials, and 8.5 Mg of flux.
The operating cycle is composed of the following distinct oper-
ations :
0 The solid recycled materials are transferred from the
stockyard and converter aisle in ladles and scoops and
charged to the converter by the overhead crane.
0 Liquid matte from the shaft furnace is charged to the
converter by the overhead crane in 21-Mg ladles.
Before pouring the first ladle, the converter is tilted
10 degrees toward the tapping position, and a blast of
15
-------
5000 Nm3/h is put on to prevent the tuyeres from clog-
ging. After pouring the matte, the blast is shut off.
This practice is continued through the pouring of the
first four ladles, which are supplied to the converter
every 5 to 10 minutes. During this period, the off-
gases, which contain 1 to 2 percent S02, are released
with natural draft through a bypass into an emergency
stack.
When the fifth ladle of matte has been poured into the
converter, the full blast is put on, the emergency
stack bypass is closed, and the converter is tilted
back to its vertical operating position. This begins
the slag blow. After 10 minutes under blast, a sixth
ladle of matte is poured into the converter and flux is
added in three batches (1.8 Mg each). This first stage
of the converting process operates with a blast inten-
sity of approximately 20,000 to 27,500 Nm3/h, with a
pressure at the tuyeres of about 8300 to 12,400 kg/m2.
During this stage, the blast rate and pressure are
unstable because of tuyere blockage, and regular punch-
ing is necessary to reduce this problem. The con-
verting process is run continuously until nearly all
iron sulfide in the matte has oxidized and liquid slag
is produced. This process is monitored by observation
of the matte and slag layer on the tuyere cleaning
punches.
When the matte has oxidized to a sufficient degree, the
converter is rolled out and the slag is drained into a
ladle. During this pouring operation, the blast is not
shut off entirely, but is instead reduced to about
10,000 Nm3/h.
After the slag pour, the converter is tilted back to
its vertical operating position, and the blast in-
tensity is simultaneously returned to its previous
level. A seventh ladle of matte is added, followed by
1.1 Mg of flux, and the slag blow is continued.
When the matte has again oxidized sufficiently, the
slag is poured off. As before, the blast is reduced in
intensity during pouring.
After the second slag pour, the second stage of the
converting process, the copper blow, begins. This
process is run continuously at a blast intensity of
approximately 22,000 to 26,000 Nm3/h and a tuyere
pressure of about 8300 to 12,400 kg/m3 until nearly all
of the copper sulfide has been oxidized. The blast
parameters in this stage are more stable than during
16
-------
the slag blow. The end of the copper blow is deter-
mined by observation of the blister on the cleaning
punches.
0 The final converting stage is the finish blow, during
which the blister is partially oxidized to decrease its
lead content. This stage is run at a decreased blast
intensity (approximately 20,000 Nm3/h) for 5 to 10
minutes.
After completion of the finish blow, the converter is
rolled forward, the blast and gas off-take to the par-
ticulate control system are shut off, and the emergency
stack bypass is reopened. The bath stands for a short
time to mature, and a final slag is drained from the
copper surface. A 2-Mg batch of flux is then added to
fix the remaining slag, which is difficult to remove
from the copper. This flux remains in the converter
aisle and is reused in the next operating cycle.
The ladle is then changed and the blister is poured off
in batches of approximately 25 Mg each. The ladles are
transferred to the refining furnace by the overhead
crane. If the refining furnace is not ready to accept
the blister, it is either held in the converter or
transferred into another converter, which acts as a
holding furnace, operating as in the copper blow, with
supplemental fuel if necessary.
0 The converter is prepared for its next operating cycle
by such operations as tuyere maintenance and breaking
off buildups around the mouth.
A THREE-CONVERTER SYSTEM
To assure steady utilization of matte from the shaft furnace
and a continuous SO2 supply to the acid plant, individual con-
verter operating cycles are synchronized to overlap the periods
of operation under blast. An operating schedule of a three-
converter system (Kl, K2, and K3) is presented here based on
measurements made over five successive operating cycles of the Kl
converter. The amount of matte processed in each converter and
operating cycle was 147 Mg, and its composition was stable and
within the typical analysis given previously in this section.
Measurements were made of the duration of each operating stage,
the blast intensity and gas flow to the particulate control
system, and the SO2 and S03 concentrations in the combined gases
from all three converters. The duration under blast was measured
separately for each operating stage. Table 3 shows the propor-
tion of the total operating cycle during which each converter was
17
-------
TABLE 3 i, CONVERTER TIME UNDER BLAST
Cycle
Operations
Charging and
waiting
for matte
(min)
Converter Kl :
1
2
3
4
5
45
60
50
55
50
Converter K2 :
1
2
3
4
5
6
42
65
40
55
45
35
Converter K3 :
1
2
3
4
5
6
52
40
45
35
47
55
Time under blast (min)
Slag
blow
117
114
120
115
118
112
115
110
118
110
116
116
-118
122
120
113
115
Copper
blow
235
230
242
238
233
228
230
236
225
240
227
232
236
230
226
244
235
Finish
blow
7
8
7
7
7
8
7
7
8
7
7
6
7
7
7
8
8
Matte pouring,
preparation
for
next cycle
(min)
120
125
110
125
117
120
108
132
119
123
137
104
124
116
110
120
122
Blast time/
total cycle
('%)
68.5
65.5
69.7
66.7
68.2
68.2
67.0
67.2
66.9
68.0
67.0
69.4
68.8
69.0
70.9
68.6
66.9
00
-------
under blast. The time under blast is relatively low for two
reasons: first, the necessity of waiting for matte from the shaft
furnace; and second, the high copper content of the matte. The
time under blast varies for each cycle for such reasons as dif-
ferences in the matte quantity and composition, tuyere blockage
and resulting different mean blast intensities, and differences
in oxygen utilization in the furnace.
Table 4 shows the characteristics of the combined gases fed
to the acid plant from the three converters during the test
period. Based upon these data, the number of converters under
blast during the test period was as follows:
3 converters 212 min 7.98 percent
2 converters 2391 min 90.06 percent
1 converter 52 min 1.96 percent
0 converters 0 min 0
2655 min 100.00 percent
It is evident from Table 4 that a system of three converters can
be optimized to assure a continuous gas supply to the acid
plant, with the simultaneous operation of two converters under
blast occurring 90.06 percent of the time. The total combined
gas stream ranged from 1400 to 1600 Nm3/min for 96.4 percent of
the test period, and from 900 to 1000 Nm3/min for the remainder.
The SO2-S03 concentration remained within the 7/0 to 8.8 percent
range for 94.4 percent of the test and was higher (up to 11.5
percent) for the remainder of the time.
Table 5, extrapolated from Table 4, shows that there were
six combinations of converter operating cycles during the test.
During nearly 90 percent of the test period, there were either
two converters in the copper blow or one each in the slag and
copper blows.
19
-------
TABLE 4. GAS CHARACTERISTICS IN A THREE-CONVERTER SYSTEM
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
Time
(min)
18
27
7
8
102
83
5
6
104
37
3
4
114
66
6
7
101
47
9
7
94
73
2
6
116
39
7
13
120
51
3
4
114
37
33
7
80
68
8
9
115
26
7
7
110
40
48
7
58
87
7
23
118
9
8
116
59
41
7
67
50
Converter Kl
Cycle
.
-
1
1
1
1
1
1
1
1
1
1
-
-
2
2
2
2
2
2
2
2
2
2
-
-
3
3
3
3
3
3
3
3
—
~
4
4
4
-
-
-
5
5
5
5
5
5
5
-
-
Blow
_
-
s
s
s
c
c
c
c
c
F
F
-
-
s
s
s
c
c
c
c
c
F
F
-
-
S
c
c
c
c
c
c
F
-
~
S
c
c
c
c
c
c
F
.
-
S
c
c
c
c
c
F
-
-
Converter K2
Cycle
1
1
1
1
-
-
2
2
2
2
2
2
2
2
2
2
-
-
3
3
3
3
3
3
3
3
3
4
~
-
5
5
5
5
5
S
5
- •
-
-
-
6
6
6
6
6
6
Blow"
C
C
C
F
-
_
S
S
s
c
c
c
c
c
c
F
-
-
S
s
s
c
c
c
c
c
F
S
S
C
c
c
c
c
F
-
-
s
c
c
c
c
c
p
-
-
-
-
s
c
c
c
c
c
Converter K3
Cycle
1
1
1
1
1
1
1
1
-
-
_
2
2
2
2
2
2
2
2
2
_
-
-
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
-
5
5
5
5
5
5
5
5
5
_
-
6
6
6
6
Blowa
S
C
C
C
C
C
C
F
_
-
_
S
S
c
c
c
c
c
c
F
_
_
_
S
s
c
c
c
c
c
F
F
S
s
s
c
c
c
c
c
F
-
S
S
s
c
c
c
c
c
F
_
_
s
s
s
c
Combined
gas
(Nm-Vmin
1400
1600
1600
1400
1400
1600
1600
1400
1400
1600
1000
1400
1400
1600
1600
1400
1400
1600
1600
1400
1400
1600
1000
1400
1400
1600
1000
900
1400
1600
1000
1400
1400
1600
1600
1400
1400
1600
1000
900
1400
1600
1000
900
1400
1600
1600
1400
1400
1600
1000
900
1400
1600
1000
1400
1600
1600
1400
1400
1600
Avg. SO 2
and SO3
cone .
(%)
8.2
6.8
11.6
8.1
8.1
8.7
11.5
8.1
8.1
8.7
7.0
8.1
8.1
8.7
11.7
8.4
8.3
8.9
11.8
8.4
8.4
8.7
6.9
7.9
7.9
8.7
7.0
7.8
8.1
8.6
6.7
7.9
7.9
8.7
11.4
8.2
8.2
9.0
7.2
8.0
8.3
8.7
6.8
7.6
8.2
8.5
11.3
8.1
8.1
8.4
6.6
7.4
7.9
8.5
7.0
8.1
8.8
11.6
8.3
8.3
8.8
S - Blag blow, C - copper blow, F - finish blow.
20
-------
TABLE 5. OPERATION OF THREE CONVERTERS UNDER BLAST
Converters
under blast
2
2
3
3
2
1
Blow*'
S, C
C, C
S, C, C
S, C, F
C, F
C
Duration
Minutes
1547
799
149
63
45
52
Percent
58.3
30.1
5.6
2.4
1.7
1.9
* S = slag blow, C »= copper blow, F = finish blow.
The production of sulfuric acid by the contact process
requires a stable gas stream of steady and strong S02 concen-
tration. Some fluctuation in feed gas is unavoidable, however,
because the individual converters are operating at different
stages of blowing and pouring or charging. The converter exhaust
fans must be regulated to minimize the differences in quantity
and composition to produce a stable combined gas. Table 6 shows
the fan drafts applied to the individual converters at different
operating stages during the test period. It shows that a con-
stant amount of draft was applied during slag or finish blows,
but that the amount of draft during a copper blow was dependent
upon the operating status of the other converters.
TABLE 6. FAN DRAFT ON .INDIVIDUAL CONVERTERS IN
A THREE-CONVERTER SYSTEM
Blow
Slag
Copper
Finish
Other converters under blast
Number
0
1
1
1
2
2
Blow"
-
-
S
C
F
S, C
S, F
-
Fan draft
(Nm3/min)
600
900
800
800
700
500
500
300
S = slag blow, C = copper blow, F = finish blow.
21
-------
DILUTION OF CONVERTER OFF-GASES
The quantity of gases released by the converting process
depends on the amount of blast air and on chemical reactions
between oxygen and the charge in the converter. The off-gases,
however, are diluted by air which is drawn in both through the
mouth and through leaks in the gas pumping, cooling, and parti-
culate removal system. It is difficult to measure directly the
amount of air dilution, although indirect estimates can be made.
Strip charts from continuous recording flow meters located on the
tuyere headers were used to determine the average blast inten-
sities. Based on these, and on process duration, the theoretical
volume of gases generated in each operating cycle was derived by
planimetry. Loss of 3 percent air at the tuyeres and utilization
of 85 to 92 percent of the oxygen were assumed, based on indica-
tions from other tests. Results of those calculations are pre-
sented in Table 7.
The effects of the oxygen/sulfur reactions on the quantity
of off-gases generated as converting proceeds can be illustrated
as follows:
Slag blow:
2FeS +
3FeS +
30,
50,
Si02 -»• (FeO)2Si02
Fe3S04
2SO,
3SO,
off-gases _ fl
blast air ~ °'90
Copper blow:
CuS +
2ZnS +
3PbS +
O,
3O,
30,
2Cu
2ZnO
2PbO
so
2S0
2S0
off-gases = Q
blast air u'yb
Finish blow:
4 Cu + 0,
2Cu20
off-gases _
blast air ~ °'76
22
-------
.TABLE 7. CONVERTER OFF-GAS PRIOR TO DILUTION
Cycle
Converte:
1
2
3
4
5
Average blast intensity (Nm3/min)
Slag blow
r Kl:
376.9
385.1
376.7
388.7
377.5
Converter K2 :
1
2
3
4
5
6
412.9
399.1
407.7
382.2
418.2
383.2
Converter K3 :
1
2
3
4
5
6
360.3
358.9
355.7
366.2
380.5
371.3
Copper blow
392.8
396.5
387.8
390.8
398.3
426.3
410.9
394.7
424.2
401.2
409.2
377.6
387.7
392.4
401.8
375.0
380.0
Finish blow
300.0
268.7
285.7
314.3
307.1
275.0
321.4
314.3
287.5
314.3
321.4
341.7
300.0
300.0
292.9
262.5
268.7
Gas volume (Nm^/min)
Slag blow
339.2
346.6
339.0
349.8
339.7
371.6
359.2
366.9
344.0
376.4
344.9
324.3
323.0
320.1
329.6
342.4
334.2
Copper blow
377.1
380.6
372.3
375.2
382.4
409.2
394.5
378.9
407.2
385.2
392.8
362.5
372.2
376.7
385.7
360.0
364.8
Finish blow
234.0
209.6
222.8
245.2
239.5
214.5
250.7
245.2
224.2
245.2
250.7
266.5
234.0
234.0
228.5
204.7
209.6
to
U)
-------
Using these factors and analytical data of S02 concentration, the
proportion of dilution air to total off-gas volume in each cycle
was calculated, as shown in Tables 8 and 9. From estimates based
on other analyses, it appears that about 30 to 40 percent of the
dilution air comes through the mouth, and the remainder results
from leaks.
OPERATING PROBLEMS
To obtain an off-gas composition approaching theoretical in
a siphon converter, precisely interrelated operation of the blast
intensity and gas removal system (fan draft) is required to yield
a pressure difference that is essentially zero. This minimizes
excessive air dilution through the mouth and other openings,
while simultaneously allowing gas removal through the siphon. In
practice, this goal is difficult to achieve because gas removal
is controlled by changes in fan speed that are slower to respond
than the rapid changes which can be made to blast intensity. The
converter gases can be pulled away too rapidly, increasing dilu-
tion" and thereby increasing gas volume and reducing S02 concen-
tration. This also has the effect of greatly reducing fugitive
gas emissions through the converter mouth. As the converter
operates in this manner, however, liquids in the furnace are
carried over into the siphon system and ductwork by the increased
gas velocity. When gas velocity decreases inside this siphon and
flow direction changes, these materials settle from the off-gases
and form deposits in the exhaust system, further reducing its
cross section and increasing the resistance to gas flow. In
extreme cases, the siphon or ductwork become so clogged that
despite increased fan speed, fugitive emissions occur through the
converter mouth. This situation could be partially corrected by
an appropriate reduction in blast intensity; however, this option
is not often viable because of the thermal process parameters and
the necessity for maintaining production. An increase in dilu-
tion also reduces the gas temperature and may result in the con-
densation of sulfuric acid in the ductwork.
The carryover of materials from the furnace into the siphon
is the major drawback of the siphon converter design. Deposition
is increased by the processing of a matte with high lead content,
as this increases the dust loading in the off-gases. The ac-
cretions occur most rapidly when the converter is filled with
matte above its horizontal axis and when it operates at high
blast intensities (more than 13 Nm3/min per tuyere). Removal of
the deposits is difficult and can only be accomplished when the
converter is cold to permit entry of men and equipment.
Disintegration of the refractory lining in the converter is
most rapid near the tuyeres, even though, to compensate for this,
24
-------
TABLE 8. CONVERTER GAS DILUTION DURING SLAG AND FINISH BLOWS
Cycle
Convert*
1
2
3
4
5
Slag blow
Fan draft 600 Nm3/min
Dilution
air
(Nm3/min)
2r Kl:
260.8
253.4
261.0
250.2
260.3
Converter K2 :
1
2
3
4
5
6
228.4
240.8
233.1
256.0
223.6
255.1
Converter K3:
1
2
3
4
5
6
275.7
277.0
279.9
270.4
257.6
265.8
Dilution
air/total
off-gas
(%)
43.5
42.2
43.5
41.7
43.4
38.1
40.1
38.8
42.7
37.3
42.5
45.9
46.2
46.6
45.1
42.9
44.3
Finish blow
Fan draft 300 Nm3/min
Dilution
air
(Nm3/min)
66.0
90.4
77.2
54.8
60.5
85.5
49.3
54.8
75.8
54.8
49.3
33.5
66.0
66.0
71.5
95.3
90,4
Dilution
air/total
off-gas
(%)
22.0
30.1
25.7
18.3
20.2
28.5
16.4
18.3
25.3
18.3
16.4
11.2
22.0
22.0
23.8
31.8
30.1
25
-------
TABLE 9. CONVERTER GAS DILUTION AT VARIOUS FAN
DRAFTS DURING COPPER BLOWS
Fan draft
Nm^/min
Dilution air
500
Converter Kl:
Cycle 1
Cycle 2
Cycle 3
Cycle 4
Cycle 5
122.9
119.4
127.7
124.8
117.6
Converter K2 :
Cycle 1
Cycle 2
Cycle 3
Cycle 4
Cycle 5
Cycle 6
90.8
105.5
121.1
92.8
114.8
107.2
Converter K3 :
Cycle 1
Cycle 2
Cycle 3
Cycle 4
Cycle 5
Cycle 6
137.5
127.8
-
-
-
—
700
.
-
327.7
324.8
317.6
_,_
305.5
321.1
-
-
-
«
-
323.3
314.3
340.0
- .—
800
422.9
419.4
427.7
424.8
417.6
390.8
405.5
421.1
392.8
414.8
407.2
437.5
427.8
423.3
414.3
440.0
435.2
900
-
-
524.8
-
.
-
-
-
-
-
.
-
523.3
514.3
540.0
-
Dilution air/total off-gas
(percent)
500
24.6
23.9
25.5
25.0
23.5
18.2
21.1
24.2
18.6
23.0
21.4
27.5
25.6
•-
-
-
-
700
-
46.8
46.4
45.4
43.6
45.9
-
-
-
-
46.2
44.9
48.6
-
800
52.9
52.4
53.5
53.1
52.2
48.8
50.7
52.6
49.1
51.8
50.9
54.7
53.5
52.9
51.8
55.0
54.4
900
-
-
58.3
-
-
-
-
-
-
-
58.1
57.1
60.0
-
26
-------
the lining is thickest in this area. Refractory wear is caused
by the effects of high temperatures, abrasion from the charge
materials, and chemical reactions.
OPERATING COST EVALUATION
The Glogow No. 1 smelter operates with three daily 8-hour
shifts. Three workers, a first furnaceman (converter operator)
and two assistants, are directly responsible for each converter.
Two additional workers are indirectly associated with each
converter in such supporting functions as materials transfer.
Other associated staff include electricians and control room
personnel (1.2 man per converter), laboratory technicians who
perform material analyses (0.3 man per converter), and the opera-
tors of the particulate control system (1.2 man per converter).
One worker is required for general maintenance and cleanup at
each converter on the first shift, and two are needed each shift
for maintenance of the particulate control equipment (with a
third assigned to each operating converter on the first shift).
Supervisory and management personnel are not included in these
requirements.
Electrical usage is recorded throughout the smelter but not
for each department, so only estimated requirements can be given
for each process. The total estimated electrical consumption for
the converter aisle and its particulate control system in 1976
and 1977 was 104 kWh/Mg blister product. The approximate dis-
tribution according to individual need was as follows:
Blast air compression 50%
Converter rotation 3%
Gas cooling 3%
Gas pumping 36%
Particulate control system 8%
Other utility requirements include natural gas, water, and
compressed air. Natural gas firing (heating value - 6500 kcal/
Nm3) is used both to reheat a cold converter and to allow a
converter to act as a holding furnace. Gas consumption is ap-
proximately 486 Nm3 per Mg blister copper. Water is used to
purge the bearings of the exhaust fans, as well as other rotating
equipment, and for general maintenance; water requirements are
1.33 m3 per Mg blister copper. Compressed air (six atmospheres)
is used for driving the tuyere punchers, operating jackhammers
and other equipment used for maintenance and refractory replace-
ment, cleaning the particulate control system, and controlling
servo-motors and monitoring equipment. The consumption of com-
pressed air is 20.8 m3 per Mg blister copper.
27
-------
SECTION 3
THE GAS CLEANING SYSTEM
CONVERTER EMISSION CHARACTERISTICS
It is difficult to fully analyze the physical and chemical
properties of the gases leaving the siphon converter because of
their_high temperatures, particulate loading, and chemical re-
activity. Some parameters have been directly measured, while
others, such as gas volume, have been determined indirectly as
described in Section 2. The test methods and equipment used are
described in Appendix C of this report.
The total volume of the off-gas after it leaves the gas
cooler on each converter ranges from 22,000 to 50,000 Nm3/h at
typical blast intensities; the amount of dilution is affected by
the condition of the siphon system, coolers, and connecting
ductwork. The operating temperature in the siphon converters
does not rise above 1350°C. The temperature of the off-gas
before it enters the gas coolers generally ranges between 800°
and 1050°C; it is primarily dependent on the blast intensity,
blowing stage, and gas dilution. The gas temperature is ap-
proximately 50° to 100°C higher during the copper blow than the
slag blow. When the blast is shut off or reduced, temperature
drops to 600° to 650°C. The temperatures measured at other
points in the gas offtake system are as follows:
gas cooler outlet 350° to 500°C
first-stage fan 450°C (maximum)
dirty gas mixing chamber 300° to 450°C
ESP inlet 280° to 400°C
ESP outlet 270° to 350°C
clean gas mixing chamber 300°C (average)
ductwork to acid plant 150° to 250°C
As detailed in Section 2, the theoretical S02 content of the
converter gas (assuming total oxygen utilization) is 14.99 per-
cent S02 for the slag blow and 20.83 percent S02 for the copper
blow. Monitoring and analysis of the gas before it enters the
coolers show approximately 13 percent S02 during the slag blow
and 18 percent S02 during the copper blow. Taking into account
dilution air, the concentration in the gases after leaving the
28
-------
cooler should equal 7.9 to 9.4 percent SC>2 (8.7 percent average)
for the slag blow, and 8.3 to 17.1 percent S02 (12.7 percent
average) for the copper blow, with an overall average of 11.4
percent SC>2.
Since the gas temperature does not exceed 1300°C, formation
of 303 results primarily from catalytic oxidation of S02. Many
substances are either true catalysts or in other ways promote SC>2
oxidation; some that have been found in the converter dusts are
listed in Table 10.
TABLE 10. SO2 CATALYSTS FOUND IN SIPHON CONVERTER DUSTS
Catalyst
Fe2°3
Fe203 + Bi203
Fe_0, + Sn02
CuO
AS203
Reaction temperature
°C
625
625
600
700
670
SC>2 converted
69.5
72.5
76.2
58.7
55.0
Studies indicate that 803 formation results largely from the
reaction of S02 with iron oxides present in the dusts entrained
in the gas stream, in accordance with reactions such as the
following:
2FeS
302
3FeS +
4Fe304
Si02 -> (FeO)2SiO2
3S02
6Fe203
This occurs principally during the slag blow, when the quantity
of iron in the converter dusts is greatest. The surface of
contact between the dusts and gases is also much larger with very
fine dusts. For example, the surface area of dusts 6 ym in
diameter is 1 m2/g, whereas it is 6 m2/g for dusts 1 ym in
diameter.
Testing showed that the S03 concentration in the converter
gases is further influenced by gas temperature. When the tem-
perature of the gases is less than 700°C at the converter outlet,
SO3 formation is inhibited and the concentration is maintained at
a low, steady level of 0.1 to 0.2 percent 803 by volume. When
the temperature of the gases is 800° to 1050°C at this point, 503
29
-------
formation continues as the gas is cooled; recorded levels were
0.7 to 0.8 percent 503, and even as high as 1 percent 503 by
volume.
THE CONVERTER GAS COOLING SYSTEM
A schematic representation of the particulate removal system
for the converter gases is presented in Figure 5. Each converter
is equipped with a gas-to-air heat exchanger, or gas cooler, and
with a first-stage fan, from which the individual gas streams are
mixed and cleaned in parallel-connected dry electrostatic precip-
itators (ESP's). The cleaned gases from the ESP's are again
combined into a final collecting chamber and from there are
picked up by second stage fans which deliver the gas to the
sulfuric acid plant about 600 m from the converter aisle. Gas to
the acid plant is still hot, 150° to 250°C.
Converter off-gases must be cooled to allow proper operation
of the particulate control system. Direct cooling was the best
answer at the Polish smelter. Air dilution was not a suitable
approach because of the necessity of maintaining a sufficiently
high S02 concentration for the acid plant. Water sprays unnec-
essarily increase gas volume, increase corrosion potential, and
cause problems at the acid plant. Despite initial difficulties,
the coolers developed by the Polish engineers have proved to be a
practical method.
The principal design problems in a cooler are cracks because
of thermal stress, dust accretion and corrosion caused by drops
in temperature below the acid dew point, and structural burnout.
The design problems are complicated by the temperature and gas
flow fluctuations, as well as the high particulate loading and
reactivity of the gas stream. Even with proper operation, gas
fluctuations can sometimes cause temperature to drop below the
acid dew point.
To minimize the effects of fluctuations in gas volume and
temperature, the off-gas streams from several individual conver-
ters are combined in a mixing chamber after passing through the
coolers. Use of this chamber ensures a relatively continuous gas
flow to the ESP's at a temperature above the acid dew point
(270°C). In addition, there is partial dust precipitation in the
collector because of the reduction in gas velocity caused by the
enlarged flow section. Dusts settle into hoppers, which are
periodically emptied.
To evaluate the operation of the gas collecting system,
detailed tests were conducted at the smaller smelter, which has
33-Mg converters. The gas removal system at this plant is
30
-------
FROM OTHER CONVERTERS r
AND FIRST STAGE FANS *
DIRTY GAS COLLECTOR
EMERGENCY
BYPASS
FIRST
STAGE
FAN
OJ
CONVERTER
COOLER
ATMOSPHERIC
AIR
I
COOLING AIR!
BLOWER |
~rOJ
HOT AIR TO
I ATMOSPHERE
1 *
t (XhJ-IXh
ESP
ESP
TRANSPORT
COLLECTOR
SECOND
STAGE
FANS
JANSEN
DAMPERS
CLEAN GAS COLLECTOR
TO ACID
PLANT
Figure 5. Flow diagram - converter gas cleaning system.
-------
identical to that used with the 80-Mg converters and similar test
results would be expected. Three converters and two ESP's were
operated during the test period.
Figure 6 charts the temperatures of the gas after it leaves
the coolers and at the ESP inlets during part of the test period.
The gas streams leaving the individual coolers are characterized
by temperature fluctuations that reach as high as 300°C. These
fluctuations, which result from changes in the blast at the
converter, are most significant during the slag blow. The rise
in temperature, which proceeds slowly, reaches the maximum value
0.5 to 1.0 hour after the initial blast begins.
During the test period, the gas temperature sometimes
dropped below the acid dew point (270°C). This typically oc-
curred for the first several minutes after the blast was put on a
cold converter. Transferring these gases from the coolers
directly to the ESP's would cause rapid corrosion of the elec-
trodes and insulators because of sulfuric acid precipitation.
The Charts reproduced in Figure 6 demonstrate that the first
mixing chamber stabilizes the gas temperature and protects the
ESP's from these hazards. The range of temperature changes at
the ESP's was less than 120°C.
The Original Cooler Design
The design of the gas coolers used with the Polish siphon
converters has been modified over the years. The first design
was an atmospheric natural-draft cooler, but it was soon replaced
by the forced-air circulation cooler shown in Figure 7. Each
side of this design had three double cooling sections, each of
which consisted of two vertical cylindrical segments containing
14 pipes (355 mm diameter) attached to the upper and lower sieve
bottoms. The first segments of the first section were attached
to an entry chamber that received gas directly from the converter
ductwork. The remaining five segments of each side of the cooler
sat on a settling chamber that had two compartments. Each of the
three sections was covered by a hood that changed the gas flow
direction. Each of the vertical segments had two or three inlets
that ensured proper draft and fresh air intake to the three
sections. The cooler was constructed of standard carbon steel
except for some elements of the first section, which were made of
a special steel. The entry and settling chambers were lined with
ceramic materials. The cooler weighed 400 Mg and was mounted on
a special steel structure with several platforms at different
levels to facilitate maintenance. The gas temperature was mea-
sured at the inlet; temperature and pressure were measured at the
outlet.
32
-------
420-
¥ 360-
«' 300-
a 240
a 180
8 120
60
0
/I
/I
1 2 3
4 5 6 7 8 9 10
l'2' 13 U lb Ib
23 4 56 78 9 1011 12 13 14 15 16 17
7 8 9 ion 1
1 23456
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Figure 6. Converter gas temperature -
cooler and collector outlets.
33
-------
ENTRY
OF COOL
AIR SECTION I
SECTION II SECTION III
ENTRY OF
GASES
A-A
LEFT
ENTRY OF COOL AIR
GAS-Uf
INLET
RIGHT
.OUTLET
OF GASES
OUTLET OF
HEATED AIR
Figure ?-. Original forced-air circulation cooler.
34
-------
The flow of converter gases and cooling air through the
cooler is presented schematically in Figure 8. The gas was
pulled through the ductwork by the first-stage fan to the entry
chamber, which was an integral part of the cooler. A bypass to
an emergency stack was located on the ductwork in front of the
cooler inlet. The gases in the entry chamber were divided into
two streams, which entered each side of the cooler. The gases
were remixed in the settling chamber after passing through the
first and second cooler section; they were also remixed in the
exit chamber after passing through the final section. Precipi-
tated dusts that accumulated in the settling chamber were removed
periodically. Cooling air was forced into the interpipe space of
the first vertical segment of each section. It flowed through
the three sections parallel with the gas and was discharged to
the atmosphere. Throttling-type control valves were used to
regulate the air flow through each section and to ensure either
partial or total heated air circulation. The operating param-
eters of the cooler are presented in Table 11.
TABLE 11. OPERATING PARAMETERS -
ORIGINAL CONVERTER GAS COOLER
Cooling surface area
Inlet gas volume
740 m 3
35,000-40,000 NnT/h
Gas temperature:
- Cooler inlet
- First section outlet
- Second section outlet
- Cooler outlet
Cooling air volume:
- Nominal
- Total
Cooling air temperature:
- First section outlet
- Second section outlet
- Third section outlet
850°-900°C
600°-650°C
500°-420°C
400°-450°C
20,000 Nm3/h
36,000-40,000 NmVh (avg)
180°-200°C
260C-280°C
300°-320°C
It was not possible to control the flow of cooling air
within individual sections of the cooler, and local overheating
occurred because of flow irregularities. This can be illustrated
by referring to Figures 9 and 10, which are simplified repre-
sentations of the temperatures of the gas, cooling air, and
cooler walls in the three sections. Figure 9 represents two
cases; in the first, air was only added to the first cooler _
section (Vp), whereas in the second, a lesser amount was added in
the first (Vp-AVp), and additional air was added in each of the
remaining two sections (AVp). The first case, in which all
cooling air enters in the first section, was clearly advanta-
geous; in the second case, wall temperatures could rise to above
the temperature limit of the construction material. In addition,
35
-------
00
o
Hi
O
o
n>
n
rt
(D
i-S
ro
(D
CO
0)
O
O
O
PJ
H-
rt
tr
n
o
c
IQ
tr
o
H
H-
IQ
H-
o
o
o
M
(D
-------
1000
SECTION I SECTION II SECTION III
900
700
S —
o
a:
CD
600
500
LOO
300
200
100
1- Vp = Vp
Kleft Vight
2- Vp = Vp
rleft Bright
\l
'
= M
Figure 9. Cooler temperatures -
symmetrical air flow.
37
-------
o
co
LU
LU
cc.
1000
900
600
700
600
500
400
300
SECTION I
200
100
V\v
\
SECTION II SECTION III
right
vp
-------
less cooling air was required for the first case. Figure 10
reproduces the steady air flow curve from Figure 9, and adds two
additional cases in which all of the flow entered the first
section, but was unevenly distributed between the left and right
sides. Total air flow was equal in both cases. These asym-
metrical air flows also resulted in temperature irregularities.
When less air was added to the left side, the temperature of the
first section walls could have risen to a hazardous level.
The gases entering the cooler were laden with dusts from the
converter, which settled on the walls of its internal ductwork
and obstructed the gas flow. To evaluate the increase in re-
sistance caused by these buildups, the pressure drop was moni-
tored from the converter exhaust to the cooler outlet. Measure-
ments were conducted for 30-minute periods just after the start
of each operating cycle for 62 days (128 cycles). The precision
of measurement was 5 mm H20. The arithmetic means of these
resistances are presented in Table 12, and the points are plotted
on Figure 11. This relationship can be described by the follow-
ing formula: p = A + Bi + Ci2. Using the test data, the param-
eters A, B, and C were calculated, resulting in the following
equation:
pi = 135 + 0.004412
in which: pi = resistance in cooler and ductwork after cycle i
i = cumulative number of cycles from beginning of
campaign
For the general case, the relationship would be as follows:
2
pi = po + 0.00441
in which po = resistance in cooler and ductwork at beginning of
campaign.
A number of problems were experienced with this cooler
design when it was operated with the 80-Mg converters. For each
cooler section, it was not possible to control the irregularities
in cooling air flow or to accurately predict the internal tem-
peratures. Gradual dust accretion occurred in the ductwork,
along with occasional acid corrosion of the structural elements
in the third section. Most seriously, the first section over-
heated and subsequently burned through, causing the converter
gases to be diluted with cooling air and the SC>2 concentration to
drop seriously. The perforated bottoms of the tubular segments
also cracked frequently because of thermal stress from the high-
temperature gases.
39
-------
TABLE 12. INCREASE IN COOLER RESISTANCE DURING CAMPAIGN
Cycle
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Average
resistance
(mm H2O)
120
120
130
100
120
140
120
100
130
100
140
160
190
190
150
160
160
180
130
140
130
140
170
120
160
170
150
140
150
130
120
140
Cycle
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
Average
resistance
(mm H_O)
150
140
130
120
150
140
150
130
150
160
110
130
130
160
130
130
200
180
160
180
170
160
150
130
140
150
190
160
100
100
120
180
Cycle
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
Average
resistance
(mm H_0)
120
160
140
150
110
120
150
160
180
190
180
200
140
120
140
190
150
160
170
180
130
160
130
190
170
200
160
130
180
170
140
200
Cycle
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
Average
resistance
(mm H20)
180
160
140
150
170
180
140
210
170
200
200
210
170
160
210
240
140
150
150
190
190
220
180
230
210
220
240
210
220
290
200
190
-------
Tfr
RESISTANCE (COOLER AND DUCTWORK),
3
O
H
(D
OJ
W
(D
O
O
O
I-1
CD
fD
CO
H-
CO
ft
o
fD
O
fD
ft
fD
H20
8
ro
O
o
co
o
o
ro
o
to
o
30
O
-n
o
to
-n
30
DO
m
tn
o
K-I CD
o
-n
O §
UD
o
o
o
ro
o
-------
The Improved Cooler Design
Using the experience gained from these initial units, the
coolers were replaced in 1977 by the new design shown in Figure
12. The most important modifications were the introduction of
atmospheric cooling in the first section and a change in the
direction of cooling air flow in the remaining sections. The
change in the first section was accomplished simply by removing
the internal tubular segments to allow the converter gases to
flow through the heat-resistant steel jacket and radiate heat to
the atmosphere. The general shape of the cooler was unchanged.
The area of the cooling surface is 633 m2, and the area reduction
from the original design is due entirely to the changes in the
first section.
A schematic of the flow of gases and cooling air through the
new cooler is presented in Figure 13. As in the previous design,
the gases are divided into two parallel streams as they enter the
first cooler section and are remixed following each section. The
vertical first-stage tubes, which are 1.75 m in diameter, protect
the remaining sections from excessively high temperatures. The
first-stage section has a total heat exchange area of 140 m-2, and
the degree of its cooling effect depends primarily on the rela-
tive gas and ambient temperatures. Based on typical gas flow
(32,000 Nm3/h) and dry 10°C weather conditions, it has been
calculated that the gases can be cooled from 950° to 841°C in the
first stage. The second and third stages operate as countercur-
rent heat exchangers, with cooling air requirements which are
greater than a comparable parallel cooler. The cooling air to
each section is divided into four streams that feed each segment;
the air is then discharged to the atmosphere. The total cooling
air flow and the flow to each segment are adjusted with throt-
tling-type control valves.
A test program was conducted to evaluate the new cooler
design, and measurements were taken at the points shown in Figure
14. Tests were conducted over a period of a month during the
middle of a converter's campaign. From 6 to 10 sets of measure-
ments were recorded during an operating cycle, and the majority
of parameters were recorded by continuous monitors. Some param-
eters were determined indirectly; for example, cooling air volume
was calculated by differentiating the function of the mean
dynamic and static pressures proportional to their midpoint
values, allowing subsequent values to be measured at their mid-
points only.
Based upon the measurements that were conducted, it was
found that the gas flow through the cooler during copper blowing
could be estimated by the following equation:
42
-------
HOT AIR OUTLET
GAS INLET
GAS
INLET
SECTION II SECTION III
GAS
C>OUTLET
Figure 12. Improved forced-air circulation cooler.
43
-------
FROM
CONVERTER
,x A -"I
TO ATMOSPHERE —
OUTSIDE AIR
Figure 13. Flow of converter gases and cooling air through improved cooler.
-------
FROM
CONVERTER
XIX XDt~
TEST PCHNT
TO ATMOSPHERES
OUTSIDE AIR
1
1
-f
1
I
o^
o
r"~™
0
LU
1
_«l
_l
O
1—
0
_^-
Figure 14. Measurement points for cooler test program.
-------
Vg =
18.6
(S00) • Va
(Eg. 1)
in which:
Vg =
S02 =
Va =
gas volume, Nra /h
average SC>2 content of converter gas,
average blast intensity, Nm3/h
percent
Calculations were based on a copper content in the matte of 64
percent, and a degree of oxygen utilization of 0.885, as had been
indicated by other tests. The amount of cooling air through the
individual cooler sections was calculated by the following
equation:
fApT
VAi _~V TT"
Va „ r—.
1/Api
iVTi
(Eq. 2)
which results from the following assumption:
Ap = 2
(Eq. 3)
in which:
VA = air volume in given time, Nm3/h;
i
AP
T
U
W
P
p Tu
VA ~ W • pu = W • T (Eq. 4)
number of individual cooler section N
pressure drop across individual section, ~~TJ
mean of air temperature at section inlet and outlet
conventional operating parameters
air rate
average air density
The stability of the coefficient of flow resistance results from
the fact that the construction of the individual segments is
identical, and therefore the operating conditions differ only
slightly between sections.
Figure 15 is a plot of the gas and cooling air temperatures
at various points in the cooler during two converter operating
cycles. The blast flow intensity is also shown to illustrate the
relationship between the converter operations and the cooler.
During converter shutdown between cycles, the cooling air was
shut off to avoid excessive temperature loss in the cooler. The
cooling air was resumed about 15 to 30 minutes after the con-
verter was rolled back to its vertical operating position, the
exact timing determined by the temperature of the gases leaving
the cooler.
46
-------
SLAG BLOW i
POURING
' \
\
SLAG
i
i
k
J\
T\
RIP
i
BREAK FC
POURING
AND
FINISH BLOW CHARGIN!
COPPER BLOW "^t /
AJ 1
1
^\f
\
k (~~
)U
' |
ST F!
I
I
OW I
\ *r
^> —
^~~~\
•»
MTENSITY /NmV
V,
h/
)R
\
SLAG
BLOW
H ^£
1
i
1
'
I
FINISH BLOW
COPPER BLOW
nn
^" 1 U
<
n/l
VV
V
\
\ ^
U
URING
SLAG
• L^
^
MOM
COOL
NG AIR TEMPERATURE /°C/
10 11 12 15
GAS TEMPERATURE:
I - BEFORE COOLER
2 - BEHIND SECT40N I
3 - BEHIND SECTION II
4 - BEHIND SECTION III
5 - BEFORE FAN
COOLING AIR TEMPERATURE:
I - BEHIND SECTION II, RIGHT SIDE
2 - BEHIND SECTION II, LEFT SIDE
3 - BEHIND SECTION III, RIGHT SIDE
4 - BEHIND SECTION III, LEFT SIDE
Figure 15. Gas and cooling air temperatures during
two converter operating cycles.
47
-------
Table 13 shows the average cooler operating parameters
obtained over a period of 1 month operations. These tests were
made during the copper blow part of the cycle, when conditions
are more stable than during the slag blow, which is also too
short to permit accurate measurements. Figure 15, however, shows
that the cooler operating parameters are similar during both
blowing periods. The air and gas temperatures in the cooler are
plotted in Figure 16; these measurements were made at the inlets
and outlets of the individual sections. The air curves are
steeper because of the greater heat capacity of the gas.
The improved cooler remains hot for about 30 minutes after
converter shutdown. If the temperature of the cooler is allowed
to drop too much, the cool converter gas at the beginning of the
next operating cycle could drop below the acid dew point. To
reduce the problem of acid formation, the cooling air would have
to be shut off until a sufficiently high gas temperature is
reached at the cooler outlet. During the test period, delays of
as much as 45 minutes were applied after the converter was rolled
back.to its operating position. This practice, however, may re-
sult in heat damage to the cooler walls. The use of a hot gas
collecting chamber, from which hot air can be automatically
distributed through the cooling air fans to retain heat in the
cooler, offers a reasonable solution to this problem.
The degree of air dilution through valves and connectors
varies inversely with the converter gas volume and the negative
pressure in the cooler. The amount of dilution can be calculated
by comparison of the S02 content in the gases at the cooler inlet
and outlet, as reported in Table 13. This leakage, which repre-
sents about 21 percent of the total gas flow leaving the cooler,
ranges from 2 to 40 percent. The amount of dilution provides a
good indicator of the structural condition of the cooler, as any
cracks or leaks will further reduce the S02 content. During the
test period, which covered the middle of a converter's campaign,
no signs of unusual wear were present.
The life of the tubular elements in the new cooler design is
difficult to estimate because of the short operating experience
since its installation. According to smelter personnel, damage
after the first campaign was insignificant and less than ex-
pected. In the original cooler design, the first and third
sections lasted 1 year, and the second 2 years. At least eight
segments had to be repaired after a campaign by repairing cracks
and changing pipes. After the first campaign with the new
cooler, however, repairs were only required on the first two
segments.
48
-------
TABLE 13. AVERAGE COOLER OPERATING PARAMETERS
Inlet gas volume 23,000 Nm3/h
Converter gas: _
- SO,, content - cooler inlet Ti'oa
2 - cooler outlet 11.3%
- Air inleakage b
- Temperature - cooler inlet
- first section outlet, left
- first section outlet, right 796°C
- second section outlet, left 593°C
- second section outlet, right 580°C
- final section outlet, left 473°C
- final section outlet, right 473°C
- cooler outlet 383°C
Cooling air: 3 ,,
- Volume - inlet 43,000 Nm/h
- second section, left 12,000 NirrVh
- second section, right 14,400 Nm/h
- final section, left 8,900 Nm^/h
- final section, right 7,700 Nm /h
- Temperature - inlet oi/L^i
- second section, left 0
- second section, right 213°C
- final section, left 236°C
- final section, right 232°C
49
-------
LU
UJ
ID
K)00
900
800
700
600
500
400
300
?nn
100
SECTION I SECTION II SECTION III
"*~^^
^>
"N.
\
\\
\\
\
--^
X^X
\\
\\
\v
i
---
— --^
X
\
\
\
\
\
\
\
\
X
1
Figure 16.. Gas and cooling air temperature.
50
-------
The accumulation of dusts in the ductwork remains a problem
even with the new cooler design. A layer of up to 10 mm thick
that forms as dusts settle from the gases both reduces heat
transfer and impedes gas flow. The dusts settle in layers that
trap air in the interstices and further increase the insulative
properties of the deposit. The settling rate depends primarily
on the gas flow and physical properties of the dusts. The
reduced cooling ability caused by the dust accumulation can be
corrected by allowing additional dilution air to enter the cooler
through flaps in the entry chamber. Since this would simul-
taneously reduce the SO? concentration, the preferred method of
correcting the problem is periodic cleaning of the gas ducts.
The tubular segments are sand-blasted between converter camj
paigns, as well as during a campaign if necessary. Sixteen to 24
hours are required for cleaning; the individual segments can be
reached through their upper lids.
The only continuous monitoring equipment presently installed
on the cooler system is a gas temperature indicator located in
front of the first-stage fan. To more properly monitor the
cooler's operation, the following measurements should also be
taken:
Cooling air temperature at each segment outlet
Cooling air flow
Gas temperature at first section outlet
Gas flow resistance within the cooler
S0» concentration at cooler outlet
Ductwork and Fan Specifications
The off-gases from each converter are routed to its cooler
through pipelines which are 2200 mm in diameter. Emergency stack
bypasses, 1200 mm in diameter, are attached. The remaining
ductwork from the cooler to the final collecting chamber is 1400
mm in diameter. All of this ductwork is made of acid-resistant
steel, and all of the pipelines and equipment from the cooler
outlet to the final collecting chamber are insulated with mineral
wool enclosed in sheet aluminum. The gas mixing chambers are
pipelines; the first has a diameter of 2008 mm, the second has a
diameter of 1750 mm. Both are made of carbon steel lined with
acid-proof brick. Dusts that precipitate in the first chamber
are usually removed manually twice a year. The final collecting
chamber, which has a diameter of 1966 mm, is also made of carbon
steel lined with acid-proof brick. It is not insulated and is
therefore equipped with a tank into which condensed sulfuric acid
drains.
51
-------
The first- and second-stage fans that pump the converter
gases are identical. Both are made of acid-resistant steel, and
they have rotors with blades shaped to prevent dust accretion.
They are driven by direct-current electric motors to simplify
speed control. Specifications of these fans and the cooling air
fans are presented in Table 14.
TABLE 14. FAN SPECIFICATIONS
First- and second-stage converter gas fans:
Manufacturer - Sturtevant (Great Britain)
Model - 24/87D
Size - 135,000 m3/h
Static pressure - 478 kg/m2
Maximum speed - 1000 rpm
Power requirements - 300 kw
Cooling air fans:
Model - WPW 80/18
Size - 831,000 m3/h
Static pressure - 260 kg/m
Speed - 985 rpm
Power requirements - 75 kW
ELECTROSTATIC PRECIPITATORS
The principal devices used for converter gas cleaning are
horizontal, four-field dry ESP's constructed under license from
Walther. They are made of acid-resistant steel to resist cor-
rosion and are insulated with wool made of blown slag covered
with sheet metal. A preheating system was constructed but it
proved ineffective because of insufficient capacity. The con-
struction of a new heater that will preheat the ESP to at least
270°C is projected. Dusts are removed from the hoppers once
daily. Table 15 presents specifications of the ESP's used on the
80-Mg converters, and the design operating parameters.
The ESP's have performed satisfactorily when all operating
requirements are met. From 95 to 99 percent of the particulates
are removed with an inlet particulate loading of 5 to 13 g/Nm3
(average 8 g/Nm3), a gas temperature equal to or greater than
300°C, and a gas flow of 15,000 to 25,000 Nm3/h. This lowers the
outlet dust concentration to 0.15 to 0.30 g/Nm3 (average 0.21
g/Nm3). This cleaning efficiency is obtained with at least 50 kV
applied to the electrodes and a corona discharge of 300 to 370 mA
(for two fields). The outlet dust loading increases to 0.3 to
52
-------
TABLE 15. ELECTROSTATIC PRECIPITATOR DATA
Model: H IV - 16/6 x 2/300 (Walther; Federal Republic of Germany)
Dimensions: 14,880 mm (1) x 3470 mm (w) x 11,420 mm (h); cross-sec-
tional area - 28.8 m2
Electrical field:
Interpole distance
Field dimensions
Collectors
Dimensions
Number
Total plate area
Rapper control
Discharge electrodes
Number per row
Total length
Support
Rapper control
Rapper frequency
Rapper drive motor
Dust hoppers: 8 x 8.0m"
Operating parameters
Maximum gas volume
Gas velocity in field
Gas residence in field
Pressure drop
Operating temperature:
Maximum 400°C
Minimum 270°C
Maximum voltage applied
Maximum corona current
150 mm
6.0 m (h) x 2.0 m (1)
plate, sigma type
6.0 m (h) x 322 mm (w)
6 x 17 rows
1576 m2
Mechanically programed
Strips with blades
16
4250 m
Quartz insulators
Continuous mechanical
0.4/min
Electric; 0.6 kW; 750 rpm
20,000 Nm /h
0.45 m/s
18 s
10 kg/m2
68 kV
375 mA
53
-------
0.6 g/Nm3 with a discharge of 200 to 300 mA, about 0.6 g/Nm3 for
100 to 200 mA, and greater than 0.6 g/Nm3 with a discharge below
100 mA.
Measurements were also conducted on the ESP's used on the
33-Mg converters, which are similar to those used at the larger
smelter. Table 16 presents these results for gases averaging 5.7
g/Nm3 at various corona discharge currents and gas velocities.
A cleaning efficiency of about 97 percent is achieved during
proper operating conditions. Since all the gas is further
cleaned before release to the atmosphere, this efficiency is
satisfactory. Table 17 through 20 present additional data
relative to ESP performance under various operating conditions.
The decrease in efficiency of particulate removal at temperatures
below 290°C or at reduced corona discharge power is clearly
illustrated. Figure 17 graphically depicts the relationship
between ESP performance and gas velocity.
Dust Analysis
Because of problems encountered in direct sampling of the
particulate content of the converter gases, the chemical composi-
tion of the dusts was determined on the basis of samples taken
from the hoppers. Table 21 presents the results of these anal-
yses, and analyses of dusts from the cooler hoppers. The content
of free sulfuric acid particles was 1.72 to 4.96 percent in the
dusts from the ESP hoppers, and 1.45 to 2.40 percent in those
from the cooler hoppers. Dusts from the ESP hoppers at the
larger smelter contained 0.93 to 4.14 percent free sulfuric acid.
The high quantities of sulfate sulfur in the dusts suggest that
the metallic compounds are principally in the sulfate form. This
conclusion was confirmed by thermographic analysis of five dust
samples from the ESP hoppers, which showed the following composi-
tion:
PbSO4 36 to 63 percent
ZnSO4 20 percent
FeS04 7 to 14 percent
CuS04 3 to 13 percent
Various trace metals are also found in the dusts, including 1.38
percent arsenic, 0.23 percent tin, and 0.05 percent bismuth. In
addition, lesser quantities of antimony, silver, selenium,
nickel, cadmium, and molybdenum were recorded. Arsenic occurs in
the oxide form, and is especially important to control because of
its toxic nature and its potential for poisoning the vanadium
catalysts of the acid plant.
54
-------
TABLE 16. ESP PERFORMANCE AS FUNCTION OF DISCHARGE POWER AND GAS VELOCITY
Current
density
(inA/m2)
Corona
discharge
power
(kW)
4 fields; temperature > 270°C
0.028
0.075
0.105
0.179
0.176
0.282
0.286
0.267
0.332
0.347
0.417
0.417
1.94
5.88
9.01
16.60
16.22
26.74
27.49
27.51
32.59
35.68
44.80
44.40
2 fields; temperature > 270°C
0.213
10.01
Gas
velocity
(m/s)
0.282
0.132
0.242
0.167
0.229
0.145
0.219
0.285
0.133
0.304
0.148
0.378
0.270
Power
requirements
(kW/1000 m3)
0.066
0.430
0.359
0.961
0.682
1.773
1.210
0.933
2.354
1.131
2.928
1.132
0.377
Outlet
particulate loading
g/mJ
0.174
0.093
0.110
0.062
0.066
0.041
0.061
0.095
0.044
0.079
0.042
0.058
0.159
g/Nm-3
0.385
0.195
0.225
0.132
0.140
0.090
0.132
0.205
0.095
0.170
0.087
0.123
0.344
Ul
-------
TABLE 17. ESP PERFORMANCE - TEMPERATURE >290°C POWER >15 KW
Operating parameters
(kV/mA)
Fields 1/2
65/110
68/90
80/320
60/180
60/180
61/180
62/170
58/140
66/240
69/240
69/240
65/230
65/230
65/210
66/200
65/210
70/280
72/290
72/280
63/200
66/250
68/260
68/270
63/110
74/340
73/320
74/300
74/290
74/290
74/330
74/310
63/150
63/150
62/155
62/160
Fields 3/4
68/270
68/270
60/320
64/320
64/300
66/300
64/260
59/160
58/180
60/200
60/200
60/230
61/220
62/210
62/190
63/200
64/200
63/230
64/220
71/210
62/300
66/310
66/310
60/240
63/340
65/325
61/330
74/320
63/320
63/330
61/320
57/100
55/100
56/100
57/100
Corona discharge power
(kW)
Fields 1/2
7.15
6.12
25.60
10.80
10.80
10.98
10.54
8.12
15.84
16.56
16.56
14.95
14.95
13.65
13.20
13.65
19.60
20.88
20.16
12.60
16.50
17.68
18.36
6.93
25.16
24.28
22.20
21.46
21.46
24.42
22.94
9.45
9.45
9.61
9.92
Fields 3/4
18.36
18.36
19.20
20.48
19.20
19.80
16.64
9.44
10.44
12.00
12.00
13.80
13.42
13.02
11.78
12.60
12.80
14.49
14.08
14.91
18.60
20.46
20.46
15.12
21.42
21.13
20.13
23.68
20.16
20.79
19.52
5.70
5.50
5.60
5.70
Total
25.51
24.48
44.80
31.28
30.00
30.78
27.18
17.56
26.28
28.56
28.56
28.75
28.37
26.67
24.98
26.25
32.40
35.37
34.24
27.51
35.10
38.14
38.82
22.05
46.58
45.41
42.33
45.14
41.62
45.21
42.46
15.15
14.95
15.21
15.62
Gas
volume
(m3/h)
48,000
46,700
15,300
15,300
15,300
15,300
17,500
16,800
23,300
23,300
16,700
23,100
23,100
23,100
11,100
11,100
28,300
28,300
29,500
29,500
34,400
34,400
34,400
20,400
35,700
35,700
35,700
43,000
43,000
31,150
44,800
22,000
22,000
24,700
24,700
Gas
velocity
(m/s)
0.463
0.450
0.148
0.148
0.148
0.148
0.168
0.162
0.225
0.225
0.161
0.223
0.223
0.223
0.107
0.107
0.273
0.273
0.285
0.285
0.332
0.332
0.332
0.197
0.344
0.344
0.344
0.415
0.415
0.300
0.432
0.212
0.212
0.238
0.238
Outlet
particulate
loadina
(g/m3)
0.060
0.078
0.052
0.047
0.044
0.054
0.036
0.062
0.061
0.034
0.018
0.088
0.038
0.088
0.064
0.045
0.069
0.065
0.085
0.095
0.088
0.080
0.083
0.057
0.041
0.051
0.055
0.060
0.059
0.048
0.062
0.064
0.066
0.101
0.057
(g/Nn\3)
0.127
0.165
0.108
0.100
0.092
0.113
0.076
0.128
0.130
0.066
0.035
0.194
0.087
0.194
0.144
0.100
0.147
0.138
0.185
0.205
0.192
0.176
0.182
0.130
0.090
0.110
0.119
0.127
0.130
0.101
0.130
0.143
0.150
0.220
0.121
Temperature
(°C)
290
290
300
300
300
300
310
310
300
300
300
320
330
330
330
330
300
300
300
300
315
315
315
340
310
300
305
300
310
290
290
330
330
305
300
(continued)
-------
TABLE 17 (continued)
m
Operating parameter*
(lev.
Fields 1/2
61/220
62/200
57/290
65/250
66/260
63/290
65/280
64/250
64/210
60/200
62/100
64/95
60/125
58/100
57/100
57/100
64/100
60/120
60/125
Mean
/mA)
Fields 3/4
62/280
60/280
57/200
67/225
65/270
69/270
58/280
58/260
60/260
61/240
57/170
58/165
59/135
60/170
60/180
60/180
61/180
60/280
58/300
• -
Corona discharge power
(kW)
Fields 1/2
13.42
12.40
16.53
16.25
17.16
18.27
18.20
16.00
13.44
12.00
6.20
6.08
7.50
5.80
5.70
5.70
6.40
7.20
7.50
-
Fields 3/4
17.36
16.80
11.40
17.08
17.55
18.63
16.24
15.08
15.60
14.64
9.69
9.57
7.96
10.20
10.80
10.80
10.98
16.80
17.40
-
Total
30.78
29.20
27.93
33.33
34.71
36.90
34.44
31.08
29.04
26.64
15.89
15.65
15.46
16.00
16.80
16.50
17.38
24.00
24.90
28.88
Gas
volume
(m3/h)
14,700
11,800
11,800
10,500
14,100
15,900
13,900
9,600
9,600
21,800
23,400
23,400
18,000
20,000
20,000
22,000
17,500
20,600
20,000
-
Gas
velocity
(m/s)
0.141
0.113
0.113
0.101
0.136
0.153
0.134
0.092
0.092
0.210
0.226
0.226
0.174
0.193
0.193
0.212
0.169
0.199
0.193
0.226
Outlet
parti cu late
loading
(g/m-J)
0.035
0.023
0.031
0.024
0.024
0.076
0.050
0.021
0.018
0.060
0.064
0.062
0.059
0.061
0.048
0.052
0.072
0.064
0.056
0.057
(g/NmJ)
0.077
0.052
0.069
0.051
0.051
0.168
0.110
0.097
0.039
0.129
0.139
0.131
0.126
0.135
0.106
0.115
0.153
0.136
0.119
0.122
Temperature
(°C)
325
320
330
305
305
315
310
315
310
300
290
290
300
320
320
320
295
295
290
-
-------
TABLE 18. ESP PERFORMANCE - TEMPERATURE >290°C, POWER <15 KW
Cn
oo
Operating parameters
(kV/mA)
Fields 1/2
51/50
50/50
_
45/20
43/20
45/20
44/20
48/50
-
53/20
50/20
_
65/115
58/70
57/70
58/50
58/50
55/50
55/50
50/50
-
60/200
60/120
59/120
Mean
Fields 3/4
48/60
50/45
61/200
64/220
46/20
44/20
45/20
44/20
48/50
58/120
46/30
43/25
55/100
55/100
52/70
50/70
50/50
52/80
50/70
51/70
51/70
60/240
-
58/130
59/120
-
Corona discharge power
(kW)
Fields 1/2
2.55
2.50
-
-
0.90
0.86
0.90
0.88
2.40
-
1.06
1.00
-
7.48
4.06
3.99
2.90
2.90
2.75
2.75
2.50
-
12.00
7.20
7.08
-
Fields 3/4
2.88
2.25
12.20
14.08
0.92
0.88
0.90
0.88
2.40
6.96
1.38
1.08
5.50
5.50
3.64
3.50
2.50
4.16
3.50
3.57
3.57
14.40
-
7.54
7.08
-
Total
5.43
4.75
12.20
14.08
1.82
1.74
1.80
1.76
4.80
6.96
2.44
2.08
5.50
12.98
7.70
7.49
5.40
7.06
6.25
6.32
6.07
14.40
12.00
14.74
14.16
7.20
Gas
volume
(m3/h)
16,800
16,800
18,800
18,800
28,000
27,700
27,700
27,700
12,600
26,800
31,100
34,100
39,200
18,800
13,900
13,900
13,700
13,700
13,700
10,000
10,000
15,900
31,500
24,000
18,000
-
Gas
velocity
(m/s)
0.162
0.162
0.181
0.181
0.270
0.267
0.267
0.260
0.122
0.258
0.300
0.329
0.378
0.181
0.134
0-.134
0.132
0.132
0.132
0.096
0.096
0.153
0.384
0.231
0.174
0.205
Outlet
particulate
loading
(g/n.3)
0.102
0.099
0.108
0.081
0.075
0.122
0.212
0.167
0.078
0.114
0.184
0.283
0.300
0.039
0.036
0.041
0.022
0.078
0.087
0.064
0.142
0.134
0.254
0.056
0.063
0.118
(g/NmJ)
0.211
0.205
0.236
0.179
0.160
0.278
0.470
0.380
0.165
0.244
0.405
0.618
0.672
0.096
0.080
0.098
0.048
0.169
0.186
0.137
0.300
0.296
0.550
0.119
0.135
0.257
Temperature
(°C)
310
310
310
320
325
340
325
340
295
300
320
315
325
330
315
315
315
305
300
300
290
315
305
300
300
-
-------
TABLE 19. ESP PERFORMANCE - TEMPERATURE 270°-290°C
Ul
Operating parameters
(kV/mAl
Fields 1/2
53/90
54/90
53/90
76/320
50/70
55/40
55/45
56/40
54/45
61/200
64/100
60/100
Mean
Fields 3/4
64/215
64/220
64/220
65/340
48/60
56/120
55/115
56/120
56/120
61/245
58/170
64/180
66/180
66/160
-
Corona discharge power
(kW)
Fields 1/2
4.77
4.86
4.77
24.32
3.50
2.20
2.48
2.24
2.43
12.20
6.40
6.00
-
Fields 3/4
13.76
14.08
14.08
22.10
2.88
6.72
6.33
6.72
6.72
14.95
9.86
11.52
11.88
10.56
-
Total
18.53
18.94
18.85
46.42
6.38
8.92
8.81
8.96
9.15
27.15
16.26
11.52
11.88
16.56
16.31
Gas
volume
(m3/h)
26,800
26,800
12,000
44,800
15,800
23,700
24,800
24,800
26,800
21,300
22,700
23,100
20,000
15,900
-
Gas
velocity
(m/s)
0.258
0.258
0.116
0.432
0.152
0.228
0.239
0.239
0.258
0.205
0.219
0.223
0.193
0.153
0.227
particulate
loadinq
(g/m3)
0.090
0.075
0.071
0.085
0.090
0.115
0.064
0.094
0.121
0.060
0.071
0.092
0.086
0.064
0.084
ig/Nin->;
0.186
0.155
0.143
0.180
0.185
0.232
0.133
0.192
0.250
0.125
0.148
0.185
0.176
0.131
0.173
Temperature
(-C)
280
280
270
285
280
270
285
280
285
280
285
275
275
275
-
-------
TABLE 2CL ESP PERFORMANCE - TEMPERATURE <270°C
Operating parameters
(kV/mA)
Fields 1/2
54/100
53/100
51/90
53/90
—
54/85
55/60
-
-
-
-
60/160
62/160
60/160
~
Mean
-
72/130
74/90
74/80
60/70
Mean
Fields 3/4
55/90
52/100
49/80
50/80
50/60
48/60
64/215
63/280
64/290
63/290
60/210
64/210
64/220
64/210
46/100
-
57/130
59/210
59/210
59/200
60/200
-
Corona discharge power
(kW)
Fields 1/2
5.40
5.83
4.59
4.77
-
4.59
3.30
-
-
-
_
9.60
9.92
9.60
~
-
_
9.36
6.66
5.92
4.20
-
Fields 3/4
4.95
5.20
3.92
4.00
3.00
2.88
13.76
17.64
18.56
18.27
12.60
13.44
14.08
13.44
4.60
-
7.41
12.39
12.39
11.80
12.00
-
Total
10.35
11.03
8.51
8.77
3.00
7.47
17.06
17.64
18.56
18.27
12.60
23.04
24.00
23.04
4.60
13.86
7.41
21.75
19.05
17.72
16.20
16.43
Gas
volume
(m3/h)
19,500
19,500
15,400
12,600
15,700
15,700
14,200
14,800
14,000
14,000
10,800
10,800
10,800
9,800
16,600
-
40,700
29,700
29,700
29,000
42,000
-
Gas
velocity
(m/s)
0.188
0.188
0.148
0.122
0.151
0.151
0.137
0.143
0.135
0.135
0.104
0.104
0.104
0.095
0.160
0.138
0.393
0.286
0.286
0.280
0.405
0.330
Outlet
particulate
loadina
(g/m3)
0.169
0.262
0.263
0.150
0.262
0.197
0.134
0.228
0.212
0.294
0.097
0.065
0.068
0.063
0.234
0.180
0.870
1.167
0.973
1.178
1.174
1.072
(g/Nm3)
0.335
0.520
0.528
0.307
0.538
0.405
0.280
0.415
0.386
0.547
0.193
0.129
0.136
0.127
0.418
0.351
0.740
2.150
1.810
2.280
2.250
2.046
Temperature
(°C)
260
260
260
260
260
260
265
255
255
265
260
260
260
265
265
-
245
220
225
245
240
-
-------
H-
•8
n
(D
CD
CD
H
l-h
O
T9
OUTLET DUST CONCENTRATION, g/mv
o
•
ro
DJ
3
O
(D
o
o
CO
Hi
O
rt
H-
O
O
HI
0)
CO
to
8
CO
I-1
O
O
H-
rt
-------
TABLE 21. ANALYSES OF DUSTS FROM ESP AND COOLER HOPPERS
Species
Lead
.Copper
Zinc
Iron
Total sulfur
Sulfate sulfur
ESP hopper dusts
Range (24 tests)
(%)
44.66-55.10
0.38-1.78
4.70-7.66
0.05-0.20
11.69-12.97
11.50-12.38
Average
(%)
48.75
0.82
6.39
0.10
12.25
12.04
Cooler hopper dusts
Range (4 tests)
(%)
33.73-47.90
2.25-8.80
2.70-5.33
1.20-5.20
11.34-14.60
10.86-14.10
Average
(%)
41.28
5.70
4.16
2.58
12.99
12.58
CTi
-------
The physical properties of the converter dusts have a direct
effect on particulate removal, collection, and transport. These
characteristics were determined based upon samples taken from the
ESP's on the 33-Mg converters; only verification tests were con-
ducted at the larger smelter because of the similarities in dust
properties. The presence of free sulfuric acid particles that
wet the dusts influences their physical properties, as shown in
Table 22. The great differences between the bulk and true
physical densities evident from these data are indicative of the
highly uneven surface of the individual dust grains.
The angle of repose of the dusts collected in the ESP is
shown i*n Figure 18. The increase in this angle in the direction
of the gas flow is characteristic. The largest quantities of
dusts precipitate in the first fields. The subsequent decrease
in grain loading favors the formation of sulfuric acid particles
that wet the dusts, causing them to adhere to the hopper walls.
This phenomenon was confirmed by analyses of the free sulfuric
acid particle content in the following samples taken from the
individual hoppers.
Field 1 0.93 percent H
Field 2 2.86 percent H
A* ~*
Field 3 2.94 percent H2S04
Field 4 4.14 percent H2SO.
The electrostatic properties of the dusts, in particular
resistivity, affect particulate removal in the ESP's. Because of
the lack of equipment for field testing, dust samples were heated
to 320° to 360°C in the laboratory (which is equivalent to tem-
peratures in the ESP), and resistivity was measured as the dusts
cooled. The results are shown in Table 23. The measured resis-
tivities were suitable for particulate removal in an ESP for
temperatures equal to or greater than 280°C. When the gas tem-
perature falls below the acid dew point, however, acid particles
precipitate on the dust grains and cause a severe drop in resis-
tivity. In such cases, the dusts become overcharged upon contact
with the settling plate and are returned to the gas stream. This
is the principal reason for the drop in cleaning efficiency at
temperatures below the acid dew point.
Fractional particle size analyses of the converter dusts
were conducted on samples taken from the hoppers, also because of
difficulties in sampling the gas stream directly. The results
are presented in Table 24. These values cannot be directly
related to the dusts present in the gas itself, however, because
of grain coagulation after settling.
63
-------
TABLE 22. DENSITY OF CONVERTER DUSTS
Density
(g/cm3)
Bulk density
Shaken density
True physical
density
33-Mg
converter
1% H2S04 in dust
Avg
0.44
0.53
4.30
Min
0.32
0.38
3.72
Max
0.61
0.74
4.84
33-Mg
converter
3-4% H2SO4 in dust
Avg
0.97
1.18
4.28
Min
0.65
0.75
3.60
Max
1.11
1.38
4.73
80-Mg
converter
l.'5% H2S04 in dust
Avg
0.58
0.69
4.12
-------
INLET
FIELD
1
FIELD
2
FIELD
3
FIELD
4
OUTLET ^
Figure 18. Angle of repose of dusts in ESP hoppers
65
-------
.TABLE 23. CONVERTER DUST RESISTIVITY
Temperature
(°C)
375
340
330
320
300
280
260
250
230
210
195
190
180
170
160
150
140
130
115
105
100
90
85
80
60
Resistivity
(ohm cm)
Sample 1
_
—
_
0.10 x 106
0.20 x 106
0.25 x 106
0.27 x 106
—
0.59 x 106
1.06 x 106
—
2.11 x 106'
—
5.28 x 106
—
1.05 x 107
_
2.11 x 107
_
5.28 x 108
_
2.95 x 109
—
8.45 x 109
2.53 x 1010
Sample 2
2.70 x 106
3.12 x 106
3.43 x 106
4.37 x 106
4.58 x 106
6.25 x 106
9.34 x 106
1.35 x 107
2.19 x 107
3.96 x 107
7.60 x 107
_
1.35 x 108
1.87 x 108
3.12 x 108
4.79 x 108
7.28 x 108
1.25 x 109
2.40 x 109
—
5.57 x 109
•_
1.46 x 1010
,_
8.85 x 1010
66
-------
TABLE 24. CONVERTER DUST GRAIN SIZE
Grain size
(ym)
>60
40-60
30-40
20-30
10-20
5-10
0-5
Fraction
(%)
24.62
12.85
9.80
9.30
13.21
11.32
18.90
Operating Problems
Frequent damage to the insulators is the most severe problem
encountered with the ESP's because servicing requires that the
power be cut off to two fields. This results in a decrease in
particulate removal efficiency and therefore an increased dust
content in the gases supplied to the acid plant. This naturally
increases operating costs. The insulators operate in more severe
conditions than any other known application, including fluctuat-
ing thermal and mechanical stresses and chemical reactivity of
the gases. During the first 6 months of operation of the ESP's
on the 33-Mg converters, 21 bushings, 19 support insulators, and
38 drive insulators were replaced. Although specific data are
not available, similar wear is likely on the ESP's on the larger
converters. The principal reason for insulator wear is the
continuous leakage of current from their surface to the dust and
acid. The problem may be mitigated by maintaining the operating
temperature above the acid dew point and the pressure between 0
to 10 mm H20. This is not always possible in practice, however,
and washing the insulators with hot, clean gas is necessary.
A second problem is the formation and precipitation of
sulfuric acid, which can occur in the cooler, ESP, wet scrubbers,
and ductwork. Sulfuric acid causes corrosion throughout the gas
cleaning system. Precipitated dusts settle out in the ductwork
and reduce heat exchange. Acid precipitation also reduces the
ESP's dust removal efficiency, damages insulators, and creates
problems with dust removal and transport from the hoppers. Acid
precipitation occurs when the partial pressure of the acid and
water vapor in the gas exceeds the saturation pressure. The
water vapor comes from either the blast air or inleakage into the
system, and it is assumed that the absolute humidity of the con-
verter gases is equivalent to that of the ambient air (3 to 16 g
67
-------
H20/Nm3, average 8 g H20/Nm3). Acid precipitates mainly on the
internal surfaces of the ESP, which are cooler than the gas. The
acid dew point can be lowered by decreasing the gas humidity
through dehumidification of the converter blast air. This would
increase operating costs, however, and would be only a partial
solution to the problem, as moist air would still enter the
system through the converter mouth and through leaks.
The acid dew point can also be lowered by a decrease in the
803 content, which can be effected by either a chemical binding
of the SO3 or a rapid cooling of the gas to about 700°C. The
results of lime addition to tie up 863 are presented in Appendix
A. Rapid cooling may be obtained by either water injection or
air dilution. Water injection has not been adopted in Polish
smelters. To cool the gases from 1250° to 700°C would require
evaporation of 0.2 kg of water per Nm3 of gas, which is about
5 m3 H2d/h at a blast intensity of 25,000 Nm3/h. Since this
approach increases gas humidity, poor cooling results, also a
drop in acid dew point of only 20° to 25°C. Allowing 80 percent
dilution through the converter mouth will lower the dew point by
40°C, but this approach cannot be recommended because of the
possibility of cooling the converter and the reduction in SC>2
concentration. Thus, the only appropriate method of reducing
acid precipitation in the ESP is by maintaining the gas tempera-
ture above 300°C.
FUGITIVE EMISSIONS
During slag blowing of the converter, many reactions occur,
including the oxidation of FeS with a simultaneous evolution of
S02 into the furnace interior over the molten bath. A negative
pressure at the gas offtake must be maintained to prevent fugi-
tive discharges through the converter mouth during this period.
Each of the several slag pours, during which the converter is
rolled forward and the blast is reduced in intensity, results in
a pressure pulsation in the converter. This produces a short-
term positive pressure that allows emissions to escape the
converter mouth.
Fugitive emissions are also a problem during the copper
blow, especially when increased air blast is applied at a depth
of over 0.5 m in the bath. At that time, the oxidation process
in the melt proceeds at a high rate, especially when the bath
temperature is 1200° to 1300°C. The rapid rate of reaction and
S02 evolution is demonstrated by the high oxygen utilization in
the blast air (90 to 95 percent) despite the short air residence
time in the bath (0.1 to 0.15 s) . The evolving SC>2 mixes with
the nitrogen in the blast air and the remaining free oxygen, and
^fills the converter interior near the tuyeres. Fugitive emis-
sions escaping through the mouth result from the pulsating
68
-------
variations in gas volume caused by the dynamic interaction of the
blast air and the gases evolving from the melt. The high-veloc-
ity gases entrain molten metal droplets, which contain both matte
and slag because of the blending of the melt caused by continual
blowing. Most of the droplets return to the bath, but others are
released through the mouth or carried over into the siphon and
ductwork by the pressure pulsations. Table 25 shows the fugitive
emissions present at various stages in the operating cycle.
Figure 19 shows the position of the converter during charging and
pouring operations.
A high blast rate is necessary to achieve adequate through-
put, and there is naturally a tendency to increase-the blast,
which results in a positive pressure in the converter. This
practice, however, conflicts with both environmental requirements
and the necessity of maintaining a high capture of SC>2 to supply
the acid plant. In addition, an upper blast limit cannot be
exceeded without causing the agitated melt to splash out of the
converter mouth. It is difficult in practice to maintain at all
times the slight negative pressure in the converter that would
prevent fugitive discharges. Because of the detrimental effects
on both the environment and the acid plant, the converter gases
cannot be bypassed to an emergency stack either during pressure
fluctuations or to prevent excess cooling of the converter during
blast reductions.
The only process monitoring at the converter is of blast air
rate and pressure, and off-gas temperature and composition.
Methods to measure fugitive emissions have not been developed by
the design engineers, and fugitive emissions measurements are
therefore not made because of the lack of a test method that
yields comparable data. This is not a problem unique to Poland.
A sulfur balance also does not yield reproducable data. Fugi-
tives are determined indirectly by measuring the levels of metal-
lic elements in the vicinity of the converters. These measure-
ments are conducted regularly; Table 26 presents the results of
tests made in 1977. These data show a considerable increase in
sulfuric acid concentrations during the slag blow; however, there
are no significant variations in SC>2, lead, and zinc because the
measurements are averages carried out over a period of time, and
therefore cannot reflect instantaneous changes or the continuous
variations in fugitives from individual furnaces. More sophisti-
cated measuring equipment located near the converter mouth would
be necessary to gain a better understanding of this problem.
Any estimates of fugitive emissions from the converters must
take into account the natural ventilation present in the con-
verter aisle that is upset by emissions from the furnace mouths.
The entire air distribution and movement in the area are there-
fore affected, and any interpretation of monitoring data must
69
-------
TABLE 25. CHARACTERIZATION OF FUGITIVE EMISSIONS
Converting stage
Fugitive emissions
Total duration
of emissions
(minutes)
Charging cold additions
Charging first five ladles
of matte
Charging sixth ladle of
matte and beginning slag
blow
Slag blowing
Slag pouring
Copper blowing
Finish blowing3
Cppper pouring
None visible
Slow fugitive emissions visible
shortly after blast is put on to
avoid clogging tuyeres; low dust and
SO2 concentration
Increasing emissions until the gas
takeoff system is regulated to the
increased blast
None visible
Visible emissions from stream of
slag leaving converter; S02 and
vaporized droplets
None visible
None visible
Very small amount of visible emis-
sions; particulate only
10-15
0
10
0
0
15
Carried to point of 0.5 percent oxygen in matte (5 minutes).
-------
CHARGING
POURING
Figure 19 . Converter -position - charging
and pouring operations.
71
-------
TABLE 26. RESULTS OF AIR SAMPLING NEAR CONVERTERS
Standard
V-
Slag blow
c
Copper blow
Concentration
(mg/m3 )
Cu
0.1
0.1343
0.2690
0.5243
0.7097
0.0978
0.1440
0.4440
0.7500
Pb
0.05
0.3868
0.4795
0.5754
2.1099
0.2250
0.3720
0.7190
0.9380
Zn
1.0
0.3325
0.3427
0.3836
0.5115
0.2610
0.3250
0.4500
0.5000
S02
20
37.50
70.50
87.00
-
60.0
62.0
67.5
84.0
H2S04
1.0
30.00
40.00
60.00
-
15.0
20.0
25.0
30.0
Maximum allowable workplace concentration established by
Polish Ministry of Health.
Air temperature 12.0°C, relative humidity 74 percent.
Air temperature 15.0°C, relative humidity 63 percent.
72
-------
consider that instantaneous changes may not be accurately re-
corded. It is likely that only average values could be obtained,
the analysis of which is subject to possible error.
The basic design of the converter assumed an air inleakage
of 30 percent of the blast air. This implied that a gas tem-
perature of 950°C in the siphon would result and also an S02
concentration of about 8 percent during the slag blow and 11 to
12 percent during the copper blow. In practice, the SC>2 con-
centrations are lower, probably because of increased dilution air
entering through the converter mouth. This increased air in-
leakage and lower siphon temperature result in the buildup of
materials in the siphon, which changes both pressure distribution
and gas flow patterns.
The chemical composition of fugitive emissions may be more
important than their quantity. The ambient SC>2 level was found
to be consistent irrespective of the converter operating cycle,
and averaged about 4 times the allowable standard for workplace
air. This indicates the need for still further control. The
sulfuric acid concentration was far higher than the workplace
standard and showed considerably more variation, confirming the
presence of fugitive discharges. Both U.S. and Polish smelters
share the SC^-K^SC^ problem, but the release of lead compounds is
a more serious problem in Poland because of the high lead content
of the concentrates. In contrast, there is very little arsenic
in Polish concentrates, and therefore no problem with this ele-
ment.
PRESSURE AND FLOW RATES
Gas pressures and flow rates in the gas cleaning system
could be determined by conducting measurements at appropriate
points. However, there are problems in conducting such moni-
toring, including the difficulty of developing instrumentation
suited to these operating conditions. Blast air pressure is
measured continuously at the tuyeres, and continuous gas pressure
measurements are made in the dirty gas collecting chamber, the
ESP's, the clean gas collector, the ductwork to the acid plant,
before the scrubbers and wet ESP's in the acid plant, and in the
single line following the wet ESP's. It is also possible to
measure the pressure of the gas when it is at the cooler outlets
and after it passes through the acid plant turboblower but
instrumentation for continuous measurement of these locations has
not been installed. Determination of flow rate is a greater
problem, and only the blast air to the tuyeres and the inlet
ductwork to the acid plant are now instrumented. The installa-
tion of additional continuous flow monitoring equipment is not
possible because of the frequent changes of direction in the
73
-------
ductwork and the resulting gas turbulence. A Prandtl tube could
be used to determine flow, but this would be a time-consuming and
rather inaccurate procedure.
A second problem in determining flow and pressure is the
highly variable operating conditions in the system; in partic-
ular, the changes which necessarily result from having a dif-
ferent number of converters operating at the same time. In
addition, the gas flow through the system varies during each
converter's operating cycle, and the leaks in the coolers and
ESP's vary between individual converters. The operating condi-
tions in the acid plant change significantly after replacement of
the catalyst bed, which in turn causes a substantial change in
pressure distribution and flow rates. Emergency situations,
operating problems, and equipment failures also affect pressure
and flow rates.
Because of these problems, a mathematical model was devel-
oped to determine the pressure and flow rates in the converter
gas cleaning system. It was developed for two specific cases:
operation of a single converter at a blast of 25,000 Nm3/h; and
a two-converter system with blasts of 24,000 Nm3/h to each
furnace. The model was initially developed based upon field test
data for pressure and flow at the points where measurements could
be made. It was then refined and verified by additional testing.
A digital computer of Polish manufacture {ODRA 1325) was em-
ployed, using the methodology described in Appendix B, which also
presents the results of these model computations.
THE GAS PUMPING CONTROL SYSTEM
The gases from each individual converter pass through the
cooler and are forced into the dirty gas mixing chamber by
first-stage fans. After the dust is removed in dry ESP's, the
gases are collected in a second chamber and are then forced by
second-stage fans to the sulfuric acid plant. Operation of the
gas pumping system is a complex operation because of the con-
flicting nature of the discontinuous gas flows from the indi-
vidual converters and the need for a continuous feed gas to the
acid plant. The acid plant is especially sensitive to changes in
the inlet S02 concentration, which should range over no more than
1 percent by volume. Fluctuations greater than this range reduce
the conversion efficiency and result in a tail gas with an S02
content that exceeds the 1150 ppiti emission limitation for the
Glogow No. 1 smelter. A considerable reduction in the feed gas
S02 content disturbs the heat balance in the contact acid plant,
which also lowers conversion rate. Variations in the flow rate
of the feed gases cause pressure fluctuations that 'can possibly
lead to gas leakage or dilution in the scrubbing towers.
74
-------
To provide an acid plant feed gas of steady quantity, pres-
sure, temperature, and S02 concentration requires a complex
operational and process control system. Many characteristics of
the process prevent simple operational and control systems from
working satisfactorily. Sudden and large changes in the gas
volume from individual converters result from the cyclical nature
of such operations as matte and slag pouring, flux addition, and
changes in the blast through the tuyeres. Fluctuations also
result from clogging and subsequent punching of the tuyeres, and
the occurrence of either fugitive emissions or air inleakage
through the converter mouth. The situation naturally worsens as
the operating cycle proceeds because of the increased hydraulic
resistance in the converter. Pressure fluctuations are caused by
the interaction of the individual converter gas streams as they
mix in the dirty gas collecting chamber. Pressures in the indi-
vidual ESP's fluctuate because of changes in the gas flow with-
drawn to the acid plant, whereas a constant slight negative
pressure is desirable to ensure an even gas flow through the ESP
and prevent dust accumulation on the insulators. The ESP tem-
perature must be maintained above 270°C to prevent acid pre-
cipitation, but below 450°C to prevent thermal damage to the
insulators and structural components.
The following conditions are necessary for optimal operation
of the gas cleaning system:
Proper scheduling of the operating cycles of the
individual converters
Prevention of fugitive emissions through the converter
mouth
Air dilution through the converter mouth equal to about
50 percent of the blast air
Gas temperature in the ESP's between 270° and 450°C
Prevention of a positive pressure in the ESP
SO,, concentration of 6 to 8 percent to the acid plant
Prevention of a positive pressure in the particulate
removal circuit in the acid plant
Effective communication between the converter operator
and the converter control room, including visual ob-
servation of the converter mouth
75
-------
The Automatic Gas Control System
Automatic process control to simultaneously achieve all
these objectives is difficult. The partially automated system
operated on the 33-Mg converters proved inadequate for several
reasons. Continuous direct measurement of the converter gas
volume was not possible because of the high dust loading and
entrained slag droplets, high temperature, and the tendency of
the lead-bearing dusts to agglomerate and stick to surfaces.
Measurements of internal converter pressure was possible, al-
though difficult, as the best location for a sensor was near the
furnace mouth, where it could be easily damaged during charging
operations. Converter off-gases could not be controlled by
valves in the ductwork because they soon became clogged with the
sticky dusts. The system pressure could not be adequately bal-
anced because changes in the gas flow from one converter caused
pressure changes in the dirty gas collector that were transmitted
to the other operating converters. A major factor which con-
tributed to the difficulties of automatic feedback control was
related to the size and configuration of the system, in which
large pieces of equipment are connected in series or parallel by
extensive ductwork. It proved slow to respond to changes in gas
volume because of the length of the duct runs.
Automatic controls for the gas cleaning system on the 80-Mg
converters have been designed based upon the experience with the
partially automated system on the smaller converters, and many of
the problems listed above have been solved. A more sophisticated
control system was desired, and it was assumed that improvements
could be made that would reduce the disturbances and upset condi-
tions that routinely occur. Figure 20 is a diagram of the sys-
tem, using Polish instrumentation designations. Automatic con-
trols would be provided at the following points:
0 Blast air to the converter
0 Cooling air intake
0 Gas pressure in the dirty gas collector
0 First-stage fan
0 Second-stage fan
0 Gas pressure in the acid plant wet ESP' s
0 Gas flow in the acid plant
Supplementary automatic override controls would also be provided
to rotate the converter from its operating position in event of
loss of blast, and also controls to monitor and synchronize the
individual elements in the system, punch tuyeres, and correct the
gas temperature at the cooler outlet. All controls would be
driven by electrically-signaled compressed air servomotors, which
have proved reliable in spite of the difficult operating condi-
tions. The following paragraphs describe the operation of each
76
-------
try ihttnttaiU
pnclpUaton
2nd ttaqt Riming MM
tani touin tttctnataOc to*tn
omfcut
apptrolt
ood rmitr
ttuignatian* : f-ftauiatl; ho- fOtfficiint tf than ; kit - auction cotfficitnt ; N~ rotations ; p -preiturt;
T- temptrotun; Hi • dltlrid iglut ; Z • position
• - adding tltmint • 0 - multiplying ettmtnt; tf- odjuttabit multlplitr; 183 -proportional controller ;
•—» rtmoti control.
Figure 20. Proposed automatic gas control system.
-------
element in the automated control system, which has been only
partially implemented at the Glogow No. 1 smelter.
The converter blast intensity control must both stabilize
the blast and reduce disturbances that arise during tuyere
punching. A conventional proportional controller of constant
proportional gain is used because the integrating elements of PI
(proportional-integral) or PID (proportional-integralderivative)
cannot respond quickly to frequent converter rotations. The
proportional controller can react immediately, and the system
should work well, providing automatic tuyere punching is provided
to prevent unstable blast flows because of tuyere clogging. The
flow rate is monitored with an orifice plate and transmitter on
the compressed air line to each converter. The flow controller
also receives an override signal from fan speed measurement. The
final control valve with its servomotor is placed on the com-
pressed air line at the converter in front of an emergency shut-
off valve. Both control valves can be activated either manually
or automatically from the control room.
The amount of cooling air pumped through the forced-air
coolers is adjusted by a damper at the fan inlet, which is also
driven by a pneumatic servomotor. The damper is actuated by a
temperature controller in the control room, which receives its
input signal from a sensor at the cooler outlet.
Interacting systems are used to control the first-stage fan
speed_and the gas pressure in the dirty gas collector, thus
assuring both proper flow from each converter and a constant
pressure in the collector. The fan speed is related to blast
intensity, and the speed of each fan is controlled in ratio to
the measurement of blast air flow to the corresponding converter.
This is a valid relationship, since all system pressures are very
close to atmospheric, and over the range of normal operation, fan
speed and flow rate are directly related. These signals are
further modified, however, by the output of a PID controller
whose input is from a transmitter measuring pressure in the dirty
gas collector. Therefore, all fans are simultaneously increased
or decreased in speed to maintain constant pressure downstream of
the gas coolers. After trimming, these signals are used to gen-
erate a signal which is a principal input to control the speed of
the acid plant turboblowers, as described in a following para-
graph.
The signal to each individual fan is trimmed once again
using a manual multiplying device. These final trimmers compen-
sate for variations in resistance in the ductwork from individual
converters, to maintain equal degrees of dilution air through
each of the operating converters. It is sufficient to make these
78
-------
adjustments manually once every several days because the resist-
ance changes slowly. To gauge the desired degree of compensa-
tion, the converter mouth must be visually monitored; the best
approach would be a control room with a clear view of the con-
verter aisle or closed-circuit television. This is the most
notable omission now in the Glogow No. 1 smelter.
Final control of fan speed is through the SCR (silicon con-
trolled rectifier) drive of the direct current fan motors, with
feedback from electronic speed measurements.
The temperature of the off-gases from the individual con-
verters, which varies with the operating cycle, is also adjusted
by signal to the first-stage fan regulators. This compensation
is made automatically in response to temperature changes measured
at the cooler inlet. To eliminate the momentary delay that would
occur in the temperature sensor when the converter blast is first
made, an overriding minimum temperature device is built into the
system that shuts off automatically when the sensor reaches that
level.
The acid plant wet ESP's are stabilized automatically at
atmospheric pressure by adjustments to the second-stage fans.
The PI controller on these fans is actuated by a signal from a
pressure sensor in the collecting chamber located in front of the
ESP's. Each fan's input signal is multiplied by the same ad-
justing coefficient; these fans and their drive mechanisms are
identical-to the first-stage fans. The pressure controller is
located in the control room.
The gas sent through the acid plant must be diluted with air
to maintain S02 concentraion between 6 and 8 percent. The total
flow of acid plant gas is also related to the total off-gas flow
from the converters. As previously described, signals are avail-
able which represent flow rate from each converter. The neces-
sary degree of dilution varies, however, depending on the blowing
stage an individual converter is in. During slag blow, dilution
ratio is 2 to 1, but during the copper blow, when more SO2 is
generated per unit volume of air, the ratio is 2.5 to 1. Instru-
mented signals are therefore generated using multiplying relays
in the control room which are manually adjusted to conform to
each converter's blowing stage. These signals are then totalized
with a summing device, and the output, with suitable scaling,
represents the desired total flow rate of offgas plus dilution
air to be sent to the acid plant. This signal becomes the set
point to a PI controller which adjusts a pneumatically-actuated_
damper in the turboblower suction. Flow of total gas to the acid
plant is measured with a venturi tube in a vertical section of
ductwork upstream of the turboblowers, thereby providing positive
feedback.
79
-------
Surge of the turboblowers is prevented by a recycle line
containing a control valve which bypasses from discharge to
suction to maintain a pre-set minimum flow rate. The bypass
valve is actuated when needed by a proportional-only controller
receiving an input signal from the venturi transmitter.
The converter must be automatically rotated to prevent
tuyere plugging in the event of insufficient blast. The con-
verters now installed rotate when the pressure in the tuyeres
drops below 0.4 kg/cm2. It has been observed, however, that
after punching, the pressure does not rise because the tuyeres
are out of the bath and there is no resistance to the blast. As
a result, the converter remains in the rolled-forward position
until the automatic system is overridden manually. To prevent
this problem, blast flow rate is to be added as a control param-
eter to the system. After cleaning, when blast flow is high, the
converter will be automatically returned to the vertical position
in spite of the low pressure caused by the lack of resistance.
Measurement of blast flow rate alone cannot provide adequate
control, however, because the plugging of a single tuyere could
then result in a perceived loss of blast that would cause auto-
matic rotation. Control by both pressure and flow is thus
required.
An automatic tuyere puncher of Polish design (patent PRL
90615; May 26, 1976) has been installed on the tuyeres to elimi-
nate _ the necessity for manual punching and to allow more stable
conditions in different operating stages. The puncher is acti-
vated by a signal when a specified level of pressure is reached
in the tuyeres. The puncher operation is flexible, and it can be
programmed to clean only every second, third, or fourth tuyere if
desired.
The interactive system of controls described in this section
allows synchronized operation of the converters, bypass stack,
parallel ductwork lines, first-stage fans, and gas flow. The
overall system would reduce disturbances or upsets, fugitive
emissions, and excess air inleakage, as well as simplify moni-
toring and control by the smelter personnel. Of particular value
would be the system's ability to react to rapid changes in blast
intensity without the delays that are characteristic with the
present manual control of the fans in a system of this size.
As this section has indicated, the design and operation of a
converter gas cleaning system are complex. Even a sophisticated
control system is unable to overcome the limitations of a poorly
designed or operated installation. Proper synchronization of the
converter operating cycles is essential. It would not be pos-
sible to have two of three converters simultaneously rotated
-forward for an extended period of time. The best approach to the
80
-------
design of an automatic control system must be based on feed-
forward of gas flow through the acid plant from the sum of the
blast air flows to the individual converters. Such a system
eliminates the necessity for any monitoring except temperature in
the converters or associated ductwork. It does not require the
use of frequently unreliable gas analyzers to control air dilu-
tion or S02 concentration. Proper operation of the ESP s is
assured by the controls in the dirty gas mixing chambers_and the
wet ESP inlets in the acid plant. The system operates with
continuously variable fans, and does not rely on the use of
valves in the ductwork. The only valves now in use control the
blast air and the turboblowers, and they are operated by reliable
servomotors driven by compressed air. The off-gases from indi-
vidual converters are controlled by use of the unique first-stage
fans, whose speed is modified by various compensating coeffi-
cients. This shifts the responsibility for their proper opera-
tion from a single operator to an integrated approach that in-
cludes the whole system.
CONVERTER PRESSURE SENSOR
Pressure fluctuations in the converter cause either fugitive
emissions or air inleakage through its mouth. Because the mouth
has a diameter of only 1.5m, changes in pressure of only several
millimeters of water can have a considerable effect. There is,
of course, no possibility of closing or blocking the converter
mouth because of the necessity of frequently charging matte,
flux, and cold scrap, as well as taking blister samples for
analysis. The pressure at the mouth should therefore be kept as
low as possible without creating excess air inleakage.
Pressure measurements within the converter are also diffi-
cult to make, however. Temperatures are high and fluctuating,
and the furnace atmosphere is unstable and reactive. Consider-
able fume and entrained particulate matter are in the gases.
Variations in the blast through the tuyeres and the amount of
obstruction in the siphon and gas removal system further affect
the internal pressure distribution. In addition, the_pressure at
the mouth changes during converter rotation and charging and_
pouring operations, during which time there is also the possi-
bility of the molten bath splashing onto an instrument. A pres-
sure sensor mounted inside the converter must therefore be able
to handle very difficult conditions, and still provide accurate
measurements over a converter campaign. For accurate pressure
measurements, a sensor must be located as near as possible behind
the converter mouth, as shown in Figure 21. At this location, it
is protected from direct contact with the molten bath or from
splashing even during full forward rotation or charging. A steel
sheet around the mouth can provide protection from damage by the
cold scrap additions.
81
-------
PRESSURE
SENSOR
TAP
Figure 21. Location of converter pressure sensor.
82
-------
During 1978, a pressure sensor was operated for 6 months on
one converter. The probe was successful in monitoring the
internal pressure near the converter mouth. Both these pressure
measurements and the blast air rate were recorded by continuous
monitors; representative strip charts are shown in Figure 22.
The probe did not require additional maintenance, and it func-
tioned properly throughout an entire campaign. It was not
obstructed by deposits, but instead the refractory lining grad-
ually wore away. During normal converter operations, the maximum
pressure fluctuations were only a few millimeters of water. To
approach the goal of zero exchange of ambient air and furnace
gases would require that these fluctuations be less than 1 'mm
H20. The test program demonstrated that changes in the blast
rate have a considerable influence on the pressure at the mouth;
an increased blast requires a corresponding increase in negative
pressure to prevent fugitive emissions. It was also determined
during the test program that the first-stage fans could be used
to stabilize the converter pressure.
During the probe test, the amount of negative pressure
required to minimize fugitive emissions was determined by visual
observation of the converter mouth. The measured value can also
be confirmed by calculations. Assuming a blast ranging from
15,000 to 30,000 Nm3/h and a bath surface of about 20 m2 (7.6 m
x 2.8 m), the velocity of the gases leaving the bath is as
follows:
v = 3d- = 15,000 to 30,000 = 76Q to 1500 m/h = 0>208 to 0.416 m/s
A 20
Assuming a gas density of 1.5 kg/m3, the pressure can be found
using Bernoulli's equation:
p = p Y! = 1.5 x (760 to 1500)2 = Q>32 to 1>3 ^ H o
Q. 2* £
The assumptions are conservative, and the negative pressure
should be kept at about 1 mm H20. The pressure measured with an
accurate sensor near the converter mouth could be compared with
atmospheric pressure near the converter, with the difference used
as a control factor for either the blast or the first-stage fan.
This represents an alternate control approach to the feedforward
system described previously.
83
-------
BLAST AIR
PRESSURE A
"THE MOUTH
0 -4 -2 -3-4 -5 -6 -7 p[mm
PRESSURE A
THE MOUTH
PRESSURE AT
FHE MOUTH
Figure 22. strip charts - converter blast
and pressure at mouth.
-------
SECTION 4
APPLICATION TO U.S. PRACTICE
As part of the Polish responsibility associated with this
research project, it is necessary to discuss how the procedures
described in the previous sections could be applied to the var-
ious types of copper smelters encountered in United States indus-
try. Although many of the techniques used in Poland differ from
conventional U.S. practice, most should be directly transferable.
The Polish copper industry has had considerable success with
the design, construction, and operation of the siphon converter.
Industrial units with capacities of 33 and 80 Mg of copper per
operating cycle have been employed for several years. The con-
verters have been equipped with several unique features of Polish
design, including automatic tuyere punchers, flux feeders with
automatic weighers, and a partially automated gas control system.
When compared with the off-gases from Peirce-Smith conver-
ters processing U.S. concentrates, the off-gases from the Polish
siphon converters have considerably higher S02, SOs, and lead
contents. The quantity of arsenic at the Polish converters is
much lower. The higher S02 content is a result of the reduced
air filtration into the system. The acid dew point is very high
(270°C) because of this high SO2 concentration and the SO3
formed during gas cooling. A comparison of the off-gas charac-
teristics of the two converter designs is presented below:
Peirce-Smith Siphon
Dilution air 150 percent 50 percent
Gas temperature 600° to 800°C 800° to 900°C
SO2 concentration 6 to 8 percent 10 to 14 percent
Despite differences in present operation as compared with
U.S. practice, no problems are expected in adapting the Polish
converter design to installations that process other types of
concentrates. Although matte containing 55 to 67 percent copper
and 8 to 14 percent iron is now processed, in the past matte with
a lower copper content (50 to 55 percent) and higher iron content
have been successfully converted. White mattes have also been
85
-------
successfully processed in the siphon converter. Oxygen enrich-
ment of blast air is necessary only when the charge is composed
of a high-copper low-iron matte and considerable cold scrap (50
to 60 percent of matte quantity).
A^comparison of the operating procedures cannot be made at
this time because of a lack of experience in Poland with the
Peirce-Smith converter. However, a stated U.S. objection to the
siphon converter is that the relatively small diameter of the
converter mouth causes problems with charging matte, cold dope,
and anode slag, or skimming slag and pouring copper. Because
spent anodes at Glogow No. 1 are sent to a refining furnace, a
converter with a large diameter mouth is not needed. A larger
mouth would naturally permit the charging of larger pieces of
cold dope. It is not believed that an increase in size would
influence the production capacity or proportion of cold additions
required for efficient operation; however, it would increase
fugitive emissions. The lengthwise diameter of the mouth is most
critical in this respect. The pressure in the converter varies
from negative at the siphon, to approximately zero at the mouth,
and positive at the far end wall, as shown in Figure 23. In-
creasing the mouth size toward the end walls upsets this balance
by exposing a larger portion of the furnace interior that is
operating under positive or negative pressure. To prevent such
an imbalance from occurring, an elliptical design should be
installed if a larger mouth size is desired.
After many years of experimentation, a particulate and SC>2
control system has been developed that consists of gas coolers,
dry ESP's, and an acid plant with wet gas cleaning, all syn-
chronized by a unique two-stage gas pumping system. This system
provides a gas stream of stable volume, temperature, and com-
position to the ESP's, thus assuring their proper operation. It
allows continuous gas withdrawal from the converters, regardless
of the number under blow, and is flexible enough to allow indi-
vidual ESP's to be bypassed if repairs are needed. The capital
costs are also lower than would be the case if each converter had
a separate particulate control system. Sampling and analytic
work has demonstrated that the use of this control system is
largely responsible for the stable gas characteristics.
Although the gas coolers still suffer some operational
damage, the newest design has proved to be more efficient,
stable, and flexible than those previously employed. This cooler
could be readily adapted to the Peirce-Smith converter with some
modifications. Because of the lower gas temperature and volume,
there is less chance for damage to the cooler with a Peirce-Smith
converter. In adapting the cooler design, there would be no
need for the initial atmospheric cooling section that is designed
to protect the remaining forced-air sections from damage. The
size of cooler necessary for this application must be defined for
86
-------
STANDARD MOUTH
ENLARGED MOUTH
ELLIPTICAL MOUTH
Figure 23. Effects of increased converter mouth size,
87
-------
the gas parameters which influence the rate of heat exchange. In
comparison to U.S. practice, the effects of Reynolds numbers and
mean logarithmic temperature difference cancel, each other, as
follows:
R2 ATm2 "
The amount of cooler surface needed thus depends primarily on the
amount of heat to be taken from the gas to reach the desired
temperature (400°C), as follows:
^2 = 600-400 ^1 ATml _
S, 800-400 R_ X AT _ U'D
l 2. m2
The cooling surface required for aPeirce-Smith should therefore
be about half that necessary for a comparable siphon unit. The
flow"cross-section of the gas cleaning system must be increased,
however, because of larger gas volumes.
To prevent excess internal gas velocity, there should be 1.5
to 2.0 times as much ESP capacity with the Peirce-Smiths. The
average particulate loading at the Glogow No. 1 ESP inlets is
8 g/Nm3. An outlet loading of at least 0.25 g/Nm3 can be ob-
tained if the ESP's are operated at 50 kV with a corona discharge
of 300 to 370 mA for two fields. The average loading behind the
ESP's is 0.21 g/Nm3, which represents a cleaning efficiency of 95
to 99 percent (average 97.24 percent). Since all gas is scrubbed
at the acid plant, better efficiency is unnecessary. The gas
temperature in the ESP's must be held above the acid dew point to
prevent acid condensation that will coat the electrodes with dust
and lead to reentrainment of the particulate and a loss in clean-
ing efficiency. ESP's are also suitable for use with Peirce-
Smith converters, although 1.5 to 2.0 times greater capacity
would be required because of the increased gas volume. Because
of the dangers of acid precipitation, a preheater should be used
to prevent the temperature from dropping below the acid dew point
during startup.
The efficiency of SC>2 absorption in the dual-contact acid
plant is about 99.0 percent, and the acid emissions in the tail
gas are within the prescribed limits. Sulfur losses in the acid
plant are thought to be about 3 to 6 percent of the total sulfur
loss, including from 1 to 2 percent in the blowdown slurry.
Increased sulfur control is possible, but would require more
efficient operation of the gas cleaning system. This would
entail an even more carefully planned converter operating sched-
ule, use of more modern ESP's, and implementation of a fully
88
-------
automatic gas control system based upon the principles outlined
in this report. With appropriate modernization and design modi-
fications, such a system could also be installed on a U.S. copper
smelter using either siphon or Peirce-Smith converters.
Proper performance of this system requires a program of
maintenance to minimize leaks of air into the system. Prevention
of leaks also requires operation of the gas removal system within
the proper temperature range, attention to gas shut-off devices
such as valves and flaps, and in particular, maintenance of air-
tight seals between the dust hoppers and the coolers, collecting
chambers, and ESP's. Leaks in the dust removal system from the
ESP's seem to have the greatest effect on the dilution problem
(where negative pressure up to 20 mm t^O can develop). In addi-
tion, leaks at this point reduce ESP efficiency as the settled
dusts become reentrained in the gas. Leaks in the gas cooling
and cleaning system can also result from improper choice of mate-
rials of construction, as well as poor design that fails to pro-
vide for the thermal stresses that will be encountered.
89
-------
SECTION 5
RECOMMENDATIONS FOR FURTHER RESEARCH
Several unresolved questions still remain concerning the
design of the siphon converter and the operation of the gas
cleaning system. The following work areas concern the converter
itself:
0 The physical and chemical characteristics of the
aerosol produced when an oxygen-enriched blast is
employed (especially at concentrations greater than 27
percent oxygen), and the effects on refractory life
° The design of the tuyeres, their angle of inclination
into, and effect on, the bath, and the degree of oxygen
utilization in the converting process
0 The effect of flux composition on the removal of trace
metals from the matte and the composition of the
furnace off-gases.
The following investigations could be made with respect to the
optimization of the gas cleaning system and the sulfur recovery
process:
0 803 formation in the converter, and techniques to pre-
vent or limit this process
0 Fugitive emission sampling and analysis
0 Trace element emissions to air, water, and solid waste
0 Implementation of the automatic control system desribed
in this report and assessment of its effect on sulfur
recovery
90
-------
BIBLIOGRAPHY
Bystron and Ma^usecki. "Swiezenie bogatego kamienia
miedziowego." (Refining of Copper Matte.) Rudy i
Metale Niezelazne. Nr. 6, 1965 r.
Bystrofi and Pieczka. "Zagadnienie swiezenia bogatych
kamieni miedziowych w konwertorach typu Hoboken."
(Problem of Refining Rich Copper Matte in Hoboken
Converters.) Sympozjum Nauk-Techn. Nowoczesna
Metalurgia Miedzi. IMN, Komitet Hutn. PAN, Komb.
Corn.-Hut. Miedzi t.I.
Czernecki, Smieszek, and Sobierajski. "Usuwanie Olowiu z
miedzi w procesach swiezenia." (Lead Removal from
Copper During Refining.) Prace IMN. 1975, t'.IV.
Dalewski, Ozog, and Skrzys. "Opyt ekspluatacii suchich
elektrofiltrow w cvetnoj metallurgi PNR." (Experience
in Operating Dry ESP's in Nonferrous Metallurgy in
Poland.) Bjull. SEV: Nauc.-Tech. Soveta po Ochr. Prir.
1977 r., nr. 3.
Grabowski and Kowal. "Konwertorowanie bogatych kamieni
miedziowych." (Converting Rich Copper Matte.) Sesja
Naukowo IMN "Rozwdj nowoczesnych technologii w
przemysle metali niezelaznych." Gliwice, 1977. t.
Ill, cz. 1.
Mscichowski, Peszat, and Babik. "Zagadnienie odpylania w
hutnictwie miedzi gazow konwertorowych, kierowanych do
fabryk kwasu siarkowego." (Problems in Particulate
Removal in Gases Supplied to a Sulfuric Acid Plant.)
Miedzynarodowe Sympozjum Nauk-Techn. Nbwoczesne Procesy
w przemysle met. niezelaznych. OPT Katowice, 1978.
Ozog and Peszat. "Skutecznosd elektrofiltr6w odpylajacych
gazy z konwertorowania kamienia miedziowego." (Effi-
ciency of ESP's in Particulate Removal from Converter
Gases.) Rudy Metale 1977 t.22, nr. 2.
91
-------
Sobierajski and Czernecki. "Swiezenie smopu
zelazo w konwertorze." (Refining of Copper-Lead-Iron
in a Converter). Sesja Naukowo IMN "Rozwdj nowoczesnych
technologii w przemysle metali niezelaznych." Gliwice,
1977. t. Ill, cz. 1.
Sobierajski and Smieszek. "Zagadnienie ogniowej rafinacji
miedzi od zanieczyszczen Metalicznych." (Problems of
Metallic Impurities in Copper Refining.) Sesja Naukowo
IMN "Rozwoj nowoczesnych technologii w przemysle metali
niezelaznych." Gliwice, 1977. t. Ill, cz. 1.
Weglorz and Warczok. "Rozwoj techniki konwertorowania
kamienia miedziowego." (Development of the Technique
of Converting Copper Matte.) Bjull. IMN 1970, nr. 3.
Wiechecki and Kopec. "Automatyzacja i mechanizacja wybranych
procesow technologicznych." (Automation and Mechaniza-
tion of Some Manufacturing Processes.) Sesja naukowo
IMN "Rozwoj nowoczesnych technologii w przemysle metali
niezelaznych." Gliwice, 1977. t. Ill, cz. 1.
92
-------
APPENDIX A
OTHER CONVERTER STUDIES
LIME ADDITION TO CONVERTER GASES
The off-gases from the siphon converters contain from 5 to
13 percent S02 (volume) at the ESP inlet. In addition, some SOs
is formed within the converter and ductwork. Table A-l presents
the results of S02-S03 testing of the converter gases, and it
indicates that the ratio of S03 to S02 ranges from 0.018 to 0.078
(average 0.035). The S03 content of the gases is of particular
importance because it is the principal determinant, along with
the steam content of the gases, of the acid dew point. Drops in
the gas temperature below this point (270°C) are undesirable
because of the resulting considerable increases in corrosion
within the system. A test program was conducted to determine the
effects of adding CaO or Ca(OH)2 to the converter gases to reduce
this problem.
Preliminary laboratory tests were first conducted to in-
vestigate the possibility of improving the physical properties of
the converter dusts. When Ca(OH)2 was added to a simulated
converter gas stream in quantities equal to 20 percent of the
total dust loading, the angle of repose was found to decrease
from 48 to 40 degrees. This indicated a reduced level of acid
precipitation.
Full-scale tests were then carried out at a 33-Mg converter
by charging lime by gravity feed through the emergency stack into
the ductwork between the converter and the gas cooler. The lime
quantity was 10 percent of the converter dust loading. The test
ran 44 hours, and was halted when it had become evident that the
dusts were not sliding down into the hoppers. Comparative anal-
yses of the dusts taken from the hoppers are presented in Table
A-2. It was not possible to draw samples from the first cooler
hopper because of sintering of the dusts and lining, but the
pattern evident from the second cooler and ESP hoppers is clear.
The increase in CaO content in these dusts is very small in
relation to the amount added. The majority of the lime simply
settled out of the gas stream in the first cooler hopper. The
free sulfuric acid content in the samples from the ESP hoppers
was then determined, with results as follows:
93
-------
TABLE A-l. SO2/SO3 CONTENT OF CONVERTER GASES AT ESP INLET
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Avg.
Converter
blast
(Nm3/h)
20,000
22fOOO
28,000
26,000
27,000
26,000
18,000
20,000
22,000
22,000
24,000
24,000
21,000
27,000
32,000
30,000
26,000
25,000
25,000
22,000
24,000
Temp.
ESP inlet
<°C)
420
400
285
325
375
380
385
325
325
325
325
325
360
400
400
400
320
350
380
400
360
S02 content
g/Nm3
389
191
251
307
421
378
399
257
296
286
307
307
362
236
278
318
197
228
322
405
320
%
12.10
5.90
7.72
9.47
13.01
11.67
12.31
7.93
9.12
8.80
9.47
9.47
11.15
7.27
8.59
9.80
6.37
7.55
10.44
12.80
9.55
SO-3 content
•*j
g/Nm3
9.04
17.40
6.41
17.84
13.39
15.17
16.34
12.95
10.27
12.38
14.62
17.10
23.11
20.54
18.44
11.13
4.13
7.10
9.88
11.40
13.43
%
0.22
0.42
0.15
0.44
0.33
0.37
0.49
0.36
0.28
0.34
0.41
0.48
0.68
0.57
0.51
0.31
0.11
0.20
0.27
0.31
0.36
S03/S02
ratio
0.018
0.071
0.019
0.046
0.025
0.032
0.032
0.045
0.031
0.038
0.043
0.028
0.050
0.078
0.059
0.032
0.017
0.026
0.026
0.024
0.035
vo
-------
TABLE A-2. CaO CONTENT IN HOPPER DUSTS
Sample location
Cooler, second hopper
ESP, first field hopper
ESP, second field hopper
ESP, third field hopper
ESP, fourth field hopper
CaO content, %
Before
test run
0.37
0.10
0.13
0.11
0.20
After
test run
0.92
0.96
0.57
0.26
0.34
95
-------
First field 0.93 percent
0 Second field 2.86 percent
0 Third field 2.94 percent
0 Fourth field 4.14 percent
Stoichiometric calculations indicate that only about 3 percent of
the total lime added is required to bind the 4.14 percent sul-
fur ic acid found in the fourth field.
The next area of investigation was to determine if lime
addition could be used to tie up the SOs in the gas stream and
lower its acid dew point. Assuming a CaO content of 60 percent
for calcined lime and 67 percent for hydrated lime, the quan-
tities required to completely neutralize the 803 were calculated
according to the following equations:
CaO + S03 -> CaSO.
Ca(OH)2 + S03 •* CaSO4 + H20
The "results are presented in Table A-3.
In the actual field tests, lime was supplied from an 18-Mg
tank and sprayed with compressed air into the ductwork between
the dirty-gas mixing chamber and the ESP on an 80-Mg converter.
The gas temperature in the ESP was held below 270°C by increased
use of cooling air. The acid dew point could not be measured
during the test because of equipment problems, and as a result,
the effectiveness of the lime addition could only be evaluated by
the formation of sulfuric acid mist on a steel rod. In addition,
the lime feeding apparatus was prone to damage. Thus the ob-
served results can only be considered preliminary. Neutraliza-
tion was not observed when lime was added at a rate of 40 kg/h
with a gas flow through the ESP of 20,000 Nm3/h. When the lime
addition was increased to 160 kg/h (with gas temperatures of
230°C at the ESP inlet and 180°C at the outlet) there was no acid
mist observed in the gas discharged from the ESP. With a gas
flow of 20,000 Nm3/h and a particulate loading of 10 g/Nm3/ this
quantity of lime is equivalent to 80 percent of the total dust in
the gases. Total daily lime requirements at this rate would be
3.84 Mg. As indicated in Table A-4, monitoring conducted during
the test program showed no increase in the ESP outlet particulate
loading as a result of the lime addition. Samples were also
drawn from the ESP hoppers and analyzed for lime and lead con-
tent; the results are presented in Table A-5. The coarser lime
particles deposited primarily in the first hoppers, which caused
some dilution of the lead content of these dusts.
96
-------
TABLE A-3. STOICHIOMETRIC LIME REQUIREMENTS
FOR S03 NEUTRALIZATION
SO £ content
g/Nm3
2.93
4.27
7.10
10.68
14.41
18.37
21.90
28.68
39.03
44.28
47.83
%
0.08
0.10
0.20
0.30
0.40
0.51
0.61
0.71
0.96
1.10
4.17
Calcined lime
requirement
(g/Nm3)
3.42
4.98
8.19
12.45
16.83
21.41
25.60
33.60
45.50
51.60
55.80
Hydrated lime
. requirement
(g/Nm3)
4.05
5.90
9.80
14.70
19.90
25.30
30.30
39.80
53.90
61.00
66.00
97
-------
TABLE A-4.
ESP EFFICIENCY DURING SC>3 NEUTRALIZATION TESTING
vo
oo
1
2
3
4
5
6
7
8
9
10
11
12
Avg.
13
14
15
Avg.
Lime
feed
(kg/h)
240
240
274
220
189
300
300
270
270
240
200
171
242
None
None
None
—
Gas flow
(Nm3/h)
14,900
13,100
15,100
13,700
14,800
18,900
19,700
21,800
17,100
24,200
24,200
24,300
22,200
24,200
26,100
18,800
23,033
Particulate loading
(g/Nm3)
Inlet
18.3
16.5
17.8
17.1
16.2
18.9
19.9
18.3
17.8
13.5
14.3
17.8
17.2
11.8
10.52
8.92
10.41
Outlet
0.31
0.31
0.36
0.48
0.36
0.29
0.31
0.37
0.27
0.36
0.21
0.24
0.32
0.24
0.52
0.18
0.31
Cleaning
efficiency
%
98.30
98.12
97.98
97.19
97.78
98.47
98.44
97.98
.98.48
97.33
98.29
98.65
98.08
97.96
95.05
97.98
96.99
Gas temperature
(°C)
Inlet
281
300
277
272
278
272
275
290
279
264
293
370
287
385
300
355
346
Outlet
260
275
244
245
263
267
268
259
271
262
283
315
267
314
290
303
302
-------
TABLE A-5. LIME AND LEAD CONTENT IN DUSTS FROM ESP HOPPERS
Sampling location
First hopper
Second hopper
Third hopper
Fourth hopper
Test 1
CaO
(%)
9.32
6.24
2.22
0.23
Pb
(%)
46.2
48.9
52.3
51.6
Test 2
CaO
(%)
13.12
10.60
3.15
0.76
Pb
(%)
38.1
48.8
49.7
48.3
Test 3
CaO
(%)
16.59
6.72
1.19
0.02
Pb
(%)
38.3
48.8
47.5
46.2
Test 4
CaO
(%)
27.46
8.62
3.25
1.00
Pb
(%)
27.5
36.7
38.9
35.1
<£>
10
-------
Because of the high lime requirements and the problems with
the charging system, further tests at the smelter were cancelled.
It is believed that lime addition is a feasible method of pre-
venting sulfuric acid condensation, but that maintenance of gas
temperature above the acid dew point is a more economical alter-
native. Lime addition can only be recommended in emergency
situations such as a sudden reduction in gas flow.
OXYGEN ENRICHMENT
The most important factor in increasing the efficiency of
the converting process is the amount of oxygen available in the
furnace. Some plants enrich their blast air with oxygen, which
requires the addition of large amounts of solid copper scrap or
concentrate to prevent excess temperatures and to protect the
refractory lining. Tests were conducted with up to 27 percent
enrichment of the blast air, and it was found that solid copper
scrap additions of up to 60 percent of the quantity of matte were
needed. Oxygen enrichment also increases the S02 content and
reduces the volume of the off-gases, which greatly facilitates
gas removal, cleaning, and sulfur recovery. The effects of
oxygen enrichment can be calculated. If a constant quantity of
dilution air is assumed to enter through the converter mouth, the
S02 content of the off-gases will increase in direct proportion
to the additional oxygen content of the blast air. This effect
is illustrated by the data in Table A-6. A constant quantity of
blast air decreases the time needed for converting a given amount
of matte in direct proportion to the degree of oxygen enrichment.
The total amount of oxygen in the blast cannot change if the
converting time remains the same; a proportional decrease in
blast air is required, as the table shows.
TABLE A-6. CALCULATED EFFECTS OF OXYGEN ENRICHMENT
02 concentration-
blast air
(%)
20.9
23
27
SO2 concentration-
off-gases
(%)
15.0
15.9
17.7
Blast air
requirements
(Nm3/hr)
24,000
21,900
18,700
Off-gases
(NmS/hr)
29,200
26,600
22,700
100
-------
APPENDIX B
PRESSURE AND GAS FLOW CALCULATIONS
The converter and its associated gas cleaning system are
capable of being studied using modeling techniques. Among the
processes that occur are gas pumping and storage, heat exchange,
chemical reactions, and particulate removal. The gas handling
system must operate in a manner that provides the sulfuric acid
plant with a feed gas of reasonably consistent volume, pressure,
temperature, S02 content, and particulate loading. Since changes
in gas flow at one point in the system affect all other points,
the whole sequence from converter to the heat exchangers and gas
preheater in the acid plant must be-considered in modeling.
Ideally, all of the gas parameters should be part of a model;
however, this is an extremely complex undertaking.
The development of a model of the Glogow No. 1 smelter was
limited to the gas pumping and storage system because of its
importance to the interactions among the system controls. The
goal in developing the model was to define changes in pressure
and gas flows through the converter and gas cleaning apparatus as
the converter operated through its cycle. Pressures in the
system are the principal determinant of the air dilution of the
gases and fugitive emissions, which in turn influence the gas
volume and SO2 content to the acid plant. Heat exchange phe-
nomena, flow resistance changes caused by temperature fluctua-
tions, chemical reactions, and particulate removal were excluded
from the model because of their minor effects on volume and
pressure. The relationships between pressure and flow rates are
nonlinear, as are the characteristics of the fans and turbo-
blowers. The model was therefore developed as a series of non-
linear algebraic equations, which were then solved at a variety
of conditions on an Odra 1325 digital computer.
A prerequisite of the modeling effort was to define scaling
factors that could be used to determine pressures and flows
within the actual system. It was assumed that all flows are
turbulent, and that the flow at a given point is related to
pressure drop as follows:
101
-------
h = R(V)2
in which: h = pressure drop (mm H_0)
V = gas flow (Nm3/h)
R = scaling factor [(mm H20)h2/(Nm3)2]
Two system configurations were selected for analysis, 1)
operation of two converters and two ESP's, and 2) operation of
one converter and two ESP's. Both cases assumed the operation of
three lines at the acid plant. In the illustrations at the end
of this appendix, Figure B-l represents the acid plant system
schematically, and Figures B-2 and B-3 represent the two cases of
converter operation. The scaling factors for the points indi-
cated on these figures were then calculated based upon actual
pressure measurements in the system and the best possible esti-
mates of gas flows. The calculated factors are presented in
Table B-l.
The fan characteristics in the system can be expressed as a
parabolic relationship, as indicated by the fan manufacturer's
typical curve, and as shown in Figure B-4. The equation of the
curve is:
h = a - bV - C(V2)
The constants a, b, and c can be derived empirically from known
values of h and V.
The following relationships define the coefficients from
only two data points:
x,-x
1 m
b = -2x
me
a =
The fan coefficients were first calculated for a speed of
1000 rpm. For both the first- and second-stage fans (1000 rpm,
AP =0, AV = 0), the coefficients were found to be: a = 435, b
1.5 x 10-3, and c = 0.0375 x 10~6.
102
-------
Changes in fan speed are accommodated by relationships which
are linear with flow change and vary by the square of the com-
pression change, as follows:
c = c
odn
b = (-
a = (-
n
n
odn
n
n
odn
"odn
2a
odn
A reduction in gas flow (AV) therefore results in the hori-
zontal shifting of the fan curve shown in Figure B-5. The fan
coefficients change is as follows:
c = c
v
b = b + 2c + AV
v
a = a - bAV - c (AV)2
Figure B-6 presents the changes resulting from a reduction in
system pressure (AP) . This vertical shifting of the fan curve is
defined by the following relationships:
f* rr; *"*
P
b = b
P
a = a - P
P
The turboblowers in the acid plant operate at a constant
speed of 1000 rpm. The fan coefficients for the smaller size
turboblowers were found to be: a = 1826, b = -53.8 x 10~3, and
c = 0.8594 x 10~6.
A program developed by Eugeniusz Kosek for calculating flows
and pressures in a system of square resistance was used to model
the gas flows in the converter system. It had previously been
used for ventilation systems in mines, as well as sanitary engi-
neering applications. For the present exercise, a program named
SUST was prepared; it was based on the original program and
written in Algol. The program utilized the Cross algorithum,
which relates pressure drops in a system according to the fol-
lowing equations:
103
-------
Ah = R(V)
AV - - -
sgn (V) - h
w
Ah
<*RV +
+ (cV)
in which:
h -> AP = pressure at primary element
hw = pressure drop across fan
R
= scaling factor, = = 1
The accuracy of the solution is ensured by specifying appropriate
values for Zn and Ev. Because the discrepancies in pressure
changes are less than In, the resulting flow changes will all be
less than Zv.
Preparation of the data for a program run consisted of the
following steps. All branches in the system diagrams (Figures
B-l, B-2, and B-3) were numbered in sequence and resistance
tables were prepared. Each branch was identified with a positive
or negative symbol to indicate a forward or reverse gas flow.
The number of independent loops was then determined by the
following equation:
L = M - N + 1
in which:
L = number of loops
M = number of branches (which may be less than
MAX)
N = number of points
Although not required, the selection of independent loops ac-
celerates program execution. The most rapid calculations are
obtained by ordering in sequence the number of branches chosen
beginning with those of lowest resistance. Those branches that
close at least one loop are excluded. The branches in which the
flows are in the direction of the loop are considered positive
and the remaining branches are noted as negative. The loops
selected for analysis are presented in Table B-2. Bypass of the
acid plant turboblowers was not considered in either the single-
or double-converter analyses.
To begin a program run, the following data are input to the
computer:
104
-------
Number of branches
Number of independent loops
Number of fans
Resistance table, with values in sequence finishing
with zero; branch number sequencing is optional
Fan table; fan number sequencing is optional
0 Loops desired; sequencing is optional, but should
finish with zero
0 The scaling factor (a) and values E and £
A completed program run prints out the gas flow, pressure drop,
and resistance in sequence for each branch. The program may be
modified to allow consideration of changes in either resistances
or fans, although changes in both require that the ending values
be given. The program can print out the conflicting loops after
each run, as well as all intermediate calculations. Preliminary
balancing flow values can be introduced to speed up the calcula-
tions. They are given in the form of resistance changes after
all data have been entered.
The calculated pressure and flow values were refined over a
series of program runs using different resistances and fan
operating parameters. After 10 iterations, values were obtained
that compared favorably with actual measurements conducted on
single- and two-converter systems. Figures B-7 through B-10
reproduce these results on schematic diagrams.
105
-------
Haat exchangers
Jas heaters
Turbo-blowers
Drying towers
** *S >'et electrostatic pracipitators
twxxs
To aulfnric
^.^ acid plant
Figure B-l. Gas flow through the acid plant system.
106
-------
TO a*?
l\ St«5e
L/ II fan
Electrostatic
prcelpitetor
Figure B-2. Gas flow through a two-converter system.
107
-------
To SAP
Electrostatic
precioitator
Converter's -out::
All blowers
Figure B-3. Gas flow through a single-converter system.
108
-------
TABLE B-l. CALCULATED SCALING FACTORS
Values of r for each numbered location:*
Sulfuric acid plant;
1)
2)
3)
4)
5)
6)
7)
8)
9)
Two
46)
47)
48)
49)
50)
One
46)
47)
48)
49)
0.0413
0.05
0.16
0.2
0.045
5
1.65
2
0.41
converters:
22.408
0.8
0.065
8.8
0.13
converter :
22.408
0.8
0.065
8.8
10)
11)
12)
13)
14)
15)
16)
17)
18)
51)
52)
53)
54)
55)
50)
51)
52)
53)
0.
0
0
5
100
100
0.
0.
1.
1.
20
0.
24
1
0.
0.
20
0.
0.
5
5
5
5
02
.3142
7
13
02
008
19)
20)
21)
22)
23)
24)
25)
26)
27)
56)
57)
58)
59)
60)
54)
55)
56)
2
2
4
4
4
4
0.045
0.175
0.05
10
0.15
30
0.02
0.008
0.008
0.1
0.08
28)
29)
30)
31)
32)
33)
34)
35)
36)
61)
62)
63)
64)
65)
57)
58)
59)
5
1.
0.
0
8
45
100
0.
1.
2
4
0.
0.
0.
0.
0.
0.
0.
0.
5
5
008
008
1
08
8
8
1
08
37)
38)
39)
40)
41)
42)
43)
44)
66)
67)
68)
69)
70)
60)
61)
62)
4
20
20
20
0.004
20
20
20
0.1
0.08
0.8
0.01
0.01
0.8
0.01
0.01
Locations refer to Figures B-l, B-2, and B-3. Scaling factors
in table have been increased by 10^ (R = r x 10 ).
109
-------
Xm
Figure B-4. Basic fan curve.
110
-------
m
j 1
Xmy Xfv xm X-f
xfv)
Figure B-5. Fan curves - reduced flow.
Ill
-------
*4
x(v;
Figure B-6. Fan curves - reduced pressure.
112
-------
TABLE B-2. LOOPS USED TO CALCULATE CONVERTER PRESSURE AND FLOW VALUES
Single-converter system:
MAX - 62. M • 58. N • 37, L • 22. a • 1, £„ • 0.01.
1) 46.-47
2) 47.48.-49
3) 47.48.50.-51
4) 55.56.-59.-58.-S4
5) 47,48,50,52,53.54.58,59,-56,-57
6) 47.48.50,52,53,54.58.-60
7) 61,-62
8) 1.3.-4.2
Two-converter system:
MAX - 70, M " 66, N - 41 . L
9) 47,48,50,52.54.58,59,61,41,1,3,5,-6
10) 47,48,50,52.53,54,58,59,61,41,1,3.5,7,9,11,13
11) loop 10.5,8.10,12,14
12) 19,21.-17,-15
13) 20,22.-18.-16
14) loop 10,5.7,9,11.15,17,23
15) loop 10,5,8,10,12,16,18,24
26. a - 1, I. - 0.01.
1) 46 .-47
2) 47.48.-49
3) 47,48.50,-51
4) 47.48.50.52.60.61,-59,-58
5) 47.48.50,52.60.61.-59.-57,-56
6) 47,48.50,52.60.61,-59.-57.-55,-54
7) 47.48,50.52,60,61 .-59.-57.-55.-53
8) 63.64,-67.-66.-62.-61
9) 47.48,50.52.60,61.66.67,-64,-65
10) 47.48.50,52,60,61,62,66,-68
11) 69,-70
12) 1.3.-4.-2
13) 47.48,50,52,60,61,62,66,67.69,41,1,3,5,-6
14) loop 13,5,7,9.11,13
15) loop 13.5.8,10,12,14
16) 19.21.-17.-15
17) 20.22.-18.-16
18) loop 13,5.7.9,11,15,17,23
16) loop 10,41,25,26,27,-28
17) loop 10,41.25.26.27,29,30,31,32
18) loop 17,31,33,34,37
19) 35,36,-34,-33
20) loop 10,5,7,-42
21) loop 10.5.8.-43
22) loop 10,41,25,26,27,29,-44
19) loop 13,5,8,10,12,16,18,24
20) loop 13,41,25,26,27,28
21) loop 20,27,29,30,31,32
22) 35,36,-34.-33
23) loop 21,31,33,34,37
24) loop 13.5.7.-42
25) loop 13,5.8,-43
26) loop 13,41.25,26,27 ,29,-44
-------
ToSAP
Stage II fan
Clean gas collector
Electrostatic
precipitator
Dirty gas collector
Converter'a mouth
Air blowers
-(flow rate) in 103 Nm3/h
-pressure in mm H_O
Figure B-7. Gas flow and pressure through
a single converter system.
114
-------
Wet electrostatic oreoipitators
-(flow rate) in 103 Nm3/h
-pressure in mm H^O
figure B-8. Gas flow and Pressure through
a single converter system.
115
-------
To SAP
1
£( t 1 Stag* II fan
•BJ-
Electrostatic
lipitator
Dirty gas collector
Converter's mouth
Air blowers
-(flow rate) in 103 Nm3/h
-pressure in mm H_0
Figure B-a. Gas flow and pressure through a
two-converter system.
116
-------
..'et •lectrostqtic precipitators
-(flow rate) in 103 Nm3/h
-pressure in mm HO
Figure B-rlO. Gas flow and pressure through a
two-converter system
117
-------
BIBLIOGRAPHY FOR APPENDIX B
Jerzykiewicz and Szczepkowicz. Algol 1204. PWN. Warszawa.
1972.
Nouvelles Techniques D1Etude et de Controle de la Ventilation
Minere. C.E.C.A., Louvain. 1969.
Vang and Hartmann. Computer Solution of Three Dimensional Mine
Ventilation Networks with Multiple Fans and Natural Ventilation
Intl. Journal Rock Mech. S.C.J. Pergamon Press. Vol. 4, 1967
pp. 129-154.
118
-------
APPENDIX C
VERIFICATION TESTING
This is a summary of the work performed by Monsanto Research
Corporation in verification testing for the Polish PL-480 project
outlined in this report. The main objective of the work was to
compare the particulate measurement method used for determining
particulate loadings and concentrations in the effluent and
process gas streams in Polish copper smelting processes to that
used in U.S. smelters. Specifically, the program provides a
comparison of the Polish method to the EPA Method 5 (EPA-5) as
published in the Federal Register.
This project was completed by performing several subtasks:
(1) A presurvey visit was conducted to observe the smelting
operation at the Polish Glogow smelter, to prepare for
sampling that operation and to observe the Polish
Particulate Sampling Method (PPSM).
(2) A sampling plan was prepared.
(3) Polish engineers and technicians were trained to use
the EPA Method 5.
(4) Comparison testing was performed at the Glogow smelter
using simultaneous sampling with PPSM and EPA-5.
(5) Analyses were performed on samples collected using both
methods.
(6) Data were evaluated from the project and included in
this final report.
A minor objective of this project was to attempt to estimate
the sulfur dioxide (SO2) emissions from the Hoboken style con-
verter that is used at the Glogow smelter by making SC>2 concen-
tration measurements of the ambient air at various points around
the converter. This minor objective was not achieved due to a
malfunction that occurred during the time of testing in the
sulfuric acid (H2S04) plant that controls the S02 emissions from
the converter. Management at the Polish smelter site decided
that sampling during this upset condition would be nonproductive
119
-------
because the concentrations during that time may not represent the
true S02 levels present during normal operation.
SUMMARY AND DISCUSSION
In the testing program, particulate measurements were made
at the inlet and outlet of an electrostatic precipitator (ESP)
dust collector that controls the particulate emissions from a
series of Hoboken converters. Simultaneous single-point runs
were made using EPA Method 5 and PPSM trains with the probe inlet
tips of each train located as close as possible to the same point
in the gas streams.
Two runs (Nos. 1 and 2) were made on the inlet stream and
seven runs (Nos. 3 through 9) were made on the outlet. Table C-l
presents the results of these runs and provides the particulate
concentrations obtained using the Polish method and those ob-
tained in both fractions of the EPA Method 5 train. The Polish
method is a single fraction, in-stack filter procedure (as
described later in this report) that is roughly comparable to the
"front half" or "noncondensable" fraction of EPA Method 5.
Therefore, a comparison factor has been calculated using the data
from these two collections, and it is shown in Table C-l. Samples
from Run No. 1 were not collected due to a broken cyclone in the
EPA Method 5 train.
During the testing, it was quickly determined that the
Method 5 equipment was not compatible with the conditions of the
converter effluent gas. The high concentration of dust in the
inlet line to the ESP caused rapid plugging of the filters and
made short run times necessary. When longer run times could be
used (at the outlet of the ESP), it was discovered that some of
the materials used in constructing the EPA Method 5 train could
not tolerate the strong acid conditions of the stack gas. It was
estimated that the sulfuric acid dew point of the stack gas was
in the range of 290° to 315°C. The silicone rubber material
which had been used to seal the filter holder and hold the filter
back-up frit was decomposed by the strong acid. Several attempts
were made to seal the filter holder with other material until a
suitable system was found. The new seal of steel and acid-
resistant rubber materials could not be heated, and the filter
was operated at just above ambient conditions. Table C-2 pro-
vides a brief description of the problems encountered with each
EPA Method 5 run (and the measures used in attempts to correct
the problems); Table C-3 briefly describes the possible effect of
these problems and suggests the direction of the bias that may
have occurred.
There was another problem that may have had a significant
impact on the results of the comparison. Because the dew point
of the stack was high, it was impossible to completely dessicate
the EPA Method 5 filters and the particulate wash residues to
120
-------
TABLE C-l. DUST CONCENTRATION OBTAINED USING BOTH TRAINS
(g/Nra3)a
Run
No.
2
3
4
5
6
7
8
9
Date
10/17/78
10/17/78
10/18/78
10/19/78
10/20/78
10/20/78
10/21/78
10/21/78
Polish
train, total
concentration
7.
5.
0.
0.
0.
2.
0.
0.
46
49
05
90
88
23
71
23
EPA Method 5
Front
half
35
15
14
19
10
19
19
11
•57.
.87
.88
.72
.31
.98
.69
.74
Back
half
0
0
1
2
1
0
1
.48
.09
.22
NAb
.55
.83
.18
.28
train
Factor
EPA front half fraction
Total
36
15
16
19
12
21
19
13
.05
.96
.10
.72
.87
.81
.87
.02
Polish
4
2
298
21
11
8
27
51
fraction
.78
.89
.9
.7
.96
.7
.04
•
Normal cubic meter at 0°C for Polish data and 20°C for U.S. data, both at
760 mm Hg pressure.
Sample lost due to an accident during analysis.
-------
TABLE C-2. DESCRIPTION OF PROBLEMS ENCOUNTERED
IN EPA METHOD 5 RUNS
Run
number
Problem encountered
1 (Inlet) Broken cyclone; sample contaminated with glass;
did not save
2 (Inlet) Filter plugged in 6 minutes; appeared to be a
good run, except that filter was wet with H2SO<,
3 (Outlet) Filter plugged in 12 minutes; silicone rubber
gasket appeared to be decomposing
4 (Outlet) Filter plugged in 39 minutes; silicone rubber
gasket destroyed (constructed gasket of Teflon)
5 (Outlet) Filter frit was broken through the Teflon gasket;
did not save back half of wash (constructed
frit holder of steel and gaskets of acid-
resistant rubber)
6 (Outlet) Appeared to be a good run, except that filter was
soaked with acid
7 (Outlet) Filter backed off of frit and was bypas$ed
8 (Outlet) Appeared to be good, except for acid on filter
9 (Outlet) Appeared to be good, except for acid on filter
TABLE C-3. EXPECTED RESULTS FOR EPA METHOD 5
Run
No.
Possible effect of problem
and suggested direction of bias
2
3
6
7
8
9
Should be representative
May have silicone rubber on filter; high front
half
Will have silicone rubber contamination; high
front and back halves
Filter bypassed part of the time; may cause low
front half
Should be representative, except for cold filter
Filter bypassed; low front half; high back half
Should be representative, except for cold filter
Should be representative, except for cold filter
122
-------
dryness. As a result, all of the EPA Method 5 values are likely
to be biased on the high side due to the mass of H_S04 on the
particulate.
The .following conclusions can be drawn regarding the
results of the tests:
(1) The EPA Method 5 can not be used to measure particulate
concentrations in the Glogow converter effluent in its
present configurations, due to the high acid dew point
and the high particulate loading.
(2) As a result of the Conclusion No. 1, the Polish method
and the EPA Method 5 can not be directly compared using
this gas/particulate stream.
(3) In the comparisons that were made, the EPA Method 5
concentrations are likely to have been biased high due
to an inability to completely dry H2S04 from the
particulate residues prior to weighing; this would
result in a high comparison factor.
SAMPLING AND ANALYTICAL METHODS
The Polish Particulate Sampling Method
The Polish Particulate Sampling Method (PPSM) used at Polish
copper smelters consists of elements and procedures similar to
U.S. EPA Method 5 for particulate measurement. In the PPSM,
preliminary tests are conducted to obtain velocity and tempera-
ture measurements of the duct to be sampled. A sample of the
duct gases is then taken extractively through filter media and a
flow metering system, so that the sample is extracted in a pseudo-
isokinetic manner. The mass of the particulate collected on the
filter media is then determined gravimetrically. A particulate
concentration is determined by dividing the total mass of partic-
ulate collected by the total amount of gas sampled, and this is
related to the total particulate loading in the duct by multiply-
ing the concentration by total flow rate.
The velocity of the gases in the copper smelter ducts is
determined by using a large P-type or standard pitot tube con-
nected to an inclined manometer. The pitot tube, approximately
2.5 cm in outside diameter and 1.8 to 2.4 m long, has an approxi-
mately 30.5-cm long pitot sensing element facing into the gas
stream. The stagnation impact area is a hole, approximately 11.5
cm in diameter, at the tip of the probe. The static openings,
which are slots rather than circular holes, are located 8 to 10
pitot tube diameters from the stagnation impact hole. The static
lines from the pitot tube are connected to a variable-angle
mercury/alcohol inclined manometer. For our tests, the manometer
123
-------
at the copper smelter contained alcohol as a pressure sensing
fluid and was set at a relatively low angle for sensitivity.
A scheme exists for laying out velocity traverse points; it
involves dividing the duct into numerous equal-area quarter-
circular regions, the number of which depends on disturbances in
the proximity of the sampling point. The centroid of each of
these areas is then sampled on two 90° traverses. This scheme,
however, was not followed at the copper smelter; in fact, only
one sampling traverse was used, and it consisted of 10 equally
spaced traverse points across the approximately 1.2-meter diame-
ter duct. Stack temperature was determined by inserting a
metal-encased glass-mercury thermometer into the port to an
immersion depth of about 20 to 25 cm. There were no provisions
for determining the moisture content of the stream. No stack gas
samples were taken for analysis to determine molecular weight;
however, a device similar to our Orsat analyzer was available at
the Institute, and it was indicated that the molecular weight of
the stack gases had been determined in the past.
The particulate sampling equipment consists of an in-stack
fiberglass plug filter, a probe, an umbilical line, a pressure
manometer, a dry gas meter, a rotameter, and an air-operated
vacuum aspirator. The filter consists of a steel cylinder with a
perforated end cap in which fiberglass is packed. Another
perforated end cap is placed on top of the fiberglass, and the
whole device is desiccated and weighed. The filter fits into a
probe nozzle head that is conical and faces into the stream to be
sampled. The leading edge of the nozzle is sharp, and the open-
ing is approximately 8 mm in diameter. This device is fastened
to the end of the sampling probe, which consists of a short
section of corrosion-resistant steel, a sweeping 90° bend, and a
section approximately 3 meters long. An umbilical tube, which
appeared to be made of reinforced neoprene rubber, is connected
to the end of the probe and directs the gases down to the inlet
of the dry gas meter. At the meter-inlet, a pressure tap is
connected to a mercury U-tube manometer to indicate pressure on
the inlet side of the meter. The dry gas meter is of the conven-
tional bellows type. The gases exit the meter into a float meter
(rotameter); from there they are connected by hose to an air-
operated vacuum aspirator.
A sampling run is started by determining the initial tare
weight of the filter cartridge in the laboratory. The filter is
then assembled into the probe nozzle, and the remaining associated
sampling equipment is connected. The probe is inserted through a
packing gland in the stack wall and faces in the direction of
flow at the first sampling point. The vacuum aspirator is started
and the control valve opened until the calculated flow volume is
'indicated on the flow meter. Each traverse point is then sampled
for a predetermined length of time, and the probe nozzle is moved
124
-------
to the next sampling point. At the completion of the run, the
probe is removed from the stack and allowed to cool. The probe
nozzle is then disassembled, and the filter element is removed to
a desiccator. The desiccator is then moved to the laboratory
where the increase in mass of the filter element is determined.
Several differences exist between the Polish method and the
U.S. EPA Method 5. The EPA Method 5 uses real-time velocity
measurements to determine sampling flow rate for isokinetic
sampling at each individual sampling point. The PPSM uses a
single sampling flow rate for all locations based on an average
of the velocities taken at an earlier time. In the EPA Method 5,
the sample stream is dried by removing moisture in- a series of
impinger condensers, and finally in a desiccant impinger, thus
assuring that only dry sample gas reaches the dry gas meter. The
PPSM uses no condenser due to the pressure drop inherent in such
condensers and the limitation in the amount of vacuum allowed on
the dry gas meter. This procedure can ..cause considerable conden-
sation in the dry gas meter and the rotameter after a short period
of operation. This pressure limitation on the dry gas meter also
limits the PPSM method to relatively short sampling periods in
highly loaded streams. Loading of the filter causes an increased
pressure drop on the entire system, thus, run times must be
shortened to accommodate the pressure increase. The EPA Method 5
suffers a similar pressure drop upon filter loading; however, the
out-of-stack filter used in EPA Method 5 can be changed when the
pressure increase becomes too high, and the meter of the EPA
Method 5 train is not in the vacuum area between the filter and
pump.
Another difference between the PPSM and the EPA Method 5
results from the temperature control of the filter media. In the
EPA 5 Method, the filter located in an external oven can be
controlled at a desired temperature. The PPSM has the filter
located in the stack, and it must be operated at stack tempera-
ture. Therefore, particulates and aerosols that are volatile at
stack temperatures may bypass the PPSM filter. This of course
was a major source of trouble in the EPA Method 5 train in that
sulfuric acid that bypassed the PPSM filter was condensed and
collected on the EPA Method 5 filter. The final difference
between the two methods stems from the operation of the PPSM dry
test meter at various pressures, starting from the start of the
run, under clean train conditions, to the end of the run when a
vacuum of 80 mm of mercury could exist at the dry gas meter. The
EPA Method 5 operates at a relatively constant meter pressure,
differing from barometric pressure only by the AH pressure which
is never more than a few inches of water above barometric pressure,
Figure C-l shows a diagram of the Polish Particulate Sampling
Method (PPSM) train, and Figure C-2 illustrates generally how the
filter container appeared to have been constructed.
125
-------
PORT CONNECTION
FLEXIBLE LINE
FLEXIBLE LINE
VACUUM
MANOMETER
Q*—FILTER CONTAINER
AIR/VACUUM
ASPIRATOR
DRY GAS METER
Figure C-l. Diagram of the Polish Particulate
Sampling Method Train.
126
-------
PROBE
L.I-*".
i.;'. • »M
">*V'X
*r • V L . t
Cw-t.-c>^----jr
U^^^vw.
-. ^7'.^'; •- -t.'^L
>^t-x.*t^-
^^X^gft-
&$$<&£#
II II H »l
SCREW CAP
GLASS WOOL
FILTER PLUG
OUTSIDE HOUSING
GLASS WOOL
CONTAINER
CYLINDER
PERFORATED END CAP
(TOP AND BOTTOM)
Figure C-2. PPSM filter holder,
121
-------
The U.S. EPA Method 5 Sampling Method
The EPA Method 5 tests were conducted according to accepted
practices and the instructions published in the Federal Register
(42:160, Thursday, 18, August 1977) with the following three
equipment modifications:
(1) A stainless-steel lined stack probe was used instead of
the conventional glass-lined model.
(2) Instead of the conventional direct hook up, a flexible,
heated, Teflon-lined probe extension was used to con-
nect the probe to the cyclone and filter oven.
(3) A filter holder back-up frit made of steel and fritted
glass was used along with two acid-resistant gaskets to
hold the filter in the conventional glass holder.
The stainless-steel probe was used to eliminate breakage
during shipment and use. The flexible probe connector was
necessary due to the angles of approach encountered in entering
the test port.
The following modifications to the procedure were necessary
due to the conditions of the stack gases:
(]) Due to the high loading of particulate and H2S04 mist,
the EPA Method 5 runs were conducted at a single point
for relatively short periods of time; no traversing was
performed.
(2) The filter oven was operated at a temperature above
ambient but below 38°C due to the inability of the
gaskets to withstand heating above that temperature.
Cleanup and analysis were performed according to protocol.
The EPA Method 5 runs were cleaned up according to accepted
practices and published instructions. Dry particulate materials
on the filter were desiccated and the weight was determined. The
acetone washes of the front half of the train (probe, cyclone,
flexible line, and front half filter) were evaporated, and the
mass of the residue was determined. The back half inpinger
contents were extracted with only chloroform (ether was not
available), the extract was evaporated, and the mass of the
residue was determined. The remaining water was evaporated at
elevated temperatures, and the mass of its residue was deter-
mined. Back-half acetone washes were also evaporated, and the
mass of the residue was determined.
Some of the associated preliminary and secondary testing
normally used in conjunction with the Method 5 testing was not
128
-------
necessary. Preliminary moisture was assumed from data of previ-
ous work. Initial velocity and temperature traverses were made
prior to the run's performance using the Polish pitot tube. A
grab sample was taken and used to determine the molecular weight
of the stack gas. Molecular weight analysis was performed by the
Fyrite technique for determining oxygen and carbon dioxide con-
tent. Because the molecular weight of the carbon monoxide (if it
existed in the stack gas) is identical to nitrogen, all remaining
gases were assumed to have the molecular weight of nitrogen,
except for the concentration of S02. SC>2 was present in a range
that will affect the overall molecular weight of the stack gas;
concentration values were acquired from plant personnel and the
molecular weight of the gas was accounted for. Final moisture
determination was accomplished using the water volumes collected
in the Method 5 train. Isokinetic calculations were made in the
field. All equipment was calibrated prior to shipment, and field
checks for y versus AH@ were performed upon arrival in Poland.
Because MRC's electric digital thermometer units are not accurate
using the 50-cycle European current, a battery operated, cali-
brated, digital thermometer was used instead.
DESCRIPTION OF THE SAMPLING SITES
The effluent from the Hoboken converters is directed through
ductwork to the outside of the converter building. At this
point, the gases are at a relatively high temperature and are
cooled by directing them through an atmospheric cooler consisting
of several passes of ductwork exposed to the ambient air. From
the heat exchanger/cooler, the gases are manifolded and separated
into six ducts, one for each ESP unit. These ducts are A-shaped,
extending approximately 7 m from the top of the cooler in an
upward leg, making an approximately 45° downward bend for another
7 m and then entering the top of the ESP. The ESP inlet sample
port is located in the downward duct just prior to entrance into
the ESP. A single 5-cm port was originally located at this
position on each ESP inlet; these ducts are approximately 120 cm
in diameter.
The outlets of the ESP consist of horizontal sections of
ductwork running from the side of the ESP building across the top
of the control room building, and manifolding into a larger duct.
These ducts are also 120 cm in diameter. Each duct originally
had a sampling port that was a single, 5-cm diameter pipe cou-
pling welded to the side of the duct at a distance of about four
diameters from the ESP building. Access is gained by climbing to
the top of the control building and erecting two sections of
scaffolding to reach the duct.
The ESP is arranged so that any section can be isolated and
repaired while the other sections are in operation. As many as
three ESP sections can be down at any time, and the three remain-
ing sections can handle the converter line load.
129
-------
MODIFICATIONS FOR VERIFICATION TESTS
Instructions for modifying the ducts for comparison and
verification testing were given to Polish engineers and referred
primarily to the installation of additional test ports in several
ESP inlet and outlet ducts. At the inlet, an additional 7 6-cm
pipe port was located so that the sampling nozzle of the EPA
Method 5 probe could be placed in the same plane and at approxi-
mately the same sampling point as the Polish probe. This port
was located a distance upstream from the 5-cm port equal to the
length of the sweeping bend section and short nozzle section of
the PPSM probe. On the outlet duct of the ESP, a similar port
was located just upstream from the current port and at the same
distance from it, as was done on the inlet duct. An additional
set of ports (two) were located two diameters further downstream
of the 5-cm port. These consisted of a pair of 7.6-cm pipe ports
located 90° apart and 45° from vertical on the bottom side of the
duct. Sufficient scaffolding was erected to gain access to the
new set of ports. Scaffolding was already in place at the old
port location. The following additional modifications were
required:
(1) An electrical extension cord was provided so that 220-
volt, 50-cycle power would be available at the test
locations. Originally, only 24-volt lighting power
existed there.
(2) Sufficient scaffolding and equipment platforms were
built.
(3) A means of raising and lowering the sampling equipment
was provided.
Sketches of the sampling locations are shown in Figures C-3
and C-4. An illustration of the U.S. and Polish probe alignment
is shown in Figure C-5.
130
-------
CONVERTER
NEW PORT
OLD PORT
ESP
Figure C-3. Sketch of sampling locations.
131
-------
-240 cm
• OLD PORT
GAS FLOW
NEW PORT
OUTLET
OF ESP
t
120cm
I
NEW PORT
ESP
CONTROL
ROOM BLDG.
Figure C-4. Sketch of sampling locations
132
-------
POLISH PROBE
CURRENT PORT
(2" PIPE)
U.S. PROBE
NEW PORT
(3" PIPE)
GAS FLOW
Figure C-5. Probe alignment.
133
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
EPA-600/2-80-072
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Evaluation of the Hoboken Converter at Glogow, Poland
5. REPORT DATE
April 1980 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Dr. Zbigniew Smieszek
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Institute of Nonferrous Metals
Gliwice,Poland
10. PROGRAM ELEMENT NO.
PL-480
11. CONTRACT/GRANT NO.
Contract #5-533-5
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Lab,
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
- Cinn, OH
13. TYPE OF REPORT AND PERIOD COVERED
1/76 to 2/79
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT
In 1975, the U. S. Environmental Protection Agency awarded a contract to the
Ministry of Smelting Poland for research to minimize emissions of fugitive pollutants
from copper smelters and to assist in the control of smelter pollutants. The project
objectives were to develop procedures for operating copper converters for steady gas
flow containing relatively high concentrations of 892; to improve cleaning and
treating of particulates in the converter gas streams to allow better operation of
S02 removal systems, such as contact sulfuric acid plants; and to show how procedures
and results that were developed could be applied to various types of copper smelters
encountered in industry. During the course of the project, portions of the PL-480
funding were utilized to evaluate the Hoboken Converter, a potential substitute
process providing stricter environmental control of fugitive S0£ and particulate from
copper converting.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution Control
Copper Smelting
Copper Converting
Hoboken Converter
Fugitive Emissions
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
146
20. SECURITY CLASS (Thispage)
22. PRICE
EPA Perm 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
134
ft U.S. GOVERNMENT PRINTING OFFICE: 1980-657-146/5641
-------
U.S. Environmental Protection Agency
Region V, Library
230 South Dasi-born Street
Chicago, Illinois 60604
-------
CMVII umuci i LCJI
Agency
IMIUI mmiui i
Cincinnati OH 45268
Environmental
Protection
Agency
EPA-335
Official Business
Penalty for Private Use, $300
Special Fourth-Class Rate
Book
Please make all necessary changes on the above label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports CHECK HERE D;
detach, or copy this cover, and return to the address in the
upper left-hand corner.
EPA-600/2-80-072
------- |