United States
Environmental Protection
Agency
Hazardous Waste Engineering
Research Laboratory
Cincinnati OH 45268
\VI, .
*... . * Xx
Research and Development
EPA/600/S2-85/068 Dec. 1985
&EPA Project Summary
Cost Comparisons of Selected
Technologies for the
Control of Sulfur Dioxide from
Copper Smelters
The U.S. nonferrous metals produc-
tion industry is a significant contributor
of sulfur dioxide, trace metal, and par-
ticulate air emissions. Most of the do-
mestic copper smelting capacity is
based on obsolescent technology that
is both capital- and energy-intensive
and hampered by considerable emis-
sion control problems. Because sys-
tems used to control sulfur dioxide
emissions also must remove particu-
late contaminants, effective control of
total particulate and trace element
emissions is accomplished as a
"byproduct" of sulfur dioxide control.
When it can be operated under au-
tothermal conditions, the surfuric acid
plant is recognized as the technology of
choice for controlling sulfur dioxide
emissions from smelters.
Unfortunately, much of the problem
of sulfur dioxide control in the nonfer-
rous industry is associated with the
weak sulfur dioxide off-gas streams.
Because weak sulfur dioxide streams
do not permit autothermal acid plant
operation, they cannot be economically
controlled by acid plant technology. In
copper smelting, the major sources of
weak sulfur dioxide off-gases are the
reverberatory furnace and muftihearth
roaster, followed by fugitive emissions
that emanate from the converter opera-
tion, matte tapping, slag tapping, and
ladle transfer. The status of develop-
ment and use of a number of technolo-
gies for wet scrubbing and for process
changes based on oxygen smelting
technologies were evaluated to deter-
mine which could be considered
promising for near-term application to
sulfur dioxide control at domestic
smelters. Cost models were developed
for those processes believed to be suffi-
ciently promising for such application.
This Project Summary was devel-
oped by EPA's Hazardous Waste Engi-
neering Research Laboratory, Cincin-
nati, OH, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).
Introduction
Primary nonferrous smelters ac-
counted for about 6 percent of the total
sulfur dioxide emitted in the United
States in 1980. These considered of 15
copper smelters (largely in the western
section of the country), 5 lead smelters
(not including Bunker Hill, which is now
closed), and 7 zinc smelters. Several of
the copper smelters currently use pro-
duction curtailment programs to limit
their emissions to comply with ambient
air sulfur dioxide standards.
The total allowable su If ur input for the
13 western primary copper smelters
(based on State Implementation Plans
and allowable sulfur dioxide emissions)
is 6572 Mg (7244 tons) per day. The sul-
fur dioxide emissions at this sulfur input
would total about 375 Mg (414 tons) per
hour, which is equivalent to 87.5
1000-MW power plants emitting at a
level of 516 ng/J (1.2 Ib of sulfur dioxide
per million Btu). These relatively high
emission levels may contribute to prob-
lems of sulfate-related visibility reduc-
tion, high concentrations of sulfur diox-
ide in the ambient air, and acid rain
generation in the West.
A typical conventional reverberatory
smelting furnace can vary in size from
30.5 to 40.2 meters long by 7.8 to 12.2
meters wide (approximately 100 to 132
feet by 25.5 to 40 feet). Typical capacity
is about 944 Mg (1040 tons) of concen-
trate per day. The reverberatory furnace
-------
can receive its copper ore concentrate
either as a "green" or wet feed or as a
calcine. The wet-feed process is one of
the most endothermic processes for
production of copper. The use of cal-
cined feed (as opposed to wet feed) re-
duces the energy consumption, gas
flow, and amount of sulfur dioxide gen-
erated in the reverberatory furnace, be-
cause the moisture and a portion of the
sulfur have been removed in the roast-
ing step.
Currently, continuous controls are not
applied to reverberatory furnaces for re-
duction of sulfur dioxide emissions. All
of the sulfur removed from the concen-
trate in the smelting furnace is dis-
charged to the atmosphere as sulfur
dioxide. As much as 38 percent of the
sulfur contained in the concentrate may
be converted to sulfur dioxide in the
conventional green charge reverbera-
tory furnace and emitted in an off-gas
stream containing approximately 1 to
1.5 percent sulfur dioxide by volume.
Fluid-bed roasting of the concentrate
before smelting in the reverberatory
smelting furnace can reduce the sulfur
emitted to about 10 to 15 percent of the
total concentrate sulfur.
The absence of continuous emission
control results from a combination of
factors, the most important of which are
1) the cost of installing and operating
emission control systems for reverbera-
tory furnace sulfur dioxide emissions,
2) the lack of profitable acid byproduct
markets, 3) the cyclical nature of copper
market prices, and 4) strong interna-
tional competition.
The contact sulfuric acid process is
the most widely used for removing sul-
fur dioxide from primary copper smelt-
er effluent gases and is considered to be
a well-demonstrated technology of rea-
sonable cost for strong sulfur dioxide
streams. Currently, 12 of the 15 active
domestic primary copper smelters pro-
duce sulfuric acid from strong sulfur
dioxide off-gases. Acid plants are not
considered economic for the control of
sulfur dioxide emissions from reverber-
atory furnaces, however, because of the
low (weak) sulfur dioxide content of the
off-gas.
Although sulfuric acid plants can be
designed to process feed streams that
contain a low percentage of sulfur diox-
ide, economic considerations usually
restrict application to gas streams that
have a sulfur dioxide content of at least
3.5 percent so that reaction tempera-
tures can be maintained by the heat re-
leased in the sulfur dioxide oxidation re-
action. Metallurgical single-stage and
dual-stage absorption sulfuric acid
plants constructed in the past were
commonly designed to operate au-
tothermally on feed streams containing
4.0 and 4.5 percent sulfur dioxide, re-
spectively. Nonautothermal operation
requires supplemental heating by natu-
ral gas and is therefore more costly.
For these reasons, primary copper
smelters using reverberatory furnaces
have not attained the emission limita-
tions specified by the State Implemen-
tation Plans as required by the Clean Air
Act of 1967. Recognizing the difficulties
of the smelting industry. Congress in-
cluded a provision in the Clean Air Act
Amendments of 1977 for the issuance of
a "primary nonferrous smelter order"
(N.S.O.) to any existing smelter apply-
ing and qualifying for such an order.
The N.S.O., which constitutes a delayed
compliance order, is unique to primary
nonferrous smelters and is provided "in
recognition of the economic circum-
stances of the industry." The law per-
mits any smelter that operates pro-
cesses from which sulfur dioxide
emissions cannot be controlled to meet
the applicable State Implementation
Plan requirements by the application of
a continuous emission control technol-
ogy to apply for and operate under the
N.S.O. until such technology is
"adequately demonstrated to be rea-
sonably available (as determined by the
Administrator, taking into account the
cost of compliance, nonair quality
health and environmental impact, and
energy consideration)." This program
places the responsibility for developing
the necessary technology upon the
smelting industry and requires compli-
ance by January 1988.
Control Approaches Studied
This study was undertaken to develop
the tools needed by the U.S. Environ-
mental Protection Agency to conduct an
independent evaluation of the "reason-
able availability" of technologies con-
sidered to have reached, or to be
approaching, an "adequately demon-
strated" status. The two primary ap-
proaches available to smelters for the
control of sulfur dioxide emissions are
1) process modifications that produce a
strong higher-concentration sulfur diox-
ide off-gas that can be controlled by
conventional acid plant technology, or
2) removal of the sulfur dioxide from
weak off-gas by wet scrubbing.
The INCO flash furnace represents a
process replacement of the reverbera-
tory furnace with an alternative smelt-1
ing technology that produces a strong
off-gas. The oxygen-fuel and oxygen-
sprinkle processes represent modifica-
tions to reverberatory furnaces that use
oxygen enhancement to produce a
stronger off-gas.
Scrubbing processes include the
lime/limestone process, which is a non-
regenerative flue gas scrubbing system
that disposes of the sulfur fixed in an
absorbent, and the magnesium oxide
and citrate processes, which involve re-
generative flue gas scrubbing that con-
verts the sulfur dioxide to a strong
stream in the regeneration process for
subsequent recovery.
Several other processes, such as the
Noranda, Mitusbishi, and Outokumpu
Oy processes, also warrant considera-
tion as possible ways of achieving re-
duced sulfur dioxide emissions and
profitable, energy-efficient operation.
Process Modifications
Modifying existing furnace opera-
tions to increase the sulfur dioxide con-
tent of the off-gas can be an effective
approach to emission control when
coupled with sulfur dioxide recovery or
a gas scrubbing system (e.g., when
there is no market for acid, scrubbing
equipment costs would be less because
of the lower-volume gas stream). Alter-
native pyrometallurgical processes are
of interest for the following reasons:
they produce a strong sulfur dioxide gas
stream that can be sent to a sulfuric acid
plant, they reduce energy consumption,
they decrease gas-stream volumes, and
they subsequently reduce operating
costs.
The full report presents descriptions
and costs of process modifications that
would enable existing copper smelters
to increase sulfur dioxide concentration
and make recovery more practicable.
Process modification principally en-
tails the introduction of oxygen into the
furnace so that oxygen-enriched air or
pure oxygen is used in the smelting pro-
cess. Descriptions of the process modi-
fications and the estimated capital and
annual operating costs for installation
and operation are presented for plant
models based on the INCO oxygen flash
furnace smelting process, the oxygen-
sprinkle process, and the oxygen-fuel
smelting process. The Outokumpu Oy,
Noranda, and Mitsubishi continuous
processes are also described, but in less
detail. Whenever possible, existing
equipment was assumed to be modified
to minimize costs; however, because
-------
the modifications are affected by
plant-specific considerations such as
equipment arrangement and layout,
costs could vary widely. In each case,
the acid or oxygen plants were assumed
to be new units. Existing converters
were utilized, and as the copper per-
centage of the matte increased, a reduc-
tion in blowing costs and labor was as-
sumed. A "greenfield" installation
would be more costly than a retrofitted
installation, as it would entail additional
costs for ore handling, blending, and
preparation facilities, as well as convert-
ers, buildings, utilities, rail facilities,
working capital, and other items. A
computer-aided cost estimation model
for process modifications developed for
this project is described.
Intrinsic benefits of this modification
vary with the extent of the replacement
of air with oxygen. The use of oxygen
can significantly reduce the energy re-
quired to produce blister copper. For ex-
ample, a smelter based on the INCO
flash furance, which utilizes pure oxy-
gen, requires only about 70 percent of
the energy input of a conventional
green charge reverberatory smelter.
The use of oxygen in smelting also re-
duces and/or replaces the fossil-fuel re-
quirements. This, in turn, increases the
sulfur dioxide concentration in the
off-gas and effects a corresponding de-
crease in off-gas flow, which reduces
the energy and equipment size required
to handle this gas. A 30-fold reduction in
gas cleaning equipment size (volumet-
ric flow rate) can be realized with pure
oxygen smelting; thus, the increased
sulfur dioxide concentration lowers the
capital cost of control equipment and
power requirements of sulfur recovery.
For example, a sulfuric acid plant that
uses a 4 percent sulfur dioxide stream
costs about 1.8 times as much as one
handling a 12 percent stream for the
same daily acid production. Power re-
quired per ton of sulfuric acid by a plant
(either single- or double-contact) using
the 4 percent stream would be 91.7 per-
cent greater than that required by a
plant using 12 percent stream.
Oxygen smelting also increases
matte grade, and higher matte grade re-
quires less time for conversion to blister
copper. For example, if the time for con-
verting matte to blister copper is "t" for
a 35 percent copper matte, the time re-
quired for a 42 percent copper matte be-
comes 0.8t and that for a 50 percent
copper matte becomes 0.65t. Obvi-
ously, fewer converters are thus re-
quired to produce the same tonnage. In
addition, a savings would be realized in
blowing time and labor, and the smelter
capacity would be increased by such a
modification.
Flue Gas Scrubbing
For weak gas streams (i.e., those con-
taining less than about 3.5 percent sul-
fur dioxide), flue gas scrubbing is re-
quired to control sulfur dioxide emis-
sions. In such processes, the smelting
furnace off-gases, after paniculate re-
moval and temperature conditioning,
are contacted with an absorbent me-
dium in an absorber tower or a scrub-
ber.
The gas scrubbing processes are cat-
egorized as regenerate or nonregener-
able. In the regenerable process, the ab-
sorbent is regenerated for reuse by
stripping (separating) the sulfur diox-
ide. The absorbent is then recycled to
the absorbing stage. Sulfur dioxide,
stripped from the liquid, is recovered by
further processing to liquid sulfur diox-
ide or sulfuric acid. In the nonregenera-
ble process, the spent absorbent is
converted to a gypsum byproduct or
discarded as waste.
Four gas scrubbing systems were
evaluated in this study. The lime/lime-
stone (nonregenerable) and the magne-
sium oxide (regenerable) scrubbing
systems have been used on full-scale
smelters in Japan. Pilot plants of the ci-
trate process (a regenerable system)
have been operated on smelters in
Sweden (by Flakt at Boliden) and in the
United States (by the Bureau of Mines).
The study evaluation of scrubbing
systems is based on a gas stream that
has passed through a waste-heat boiler
and an electrostatic precipitator for par-
ticulate removal. Most smelters include
both of these operations. A generalized
scrubber system flow diagram shows
the gas pretreatment options and the
gas/liquid system. If the temperature of
the gas stream is below 205°C (400°F),
the stream can go directly to the ab-
sorber. If further particulate removal is
required, the stream can be "treated" in
a venturi scrubber, after which the
cooled gas goes to the absorber.
The estimated capital and annual op-
erating costs for each scrubbing system
addressed in this study were prepared
by utilizing a standardized format.1 The
costs reflect a "study estimate" with a
reliability of about ±30 percent. A
HJhl, V. W., A Standard Procedure for Cost Analysis
of Pollution Control Operations. U.S. Environmen-
tal Protection Agency. EPA-600/8-79-018a. June
1979.
computer-aided cost-estimating tech-
nique used to generate these costs is
also described.
Costs for gas scrubbing systems vary
widely depending on the gas pretreat-
ment required and the sludge treatment
and disposal alternatives selected. An-
cillary items such as bypass ducting, a
new stack, sludge pond, etc., will also
affect the capital cost. The computer-
aided cost-estimating system allows the
user to select input parameters for a
specific case.
Conclusions
A comparison of the incremental
costs for control of sulfur dioxide emis-
sions by process change with those for
the addition of flue gas scrubbing
shows the kind of decisions the industry
faces. For a plant producing 90,720 Mg
of copper per year, the approaches
based on process change (ranging from
$67 to 87 million) and regenerable flue
gas scrubbing (at about $70 million) are
the most capital-intensive, whereas the
approaches based on lime or limestone
scrubbing ($36 to 52 million) are much
less capital-intensive. The capital costs
presented for regenerable flue gas
scrubbing are based on the citrate sys-
tem. Although the sulfur dioxide ab-
sorption and stripping components of
this sytem have been well demon-
strated at pilot scale on smelter off-gas,
the developer has not yet demonstrated
that a full-scale citrate system can be
operated economically. Also, the case
presented for citrate scrubbing with the
stripped sulfur dioxide fixed as liquid
sulfur dioxide, which offers very attrac-
tive capital and annualized costs, is
based on the marketability of the liquid
sulfur dioxide produced. Such a market
is unlikely for smelters located in the
southwestern United States. On the
other hand, a greenfield smelter located
near a paper producing area probably
could tap a sulfur dioxide market of suf-
ficient size.
In contrast, the estimated annualized
costs for process change (which range
from 28 to 36 cents per kilogram of cop-
per) appear to make this approach more
attractive than these approaches based
on add-on flue gas scrubbing (which en-
tail costs of 22 to 29 cents for the lime or
limestone processes, and 14 to 17 cents
for the citrate process). The advantage
of the former is that the annual operat-
ing costs for a process change include
the costs of operating the smelting fur-
nace and production improvements,
whereas those for an add-on flue gas
-------
scrubbing system represent an incre-
mental cost in addition to those for op-
erating the smelting furnace.
The decision then becomes one of
trading off capital costs for operating
costs. The trade-off analysis is difficult
because of the variable price of copper,
the marketability and price of the sulfur
byproduct, the life of the smelter, and
the cost of capital. It would be more eco-
nomical to use a throwaway flue gas
scrubbing system in cases where the
cost of capital was high and the life of
the smelter was limited by its present
condition, market viability, or the secu-
rity of the concentrate supply. Where
the life of the smelter is judged to be
longer, continuous control of sulfur
dioxide would be achieved most eco-
nomically through process change.
The Project Report was authored by personnel of PEI Associates, Inc.. Cincinnati,
OH 45246; the author of this Project Summary John O. Burckle (also the EPA
Project Officer, see below) is with the Water Engineering Research Laboratory,
Cincinnati, OH 45268.
The complete report, entitled "Cost Comparisons of Selected Technologies for the
Control of Sulfur Dioxide from Copper Smelters," (Order No. PB 85-215
705/AS; Cost: $23.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, MA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S2-85/068
U $ ^NVIR PROTECTION AGENCY
H&Toi»»858lWT
CHIC&GO It 60604
------- |