TECHNICAL DATA IN SUPPORT OF REGULATIONS
FOR CONTROL OF SULFUR DIOXIDE EMISSIONS AT
ARIZONA COPPER SMELTERS
October 1975
I. Background
On July 27, 1972 EPA disapproved Arizona's control strategy for the
attainment of the National Ambient Air Quality Standards (NAAQS) for
sulfur oxides and granted an 18 month extension for the promulgation of
regulations to include measures that would insure attainment of these
standards.
The available ambient air quality data prior to May 1973 measured
in the vicinity of the major sources of sulfur oxides within the state,
the seven copper smelters, had been challenged in court and in public
hearings. EPA therefore established a network of ground level monitoring
sites for the collection of sulfur oxides data to be used in evaluating
control strategies and determining the amount of control required of
each smelter in order to achieve the standards.
The emission limitations in paragraph (e) of the proposed regulations
are based on the use of the proportional model that assumes a direct
relationship between source emissions and ambient air quality. Source
emission data have been supplied by each of the smelters and the ambient
air quality data used in the model was obtained from the EPA monitoring
sites and from data collected by the smelter's monitoring networks and
from data collected by the State of Arizona.
Emission limitations of paragraph (f) are based on an evaluation of
the configuration of each smelter and on sulfur balance and operating
data obtained directly from the smelter. Appendix A contains the results
of an evaluation of the possible control techniques available to each
smelter. Section V of Appendix A contains an evaluation of the amount
of positive control possible through reasonably available retrofit
control technology.
Under the provisions of applicable State regulations, each smelter
has embarked on a control program to reduce the effect of their emissions
on the ambient levels of sulfur oxides. These control programs include
a wide range of permanent control measures and dispersion techniques.
The amounts of positive control anticipated through these control programs
range from approximately 42% reduction in emissions with production
curtailment to well over 90% reduction in emissions with process
-------
changes and sulfuric acid-producing facilities. The control programs
are in varying stages of completion. The emission limits established in
this document are based on the smelter configuration at the time the
State implementation plan was submitted to EPA (January 31, 1972) and
are based on the maximum production capacity of the smelter and current
sulfur balance information obtained from each smelter.
The regulations contain 3 paragraphs relating specifically to
control measures. Paragraph (d) pertains to the control of fugitive
emissions.
Paragraph (e) establishes an emission limit that will allow attain-
ment of the NAAQS. EPA has determined that in most cases this emission
limit is hot achievable through the application of currently reasonably
available retrofit control technology but could be achieved through weak
gas stream treatment process changes or permanent production curtailment.
EPA is not now requiring process changes or permanent production curtailment
but anticipates that as retrofit control technology improves this emission
limit can be met.
Paragraph (f) provides for interim measures until such time as the
emission rate of paragraph (e) can be met. Paragraph (f) establishes an
emission limitation that can currently be met with reasonably available
retrofit control devices (acid-producing facilities) and allows the use
of dispersion techniques and other measures to avoid ground level
concentrations in excess of the standards on an interim basis.
The methods used to determine the amount of control required and
the emission limitations are included in this document.
II. Air Quality Data
EPA established monitoring stations in the vicinity of each of the
copper smelters in Arizona in the late summer or fall of 1973. The
purpose of these stations was to collect short term ambient air quality
data to be used in the determination of required emission limitations.
Both continuous monitoring systems and 24-hour bubbler sampling systems
were installed. The location of the stations was selected to either
approximate the location of potentially high concentration areas or to
supplement the existing smelter and/or State networks. In many instances
the availability of power or property access precluded locating in the
desired area.
Several of the smelters also operate monitoring networks. Some of
the networks have been in operation for several years (Douglas) while
others are more recent (Hayden). All of the smelter companies have
provided past data from their monitoring sites to EPA and several
continue to submit data on a routine basis.
-2-
-------
The State of Arizona also operates at least one monitoring site
near each smelter.
Data from the EPA network and, where possible, from the smelter
networks, has been summarized in Table 1. The highest two 3-hour and
highest two 24-hour concentrations are included. The period of record
available to EPA of individual smelter network data varied from only 4
months in the Hayden area to 21 months of data from several Phelps-Dodge
stations*. The EPA stations were operational in the fall of 1973 and
continued through October 1974. The continuous data is available through
most of October 1974 and 24-hour bubbler data were obtained through
October 1974 and in some cases a few November samples were taken.
Emission Reduction Requirements .
The amount of control, or reduction in emissions required to meet
the ambient air quality standards, is determined using the following
relationship between ambient air quality levels and emissions:
100 A-B = percent reduction required
A-C
where A is measured air quality level
B is the ambient air quality standard value
C is the background level
(considered to be zero in smelter locales)
and
(current emissions)(100 - percent reduction required) =
100
maximum allowable emissions
The percent reduction in emissions required to meet all NAAQS is
calculated for each smelter area and is also given in Table 1. This
method assumes a direct relationship between the total plant emissions
and the ambient S02 concentrations in the vicinity of the smelter. Such
an assumption is not unreasonable where the smelter is the single source
of S02 in the area. In the case of the Hayden area, where there are two
smelters, some additional assumptions are required and are given below.
The application of this formula to data from each area is given in
subsequent paragraphs.
* Data is continuously supplied by the smelters, including KCC and
ASARCO at Hayden, and the listings are available but not summarized
or analyzed.
-3-
-------
TABLE 1
Date
Concentration
(ppm 3-hour average)
(um/tn^ 24-hour average)
Reduction required to*
averaging meet standards
time (percent)
(hours)
Phelps Dodge/Ajo
11 Nov 1973 1.65
2 Nov 1973 1.35
1 Nov 1973 784
8 Jan 1974 662
Phelps Dodge/Douglas
17 Jan 1974 3.81
8 Dec 1973 3.37
20 May 1974 1395
8 Dec 1973 1031
Phelps Dodge/Morenci
10 Nov 1973 4.10
11 Feb 1974 3.51
21 Dec 1973 2335
14 Jan 1974 1502
Kennecott and ASARCO/Hayden
20 Oct 1973 3.17
13 Oct 1973 2.67
4 Jun 1974 2.33
8 Nov 1973 2319
Magma/San Manuel
12 Oct 1973 2.32
20 Oct 1974 1.98
14 Nov 1974 .39
12 Oct 1973 .36
Inspiration
10 Sep 1974 4.43
12 Mar 1974 3.67
5 Nov 1974 2193
2 Nov 1974 1617
3
3
24
24
3
3
24
24
3
3
24
24
3
3
3
24
3
3
24
24
3
3
24
24
69.7
63
53.4
44.9
86.9
85.2
73.8
64.6
87.8
85.8
84.4
75.7
84
82
79
84
78.4
74.9
64.1
61.1
88.7
86.4
83.4
77.4
* Calculated from proportional rollback
-4-
-------
The rollback or proportional model is not adequate for the purpose
of determining the effect of certain dispersion techniques such as
taller stacks or intermittent controls. It cannot take into account the
fact that emissions may come from different heights and locations within
the smelter, thus it cannot be used, for example, to evaluate the effect
of changing the emission points of fugitive emissions from building
emissions to venting them through a stack. Use of a diffusion model
would be preferable to use of the rollback model in relating emissions
to air quality, however most of the smelters are located in areas with
elevated terrain features and current diffusion models do not, with
precision, handle the effects of various topographic elements. Also
meteorological variables in areas of rough terrain are not easily quantified.
The meteorological input data for diffusion modeling must be extrapolated
from existing National Weather Service stations located at considerable
distances from the smelters, since complete data are not available
locally. For these reasons it is not surprising that modeling results
vary considerably with different modeling techniques and do not generally
compare favorably with actual measured short term data. The rollback
model was therefore accepted as containing fewer assumptions and as
providing reasonable results for total source control.
Operating data were supplied to EPA by each smelter. These data
contained average production figures and sulfur balance data as well as
specifics for particular days of interest. In some cases actual emission
data were supplied by the smelter, while in other cases emissions were
calculated from the sulfur balance figures supplied.
In the subsequent paragraphs the calculations of the emissions that
will achieve the NAAQS are discussed. All calculations are in terms of
tons of sulfur and have been calculated for an averaging time of one
day. The regulations refer to emissions in terms of pounds per hour
averaged over a six hour period. This is required for this compliance
testing program where 24-hour tests are impractical. Had hourly operating
or emission data been available from the smelters, the calculations
would have been based on the shorter averaging period. The variable
nature of the smelting process is recognized and hourly data would have
been preferable. However, use of the emission data for the same day as
the air quality data provides as accurate a relationship as is available
from past records.
-5-
-------
A. Hayden, Arizona
There are two smelters in relatively close proximity at Hayden,
Arizona, one operated by ASARCO, Incorporated and the other by the
Kennecott Copper Corporation. In this discussion the first will be
identified as ASARCO, the second as KCC.
EPA operated two continuous and four 24-hour bubbler S02 monitoring
sites in the vicinity of the two smelters from mid-1973 until October
1974 when the bubbler sites were discontinued and the continuous monitors
turned over to the State of Arizona.
The two smelters operate a Joint Control Center for meteorological
and S02 monitoring and maintain 7 sampling sites in the Hayden vicinity.
High S02 concentrations have been recorded at all sites on occasion
but the two sites where high concentrations most frequently occur are
the Globe Highway Joint Control site and the monitoring site at Reese-
Montgomery Ranch to the WWW of the smelter complex. The highest concen-
trations on record in the EPA office have occurred at the Reese-
Montgomery Ranch site.
The highest 24-hour average concentration occurred on November 8,
1973 when the 24-hour bubbler sampler indicated a concentration of
2319 ug/iP. Based on the proportional rollback formula, this concen-
tration, when compared to the primary NAAQS for S02, indicates a reduc-
tion of 84.3% is required in S02 emissions in order for the standard to
be met in the Hayden area. On the 8th of November 1973 the emissions
from the two smelters were as follows:
KCC 32.85 T/D sulfur
ASARCO 461 T/D sulfur
These numbers are based on actual daily estimates of emissions in the
case of ASARCO and on average November sulfur balance data and actual
daily production rates for KCC. They represent the best available
information. Applying the 84.3% reduction requirement to the total
emissions for that day yields a value of 77.8 T/D allowable sulfur
emissions ((1 - .843)494 =77.8 T/D S).
Each smelter has a different production capacity, has different S02
control systems, emits different quantities of S02 and is obviously
located differently with respect to the sampling site. Also emissions
come from different locations within the smelter complex and from
different heights. These facts tend to make the use of the rollback
formula and the assumption of a direct proportionality of emissions to
ambient concentrations somewhat doubtful. Several attempts were made to
find a suitable diffusion model for use in this situation where both the
configuration of the smelters and their emissions and the meteorology
-6-
-------
could be used to more accurately predict the ambient concentrations and
the effect of various changes. In addition, a five day intensive field
investigation was conducted in April 1975 to attempt to describe the
input parameters to the model. The results of these efforts indicate
that the currently available diffusion models do not accurately describe
the behavior of the emissions from the two smelters in Hayden. The
field investigation led to the conclusion however that the Globe Highway
station is occasionally subject to high concentrations due in large part
to fugitive or low level emissions and that while there may be fugitive
emissions traveling as far as the Reese-Montgomery Ranch site, the
concentrations there are probably primarily due to emissions from the
stacks. Diffusion modeling produced concentrations that were not realistic
based on actual measured data from the available sites. In general the
modeling results were much lower than actual measurements unless adjustments
were made for topography and with those adjustments, the results appeared
to be unrealistically high. As a result, the proportional model or
direct relationship of emissions and ambient concentrations appears to
be the most reasonable representation. The rollback model required some
adjustment to address the contribution from each smelter. This was done
by apportioning the reduction requirements according to the maximum
production capacity of each smelter. The amount of input sulfur at
maximum concentrate feed rate for Kennecott is 1500 T/D at 30% S equals
450 T/D input S. For ASARCO the maximum concentrate feed rate is 2000
T/D at 29.4% S equals 588 T/D input S. Thus KCC would contribute about
43% of the uncontrolled sulfur emissions and ASARCO 57%. These same
figures applied to the allowable emissions results in an emission rate
of 34 T/D S limit applied to KCC and 44 T/D S limit for ASARCO in order
to meet the primary standard.
Annual averages determined from the four bubbler sites indicate
that less control is required to meet the annual average NAAQS than the
24-hour standard.
The short term 3-hour standard is 0.50 ppm or 1300 ug/m^. Data
from Reese-Montgomery Ranch site supplied the highest 3-hour averages.
A concentration of 3.17 ppm was recorded on October 20, 1973. Using the
rollback formula a reduction of 84.2% is required to meet the 3-hour
standard. KCC emissions on that day were estimated from average October
sulfur balance data for KCC and the production rate on that day to be
97.6 T/D S. ASARCO emission were supplied by the smelter and were 223
T/D S. Without controls the emissions would have been 307 T/D S from
KCC and 321 T/D S from ASARCO. Applying the 84.2 % reduction requirement
to the actual emissions results in an emission rate of 50.6 T/D S needed
to meet the secondary standard. These emissions were again apportioned
on the basis of the relative smelter capacity resulting in an allowable
emission rate of 22 T/D S applied to KCC and a rate of 29 T/D S applied
to ASARCO.
-7-
-------
B. San Manuel, Arizona
The Magma Copper Company operates a smelter at San Manuel. They
also operate an extensive monitoring network in the vicinity with
locations as far away as Oracle and Reddington. Ambient air quality
data were made available to EPA by the Magma Copper Company. Data were
collected at monitoring sites in the vicinity of the smelter and the
period of record used in this analysis was generally from February 1973
through November 1974. Additional data are available but have not been
analyzed.
EPA collected data at four monitoring sites in the vicinity of this
smelter from September 1973 until October 1974. Those data have been
used in conjunction with the smelter's data to determine the S02 concen-
tration distribution.
The emission limitation established to meet both the primary and
secondary standards was determined using only measured S02 data and
applying the rollback formula to determine the emission reduction
requirements. Diffusion modeling results prepared by EPA, and also a
modeling study completed for Magma Copper Company by Stanford Research
Institute, were reviewed and an attempt was made to reconcile the
modeling results with the measured data. Model predictions were quite
good for 24 hour and longer averaging periods but did not compare well
with short term concentrations.
High concentrations have been recorded at all stations within five
miles of the smelter. The highest values were reported at the smelter's
Townsite and Golf Course locations and the EPA Slag Pile site. On the
14th of November 1974 a concentration of 0.39 ppm 24-hour average was
reported at the smelter's Golf Course station. This is the highest 24-
hour average for the period of EPA's record.
Information was supplied by the Magma Copper Company for the actual
production and emission rates on specific dates. These data were used
to determine the emissions that could be allowed and still meet the
NAAQS. On the 14th of November 1974 the total sulfur emitted to the
atmosphere was 394 tons. Applying the rollback formula to the 24-hour
measure concentration on that day results in a required of reduction of
64.1%. Reducing the emission rate on that day by 64.1% results in an
emission rate of 141 T/D required to meet the primary standard.
The highest measured 3-hour concentration was 2.32 ppm measured at
the Townsite location on October 12, 1973 and the next highest 1.98 ppm
occurred at the EPA North Slag site on October 20, 1974. Calculations
using rollback and applied to actual emissions on those days resulted in
-8-
-------
the use of the October North Slag data to specify the allowable emission
rate. On October 20th the total sulfur emitted was 306 tons. The
proportional reduction requirement determined by comparing the 1.98 ppm
to the .5 ppm standard is 74.7%. Reducing the 306 tons of emissions by
74.7% results in an allowable emission rate of 77.3 T/D (306(1 - 0.747) = 77.3),
Thus, the 77.3 T/D emission rate is the one that must be met to achieve
all NAAQS.
-9-
-------
C. Douglas
The Phelps Dodge Corporation operates a smelter at Douglas, Arizona.
They also operate an extensive monitoring network in the vicinity and
have provided EPA all of the data from their network. EPA operated 3
stations in the vicinity from fall 1973 until October 1974. Data from
both networks have been reviewed and form the basis for the emission
calculations.
The smelter at Douglas is located in open relatively flat terrain.
A number of modeling studies have been done for the Douglas area.
Contrary to modeling studies for other smelters located in areas with
elevated topographic features, modeling results here tend to under-
estimate the actual measured concentrations indicating that the model
does not adequately account for the specific meteorological conditions
that produce high ground level concentrations of S02- It is also quite
possible that short term emission data are not sufficiently accurate to
allow realistic short term model predictions.
The highest reported short term concentrations in the Douglas area
have been measured at the smelter's Pirtleville site. The maximum 3-
hour concentration was 3.81 ppm measured on January 17, 1974. Using the
rollback formula a reduction of 86.9% is required when this concentration
is compared to the 3-hour NAAQS. Phelps Dodge Corportaion supplied to
EPA data concerning production and emission rates on specific dates. On
January 17, 1974 the total sulfur emissions were listed as 597 tons.
Applying the reduction requirement of 86.9% to this 597 T of S results
in an allowable emission rate of 78.2 T/D of S to achieve the secondary
NAAQS for S02 (597(1 - .869) = 78.2).
Similar calculations were made to determine the requirements needed
to meet the primary standards. The highest 24-hour average concentration
was also measured at the smelter's Pirtleville station on December 9,
1973. This concentration was 2437 ug/m^. Rollback indicates an 85%
reduction requirement. Applied to the emission rate of 669 tons of
sulfur on that day, the allowable emission rate is 100 T/D of sulfur.
Less reduction is required to meet the annual standard.
-10-
-------
D. Morenci
The Phelps Dodge Corporation operates a smelter at Morenci.
Ambient monitoring data are available from nine smelter monitoring sites
and for. a one year period from EPA monitoring sites. These data were
used to develop the emission limitations.
This smelter is located in an extremely rugged terrain area.
Monitors that are located less then two miles from the smelter may
differ in elevation by nearly 1500 feet. The rugged terrain features
and the possible effects of mountains and valleys on the meteorological
model inputs resulted in a decision to use actual measured data.
The highest measured concentrations occurred at the smelter's
monitoring location called Cadillac Point, and the highest reported
3-hour average of 4.10 ppm was measured on November 10, 1973. Comparing
this concentration to the 3-hour secondary standard using the rollback
formula results in a required emission reduction of 87.8%. Phelps Dodge
Corporation supplied operating data for the smelter on specific dates
from these data it was determined that the emission rate on November 10th
was 725 tons of sulfur. Reducing this emission rate by the required
87.8% results in an allowable emission rate of 88.45 T/D of sulfur to
meet the secondary NAAQS (725(1 - .878) = 88.45).
The highest 24-hour average concentration also occurred at Cadillac
Point. The concentration was 1.29 ppm and was recorded December 22,
1973. Comparing this value with the primary standard yields a reduction
requirement of 90.7%. An emission rate of 996 tons of sulfur was calculated
from data supplied by the smelter for this data. A 90.7% reduction in
emissions leaves an allowable of emission rate of 92.6 tons of sulfur to
meet primary standard.
Less reduction is required in order that the annual standard be
attained.
-11-
-------
E. Ajo
The Phelps Dodge Corporation operates a smelter in Ajo. EPA established
a monitoring network in 1973 composed of four 24-hour bubbler sites and
two continuous SC>2 monitors. These sites were discontinued or turned
over to the State in October 1974. The smelter has operated 4 stations
with continuous monitors since mid-summer 1973. Data from all sites
have been made available to EPA. These data were used to determine the
allowable emission rates.
Diffusion modeling was attempted to provide a more accurate repre-
sentation of the relationship between air quality levels and emissions
than can be obtained through the use of the proportional or rollback
model. Results were compared with actual measured data. Model predictions
were generally lower than measured data except where topographic considera-
tions were entered in the model calculations and those results appeared
to be higher than might be anticipated. There did not appear to be a
simple calibration or adjustment technique to apply to the model to
compensate for the difference. For that reason, actual measured data
was used.
The highest measured 3-hour concentration of 1.65 ppm was reported
on November 11, 1973 at the South Tailings Pond Phelps Dodge monitoring
site. The second highest concentration of 1.35 ppm was measured at the
EPA Oxidation Pond site on the 2nd of November 1973.
Operating data for the smelter on specific dates was supplied by
the Phelps Dodge Corporation and this data indicated emissions to the
atmosphere of 132.3 tons of sulfur on the llth of November and 49.2 tons
of sulfur on the 2nd.
Using the rollback model a reduction in emissions of 69.7% would be
required to meet the secondary standard on the llth and 63% reduction
required for the 2nd. Applying these values to the actual emission
rates on those days resulted in allowable emission rates as follows:
11 Nov 73: (1 - .697)132.3 = 40.1 T/D S
2 Nov 73: (1 - .63)49.2 = 18.2 T/D S
The lower emission rate would of course apply and becomes the maximum
emission rate allowable.
-12-
-------
Reduction requirements to meet the 24-hour standard are based on
the concentration of 784 ug/m-' measured on the 1st of November 1973 at
the EPA Oxidation Pond station. Using the rollback formula a reduction
of 53.4% is required. Emissions on that day were calculated from the
actual concentrate feed processed and average sulfur balance data, and
totalled 65.96 tons of sulfur emitted to the atmosphere. The allowable
emissions to meet the primary standard are then equal to 30.7 T/D sulfur
((1 - 0.534)65.96 = 30.7).
Annual average concentrations resulted in lower emission reduction
calculations.
-13-
-------
F. Miami
The Inspiration Copper Company operates a smelter at Miami. Air
quality data in the area are limited to the data from the EPA smelter
monitoring sites and from the State monitoring sites. Only the EPA
monitoring data were used in this analysis. EPA operated one continuous
monitor and 3 bubbler samplers in the vicinity of Miami from approximately
October 1973 through October 1974. The sampling site at Jones Ranch was
considered to be in an area where stack emissions would result in maximum
concentrations and the measured data from that site are used to establish
emission limits.
This smelter is located in an area of extremely complex topography.
Diffusion modeling results did not compare favorably with actual measured
data due in part to the lack of appropriate emission and meteorological
input data.
The highest measured 3-hour concentration was reported on September 10,
1974. This was 4.43 ppm. The second highest value, 3.67 ppm, occurred
on March 12, 1974.
Production data were made available by the smelter for specific dates
and average monthly input sulfur data were used to calculate emission
requirements.
Based on the high 3-hour average concentration measured September 10,
1974, the rollback determined reduction required is 88.7%.
Applying this figure to the September 10 concentrate change rate
of 546.5 tons and the average September sulfur content of the concentrate
(21.82%) results in an allowable emission rate of 13.5 T/D S
(119(1 - .887)) = 13.5) required to meet the secondary standard.
The highest 24-hour average of 2193 ug/m^ occurred on November 5,
1974. A rollback reduction requirement of 83.4% is indicated from this
value. Emissions on this date were determined from the concentrate
processed, average November sulfur content of the concentrate and an
estimate of the amount of sulfur removed through limited operation of
the new acid plant. Based on information supplied by the smelter,
approximately 57% of the furnace and converter off-gases were processed
-14-
-------
through the acid plant. An efficiency of 99.5% was assumed for the acid
plant. Process losses were estimated at 7% of the input sulfur (5% in
slag and 2% in acid sludge). Total emissions were calculated from total
input sulfur:
436.9 x 24.45% = 106.82
process loss - 7.48
99.34
processed in acid plant x .57
56.62
converted to acid 56.34
emitted to atmosphere 99.3 - 56.34 = 42.96
The reduction required to meet the primary standard is 83.4%, thus:
42.96 emissions November 5
- 35.83 reduction required
7.13 T/D emissions allowable
The estimated annual average concentration of sulfur oxides indicates
that less control is required to meet the annual average NAAQS.
-14-A-
-------
IV. Emissions allowable through use of RACT
The following paragraphs contain the determination of the amount of
emissions that can be controlled through the installation of a sulfuric
acid producing facility designed to process all gaseous emissions from
the converters in each smelter. The calculations are based on the
processes used and the stacks in existence on January 31, 1972 and the
latest information on production rates and sulfur balance as provided by
the individual smelter.
All calculations are based on the maximum operating capacity of the
smelter in terms of daily concentrate feed rate. The acid plant size
required to process all converter emissions at the maximum production
rate is determined and used as a basis for calculating the allowable
emissions with this control system. Where an acid plant is used for the
treatment of other than converter emissions it is assumed that this will
continue in operation and will operate at the same efficiencies as
prescribed for converter off-gases.
The emission rates determined in the subsequent paragraphs are
based on the utilization of a double absorption acid plant where no
acid plant is currently installed or where such an acid plant exists.
The assumed conversion efficiency for such a plant is 99.5%. Where a
smelter has installed a single absorption acid plant this plant has an
assumed conversion efficiency of 97%. Emission rates from these plants
have been set at 650 ppm for the double absorption plant and 2600 ppm
for the single absorption plants.
Studies of existing operational sulfuric acid plants indicate that
the efficiency of conversion of SC>2 to SO^ decreases due to catalyst
deterioration between screenings. EPA tests conducted on two single
absorption acid plants indicate that with a twelve-month catalyst cleaning
cycle the difference in emissions due to deterioration appears to be of
the order of magnitude of 30 percent.1 The value of 30 percent was
calculated as the difference between the performance of two plants—one
at each end of the catalyst cycle. This difference in emissions reflects
not only catalyst deterioration but other factors as well, such as a
difference in emissions due to design or construction variations between
the two acid plant vendors. More recent results of tests on a double
absorption acid plant indicate no deterioration. However, discussions
with an acid plant designer indicated that up to a 10 percent increase
in emissions was expected before renewal of the catalyst.^ Catalyst
deterioration, therefore, has to be taken into account when predicting
the expected emissions concentration from an acid plant.
^Background Information for New Source Performance Standards;
Primary Copper, Zinc and Lead Smelters, Vol 1: Proposed
Standards, EPA, OAQPS, RTF NC, October 1974, p. III-3.
2Ibid, p. VI-19.
3Ibid, p. VI-20.
-15-
-------
The typical conversion efficiency for a single absorption acid
plant at the beginning of the catalyst cleaning cycle is 97 percent.
Although, as noted above, the conversion efficiency can deteriorate
between catalyst cleanings, it is reasonable to expect no more than a 10
percent change.
The overall acid plant efficiency also varies with the volumetric
flow rate and the sulfur dioxide concentration of the gas stream to the
acid plant. EPA tests indicate instantaneous sulfur dioxide emission
concentrations may vary from 1,000 ppm to 7,000 ppm during normal operating
conditions due to these flow rate and gas stream concentration variations.
These tests indicated that, by providing an allowance of 10-20% over the
vendor's guaranteed emission concentration, and averaging the concentrations
over a 6-hour period, the average emission concentration will be within
specified limits.^
Increasing the averaging time beyond six hours does not significantly
affect the average emission concentration. A 10% allowance for catalyst
deterioration and a 20% allowance for variations in input 862 concentra-
tions and volumetric flow rates increases the allowable emission concen-
tration from the acid plant from the vendor guarantee of 2000 ppm to
2600 ppm. This allowance is reflected in the emission rates established
in paragraph (f) of the regulations. Similarly the allowable emissions
for a double absorption acid plant are increased from 500 to 650 and the
corresponding efficiency is assumed to be 99.5%
The allowable emissions after the installation of an acid producing
facility of sufficient capacity to process all converter emissions were
calculated according to the following general formulae:
a. Calculation of total input sulfur that can be processed through
acid plant operating at maximum capacity:
acid plant capacity f (98/64) = daily tonnage S02 converted to t^SO^
SO? converted to H2SO/, = tons S02 daily to acid plant
acid plant efficiency
(ratio S to SO?)(tons SO? to acid plant) = total sulfur input to smelter
(percent of 862 into smelter that can be processed at
emitted by converters) at maximum capacity of acid plant
4Ibid, p. 4-7.
-16-
-------
b. Method of calculating total plant emission limitation
Emission limit = 2(1 x (R+F)) + A
where I = total sulfur input into plant at maximum design
capacity of acid plant
R = fraction of total sulfur emitted from reverberatory
furnaces
F = fraction of total sulfur emitted as "ground smoke"
or fugitive emission
A = allowable SC>2 emissions from acid plant and is
determined by subtracting the SC>2 in H2S04
produced from the total S02 input
The emissions calculated in this fashion allow for all reverberatory
and fugitive emissions to be released through a stack serving the facility.
The requirements of paragraph (d) are presumed to have been met in this
calculation and the assumption here is that all fugitive emissions will
be vented to a stack. In reality it is doubtful that all fugitive
emissions can be so vented and the emission limit thus calculated is
therefore higher than what will in fact be emitted from the stacks. The
amount of this difference varies with the percentage of total input
sulfur emitted as fugitive and the amount that will be controlled or
vented to a stack in accordance with paragraph (d) of the regulation.
An allowance of 5% of the input sulfur is made in the subsequent calcula-
tions for the fugitive emissions vented to a stack. Estimates of losses
from buildings are not included.
The percentage of total sulfur emitted from the converters and
reverberatory furnaces, the percentage emitted as fugitive and the
amount of sulfur lost in the process such as in slag, blister copper or
scrubber sludge were all based on information provided by each smelter
or by the smelter owner. It is recognized that these are estimates and
may vary from day to day, and certainly may vary from hour to hour.
This makes it difficult to relate these estimates to the hourly average
emission limitation established in paragraphs (e) and (f) unless assumptions
are made that the sampling time for compliance testing and the allowances
that have been made by including fugitive emissions, by use of a conversion
efficiency in the acid plant less than what a vendor would normally
guarantee, all provide an increase in the emission limit that mask
nearly all the large fluctuations that might be expected over the short
time periods.
-17-
-------
The subsequent paragraphs discuss the calculations and emission
rates determined for each smelter where the current operating procedures
and installed control equipment are not sufficient when operating at
design specifications to meet the emission limitations required to
achieve the NAAQS.
The Inspiration Copper Company smelter at Miami has been completely
redesigned and treatment of all electric furnace and converter off-gases
should provide sufficient control to enable them to meet the specified
emission limit.
The Kennecott smelter at Hayden operates a double absorption acid
plant to process converter and roasting off-gases and currently available
emission data indicate they can operate well within the specified emission
limitation.
The Phelps Dodge smelter at Ajo operates an absorption system for
weak gas streams and with this system the emissions will be within
specified limits.
The remainder of the smelters will need some control in addition to
the installation of an acid plant to process converter off-gases. For
these smelters the regulations provide an alternative to meeting the
emission limits by a specified time. This alternative allows the
smelter to make application for use of such supplementary control
systems as may be required to meet the NAAQS until such time as the
control of weak gas streams becomes reasonably available. A condition
for the approval of such supplementary systems is the installation and
operation of an acid producing facility designed to process all converter
off-gases. The following paragraphs specify the emission limits from
such facilities. Where a smelter has additional acid producing facilities
to accomodate other than converter off-gases, it is required that these
remain in operation, and the emission limits include the current reduction
from such facilities.
The data supplied by each smelter included maximum tons per day of
concentrate processed, average sulfur content of the concentrate and
estimated average sulfur balance data including the percentages of
sulfur emitted from the converters, roasters and reverberatory furnaces
as well as process losses and the amount of sulfuric acid produced. The
estimated amount of fugitive emissions varied greatly from smelter to
smelter and most smelters indicated the figures were only gross approxi-
mations. For the purposes of the calculations of emissions allowable
after installation of an acid producing facility and assuming most
fugitive emissions are vented to a stack, a figure of 5% of the input
sulfur was selected as being representative of the vented fugitive
emissions. The emissions from the converters, roasters and reverbera-
tory furnaces were then adjusted to account for this 5% fugitive emission.
The adjustment was proportioned to each source according the percentage
of the total emissions contributed from each source.
-18-
-------
A. ASARCO/Hayden
1. Operational data from ASARCO/Hayden smelter
maximum production rate
(concentrate feed rate) 2000 T/day
percent sulfur concentrate (average) 29.4%
input sulfur at maximum production rate 588 T/day
2. 1973 Average sulfur balance data
% input tons/day at
Emissions sulfur max, production
converter 28.2 165.7
reverbertory furnaces 16.2 95.0
roasters 23.3 136.9
fugitive 5.0 29.4
Process Losses
slag, dust, blister 3 18
Converted to acid 24.4 143
TOTAL 100.1 588
3. Acid plant capacity required to process all converter off-gases:
Total converter off-gas = percent of input converted to
^2^0^ plus percent converter
emissions (24.4 + 28.2)
Total sulfur available from
converter off-gases = 52.6% (588 T/day S) = 309.3 T/D S
Assume all S emitted as S02 = 618.6 T/D S02
Acid plant capacity required to process all converter off-gases
with an acid plant conversion efficiency of 96.1%:
618.6 x .961 x £8 = 910.3 T/D H2S04
64
A 910.3 T/D acid plant will convert 297.2 T/D S to H2S04
4. Allowable emissions from acid plant:
AE = amount of input sulfur not converted to acid
= 309.3 - 297.2
= 12.1 T/D S = 24.2 T/D S02
-19-
-------
5. Total allowable emissions
= 2(total emission from reverberatory furnace + emissions
from roasters + fugitive emissions + acid plant
emissions)
= 2(95.0 + 136.9 + 29.4 + 12.1)
= 2(273.4) = 546.8 T/D S02
546.8 x 2000 Ibs = 45,567 Ibs/hr
24 hrs
-20-
-------
B. Magma Copper Co., San Manuel
1. Operational data
Estimated maximum concentrate feed rate 3150 T/D
Average sulfur content of concentrate 32%
Input sulfur at maximum feed rate 1008 T/D
2. 1973-74 Sulfur Balance
% of tons/day at
Emissions input sulfur max, prod, rate
converters 61.8 623.4
reverberatory furnaces 29.4 295.9
fugitive emissions 5.0 50.4
Process Losses
slag & blister 3.8 38.2
3. Acid plant capacity required to process all converter off -gases
with acid plant conversion efficiency of 96.1%.
S02 converted to acid = acid plant efficiency x (tons of S02
from converters - sulfur removal in scrubber)
4 tons/day of sulfur is removed in scrubbing process.
S02 converted to acid = .961 x 2(623.4 - 4)
= .961 x 2(619.7)
= 1190.5 tons/day
Acid plant capacity = Tons S02 converted x ratio of H2S04
to S02
= 98 x 1190.5
64
= 1823 Tons per day acid plant capacity
4. Allowable emissions from acid plant
Ag = amount of input SC>2 not converted to acid
= input from converter - amount converted and removed
= 1246.8 - (8 + 1190.5)
= 48 Tons/day S02 allowable emissions from acid plant.
-21-
-------
5. Total allowable emissions:
%AGMA = 2(total sulfur emitted from reverberatory furnaces +
fugitive emissions + acid plant emissions)
= 2(295.9 + 50.4) + 48.3
= 2(346.3) + 48.3
= 741 T/D S02
741 x 2000 Ibs = 61,750 Ibs/hr
24 hrs
-22-
-------
C. Phelps-Dodge/Morenci
1. Operational data
maximum production rate
(concentrate feed) 2778 T/day
percent sulfur in concentrate 38.46%
input sulfur from concentrate 1,068.4
2. Sulfur balance data (as of 1974)
% of tons /day at
Emissions input sulfur max, prod, rate
converters 47.9 511.7
reverberatory furnaces 26.9 287.4
fugitive 5.0 53.4
roaster acid plant 0.8 8.5
Process Losses
slag & blister 2.35 25.1
roaster acid sludge 0.42 4.5
Sulfur converted to KSO 16.62 177.6
An additional 20 Tons /day of sulfur is input directly to the
converters in scrap increasing the total emissions from the
converters to 531.7 T/day.
3. Acid plant capacity required to process all converter off -gases.
efficiency of a single absorption acid plant = 96.1%
S02 converted to acid = acid plant efficiency x tons of 862 available
S02 converted to acid = .961 x 2(531.7 - 8.5 T/day to sludge)
S02 converted to acid = .961 x 2(523.2)
= 1005.6
Acid plant capacity = 9^5 x 1005.6 = 1539.8 T/D
64
4. Allowable emissions from acid plant (Ag)
AE = amount of input S02 not converted to acid
= 1068.4 - 1005.6
=62.8 T/D S02
5. Total allowable emissions
%ORENCI = 2 (total sulfur emitted from reverberatory furnaces
and as fugitive + roaster acid plant emissions)
+ allowable emissions from converter acid plant (Ag)
= 2(287.4 + 53.4 + 8.5) + 62.8
= 761.4 T/D S02
761.4 x 2000 Ibs = 63,450 Ibs/hr
24 hrs
-23-
-------
D. Phelps-Dodge/Douglas
1. Operating data
maximum production rate
(concentrate feed) 2368 T/day
average sulfur content of concentrate 37.93%
input sulfur at max. prod, rate 898.2 T/day
2. Sulfur balance estimates
% of tons/day at
Emissions input sulfur max, prod, rate
converters 38.72% 347.8
reverberatory furnaces 8.34 74.9
roasters 45.58 409.4
fugitive 5.0 44.9
Process Losses
(slag, blister, etc.) 2.36 21.2
3. Double contact acid plant required to process sulfur emitted
from converters.
Estimated efficienty of double contact acid plant = 99.5%
SC>2 converted to acid = acid plant efficienty x tons
SC>2/D from converters
= .995 x 2(347.8)
= 692.1 T/Day
Acid plant capacity = tons SC>2 converted x ratio I^SO^ to SC>2
= 692.1 x 98
64
= 1059.8
4. Allowable emissions from acid plant
AE = amount of S02 not converted to
= 695.6 - 692.1
= 3.5 T/D
-24-
-------
5. Total allowable emission rate
= 2(total sulfur emitted from roasters, reverberatory
furnaces and as fugitive) + allowable acid plant
emissions Ag
= 2(409.4 + 74.9 + 44.9) + 3.5
= 2(529.2) + 3.5
= 1061.9 T/Day S02
1061.9 x 2000 = 88,490 Ibs/hr
24
-25-
-------
V. Designated Liability Area
A designated liability area is defined as the geographic area
within which emissions from a source may significantly affect ambient
air quality. A simplified modeling approach was used to determine the
extent of the area surrounding each smelter where violations of the
standards could conceivably occur. No attempt was made to incorporate
terrain effects. From this model Figure 1 was developed which relates
the emission rate of the source to the radius of the area within which
exceedances of the standard can be expected. The radius defines the
designated liability area as distance from the stack (or stacks) serving
the facility. Based on Figure 1, the radius of the designated liability
areas for each source is as follows:
Emission Rate Radius
ASARCO/Hayden 23 T/hr 10 miles
MAGMA/San Manuel 31 15 miles
PD/Morenci 32 15 miles
PD/Douglas 44 22 miles
For the purposes of application to use supplementary control systems
an "isolated source" is defined as a source that can and will assume
legal responsibility for all violations of the applicable national
standards in the designated liability area.
These designations should apply as long as the existing SC>2 sources
in each area remain the only SOo sources. Should new sources locate in
these areas, the designated liability area will be subject to re-evaluation.
-26-
-------
500—
i-
CJ
o.
oJ
o
P 50-
0)
ce
c
o
•r-
CO
in
10--
5.J-
1
5 10 20 50 100
Distance from stack (miles)
FIGURE 1. Radius of Designated Liability Area
-27-
------- |