United States Office of Air Quality EPA-450/3-83-014
Environmental Protection Planning and Standards August 1983
Agency Research Triangle Park NC 27711
Air
Review of New
Source
Performance
Standards for
Petroleum
Refinery Claus
Sulfur Recovery
Plants
-------
EPA-450/3-83-014
Review of New Source Performance
Standards for Petroleum Refinery
Glaus Sulfur Recovery Plants
Emission Standards and Engineering Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
August 1983
-------
This report has been reviewed by the Emission Standards and Engineering Division of the Office of Air
Quality Planning and Standards, EPA, and approved for publication. Mention of trade names or commercial
products is not intended to constitute endorsement or recommendation for use. Copies of this report are
available through the Library Services Office (MD-35), U. S. Environmental Protection Agency, Research
Triangle Park, N.C. 27711, or from National Technical Information Services, 5285 Port Royal Road,
Springfield, Virginia 22161
-------
TABLE OF CONTENTS
Page
1. SUMMARY ...............................
1.1 CONTROL TECHNOLOGY .....
1.2 ECONOMIC CONSIDERATIONS AFFECTING 'THE 'NSPS .................. 11
1.3 OTHER FINDINGS ........................ ....'.'.'. .............. J" {
2. INTRODUCTION .................................
2.1 NSPS AND MSPS REVIEW
2.2 BACKGROUND INFORMATION . .................................... ?~7
2.3 SULFUR RECOVERY IN REFINERIES .' ! ............................ o,
2.4 REFINERY SULFUR PLANT STATISTICS ..'.'.' ....................... . 7
FOR NSPS CONTROL ' .' .' .' .' ! .' .' .' .' .' .' .' .' .' .' .' £fl
o o
................................ c-o
3. CURRENT STANDARDS FOR REFINERY SULFUR PLANTS ............... 3_!
3,1 AFFECTED FACILITIES ........
3.2 CONTROLLED POLLUTANTS AND EMISSION 'iFVELS ................... ^ i
3.3 STATE REGULATIOMS .......... .................. -fj
3.4 TESTING AND MONITORING REQUIRFMFNTS ......................... 77
3.5 REFERENCES ....... " ......................... 4~3
.......................................... 3-4
STATUS OF CONTROL TECHNOLOGY
,
f —
1.1 EXTENDED CLAUS REACTION PROCESSES /• ,
4.2 TAIL GAS SCRUBBING PROCESSES .." .......................... „,
4.3 COMMERCIAL STATUS OF EMISSION COMTROi's'!! ................... In
4.4 REFERENCES ..... .................... 4"13
............................................ 4-19
5. COMPLIANCE STATUS OF REFINERY SULFUR PLANTS ................. 5_!
5.1 AFFECTED FACILITIES ... c .
5.2 COMPLIANCE TEST RESULTS .. .................................. c 1
5.3 OPE.RABILITY OF NSPS UNITS . ............................... 5 ,
1.4 STATUS OF EMISSION MONITORS . ............................... c c
5.5 EMISSION TESTING ...... ............................ f'^
5.6 REFERENCES ........ ................................... T6
....................................... 5-6
6. MODEL PLANTS AND COST ANALYSES ... K ,
........................ o-l
6.1 MODEL PLANTS .......... .
•5.2 CONTROL LEVELS ..... .................................. 5"1
6.3 COST ANALYSIS ...... '.'.[ ..................................... ^~5
6.4 REFERENCES .. ...................................... 5"5
........................................... 6-18
-------
TABLE OF CONTENTS (continued)
Page
7. OTHER IMPACTS REVIEWED ..................................... 7-1
7.1 NON-AIR ENVIRONMENTAL IMPACTS .............................. 7-1
7 .2 ENERGY AND ENERGY-RELATED IMPACTS .......................... 7-2
7.3 OTHER IMAPCTS ...................... . ................. 7-3
7 .4 REFERENCES ................................................. 7-4
3
RECOMMENDATIONS ............................................ 8-1
8.1 REVISIONS TO NSPS .......................................... 8-1
8.2 REVISIONS TO MONITORING REQUIREMENTS ....................... 8-3
3.3 REVISIONS TO COMPLIANCE TESTING REQUIREMENTS ............... 8-3
3.4 SUMMARY OF RECOMMENDATIONS ................................. ft-4
8.5 REFERENCES ................................................. 8-4
-------
1. SUMMARY
1.1 CONTROL TECHNOLOGY
Based on emissions data obtained in the original NSPS study and
recently obtained emissions and reliability data from an industry survey
the most effective emission control technology for refinery Claus sulfur'
plants are systems capable of achieving 99.9 percent overall sulfur
recovery. These systems cost as much as the parent Claus plant, but
have shown good reliability and have been successfully integrated into
refinery operations at 70 sites, with another 19 planned or under construction
These systems include S02 scrubbing (Well man-Lord), reduction-Stretford
sulfur recovery (Beavon), and reduction-amine absorption (SCOT ARCO
and BSRP/MDEA). All systems subject to the NSPS levels of 25o'ppmv S02
or 300 ppmv total sulfur have successfully complied to date.
1.2 ECONOMIC CONSIDERATIONS AFFECTING THE NSPS
The primary issue involving review of the NSPS is the cost of
controls. To determine cost trends, facilities of 10.16, 50.8, and 101 6
megagrams per day (Mg/D) were modelled. At 10.16 Mg/D, the cost-effectiveness
of control was assessed at 2,125 dollars per megagram of sulfur dioxide
(S02) removed. At 50.8 and 101.6 Mg/D, the corresponding cost-effectiveness
indeces were found to be $880/Mg and S675/Mg, respectively. The current
NSPS would then require a maximum expenditure of about 51,430/Mg (at the
20.32 Mg/D cutoff), but more typically would be considerably less than
S900/Mg S02 based on current and planned sulfur plant capacities.
1.3 OTHER FINDINGS
No significant adverse environmental impacts were noted for the
control technologies. Control systems energy consumption is significant
and accounts for 5 to 13 percent of total sulfur plant operating costs,
for the models examined.
1-1
-------
For systems with tail gas incineration, EPA Method 6 and continuous
S02 analyzers are used for initial compliance testing and monitoring,
respectively. For systems without tail gas incineration, a modified
EPA Method 15 has been used and possible changes to this method for
measuring reduced sulfur compounds may be forthcoming. Continuous
monitors for total reduced sulfur have recently been introduced and are
currently being evaluated by the EPA. No satisfactory hydrogen sulfide
(H?S) monitors have been identified.
1-2
-------
2. INTRODUCTION
2.1 NSPS AND NSPS REVIEW
The United States Environmental Protection Agency (EPA) proposed
new source performance standards for petroleum refinery sulfur plants
under Section 111 of the Clean Air Act on October 4, 1976, (41FR43866).
These regulations were promulgated on March 15, 1978, (43FR10866) and
amended on October 25, 1979, (44FR61542). The regulations applied to
Claus sulfur recovery plants greater than 20 long tons per day (LT/D)
capacity, the construction or modification of which commenced after
October 4, 1976.
The Clean Air Act Amendments of 1977 require that the Administrator
of the EPA review and, if appropriate, revise established standards of
performance for new stationary sources at least every 4 years. The
purpose of this report is to review and assess the need for revision of
the existing standards for refinery sulfur plants based on developments
that have occurred or are expected to occur within the petroleum refining
industry. The information presented in this report was obtained from
reference literature, discussions with industry representatives, trade
organizations, control equipment vendors, EPA regional offices, and State
and local agencies.
2.2 BACKGROUND INFORMATION*
Petroleum refineries convert naturally occuring "crude" petroleum
liquids into marketable fuels such as heating oil and gasoline in a number
of chemical processes. During this processing, impurities such as sulfur
are liberated as gaseous hydrogen sul fide (H2S) and are collected with
Plant gases known as process or fuel gas. To satisfy air pollution
regulations which effectively limit the sulfur in fuel gas, and to reduce
corrosion problems, refineries "sweeten" or remove hydrogen sul fide from
the fuel gas before burning it in process heaters and boilers.
Sweetening processes currently used in petroleum refineries consist
of scrubbing the sour gases with liquids which preferentially absorb
hydrogen sulfide and carbon dioxide over other species. Regeneration of
2-1
-------
the scrubbing solutions evolves a secondary gas stream containing
concentrated hydrogen sulfide with lesser amounts of carbon dioxide,
water vapor, and hydrocarbons.
Refinery process water may also contain dissolved gases such as
ammonia and H2S, which require removal before the water may be reused or
discharged. The water is subjected to thermal or steam stripping which
liberates the dissolved gases into c gas stream consisting of water vapor,
hydrogen sulfide, hydrocarbons, and ammonia.
In many instances, the choice cf disposition of this gas stream is
to route it to sulfur recovery with other H?S-rich streams. Alternatively,
the sour water stripper overhead may be incinerated where sulfur dioxide
regulations oermit.
2.3 SULFUR RECOVERY IM REFINERIES
At one time, many refineries sold the HjS-ric^ gas streams to
neighboring chemical plants, or "scavengers", as feedstock for sulfuric
acid or elemental sulfur production. Recent trends, however, are to
convert the H?S on-site to marketable liquid sulfur via the Glaus process.
2.3.1 Glaus Process2
Figure 2-1 is a representative process diagram of the Claus process.
Basically, the overall chemical reaction is a thermal and catalytic
oxidation of H2S to elemental sulfur in the gaseous phase:
(1) H?S + 1/2 02 ->• H20 + S
The reaction is exothermic in that considerable heat is generated by
the Claus process. Additionally, one mole of water vapor and one mole of
sulfur vapor are formed for each mole H2S converted.
Actual Claus reactions occur in stages as shown in Figure 2-1.
The sour gases are initially combusted in a furnace where sufficient
air is admitted to convert one-third of the H2S to S02:
(2) H2S + 3/2 02 -" S02 + H20
Then the remaining 2/3 rbS and the 1/3 S02 react:
(3) 2H2S + S02 + 3S + 2H2n
Combining reactions (2) and (3) yields the overall Claus reaction (1).
2-2
-------
IIP Steam
ro Acid C«i
KO
Drum
Reaction
furnice
c
Converter
Ha "
LP Steim
Condenser
No. 1
./ v
M
Blower
DfH
Sulfur
/^>Jleheater
r-H JNo. 2
>/
LP Steam
Condenser
No. 2
BIM
LP Steam
Condenser
No. 3
Sulfur
nrw
Convtrttr
No. 3
J
LP 5t«.m
I I
Gas
Condenser
No. 4
Sulfur
IHW
Sul fur
Sulfur Pit
Pump I
Liquid
"Su1fur
2_K now dlagra., for a throe-stafle ciaus sulfur
recovery facility
-------
Since the above reactions are exothermic, the conversion of H2S to
elemental sulfur is promoted by removal of heat via shell and tube heat
exchangers; therefore, the Glaus p'ant is a net exporter of steam as well
as sul fur.
Reaction (1), in addition to being favored by lower temperatures, is
also promoted by catalysts and removal of sulfur vapor. Therefore, upon
leaving the furnace (where up to 60 percent of the Glaus reaction has
taken place), the gases are subjected to successive catalytic stages and
sulfur condensers, with each successive catalytic stage operated at lower
temperatures. In lieu of emission regulations, the Glaus plant is normally
operated with two or three catalytic stages, depending on economic
considerations, with the final condenser outlet routed to an incinerator.
2.3.2 Claus PI ant Emissions
The only significant source of emissions is the Glaus incinerator;
fugitive sulfur emissions are possible due to leaks and atmospheric
venting of liquid sulfur storage arid transfer areas. Emissions are
typically sulfur dioxide where incinerators are operated at temperatures
of 650°-800°C, sufficient to destruct sul fides and elemental sulfur
vapor. Lower oxidizer temperatures of 540-650°C may be adequate to
destruct gaseous sulfides where the sulfide concentration has been
significantly reduced upstream by tail gas treating. Emissions are a
direct function of the Glaus conversion efficiency, which will be discussed
in the next section. For a typical Glaus plant operating at 96 percent
conversion efficiency, emissions are 8 percent by weight of the incoming
sulfur feed.
Other emissions from the Glaus incinerator are small amounts of
hydrocarbons, nitrogen oxides, and carbon monoxide, all of which are
dependent upon fuel combustion parameters and generally unrelated to
Glaus plant operation.
2.3.3 Factors Affecting Sulfur Dioxide Emissions3*4'5
Design of the Glaus plant is important, as the type of catalyst,
number of catalytic stages, and process controls all influence emissions.
Obviously, the number of catalytic stages determines to a great extent
the ultimate sulfur recovery efficiency. A Glaus furnace may operate at
2-4
-------
60 percent conversion, while successive catalytic stages may increase
conversion to 35-90 percent for one, 92-95 percent for two, and 96-97
percent for three stages. The type of catalyst is also important, as
newer alunina catalysts show 1 to 2 percent improvement over the conventional
bauxite catalysts. Finally, the Clans plant requires both upstream
monitoring of acid gas feed and downstream monitoring of tail gas sulfur
species to enable operation at optimum conditions.
Glaus plant operation is heavily influenced by the feedstock
composition. The presence of hydrocarbons, carbon dioxide, and ammonia
all adversely affect Claus plant performance, first by the dilution of
reactive H^S and SO? in the Claus plant, but nore importantly by
adverse side reactions. Hydrocarbons and ammonia if not properly combusted,
form solid compounds which rapidly degrade catalyst surfaces and Claus
performance. Carbon dioxide also reacts with hydrogen sulfide, thereby
Hi(iinishing sulfur recovery:
(4) C02 + H?S -»• H20 + COS
(5) COS + H2S ->• HoO + CS2
Thus, two additional sulfur compounds, carbonyl sulfide (COS) and
carbon disulfide (CSg) are formed in the Claus furnace and, though hydrolyzed
in the subsequent catalytic stages, are significant contributors to Claus
emi ssions.
Hydrocarbons may also react in the Claus furnace to form CS£:
(6) CH4 + 2$2 -»• C$2 + 2H2S
Operator control of the process is the most influential factor
af-Fecting emissions. In order to maximize sulfur conversion, the following
parameters must be controlled:
0 stoichiometric ratio of H2S to S02
furnace, catalyst bed, and condenser temperatures
0 catalyst activity
Figure 2-2 illustrates the imnortance of maintaining the H2S-S02
ratio at 2 to 1. This is accomplished by metering the air flow to the
furnace to convert exactly one-third of incoming H2S to SO?. Air control
is complicated by variable feedstock flow rates and changes in composition,
both of which affect furnace stoichiometry. If air to the furnace is
deficient, the H2S-S02 ratio is too high and sulfur recovery diminishes;
2-5
-------
36(
25
o
u.
u_
UJ
e:
UJ
UJ
1.0 2.0 3455
H2S/S02 >«OL SATJO
(TAIL GAS)
Figure 2-2. Theoretical Claus sulfur recovei-y efficiency vs. Mole Ratio.
7 I 3
2-6
-------
if air is excessive, too much 502 is formed, the ratio becomes less than
2 to 1, and recovery again diminishes.
Temperatures must be maintained at optimum levels; high temperatures
decrease reaction equilibrium and sulfur condensation while low temperatures
may promote adverse reactions on catalyst surfaces. The final condenser
must especially be maintained at a 1 ow temperature to minimize sulfur
vapor 1osses.
Catalyst activity is maintained by periodic regeneration or replacement,
which require either a period of suboptimum operation or plant shutdown.
Operation of a Glaus plant at low loads may adversely affect
performance. One vendor reported a 2 to 3 percent loss in recovery at
20 percent load. Operation from two-thirds capacity up to 120 percent
capacity is reported with no loss in recovery.
2.4 REFINERY SULFUR PLANT STATISTICS^,8,9,10,11
In 1973, total Glaus sulfur capacity in U.S. refineries totalled
8,000 megagrams per day (Mg/D). 1974 construction was estimated at over
1,000 Mg/D. Since statistics have not been kept on whether the growth
since 1973 has been due to new facilities or replacements, the actual
Glaus capacity is not known, but is considerably greater than 10,000
Mg/D. Recent construction announcements show that for 1981, nine sulfur
plants were installed totalling 800 Mg/D, with a tenth plant of unspecified
capacity constructed. In 1982, eight plants having 516 Mg/D were scheduled
for completion, with two others of unspecified size due to start up.
Vendor announcements indicate that at least 13 new Glaus facilities
will be constructed in 1983, totalling 2,009 Mg/D capacity (See Table 4-2).
Construction announcements in Hydrocarbon Processing for early 1983
project that 28 new Glaus plants will be constructed in the 1983-85 time
frame, 25 of which will total 5,184 Mg/D. Of these, 19 individual plants
totalling 5,083 Mg/D will be sized greater than 20.32 Mg/D, capacity. Six
plants of 101 Mg/D total capacity will be constructed that are not subject
to Federal NSPS.
These figures indicate that strong growth in sulfur plant construction
will continue, the average size unit will be large (~200 Mg/D), and the
total capacity of units not covered by NSPS will account for less than
2 percent of new plant growth.
2-7
-------
2.5 SELECTION OF SULFUR PLANTS FOR NSPS CONTROL
Refinery sulfur plants were originally selected for NSPS development
because of their potential for emissions of sulfur dioxide in significant
quantities. Though the actual emissions from Cl aus plants have likely
decreased significantly from the estimated 306,715 megagrams annually in
197311 due to replacements with NSPS units and considerable retrofitting of
existing units, the potential for emissions from Claus plants without
controls remains. For example, a 101.6 Mg/D plant operating at 96 percent
conversion for 350 days per year at rated capacity could emit 2,845 megagrams
per year sulfur dioxide, a criteria pollutant.
The widespread use of emission controls on Cl aus plants, hereafter
referred to as "tail gas units", on many retrofitted existing Claus plants
and practically all refinery Cl aus plants installed since 1975, indicates
that the technology for Cl aus emissions control is well established and
generally accepted by industry. Therefore, the ingredients for NSPS
development—growth, emission potential, and demonstrated control technology—
that were present prior to development of the NSPS, persist at this time.
2.6 REFERENCES
1. Standards Support and Environmental Impact Statement Volume 1:
Proposed Standards of Performance for Petroleum Refinery Sulfur Recovery
Plants, EPA 450/2-76-016a, September 1976, pp. 3.1-3.2.
2. Reference 1, pp. 3.2-3.9.
3. Reference 2.
4. Parnell, David C., "Differences; in Design of Cl aus Plants for Various
Applications", Paper Number 22d, Spring National AIChE Meeting, April 9,
1981.
5. GPA Panelist Outlines Cl aus Process Improvements in Sulfur Recovery,
Oil & Gas Journal, p. 9299, August 7, 1978.
6. Reference 1, pp. 3.1-3.2.
7. "HPI Construction Boxscore", Hydrocarbon Processing, October 1981,
pp. 3-18.
8. Letter, W. T. Knowles, Shell Oil Company to Charles B. Sedman, U.S. EPA,
August 24, 1982.
2-8
-------
9. Letter, M. A. Peterson, Union Oil Co. of California, to C. Sedman
U.S. EPA, September 15, 1982.
10. Letter, J. C. Brocoff, Ralph M. Parsons Co., to S. T. Cuffe U S EPA
dated February 16, 1983. ' '
11. "HPI Construction Boxscore," Hydrocarbon Processing, February 1983.
12. Reference 1.
2-9
-------
3. CURRENT STANDARDS FOR REFINERY SULFUR PLANTS
3.1 AFFECTED FACILITIES
Existing new source performance standards (NSPS) for new, modified,
and reconstructed refinery sulfur recovery facilities limit sulfur
emissions from Claus sulfur recovery plants of greater than 20.32 megagrams
per day (Mg/D) capacity. A Cl aus sulfur recovery plant is defined as a
"process unit which recovers sulfur from hydrogen sulfide by a vapor-
phase catalytic reaction of sulfur dioxide and hydrogen sulfide".1
3.2 CONTROLLED POLLUTANTS AND EMISSION LEVELS
The NSPS limits emissions of reduced sulfur compounds, hydrogen
sulfide, and sulfur dioxide as follows:
Reduced Sul fur Compounds
Reduced sulfur compounds from Claus plants are defined as hydrogen
sulfide, carbonyl sulfide, and carbon disulfide. These are limited to
0.030 percent (300 ppmv) by volume at zero percent oxygen on a dry basis.
These are measured only if the emission control system is a reduction
system not followed by an incinerator. This is roughly equivalent to 99.8-
99.9 percent sulfur recovery.
Hydrogen Sulfide
Hydrogen sulfide emissions are limited to 0.0010 percent (10 ppmv)
by volume at zero percent oxygen on a dry basis. Hydrogen sulfide
measurements are required only if the emission control system is a
reduction system not followed by an incinerator.
Sul fur Dioxide
Sulfur dioxide emissions are limited to 0.025 percent (250 ppmv) by
volume at zero percent oxygen on a dry basis if emissions are controlled
by an oxidation control system or a reduction control system followed by
incineration. This is comparable to the 99.8-99.9 percent control level
for reduced sulfur.
3.3 STATE REGULATIONS
In 1976, when NSPS were proposed, most States having petroleum
refineries generally required 99 percent sulfur removal for new Claus
Plants.2 The Environment Reporter reveals some recent changes, but in
3-1
-------
general, the States having the majority of refineries still require
99 percent sulfur recovery (equivalent to about 1300 ppmv S02 at stack
conditions).3 Table 3-1 summarizes; selected 1972 and 1982 standards
for refinery sulfur plants. One noticeable omission is for California
which has standards set by local a-'r pollution control districts. (One
district having refineries generally requires control equivalent to the
NSPS.) Hydrogen sulfide regulations were generally based on ground
level concentrations. The listing in Table 3-1 may understate the
ultimate control requirements, as other State regulations such as best
available control technology (BACT) or prevention of significant
deterioration (PSD) mandates may well supercede emission codes.4
Table 3-1. SELECTED STATE REGULATIONS FOR NEW SULFUR RECOVERY PLANTS AT 101.6 Mg/
State 1972 1982
Delaware 2000 ppmv (98.5%) Process Wt. (93.4%)
Illinois — 2000 ppmv (98.5%)
Louisiana .01 Ib/lb S input (99%) .01 Ib/lb S input (99%)
New Jersey 15000 ppmv ( 90%) 15000 ppmv ( 90%)
Ohio .01 Ib/lb S input (99%) Process Wt. (99.2-99.4%
for 101.6 Mg/D)
Oklahoma .01 Ib/lb S input (99%) .01 Ib/lb S input (99%)
Pennsylvania Process Wt. (93.4%) 500 ppmv ( 99.6%)
Texas* Process Wt. (87.6%) Process Wt. (2200 ppmv
or 98.4%)
*In most instances superceded by BACT requirements (Reference 4).
Since most refineries are located in industrialized urban areas, and
because essentially all sulfur plants potentially emit greater than 90.74
megagrarns per year and are subject to additional regulations such as
BACT/PSD mentioned above, essentially all sulfur plants installed within
the last 5 years have been required to install tail gas treaters. The
only exceptions have been small sulfur plants in rural areas. States
contacted generally require tail gas treaters as best available control
technology (BACT) unless the source is shown to have a negligible impact
on air qua!ity.5>6
3-2
-------
3.4 TESTING AMD MONITORING REQUIREMENTS
3.4.1 Testing Requirements
Performance tests to verify comoliance with the standards for refinery
sulfur plants must be conducted within 60 days after achieving full
capacity operation, but not later than 180 days after the initial startup
oP the facility. This is a uniform requirement for all affected facilities
under 40 CFR 60.8. The EPA reference methods to be used in connection
with the affected facilities include:
1. Method 4 for moisture content
2. Method 6 for SO?
3. Method 15 for H2S and reduced sulfur compounds
For Method 5, a series of three runs each spanning a minimum of four
consecutive hours is reouired. For Method 15, three runs each consisting
of 15 samples taken over a minimum of three hours is required. Reference
Method 4 is conducted simultaneously with Method 15, sampling at a rate
-ronortional to the gas velocity for a minimum of four continuous hours"
samoling for each run.
Total reduced sulfur is expressed as S02 eauivalent under Method 15
by the following formula:
SO? equivalent = i (H2S, COS, 2CS2)d
where: S02 equivalent = the sum of the concentration of each of me
measured compounds expressed as sulfur dioxide in opm
H2S = hydrogen sulfide, ppm
COS = carbonyl sulfide, opm
C$2 = carbon disulfide, ppm
d = dilution factor, dimensionless
M
3-1 Average S02 equivalent = z S02 equivalent i
1 = 1
HCl-Swo)
where: average S02 equivalent = average S02 equivalent in ppm, dry basis as
S02 equivalent = S02 in ppm as determined in equation 3-1
XJ = M umber of analyses performed
3wo = Fraction of volume of water vapor in the gas stream as
determined by Method 4
3-3
-------
3.4.2 Monitoring Requirements
A continuous monitoring system is required under the MSPS to monitor
and record the concentration of 502 or alternatively, reduced sulfur and
rl?S compounds, on Claus tail gas exhaust to the atmosphere. Specifications
for continuous sulfur dioxide monitors were promulgated in Appendix 3,
40 CFR Part 60.
3.5 REFERENCES
1. Federal Register, Wednesday, Ma^ch 15, 1978, Part III 10866-10873.
2. Standard Support and Environmental Impact Statement Volume 1: Proposed
Standards of Performance for Petroleum Refinery Sulfur Recovery Plants,
EPA 450/2-76-016a, September 1976, an. 3.13-3.15.
3. Environment Reporter, State Air Laws, Bureau of National Affairs
(updated to 7/9/32), pp. 201:001-555:0523.
4. Letter from Sam Crowther, Texas Air Control Board, to S.T. Cuffe,
!'.3. EPA, dated January 7, 1983.
r->. Telephone Conversation: C. Sedman, EPA, to Sam Crowther, Texas Air
Control Board, March 16, 1982.
6. Telephone Conversation: C. Sedman, EPA, to Jim Stone, Louisiana
Bureau of Environmental Services, March 17, 1982.
3-4
-------
4. STATUS OF CONTROL TECHNOLOGY
The total sulfur emissions from a Claus sulfur plant were established
in Chapter 2 as a direct function of the extent to which the Claus reaction
reaches completion. Themodynamically, the Claus reaction is limited at
normal operating temperatures and pressures to 97-98 percent recovery,
but in actual practice is reduced by process limitations such as unsteady
state operation and catalyst aging.1 Therefore, to reduce emissions
to the atmosphere, the Claus process must be augmented by (1) extending
the Claus reaction into a lower temperature liquid phase, or (2) adding
a scrubbing process to the Claus exhaust stream.
4.1 EXTENDED CLAUS REACTION PROCESSES
There are at least five processes currently available to augment or
extend the Claus reaction beyond the recoveries normally achieved in a
conventional Claus with three catalytic stages. These are the BSR/Selectox,
S.jlfreen, Cold Bed Absorption, Maxisulf, and IFP-1 processes. Of these
?our, the only domestic refinery applications to date involve the IFP-1
Process; therefore, only the IFP-1 will be discussed in detail. The
other processes are briefly described herein as aoplicable.
4.1.1 BSR Selectox2*3
The BSR/Selectox I process, recently developed by Union Oil of
California and the Ralph M. Parsons Company, is designed to provide a
sulfur recovery efficiency in the range of 99 percent, in conjunction
with a three-stage Claus.
The BSR/Selectox I is a fixed bed catalytic process consisting of
two steps. In the first step, tail gas from the second stage of the
Claus plant is heated to above 238'C (500°F) in a reducing gas generator
fueled by substoichiometric air and refinery fuel gas. The hot gases arp
oassed over a catalyst bed where all sulfur species are converted to
hydrogen sulfide. The gas is cooled, reheated, and passed over a
proprietary catalyst to oxidize the H2S to elemental sulfur. Sulfur is
condensed out with the remaining tail gas passed to the final Claus stage
4-1
-------
Close control of H2S:S02 ratio in tha Glaus plant is not as critical as
with Claus and other extended Glaus reaction schemes. Up to 99 percent
sulfur recovery is reported on an overseas refinery application.
4.1.2 Sulfreen4-5
The Suifreen process converts H2S and S02 contained in Claus tail gas
to elemental sulfur at temperatures of 127°C to HO°C (260°F to 300°F) by
extension of the Claus reaction.
Claus tail gas is first scrubbed with liquid to wash out entrained
sulfur liquid and sulfur vapor. The tail gas is then introduced to a
battery of reactors where the lower temperatures push the Claus reaction
toward completion on the surfaces of a special alumina catalyst. A
regeneration gas, usually nitrogen, periodically desorbs the sulfur-laden
catalyst beds, first driving off water vapor and carbon dioxide at 300°C
(572°F) and then sulfur at 400°C (752°F). The sulfur is condensed out of
thp carrier gas, the carrier gas scrubbed in a sulfur wash, and then
returned to the regeneration cycle.
A Sulfreen unit may consist of as little as two reactors, one in
absorption and one in desorntion service. The gases fron the reactors
being desorbed are incinerated before discharge to the atmosphere.
4.1.3 Amoco CBA6
The cold bed adsorption (CBA) process, developed by Amoco Production
Company, is essentially the same coicept as the Sulfreen process, except
low temperature acid gas feed is used as the regeneration gas. A recent
study assesses the CBA capability on a two-stage Claus plant at 98 percent
recovery. Currently, three units (one on a natural gas plant in the
United States) are in operation with capacities from 15 to 900 metric
tons o^ sulfur per day.
4.1.4 Maxisulf7
The Maxisulf process,recently developed by Davy McKee, is similar
in principle to the Sulfreen and Amoco CBA processes and features a
cyclic, two-reactor process, one absorbing and one desorbing. The key
feature is that a heated slipstream of Claus tail gas is used for the
desorbing gas, then recombined with tail gas, entering the absorbing
reactor. Thus, a closed loop, forced circulation desorption scheme is
avoided. Efficiencies of 99 percent on refinery application are cited by
the vendor. Two units are scheduled for construction in Germany.
4-2
-------
4.1.5 IFP-l8,9
The IFP-1 (Instltut Francais du Petrole) process is the only Claus
extension type of tail gas process to be successfully applied on U.S.
refinery Claus plants. It was initially applied at two refineries in
1973 as a retrofit second-stage to one-stage Claus plants. Larger
installations followed as shown later in Table 4-1.
The IFP-1 process is essentially a liquid-phase Claus reactor which
accepts Claus tail gas directly with no conditioning. The reactor is a
packed column with a specially designed "boot" for collecting liquid
sulfur. Metal salts catalyze the reaction which takes place in a high
boiling point solvent, polyethylenglycol (PEG), above the melting point
of sulfur—In the ranqe of 121-126"C (250-2SO°F). The metal salts form a
complex with M2S and S02 in the feed gas, which in turn reacts with
additional H2S and S02 to form elemental sulfur and regenerate the
catalyst. Sulfur coalesces and settles into the boot of the reactor,
•Von */hicn it is drawn as a molten product.
Gas typically leaving the reactor contains about 1500-2500 ppmv
sulfur which includes essentially all COS and CS2 formed in the Claus
ilant, about 300 opmv sulfur vapor (the equilibrium concentration of
sulfur vapor at 126°C), and the unreacted H2S and S02. Conversion
efficiencies on a nonrefining application of 99.3 percent have been
reported. The reactor exhaust containing 1500-2500 ppmv sulfur is
incinerated before discharge to the atmosphere. This represents overall
control of roughly 99.0 percent.
Conversion efficiencies are maximized by (1) operating the IFP at
H2S to S02 ratios of as near 2:1 as possible and (2) operating the first
Claus reactor at a higher temperature than normal to minimize COS/CS2
formation.
Operation slightly above the 2:1 H2S to S02 ratio is practiced due
to the adverse effects of operation below 2:1. When the Claus tail gas
is deficient in H2S to carry the Claus reaction toward completion, the
IFP solvent evolves absorbed S02 which decreases efficiency and increases
sulfur emissions. Operation at long periods under H2S deficient conditions
may result in deterioration of the solvent/catalyst complex, where
emissions increase until the unit is shut down and IFP solvent regenerated
or completely replaced.
Figure 4-1 illustrates the IFP-1 process.
4-3
-------
STEAM
ARSORPTION
TOWER
>- TAIL GAS TO
INCINERATION
CATALYST
-SULFUR
CLAUS PLANT
TAIL GAS BEFORE
INCINERATION
Figure 4-1. Flow diaciram for IFp-1 Claus tall-gas clean-up process
-------
4.2 TAIL GAS SCRUBBING PROCESSES
There are essentially two generic types of tail gas scrubbing
processes—the first where Glaus tail gas is oxidized and the oxidized
sulfur (S02) absorbed by caustic scrubbing and the second where Claus
tail gas is reduced, and the reduced sulfur (H2S) absorbed by scrubbing
with solvents or caustic reagents. Initially, the first tail gas scrubbers
were mainly the sulfur dioxide/caustic type. Subsequently, the vast
majority have been the reduction scrubber variety. For subsequent
modelling and analyses, the reduction scrubber systems have been chosen
as representative technologies. Both processes are described herein as
demonstrated technologies.
4.2.1 Oxidation Tail Gas Scrubbers
At least three processes were developed to scrub S02 from incinerated
Clans tail gas and recycle the concentrated SO? stream back to the Claus
for conversion to elemental sulfur or, alternatively, send the concentrated
30? to a sulfuric acid plant. These were the Wellman-Lord, Stauffer
Aouaclaus, and IFP-2. Since only the Wellman-Lord has been applied
successfully to U.S. refineries, it is the only process of its type
examined.
4.2.1.1 The Wellman-Lord Process 10.H The Wellman-Lord process
was developed by Wellnan-Power Gas Incorporated and has been applied to
various industrial S02 sources.
Figure 4-2 illustrates the Wellman-Lord process as applied to Claus
tail gases. The Wellman-Lord system uses a wet regenerative process to
reduce stack gas sulfur dioxide concentration to less than 250 ppmv or
approximately 99.9 percent sulfur recovery.
Claus plant tail gas is incinerated and all sulfur species are
oxidized to sulfur dioxide. Gases are then cooled and water quenched to
remove excess water and lower gas temperatures to absorber conditions.
The SOo-rich gas is then contacted countercurrently with a solution of
sodium sulfite (Na2S03) and sodium bisulfite (NaHS03) which reacts with
the S02 to form the bisulfite:
S02 + Na2S03 + H20 - 2NaHS03
The off-gas is reheated (where required) and vented to the atmosphere.
4-5
-------
EVAPORATOR
AND STEAM STRIPPING
DISSOLVING
TANK
QUENCH AND GAS
CnOLIMG SECTION
I
en
CLAUS PLANT
TAIL GAS
AITER
INCINERAIIOM
RECYCLE
QUENCH HATER
•rf1- .
-STACK
ACID WATER PURGE
TO NEUTRALIZATION
Nail SO-,
SOLUTION
J
1
NaOII
MAKE-UP
N32MJ3
SLURRY I
PURGE
Na2S03 SOLUTION
0
ZD
o
o
o
v
PRODUCT SO?
RECYCLE TO"
CLAUS PLANT
H20 RECYCLE
figure 4-2. Flow diaqram fmr the Hellman-Lord S02 recovery process.
-------
The bisulfite solution is boiled in an evaporator-crystal 1 izer,
where the bisulfite solution decomposes to S02 and H20 vapor and sodium
suKite is precipitated:
heat
2MaHS03 -»• Na2S03^ + H20 + S0o +
Sulfite crystals are separated and redissolved for reuse as lean
solution to the absorber. The wet S02 gas is directed to a partial
condenser where most water vapor is condensed and reused to dissolve
sulfite crystals. The enriched S02 stream is then recycled back to the
Glaus plant for conversion to elemental sulfur or sent to an acid plant
for conversion to sulfuric acid.
The Wellman-Lord process has heen operating in U.S. rpfin^ries since
1972.
4• 2•?- Reduction Tail Gas Scrubbers
At least four processes have been developed for tail gas sulfur
renoval. These processes convert the tail gas sulfur species to HoS by a
-eduction steo, then scrub the H2S from tail gases prior to venting.
These are the Beavon, Reavon MDEA, SCOT, and A?CO processes. The Beavon
process is unique in that the H2S is converted to sulfur outside the
Claus unit using a lean H2S-to-sulfur process called Stretford. The
other three processes utilize conventional amine scrubbing and regeneration
to remove the H23 and recycle back as Claus feed. Since the Beavon MDEA
SCOT, and ARCO processes are similar and the SCOT process the most commonly
used, the SCOT process will be described in more detail, with the Beavon
MDEA and ARCO descriptions minimized to point out the deviations from the
SCOT.
Also, since all processes utilize a reduction step, this step is
described first as a common process.
4'2-2-1 The Reduction Step. All generic reduction tail gas
processes utilize a reduction step in which sulfur species are'converted
essentially to H2S by hydrogenation and hydrolysis under moderate conditions
of temperature and pressure. Before the tail gas enters a packed bed
hydrogenation reactor, fuel gas is combusted substoichiometrically in an
4-7
-------
inline burner to produce the reducing conditions necessary to convert
sulfur gases to H^S. The combustion products, primarily carbon monoxide
(CO), nitrogen, and water vapor (H20), are mixed with the tail gas to
provide a reducing atmosphere. Extra hydrogen may be required upstream
of the burner, depending on the hydrogen content of Claus tail gas. A
cobalt-molybdenum catalyst promotes the hydrogenation and hydrolysis
reactions as follows:
SB + 3H2 > 8H2S
S02 + 3H2 * H2S + 2H20
COS + H20 > H2S + C02
CS2 + 2H20 * 2H2S + C02
After hydrogenation and hydro!/sis, the tail gas is cooled and water
removed.
4.2.2.2 Beavon Process.^2'13 The Beavon process was developed by
the Raloh M. Parsons Company and Union Oil Research.
In the Reavon or 3eavon/Stretford process, the cooled gas is directed
to a Strstford sulfur plant, where it is contacted countercurrently with
a sodium solution and absorbed. The absorbed H^S is oxidized and
precipitated out of the solution as elemental sulfur solids, and the
sodium values regenerated by the following reactions:
(a) Absorption of H2S
H2S + Ma£C03 > NaHS + NaHCOs
(b) Precipitation of sulfur
2NaV03 + NaHS + NaHCOs * S4- + N32V205 + HaeCOs + HaO
(c) Regeneration of sodium varadate (NaV03)
N32V205 + ADA* (oxidized) > 2NaV03 + ADA (reduced)
* Anthraquinone Disulfonic Acid
Air is then blown through the solution to froth out the sulfur and
regenerate the ADA:
ADA (reduced) + 1/2 02 -» ADA (oxidized)
Sulfur froth is then collected, filtered, and remelted to be combined
with Claus sulfur.
4-8
-------
The overall reaction is the Claus reaction; hence, no chemicals are
consumed in theory. Actually, adverse side reactions occur due to
temperature excursions in the presence of trace oxidizing species in the
tail gas, and result in the buildup of sodium thiosul fate and related
compounds in the circulating liquor. This requires a periodic or continuous
purge stream to keep dissolved solids to a desired level.
A new variation of the Beavon process involves replacement of the
Stretford process with the Unisulf proces; although similar to the
Stretford, the Unisulf reportedly requires no purge of solution under
normal operating conditions.
Figure 4-3 is a typical flow diagram for the Beavon process.
Stretford absorber off-gases, typically containing 20-80 ppmv carbonyl
sulflde and trace species of other sulfur gases, do not require incineration
and are normally vented to the atmosphere without further processing. A
stand-by incinerator is normally available, however, to handle process
upsets where H2S emissions exceed a given level, usually 10-20 ppmv in
stack gases.
The Beavon process has been operating in U.S. refineries since 1973
4.2.2.3 The SCOT Process. 14.15 The Shel 1 Clau$ Qff_gas ^^
(SCOT) process scrubs the cooled reactor gas with an alkanolamine solution
in an absorber. The solution selectively absorbs H2S over S02. Absorbed
acid gases are 1 iberated from the amine solution by stripping with steam
in a regenerator and are recycled to the gas inlet of the Cl aus unit.
Amine absorber off-gas containing about 200-300 ppmv H2S requires
incineration, but at a lower temperature (~540'C) than a typical Claus
incinerator. A typical performance guarantee for the SCOT is 250 ppmv
S02 in the incinerated off-gas, though guarantees as low as 150 ppmv have
been given.
The SCOT process commonly uses diisopropanol amine (DIPA) a secondary
amine or methyldiethanol amine (MDEA), a tertiary amine, which are more
selective than amines used for refinery gas treating. Other solvents may
be used, but the final choice depends on process economics.
Figure 4-4 schematically represents a typical SCOT process. The
SCOT process has been operating in U.S. refineries since 1973.
4-9
-------
run GAS
~i
AIR——
GLAUS PI AMI
TAIL GAS RITORE
INCINERATOR
nxrn orn
REACTOR
BURNER
Low pressure
steain
IIYOROGENATED
TAIL GAS
COOLINf
TOWER
ABSORBER OFF-GAS TO INCINERATION OR STACK
< Caustic
SIR
r«-
LIQUID RETURN
j STRETFORD SOLUTION
AIR
RD
•SULFUR FROTI
GAS PURIFYING OXIDIZE
TOWER ABSORBER
SOUR HATER
SULFUR
HELTER
(TO WASTt
TREATMENT)
FILTER
OR,
CENTRIFUGE
I
SULFUR
PURGE STREAM
Figure 4-3. Flow diagram for the Beavon sulfur removal process.
-------
Reducinn
Line Heater
f 1,1115 nlant tail nas
nrior to incinerator
Air
Cool inr| Tower
Packed or Tray
Tall gas to incinerator
Reactor
Fixed-bed
reducinq
catalyst
I.P. steam
i
rn
1
Air or
water
Caustic
•*—Lean amine from regenerator ;
Tray Tower Absorber
. Clans Unit
Sour-water I Rich
to Amine
exlstinq sour-
water stripper
Fiqure 4-4. FLOW OIAfiRAM FOR THE SHELL CLAUS OFF-GAS TREATING PROCESS .
A~*
! ea t
Exchange
i
».
L
l>
•
ean
T i
V
0
4-1
td
c
-------
Figure 4-5.
Flow
Glaus
Tall Gas
o
1'lnnt
fuel C.13
o >
»
O
Coir.busLlon Air
In Lino
II cat c r
RunctoiS
S loam
-C26
Tnil Gn.i
To Incinerator
Kc;
Absorber
Qucnci
sure
— >_
i
rr
v^ J L
" —
~^--,
it Cooler
^'
L
-< — '
*
-er
u
s
J
a\/-'i^j"^
C\ i "
()ucnch
Hater to
Sour Uatcr
Lean Solvent
: Drum
cw
/^^\
Lou I'n
-------
4.2.2.4 ARCO Process.15 Conceptually similar to the SCOT process
described above, the ARCO process is based upon amine absorotion of H2S
and recycle to the Claus plant. Design performance levels of 250 ppmv
S02 in the incinerated absorbed off-gas have been common for the ARCO
process. Figure 4-5 is a representative ARCO process scheme. It has
been installed in U.S. refineries since 1975.
4.2.2.5 Beavon/MDEA.17 A recently announced option to the Beavon
process previously described is substitution of the Stretford sulfur
recovery plant with an amine absorber/regenerator with H2S recycle to
the Claus similar to the SCOT and ARCO orocesses. A representative
schematic is not presented here, but it is assumed similar to the SCOT
and ARCO processes, with associated performance guarantees. The 3eavon/MDEA
uses .ethyldiethanol amine (MDEA), a tertiary amine, which is more selective
for H2S than the secondary amines frequently used in amine tail gas
-ocesse,. Also, the licensors prefer to generate all needed h^roaen in
MP reducing aas generator, obviating an external source of hydrogen.
4.3 COMMERCIAL STATUS OF EMISSION CONTROLS FOR REFINERY SULFUR PLAHTSlS,19,20,21,22
The first comnercial tail gas treater installed in 1972 in a IJ S
refinery was the Wellman-Lord process. The Beavon, SCOT and IFP-l' '
processes were installed at U.S. refineries the following year. In 1975
the first ARCO process was installed. Since 1976, when the MSPS for
refinery sulfur plants was announced, all sulfur plants subject to the
MSPS have chosen the SCOT, Beavon, or the ARCO processes, although one
non-MSPS Wellman-Lord unit was installed in 1981. Table 4-1 lists the
tail gas units installed in U.S. refineries as of 1982. Units planned
or under construction are listed in Table 4-2. Each "unit" refers to a
separate tail gas process sequence as shown in Figures 4-1 through 4-5
A unit may serve one or several Claus units. Capacities shown in Table 4-1
are for total Claus capacity served.
4-13
-------
Table 4-1. TAIL GAS TREATERS INSTALLED IN U.S. REFINERIES
-P-
I
Unit
ARCO
ARCO
ARCO
ARCO
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Beavon
Location
California
Texas
Washington
Pennsylvania
California
Cal i fornia
Pennsylvania
Cal ifornia
California
Louisiana
Louisiana
Louisiana
New Jersey
Texas
Texas
111 inois
Louisiana
New Jersey
Texas
Missouri
Indiana
Onstream Date
No. of Units
Total Sulfur Plant
Capacity. Mg/D (LT/D)
1975
1976
1977
1982
1973
1973
1973
1974
1975
1975
1976
1976
1976
1977
1977
1977
1978
1980
1980
1981
1981
1
1
1
1
2
2
1
1
3
3
1
1
1
1
1
2
1
2
1
1
1
— — — i_,— ^.-..jn
182.9
320.1
122.0
172.8
203.2
304.8
142.3
355.7
249.0
312.0
304.8
235.8
304.8
304.8
829.3
304.8
235.8
274.4
101.6
233.8
396.4
-. -" . --
(180)
(315)
(120)
(170)
(200)
(300)
(140)
(350)
(245)
(307)
(300)
(232)
(300)
(300)
(816)
(300)
(232)
(270)
(100)
(230)
(390)
-------
Unit
Beavon
Beavon
Beavon
Beavon
IFP-1
IFP-1
IFP-1
IFP-1
IFP-1
IFP-1
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
Location
Cal 1 form' a
Cal i form' a
Louisiana
Cal i form" a
Texas
Texas
Texas
Cal i form' a
Texas
Texas
Cal i form' a
Cal i form' a
Pennsylvania
Michigan
Okl ahoma
Louisiana
Texas
Louisiana
Texas
Texas
Okl ahoma
Table 4-1. (Continued)
Onstream Date
1981
1981
1982
1982
1973
1973
1976
1976
1976
1977
1973
1973
1974
1975
1975
1975
1975
1976
1976
1977
1977
No. of Units
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Total Sulfur Plant
Capacity. Mg/D (LT/D)
122.0
152.5
203.2
39.6
45.7
45.7
101.6
182.9
406.4
254.1
15.2
35.6
162.6
81.3
29.5
42.7
318.1
43.7
196.2
152.5
63.0
(120)
(150)
(200)
( 39)
( 45)
( 45)
(100)
(180)
(400)
(250)
( 15)
( 35)
(160)
( 80)
( 29)
( 42)
(313)
( 43)
(193)
(150)
( 62)
-------
Unit
en
Location
Table 4-1. (Continued)
Onstream Date
No. of Units
Total Sulfur Plant
Capacity, Mg/D (LT/D)
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
Texas
Pennsylvania
Louisiana
Cal ifornia
Illinois
Wyoming
Texas
Ohio
Ohio
Cal ifornia
Cal ifornia
Kentucky
Texas
Louisiana
Louisiana
Louisiana
Texas
Al abama
Texas
Al abama
1977
1978
1979
1979
1979
1980
1980
1980
1980
1980
1981
1981
1982
1982
1982
1982
1982
1982
1982
1982
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
•-
233.8
46.8
152.5
10.2
457.4
50.8
381.1
122.0
101.6
7.4
177.9
203.2
115.9
127.0
61.0
8.1
14.2
40.7
18.3
53.9
(230)
( 46)
(150)
( 10)
(450)
( 50)
(375)
(120)
(100)
( 7.3)
(175)
(200)
(114)
(125)
( 60)
( 8)
( 14)
( 40)
( 18)
( 53)
-------
Table 4-1. (Continued)
Unrt
Well man-Lord
Well man-Lord
Well man-Lord
Well man-Lord
Well man-Lord
Location
Onstream Date
Total Sulfur Plant
Cal i form' a
Cal i form' a
Cal i form' a
California
Cal i form' a
1972
1975
1976
1977
1981
i«u . u 1 Ull 1 LS
1
1
1
1
1
bdpdC It
457.4
330.3
304.8
330.3
203.2
y, Mg/u ur/D)
(450)
(325)
(300)
(325)
(200)
-------
Table 4-2. REFINERY TAIL GAS UNITS PLANNED OR UNDER CONSTRUCTION
Unit
Location
Onstream Date
No. of Units
Total Sulfur Plant
Beavon/MDEA
Beavon/MDEA
Beavon
Beavon
Beavon
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
SCOT
Louisiana
Louisiana
Alaska
Kansas
Cal i form' a
Tennessee
Texas
Cal i form' a
Texas
Texas
Minnesota
Washington
Texas
Cal i form' a
Del aware
Louisiana
Texas
Louisiana
Ohio
Texas
Texas
1983
1983
--
—
—
1983
1983
1983
1983
1983
1983
1983
1983
1983
1984
1984
1985
—
--
—
—
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
i
365.9
233.8
229.7
10.5
30.5
91.5
4.6
304.8
252.1
255.1
304.8
50.8
79.3
66.1
241.7
132.1
1,016.4
38.6
32.5
162.6
177.9
•j ^ ^i ' » — •/—»
(360)
(230)
(226)
( 10.3)
( 30)
( 90)
( 4.5)
(300)
(248)
(251)
(300)
( 50)
( 78)
( 65)
(235)
(130)
(1,000)
( 38)
( 32)
(160)
(175)
-------
As shown in these tables, there are 76 reported tail gas treaters
operating in domestic refineries with an additional 24 units planned or
under construction. These figures do not account for units that have
been replaced or are currently inoperative.* Total sulfur plant capacity
controlled by these units is 12,514 Mg/D with an additional 4,109 Mg/D
Planned or under construction. Thus, the average tail gas treater currently
operating handles 165 Mg/D of Cl aus plant capacity, while planned units
average 179 Mg/D Glaus capacity.
*Also not included is a hybrid 34.9 Mg/D tail gas unit which is not
commercially available.
4.4 REFERENCES
process-
of Canada- IK- to
' Claus
C. s«tan. U.S. EPA,
2e Claus
Ap'n'/is! W?r ' "Vl'Slt t0 IFP Sulfur Recm^ ""If. C. Se
-------
11. Reference 3. p. 4-25.
12. Reference 2. p. 4-5.
13. Reference 3.
14. Reference 2. pp. a-14 and 4-15.
15. Kuijpers, N.S.M.J., "The Shell Off-Gas Treating Process" - presented
at the Gas Sweetening and Sulfur Recovery Seminar, Amsterdam, The
Netherlands. November 9-13 , 1981.
16. Trip Report - "Visit to ARCO Refinery, Pasadena, Texas", Charles Sedman,
U.S. EPA. September 20, 1982.
17. Reference 2. n. 4-3
18. Letter, '4. T. Knowles, Shell Oil Companv, to Charles 3. Sedman,
U.S. EDA. August 24, 1982.
19. Letter, M. A. Peterson, Union Oil Companv of California, to
Charles 3. Sedman, U.S. EPA. September 15, 1982.
?:). "Survey Report on SO? Control Systems for Non-Utility Combustions
and Process Sources - May 1977", prepared by PEDCo Environmental, Inc.
Contract Mo. 63-02-2603.
21. Letter, H. J. Grimes, ARCO Petroleum Products Co. to C. Sedman,
U.S. EPA, dated October 5, 1982.
22. Letter, D. H. Oil worth, Davy McKee, to C. Sedman, U.S. EPA, dated
October 5, 1932.
4-20
-------
5. COMPLIANCE STATUS OF REFINERY SULFUR PLANTS
5.1 AFFECTED FACILITIES
Of the 43 sulfur plants constructed during the period 1977-1982 in
domestic petroleum refineries, only 17 are subject to the sulfur plant
NSPS. Of the 26 non-MSPS units, only 4 were exempted due to size (less
than 20 long tons per day capacity). The remaining 22 units were contracted
for prior to October 4, 1976, and were "grandfathered" as an existing
facility at the time of NSPS proposal.
Of the 17 units subject to the NSPS, 7 are in start-up and hav, not
oeen compliance tested. Emission test results fro* the 10 certified NSPS
facilities are presented and discussed in the followina section Unless
otherwise noted, all results are based on three test runs using EPA
methods discussed in Chapter 3.
5.2 COMPLIANCE TEST RESULTS
5-2-1 Deduced Sulfur and Hydrogen Sulfide1,2,3,4
As discussed in Chapter 3, reduced sulfur compounds and hydrogen
sulfide limits are enforced wherever a reduction tail gas system is used
and the tail gas not incinerated after treatment. Four Beavon tail gas
units are operating under these restrictions, and the compliance test
results are summarized in Table 5.1.
Table 5.1 illustrates the effectiveness of the Beavon process
especially the Stretford H2S absorber. Of the four units tested all ar*
TH compliance, being well under the 300 ppmv reduced sulfur and 10 opmv
H2S restrictions. Typically, the only measurable sulfur compound oresent
in Seavon exhaust gases is carbonyl sulfide (COS).
5-2.2 Sulfur Dioxide5,6.7,8,9,10
Units which incinerate tail gases are subject to sulfur dioxide
Units of 250 ppmv dry basis, corrected to zero oercent oxyaen Six SCOT
treaters which incinerate tail gas after treatment are currently operating
under these rules, and the associated emission test results are'present-d
in Table 5-2.
5-1
-------
Table 5-1. NSPS COMPLIANCE TEST RESULTS FOR
REDUCED SULFUR
-------
As shown Tn Table 5-2, the SCOT emissions are somewhat higher than for
Beavon units, and somewhat less predictable, reflecting the effect of
process conditions upon the amine absorbers. Of the six units tested,
average emissions range from approximately 100 to 200 ppmv S02. All six
units are in compliance.
These short-term tests represent the only emission data gathered
during this study. Although S02 and reduced sulfur monitors are generally
installed on these NSPS units, data are not recorded and reported to
agencies and are, therefore, not available for analysis.
5.3 OPERABILITY OF NSPS UNITS1-0,!! ,12,13,14,15,16
Through EPA and API surveys, a total of 7 NSPS and 16 non-NSPS
-fineries responded to questions concerning operabillty and maintenance
oroblems encountered in tail gas treaters.
^•rom the surveys, it is evident that most problems in tail gas
heaters are preceded by upsets in the Glaus plant, which can send
excessive amounts of either S02 or H2S into the tail gas reactor. For an
anine tail gas system, unchecked breakthrough of S02 tlrough the reactor
into the absorber causes no immediate excess emission because the amine
combines irreversibly with S02. However, permanent loss of solution
activity ensues, the solution becomes corrosive, and requires discarding
A breakthrough of H2S beyond the design capacity of the absorber causes
excess emissions of H2S, but solution performance returns to normal as
soon as the breakthrough is stopped.
A short-term breakthrough of S02 into the Stretford system causes no
excess emissions because the Stretford solution also reacts irreversiblv
with S02 causing an increase in chemical consumption and more frequent "
system purge. The same is true of short-term H2S overloads above design
capacity, but prolonged overloads cause tower plugging and adverselv
affect Stratford chemicals which may take several days to return to"
normal operation.
5-3
-------
The above helps to explain the survey results which show:
0 older, non-NSPS units to be more reliable than MSPS units
(increased reliability with system age)
0 most problems directly attributable to SO? breakthrough
Common problems reported for anine systems included excess solvent
foaming, quench water filter plugging, quench column level control, and
catalyst bed plugging. Less frequent problems included heater tube leaks,
oump failures, and blower failures, all of which appear unrelated to the
nrocess itself.
Similar reactor and quench tower problems were reported for Stretford
units, along with the less routine pump, compressor, and heat exchanger
failures. Additionally, the Stretford portion of some units using direct
nelting of sulfur slurry has caused less severe, hut more consistent,
maintenance probes. Plugging of lecanter and me! ters along with general
solids accumulation have been reported.
Generally, the survey indicates the most important factor in successful
tail gas plant ooeration is experience. For units with more than 3 years
operating experience (mostly non-MS^S units), system reliabilities approach
100 percent in many cases. Both amine and Stretford units received praise
from operators. However, the vast majority of problems and somewhat less
enthusiastic responses to the survey came from NSPS units.
Most SOj and HoS breakthrough-related problems (quench tower plugging
and corrosion, high chemical consumotion) appear corrected by closer
attention to the built-in safeguards in tail gas treaters. The alkaline
guard (quench tower pH control) and level control should alleviate most
downstream corrosion, plugging, and chemical degradation problems.
Qoerating at a H?_S:S02 ratio slightly above 2 to 1 allows for a greater
margin of operating error without irreversible loss of solution activity
or onset of corrosion problems.
Reactor problems appear due tc the introduction of unsaturated
hydrocarbons via fuel gas to the heater and should be alleviated by better
quality control of fuel.
Degradation of amines and excess foaming have been alleviated by
installation of carbon absorption units and use of anti-foaming agents.
5-4
-------
Stretford problems involving plugging and solids accumulation have
been alleviated by replacement of level controllers and more operator
attention. Stretford solutions outfitted to filter and rinse sulfur
before melting have been more successful, and the licensor is exclusively
using filters in new plants under design.17
5.4 STATUS OF EMISSION MONITORS18
5.4.1 SO? Monitors
Where incinerators are used to oxidize tail gas, sulfur dioxide
monitors have been installed on all new units surveyed. Practically all
existing tail gas installations with incinerators also use S02 monitors.
Both in-stack and extractive type S02 monitors, identical to those found
on hoilers, are currently operating. Problems encountered are similar to
t'tose on boilers, and include:
Pegged sampling lines on extractive systems
orobe failures on extractive svstems
sample conditioning system on in-stack monitors
factory servicing of in-stack monitors
Most in-stack monitors installed prior to 1980 performed very noorly
in field applications and required reservicing at the factory or replacemPnt
with more durable instrumentation. Vendors have also made improvements
in sample extraction and conditioning components, as evidenced by the
inoroved reliabilities reported by more recent installations.
Extractive monitors have experienced initial problems with the
sampling lines and probes. Installation of probe shields and higher
pressure backflush systems in sample lines have alleviated these problems.
5-4.2 Reduced Sulfur and H?S Monitors
Reduced sulfur monitors are relatively new and were found on onlv
two operating facilities. In both cases, the systems were reported as
unsatisfactory due to high maintenance and poor operability. Problems
encountered include probe and sample line plugging, and several failures
of the computer software which required reprogramming.
Hydrogen sulfide monitors are generally the lead acetate tape monitors
which are used in conjunction with an H2S alarm system tied to a standbv
mcinerator. As such, these monitors are more qualitative than quantitative
and would not meet stringent performance criteria. Problems reported are
minimal and often were due to lack of periodic maintenance.
5-5
-------
5.5 EMISSION TESTING
One small consideration should be noted with regard to EPA Method 15—
determination of reduced sulfur compounds. Most recent emission tests
have been performed using a modified EPA Method 15, where acetate buffer
and improved chromatographic separation columns have simplified the sample
conditioning requirements of Method 15.1-9
5.6 REFERENCES
1. Letter, R. T. Denbo, Exxon Company, U.S.A., to Don R. Goodwin,
U.S. EPA, dated June 11, 1982.
?. Letter, G. E. Lowe, Marathon Petroleum Company, to Don R. Goodwin,
U.S. EPA, dated September 17, 1982.
3. letter, R. J. N'iederstadt, Mobil Oil Corooration, to Don R. Goodwin,
'J.S. EPA, dated June 15, 1982.
Letter, Steven Feeler, Missouri Department of Natural Resources, to
'I. 1. Gednan, U.S. EPA, dated September 24, 1982.
5. Letter, C. M. Tyler, SOHIO, to Don R. Goodwin, U.S.. EPA, dated
July 15, 1Q82.
6. Letter, J. P. Gay, Ashland Petroleum, to Charles 8. Sedman,
U.S. EPA, dated September 27, 1982.
7. Letter, 8. F. Ballard, Phillips Petroleum, to Don R. Goodwin,
U.S. EPA, dated July 13, 1982.
8. Letter, Richard Grusnick, Alabama Department of Environmental
Management, to Charles B. Sedman, U.S. EPA, dated October 15, 1982.
3. Letter from G.J. Vetter, GHR Eneray Corporal ton, to C. Sedman,
U.S. EPA, dated January 28, 1983.
10. Letters, E. P. Crockett, American Petroleum Institute, to Charles 3.
Sedman, U.S. EPA, dated June 15, June 30, and July 14, 1982.
11. Reference 1.
12. Reference 2.
13. Reference 3.
5-6
-------
14. Reference 6.
15. Reference 8.
,1,6: ^tter' L- M- Lovell, Amoco Oil Company, to Don R. Goodwin
U.S. EPA, dated June 23, 1982.
13. References 11-17.
19
19 T^ePho^°nv^sation B Ferguson, Harmon Engineering and Testing,
inc., to c. Sedman, U.S. EPA, dated November 18, 1982.
5-7
-------
6. MODEL PLANTS AND COST ANALYSES
This chapter defines model plants which represent typical refinery
sulfur plant alternatives for new installations and presents estimated
costs of those alternatives.
6.1 MODEL PLANTS
In order to have a common basis for comparing costs of emission
controls to meet the existing NSPS, model plants are selected. Resource
requirements, dollar costs, and environmental impacts are then determined
for each model plant. From these assessments, the relative impact and
appropriateness of NSPS for various size sulfur plants may be weighed.
6.1.1 Model Plant Size
In Chapter 4 it was shown that sulfur plants constructed with tail
gas treaters since 1972 have ranged from 7.4 to 457.4 megagrams per day
(Mg/D) capacity. Actual individual sulfur plants up to 400 Mg/D have
been constructed. Planned tail gas units range from 4.6 to 1,016 Mg/D
with single Cl aus plants of up to 508 Mg/D forecasted. Tail gas units'
constructed in the United States have been either the extended Claus
systems (IFP) or add-on absorbers (Wellman-Lord, SCOT, Beavon, or ARCO)
All planned tail gas units are essentially the reduction/absorption type
with the amine scrubbing variation representing the majority choice.
For the economic modelling and comparisons, Claus plants at 10.16,
50.8, and 101.6 Mg/D have been selected for model analyses.
6'1'2 Choice of Representative Control System
The NSPS control cases are represented by the reduction/amine
absorption process for simplicity. Al thoug-h the oxidation (Wellman-
Lord) system is clearly an alternative, the reduction systems have been
the overwhelming choice for NSPS Claus plants. The Beavon-Stretford
process has certain advantages over the amine (SCOT/ARCO/Beavon-MDEA)
systems with respect to increased size and decreased H2S content in
Claus feed; however, for typical refinery applications in the 10 to
100 Mg/D range, amine systems are the majority (18 of 20 operating units)
ch01ce for new installations (see Appendix A, pg. A-3 for more discussion)
6-1
-------
6.1.3 Assumptions of Modelling Parameters
Table 5-1 presents process parameters of model plants chosen. These
model plants were developed using reported process data from MS PS plants,
technical data from vendors, and previous studies of sulfur recovery
plants by EPA.1.2 Details of each model are discussed in Appendix A.
All cases handle acid gas consisting of 80 percent hydrogen sulfide,
10 percent carbon dioxide, 4.5 percent ammonia, 0.5 percent hydrocarbons,
and 5.0 percent moisture. The acid gas streams are assumed saturated at
42.9°C (109°F) and 170 kilopascals (24.7 psia). Sour water streams
containing the bulk of hydrocarbons and all ammonia are completely
combusted in the first combustion stage, with amine off-gases combusted
in the second stage.
Glaus plants are assumed to use high efficiency alumina catalysts
for maximum sulfur recovery: the 101.6 LT/D case uses two Glaus stages at
95.1 percent recovery, while the 50.8 and 101.6 Mg/D cases use three Glaus
stages at 96.6 percent recovery.
Tail gas units are sized at twice the anticipated feed rate, and
Glaus plants are sized to accomodate the additional recycle stream. For
example, the model plant 3B features a 105.0 Mg/D Glaus plant (101.6 Mg/D
feed, 3.4 Mg/D recycle, 0.1 Mg/D emission rate) and a tail gas unit
sized at 6.8 Mg/D. Since the recycle stream is more dilute with respect
to H2$, the Glaus size (based on gas flow) actually increases by 50
percent in the 3-stage cases and 7.6 percent in the 2-stage case.
All Glaus plants consume 4,300 Kp steam and generate 1,760 Kp and
106 Kp steam, with 3-stage plants also generating 352 Kp steam. Boiler
feedwater is available at 2,255 Kp and 110°C, while cooling water is
available at 29°C and returned at 43°C. Incinerators are designed to
operate at 649°C (1200°F), 25 percent excess air for the Glaus only
cases, and the Glaus/tail gas/incinerator heat recovery case. Incinerators
operate at 538°C (1000°F), 25 percent excess air for tail gas treating
with no incinerator heat recovery. Only for the 101.6 Mg/D case is waste
heat recovery employed at the incinerator.
6-2
-------
Table 6.1. MODEL PLANT PARAMETERS
MODEL PLANT 1A
1. Sulfur Intake: 10.16 Mg/D (10 LT/D)
2. Sulfur recovered: 9.66 Mg/D (95.10% efficiency)
3. Plant description: Glaus furnace, two catalytic stages + incinerator
4. S02 emission rate: 348.6 Mg/Y (384.2 T/Y)
5. Operating schedule: 350 D/Y
MODEL PLANT IB
1. Sulfur intake: 10.16 Mg/D (10 LT/D) + 0.49 Mg/D recycle
2. Sulfur recovered: 10.15 Mg/D (99.90 percent efficiency)
3. Plant description: Glaus furnace, two catalytic stages, one catalytic
reactor, armne absorption and regeneration
incinerator '
4. S02 emission rate: 7.1 Mg/Y (7.84 T/Y)
5. Operating schedule: 350 D/Y
MODEL PLANT 2A
1. Sulfur intake: 50.80 Mg/D (50 LT/D)
2. Sulfur recovered: 49.09 Mg/D (96.64% efficiency)
3- Plant description: Glaus furnace, three catalytic stages + incinerator
4. S02 emission rate: 1,209.4 Mg/Y (1,332.8 T/Y)
5. Operating schedule: 350 D/Y
MODEL PLANT 2B
1. Sulfur intake: 50.80 Mg/D (50 LT/D) + 1.68 Mg/D recycle
2. Sulfur recovered: 50.75 Mg/D (99.90% efficiency)
3. Plant description: Glaus furnace, three catalytic stages one catalytic
"ncinerator1'"6 abs°rptfon and ^generation, y
6-3
-------
Table 6.1. MODEL PLANT PARAMETERS (continued)
4. S02 emission rate: 35.56 Mg/Y (39.20 T/Y)
5. Operating schedule: 350 0/Y
MODEL PLANT 3A
1. Sulfur intake: 101.6 Mg/D (100 LT/D)
2. Sulfur recovered: 98.15 Mg/D (99.64 percent efficiency)
3. Plant description: Glaus furnace, three catalytic stages, incinerator
with heat recovery
*. S02 emission rate: 2,418.9 Mg/Y (2,665.5 T/Y)
5. Operating schedule: 350 D/Y
Mnnn_ PLAMT 3^
1. Su1*'.ir intake: 101.fi Mg/D (100 LT/D) + 3.35 Mg/D recycle
?. Sulfur recovered: 101.5 Mg/0 (99.90% efficiency)
3. Plant description: Glaus furnace, three catalytic stages, one catalytic
reactor, amine absorotion and regeneration,
incinerator with heat recovery
4. SO? emission rate: 71.12 Mg/Y (78.40 T/Y)
5. Operating schedule: 350 D/Y
5-4
-------
With waste heat recovery, 600 psig steam is also generated, while tail
gas treatars are net consumers of 50 psig steam. Complete utility
consumption and generation balance sheets are presented in Appendix A
to this document.
6.2 CONTROL LEVELS
Basically, the control levels are represented by the two sulfur
recovery levels currently achieved in actual practice-96.6 percent
recovery or control for the basic 3-stage Claus with alumina catalysts
and 99.9 percent recovery for 3-stage Claus with state-of-the-art tail
gas controls represented by amine absorption/recycle processes, (^or
2-stage smaller sulfur plants, 95.1 percent recovery is achieved with a
oroportionally larger tail gas system to achieve 99.9 percent overall
control.) Henceforth the Claus-only case will be referred to as baseline
control and the Claus and tail gas treatment as MSPS control.
3.3 COST ANALYSIS
The model plants described in Section 6.1 were analyzed for economic
impacts of controls by estimating fixed capital costs, annualized costs,
emission reductions, and cost-effectiveness of controls. THe estimates'
are based upon previous sulfur plant studies and the data from actual new
installations as gathered by EPA specifically for this study. Detailed
cost analyses are presented and discussed in Apoendix A to this report.
o.3.1 Assumptions
Fixed capital costs were estimated from an analysis of capital cost
data furnished by individual operating plants and eauipment vendors. The
range of operating variables examined were so great that a composite
model facility was selected with distinct modelling and economic assumptions
Modelling assumptions were presented in Table 5.1. Table 5.2 lists key
economic assumptions used to determine representative annualized costs.
The most difficult economic parameter to gauge is the assignment of
maintenance and repair costs. Previous studies have used vendor projections
of maintenance costs at 3.5% of fixed capital costs;l,2 wniie the background
document to the original MSPS estimated maintenance costs at 3 percent of
fixed capital for tail gas treaters.3
6-5
-------
Table 6-2. ECONOMIC ASSUMPTIONS USED TO CALCULATE ANNUALIZED COSTSa
I. Util ity prices:
1. 4,300 Kp steam S15.98/Mg (57.2571,000 Ib)
2. 1,760 Kp steam $14.88/Mg ($6.7571,000 Ib)
3. 352 Kp steam $12.68/Mg ($5.75/1,000 Ib)
4. 106 Kp steam $ 9.92/Mg ($4.50/1,000 Ib)
5. boiler feedwater $ 3.31/Mg ($1.50/1,000 Ib)
6. steam condensate $ 2.76/Mg ($1.25/1,000 Ib)
7. cool ing water $13.21/103m3 ($ .05/1,000 gal )
8. catalyst:
a. alumina $500/m3 ($17/ft3)b
b. cobalt-molybdenum (Co/Mo) S5,000/m3 ($170/ft3)b
9. Chemicals:
a. diisopropanolamine $0.49/Kg ($1.07/lb)c
b. soda $330.6/Mg ($300/ton)c
10. fuel gas $3.64/109/J ($3.50/106 Btu)d
11. electric power $0.05/KWH
12. sulfur $9&.42/Mg ($125/LT)e
II. Labor (8,720 hours per year basis)
1. operators: ($14.50/hr)
2/3 per shift for Claus
2/3 per shift for tail gas treater
2. supervision: (S18,80/hr)
1/4 per shift for sulfur recovery facility
III. Maintenance and Repair
Labor and materials: 3.0 percent of fixed capital
Costsc
IV. Other Miscellaneous Costs
1. Operating supplies: 10 percent of operating labor
2. Laboratory charges: 10 percent of operating labor
V. Fixed Charges
1. Capital charges = fixed capital costs x
= .13147 for n = 15 years, i = 10 percent
= .171059 for n = 15 years, i = 15 percent
= .213821 for n = 15 years, i = 20 percent
2. Local taxes - 1 percent of fixed capital costs
3. Insurance - 0.6 percent of fixed capital costs
6-6
-------
Table 6.2. ECONOMIC ASSUMPTIONS USED TO CALCULATE ANNUALIZED COSTS^ (continued
VI. Overhead
1. plant overhead - 25 percent of operating labor + 25 percent of
maintenance and repair
2. administrative - 1 percent of annualized costs
3. distribution and marketing - 1 percent of annualized costs
values assigned from Reference 1 unless otherwise
Reference 2 C°nSUmptl'0n fl^ures for ™del P^nts 'rom EPA survey and
b Telephone conversation with Mr. R. E. Warner of Ralph M. Parsons Co.
"60. j. j 1983. ^ '
c Chemical Market Reporter. October 4, 1982.
'1 Mej,iorand,«n: R. E. Jenkins to C. B. Sedman, EPA, dated September 7,
3 Average of EPA survey.
5-7
-------
Actual maintenance costs gathered by EPA for this study showed Glaus
costs ranging from 2.3 percent to 6.1 percent of fixed capital costs and
2.1 to 6.3 percent of fixed capital for reduction-based tail gas units.
Estimates chosen for this study estimated maintenance costs at 3.0 percent
for all cases, corresponding to the average of actual data based on data
submitted by operators.
Other assumptions presented in Table 6.2 generally agree with previous
studies, except that cost of chemicals, utilities, and labor have been
indexed to current levels. (See footnotes, Table 6.2.)
6.3.2 Results of Cost Comparison
Table 6.3 presents the line -tern cost estimates for the models
discussed in Section 6.1 for interest rates of 10, 15, and 20 percent.
Table 6.4 compares the costs, pollutant removal rates, and cost-effectiveness
of control as expressed in dollars per ton of sulfur dioxide removed.
All discussion herein will assume a 10 percent interest rate.
Table 6.3 demonstrates the economics of scale of sulfur plant
operations. Generally, the most important cost, that of the cost of
capital , increases fractionally with increased size.
Maintenance and repair, plant overhead, and other nonl abor operational
costs show similar economics of scale, while direct labor costs are
practically fixed regardless of plant size. Labor is, however, related
to the number of unit operations controlled; therefore, addition of a
tail gas treater effectively doubles the labor requirement.
Credits for steam, condensate, and sulfur play a large role in
determining the economic viability of a sulfur plant. Since these credits
are a direct function of plant size (for a given H2S/C02 acid gas feed),
the profit margin is heavily favored for increasing plant size.
Table 6.4 illustrates that d 10.16 Mg/D plant operates at a deficit
even without tail gas controls. Tail gas controls turn a highly profitable
50.8 Mg/D plant into a break-even venture, while at 101.6 Mg/D, the tail gas
treater halves the profits, but the system still returns a substantial
annual surplus.
6-8
-------
Cost-effectiveness of tail gas control indicates a similar trend,
showing typically $2,125 per Mg SOe cost at 10.16 Mg/D, $880/Mg at
50.8 Mg/D, and $675/Mg at 101.6 Mg/D. Interpolating these figures to the
current NSPS cutoff at 20.32 Mg/D indicates that the maximum cost per
megagram currently incurred (in 1982 dollars) is about $l,430/Mg, while
the more typical cost of a new facility greater than 100 LT/D is considerably
less than $900/Mg. (See Figure 6-1.)
6-9
-------
Table 6.3. LINE ITEM COSTS FOR MODEL PLANTS
MODEL 1A (10.16 Mg/d)
Capital cost - $2.54 x
Direct operating cost
A. Utilities & Chemicals
1. 4,300 Kp steam
2. treated boiler feedwater
3. electric power
4. fuel gas
5. catalyst
B. Labor
1. Operators
2. Supervision
C. Maintenance and Repair
D. Supplies and laboratory charges
Fixed Charges:
A. Capital
B. Taxes
C. Insurance
PI ant Overhead:
General Expenses
A. Administrative
B. Distribution and sales
Total Annuali zed Costs
Credits
1. 1,960 Kp steam
2. 106 Kp steam
3. steam condensate
4. sulfur
Total Credits
Net Annual Operating Cost for Case 1A
1 = 15%
$ 6,395
21,615
21,210
17,640
655
$84,680
41,170
$76,200
$16,940
$434,490
25,400
15,240
$40,220
$ 8,020
$ 8,020
$817,895
$ 87,320
5,670
8,558
399,420
$499,265
$320,439
i = 10%
$ 6,395
21,615
21,210
17,640
655
$84,680
41,170
$76,200
$16,940
$333,960
25,400
15,240
$40,220
7,160
7,160
715,645
$ 87,320
5,670
8,558
399,420
$499,265
$218,189
i = 20%
$ 6,395
21,615
21,210
17,640
655
$84,680
41,170
$76,200
$16,940
$543,105
25,400
15,240
$40,220
9,100
9,100
928,670
$ 87,320
5,670
8,558
399,420
$499,265
$431,214
6-10
-------
Table 6-3. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL IB (10.16 Mg/d)
Capital Cost - $4.96 x 106
Direct operating cost
A. Utilities & Chemicals
1. 4,300 Kp steam
3. Labor
C. Maintenance 4 Repair
D. Supplies & Lab Charges
Fixed Charges
8.'
General Expenses
A. Administrative
3. Distribution ,«, s.!.,
TOW Annual fzed Costs
Credits
3. st.» condensate
i -
j 7jl25
sis
$ 7,125
5S
?
'is
i = 20%
5 7,125
S148,800 $148,800 S148,800
$33,870 $ 33,870 j 33,870
652-140 1.060,545
79,540
1 /i ^m
}J;||g
79,540
}|;«0
H.690.065 1,439,945 1,906,550
"".580 419,580
5547,465 1547,465
5547,465
Net Annual Operating Cost for Case IB $1,143,600 $942,480 $1,359,085
6-11
-------
Table 6.3. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 2A (50.8 Mg/D)
Capital Cost - 54.33 x 106
Direct Operating Cost i = 15% i = 10% i = 20%
A. Utilities & Chemicals •
1. 4,300 Kp steam $ 53,290 $ 53,290 $ 53,290
2. treated boiler feedwater 155,310 155,310 155,310
3. electric power 52,500 52,500 52,500
4. fuel gas 88,200 88,200 88,200
5. catalyst 4,005 4,005 4,005
B. Labor
1. Operators 84,680 84,680 84,680
2. Supervision 41,170 41,170 41,170
C. Maintenance and Repair 129,900 129,900 129,900
D. Supplies and Lab Charges 16,940 16,940 16,940
Fixed Charges
A. Capital 740,690 569,310 925,840
B. Taxes 43,300 43,300 43,300
C. Insurance 25,980 25,980 25,980
Plant Overhead 53,645 53,645 53,645
General Expenses
A. Administrative 15,000 13,300 16,850
B. Distribution and Sales 10,000 13,300 16,850
Total Annual ized Costs $1,519,610 1,344,830 1,708,460
Credits
1. 1,960 Kp steam $425,250 $425,250 $425,250
2. 352 Kp steam 15,940 15,940 15,940
3. 106 Kp steam 23,435 23,435 23,435
4. steam condensate 46,200 46,200 46,200
5. sulfur 2,028,600 2,028,600 2,028,600
Total Credits $2,539,425 2,539,425 2,539,425
Net Annual Operating Cost for Case 2A ($1,019,815) (1,194,595) (830,965)
6-12
-------
Table 6.3. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 28 (50.8 Mg/D)
Capital Cost - $7.83 x 106
Direct Operating Cost f
A. Utilities & Chemicals L
1. 4,300 Kp steam $
B. Labor
1. Operators
C. Maintenance & Repair
D. Suppl ies 5 Lab Charges
Fixed Charges
B.' Tails'1
C. Stance
Plant Overhead
General Expenses
A. Administrative
8. Distribution *M.,
Total Annual ized Cost
Credits
Net Annual Operating Cost for Case 28
1 = m
*.
i
•
:
169,360
82,340
234,900
33,870
IbIS
234,900
33,870
lil:l%
234,900
33,870
1.339,400 1,029,490 1,674,210
JfsBO 2'SS 78'3°°
46,980 46,980 46,980
1m nfic
101,065
,_, n^r
101,065
0/1 ccn
24,650
101,065
31,100
2,842_150
52,684,560 52,684,560 $2,684,560
$ 157,590 ($158,520) $ 499,100
6-13
-------
Table 6.3. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 3A (101.6 Mg/D)
Capital cost - $6.26 x
Direct Operating Cost i = 15% 1 = 10% i = 20%
A. Utilities & Chemicals ~ -
1. treated boiler feedwater $402,575 $402,575 $402,575
2. electric power 89,040 89,040 89,040
3. fuel gas 176,400 176,400 176,400
4. catalyst 8,010 8,010 8,010
B. Labor
1. Operators 84,680 84,680 84,680
2. Supervision 41,170 41,170 41,170
C. Maintenance & Repair 187,800 187,800 187,800
D. Supplies & Lab Charges 16,940 16,940 16,940
Fixed Charges
A- Capital 1,070,835 823,065 1,338,515
B. Taxes 62,600 52,600 62,600
C. Insurance 37,560 37,560 37,560
Plant Overhead 68,120 58,120 68,120
General Expenses
A. Administrative 22,460 19,980 25,135
B. Distribution & Sales 22,460 19,980 25,135
Total Annual ized Costs $2,290,650 2,037,890 2,563,680
Credits
1. 4,300 Kp steam 280,140 280,140 280,140
2. 1,960 Kp steam 92,460 92,460 92,460
3. 352 Kp steam 291,730 291,730 291,730
4. 106 Kp steam 46,870 46,870 46,870
5. steam condensate 35,910 35,910 35,910
6. sulfur 4,057,200 4,057,200 4,057.200
Total Credits $5,604,310 5,604,310 5,604,310
Net Annual Operating Cost for Case 3A ($3,313,660) ($3,566,420) ($3,040,630)
6-14
-------
Table 6.3. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 3B (101.6 Mg/D)
Capital cost - $10.60 x 106
Direct Operatinq Cost ,• 15. .
A. Utilities i Chemicals —I-i£i 1 = 10%. t = 20%
ii i:^" if II
7- ~" 37:91o 1;^ 1;95880°
Labor
C. Maintenance a Repair 318,000 318,000 318,000
0. Supplies* Lab Charges 33,870 33,870 33,870
Fixed Charges
1'?J,3'235 1,393,690 2,266,490
C. Insurance ^'^ ™> 106,000
63,600 63,600 63,600
Plant Overhead ,,. Q/in
121,840 121,840 121,840
General Expenses
A. Administrative ™ ?(-n .c ccn
B. Distribution » Sales $™ ^5,550 44,280
Total An™,, zed Cost $4,088>745 3j660_800
Credits
1. 4,300 Kp steam
4 st- densate « « 4838!
".^5,800 4,195.800 4,195.800
S S5.642.61S 5,642,615 5,642,615
Net Annual Operating Cost for Case 3B ($1,553,870) (Jl,981,815) (51,091,555)
6-15
-------
Table 6.4. COST & COST-EFFECTIVENESS OF NSPS CONTROLS
1 = 10 percent
Base Case Annual Cost, $
Base Case S02 Removed, tons/yr
NSPS Case Annual Cost, $
NSPS Case S02 Removed, tons/yr
Cost-Effectiveness, S/ton
Plant Size, I.T/D
10
218,189
6,765.74
$942,480
7,107.22
$2,126
5U
($1,194,595)
34,362.3
($158,520)
35,536.1
$882
1UU
($3,566,420)
68,724.5
($1,981,815)
71,072.2
$674
i = 15 percent
Base Case Annual Cost, $ 320,439
Base Case S02 Removed, tons/yr 6,765.74
NSPS Case Annual Cost, $ $1,142,600
NSPS Case S02 Removed, tons/yr 7,107.22
Cost-Effectiveness, $/ton $2,413
;$1,019,8151
34,362.3
$157,590
35,536.1
$1,002
($3,313,660)
68,724.5
($1,553,870)
71,022.2
$749
i = 20 percent
Base Case Annual Cost, $ $431,214 ($830,965)
Base Case S02 Removed, tons/yr 6,765.24 34,362.3
NSPS Case Annual cost, $ $1,359,085 $499,100
NSPS Case S02 Removed, tons/yr 7,109.22 35,536.1
Cost-Effectiveness, $/tcm $2,723 $1,133
($3,040,630)
68,724.5
($1,091,555)
71,072.2
$829
6-16
-------
10,000 ., .
Figure 6.1. Cost-Effectiveness of NSPS Control
Sulfur Plant Size, Mq/D
5S
100,
0
6-17
-------
5.4. REFERENCES
1. "SO? Emissions in Natural Gas Production Industry—Background
Information for Proposed Standards," EPA 450/3-82-023a, January 1983
Chapters 6 and 8. '
2, Sulfur Recovery Study - Onshore Sour Gas Production Facilities
Ralph M. Parsons Company, August 20, 1981. '
3. Standards Support and Environmental Impact Statement Volume 1:
Proposed Standards of Performance for Petroleum Refinerv Sulfur
Recovery Plants, U.S. EPA, September 1976. Chapter 3.
6-18
-------
7. OTHER IMPACTS REVIEWED
7.1 NON-AIR ENVIRONMENTAL IMPACTS
7-1.1 Water Pollution Impact
Of the control technologies examined which can meet NSPS requirements,
mtle if any impact upon water quality is foreseen. The amine absorption/
regeneration processes generate significant quantities of reusable process
water normally filtered and sent to the sour water stripper. Only if
significant S02 breakthrough occurs, does the water form soluble sul fates
and sulfites, in which case the water may be sent to the plant water
treatment facility. For an integrated refinery, this would represent
substantially less than 1 percent of total water treated. It is presumed
that this condition occurs infrequently, based on results of EPA's survey.1,2,3
The oxidation process does produce process water containing dissolved
sulfates; however, this process is not planned on any NSPS units at this
time.4
The reduction/Stretford process should produce identical sour Water
streams as the amine absorption process. The vendor of this process
recommends two-stage quench towers, ensuring that only small amounts of
water require treatment for sulfites/sul fates, with the majority reporting
to the sour water strippers for re-use.5
The Stretford process itself can become a potential source of water
pollution, since by-product sulfates and thiosul fates require periodic
purging. Disposal methods of this purge stream involve recovery of
sodium value by evaporation or spray drying, biological degradation or
oxidative combustion.6 After salt recovery, the solid residue may be
landfilled. The next section discusses another alternative which results
in no liquid waste purge.
7-1
-------
7.1.2 Sol Id Waste Impacts
The potential solid wastes from NSPS control systems consist of
spent reduction catalysts (cobalt-molybdenum) and solid residue from
Stretford purge systems. The spent catalysts have market value and have
historically been returned to the vendors for credit when replaced. One
recent study concludes that spent Stretford solution residues are very
small in volume and have an insignificant solid waste impact.7 Another
opinion, however, is that any solid waste, no matter how small, presents
disposal problems in some locations. The vendor for this system indicates
that an alternative sulfur recovery step is now available which will not
require purge and disposal of the absorbing solution.8
The conclusion is that NSPS controls may precipitate a minor solid
waste problem, but in the near future iiay diminish as new operations
choose waste-free technologies.
7.2 ENERGY AND ENERGY-RELATED IMPACTS
The most significant negative impact of applying tail gas treatment
results from the additional steam, hydrogen, electricity, and fuel gas
consumed. In all processes examined capable of achieving NSPS levels,
low pressure (352 kilopascal) stee.m and electricity are consumed. Fuel
gas consumption is also significant where final incineration is required;
however, the reduction/Stretford option results in fuel gas savings.
Hydrogen consumption depends upon Glaus operation and Claus feed
characteristics; in some cases, no hydrogen is consumed while others
require nominal amounts of hydrogen,9,10,11
For the 101.6 Mg/yr model plant, incremental annual energy consumption
(NSPS case less the Base Case) is as follows:
electricity 1.411 x 10^ KWH or 5.08 x 1Q12 joule (j)
fuel gas/hydrogen 56.22 x 1012 j
9,300 Kp steam (1.60 x 1012 j)
1,760 Kp steam (5.53 x 1012 j)
352 Kp steam 127.56 x 1012 j)
106 Kp steam (1.12 x 10*2 j)
Net Consumption: 180.61 x 10*2 j/yr
7-2
-------
Since the sulfur plant emission controls account for an annual
reduction of 2,316.2 Mg/y (2,552.45 t/y), the energy cost is about
78 x 10 joule per Mg S02 removed. The secondary impact of energy
consumption, air emissions generated to replace energy loss, may be
calculated based on a coal-fired utility boiler assumption. This worst-
case scenario indicates that the 78 x 109 joule of coal heating value
expended to convert one megagram of S02 into one-hal f megagram of salable
sulfur would generate .045 Mg S02, .001 Mg particulate matter, .002 Mg
NOX, and 0.25 Mg of solid waste.
7.3 OTHER IMPACTS
The only other impacts of significance incurred by NSPS controls
involve the additional labor requirements and the overall reliability of
sulfur plant operations. In Chapter 6, a 2/3 man-per-shift incremental
impact was assigned for addition of tail gas controls. In actuality the
sulfur recovery unit would likely already have two operators per shift
assigned to the amine and Claus units. Addition of a tail gas unit would
be integrated into the control scheme such that the two operators would
devote one-third of their time to tail gas controls and, therefore less
time to their other responsibilities. This would likely require more
reliance on automated controls for other processes and improved data
retrieval and storage at the control panel. These phenomena are in fact
taking place as sulfur recovery areas undergo replacement and expansions
of existing facil ities. 12
Reliability of the sulfur plant is typically 95 percent at new tail
gas installations; however, for the older tail gas installations, reports
indicate reliabilities of near 100 percent and maintenance costs less
than or equal to Claus plant level s.13,14,15 Hencej for the
facilities modelled in this study, it can be argued'that reliability
overall could not have decreased more than 5 percent. In fact the
Claus/tail gas failures often occur together, thus, the conclusion is
that reliability of a properly designed and operating tail gas unit does
not significantly impact sulfur plant operations.
7-3
-------
Overall, the impact of tail gas controls on refinery operations is
a reworking of operator schedules to include 1/3 time per operator devoted
to tail gas controls, and a near doubling of anticipated maintenance
labor on the sulfur plant, the majority of which would occur simultaneously
for Glaus and tail gas treaters.
7.4 REFERENCES
1. Confidential letter, E. P. Crockett, American Petroleum Institute, to
C. B. Sedman, U.S. EPA, dated June 30, 1982.
2. Sedman, C. B., U.S. EPA, Trip Reoort - ARCO Refinery, Houston, Texas,
dated September 20, 1982.
3. Letter, C. M. Tvler, Standard Oil Company of Ohio, to Don Goodwin,
U.S. E°A, dated July 15, 1982.
4. Letter, D. H. Oil worth, Davy-McKee, to C. 8. Sedman, U.S. EPA, dated
October 5, 1982.
5. Telephone conversation, C. B. Sedman, EPA, and R. E. Warner, R. M. Parsons
Company, October 19, 1982.
6. "SO? Emissions in Natural Gas Production Industry - Background Information
cor Proposed Standards", EPA 450/3-82-023a, January 1983, DO". 7-9 to 7-12.
7. Reference 6.
8. Letter, J. C. Brocoff, R. M. Parsons Co., to S. T. Cuffe, U.S. EPA,
February 16, 1983.
9. Reference 2.
10. Sednan, C. 8., U.S. EPA, Trip Report - Phillies Petroleum Refinery -
Sweeny, Texas, dated September 27, 1982.
11. Sedman, C. B., U.S. EPA, Trip Report - Mobil Oil Refinery - Beaumont,
Texas, dated October 15, 1982.
12. Reference 10.
13. Reference 1.
14. Reference 3.
15. Confidential letter, G. E. Lowe, Marathon Petroleum Company, to D. R.
Goodwin, U.S. EPA, dated September 15, 1982.
7-4
-------
8. RECOMMENDATIONS
8.1 REVISIONS TO NSPS
8.1.1 Sulfur Emissions
From the previous chapters, it is shown that the only significant
disadvantage of requiring NSPS controls is cost, both capital and operating
Capital costs are essentially doubled to remove the final four percent of
potential S02 emissions. Operating costs are essentially doubled, since
labor, maintenance, and cost of capital are doubled. Steam and sulfur
credits are not significantly affected.
Potential revisions to the standard could include lowering allowable
emissions to, say 125 ppmv, or relaxing the requirements to 500, 1 000
or 1,500 ppmv (corrected to zero percent oxygen). Raising or lowering'
to the above levels would accomplish very little from a cost standpoint
since the same systems as found in NSPS application would be used.l
Therefore, capital expenditures would not be significantly affected and
only the energy portion (and possibly maintenance costs) of operating
costs would be noticeably affected.2,3
To make a significant impact on capital and operating cost, the NSPS
would either have to be revised to allow the Glaus extension processes
or dropped altogether. Claus extension processes have not been subjected
to modelling and analysis, but current experience indicates that the
typical control level is 98.6 percent efficiency.4 Hence, for a 101 6
Mg/d facility, the additional operating cost would be about S578 000 for
a cost-effectiveness of $395/Mg S02 removed. The Glaus plant would
remain a major S02 source, emitting nearly 1,000 megagrams S02 per year
With full tail gas control at $750/Mg, the facility emits less than 100
megagrams S02 annually and could be considered less than a major emission
source.
A problem not mentioned in this study surfaced during the review of
this document in draft form. Briefly, the NSPS assumes all sulfur species
in incinerators to be converted to S02; hence, only S02 is regulated.
8-1
-------
One State agency has commented that temperature and 02 monitoring of
incinerators are needed to ensure total sulfur oxidation to $02. It is
recommended that the EPA pursue this problem in conjunction with other
potential changes to be discussed.
8.1.2 Lower Capacity Cut-off
Another way of reducing costs of NSPS would be to raise the lower
capacity exemption of 20.12 Mg/D to some other level , say 50.8 Mg/D.
As shown back in Chapter 4, Table 4.2., only 3 of 24 planned units are in
the 20 to 50 Mg/D range. Additionally, Chapter 6, Figure 6.1 suggests
that the cost-effectiveness at 20.32 Mg/D is not significantly different
at 50 Mg/D. Only at less than 10 LT/D capacities do the cost-effectiveness
curves become steep enough to convincingly serve as an economic basis for
less stringent regulations. Unless some arbitrary cost-effectiveness
value is chosen as a guide for determining regulatory levels, the recommended
path is to retain the 20.32 Mg/D capacity cut-off.
8.1.3 Other Emissions
Since most sulfur plants are subject to State and local regulations,
emission tests are frequently conducted for other pollutants such as
carbon monoxide, particulate matter, nitrogen oxides, and hydrocarbons.
No specific control techniques for these pollutants have been identified,
so it is assumed that the basis for regulation is good operation of the
process. Examination of emission test results shows that emission levels
of nonsul fur species other than carbon monoxide are well below the NSPS
sulfur level. Table 8.2 contains these emissions and suggests that
regulation of other sulfur plant emissions are not warranted on a national
basis.
Table 8.2. TYPICAL SULFUR PLANT EMISSIONS WITH TAIL GAS CONTROLS,6
With Incineration Without Incineration
CO ppmv 650 300
CH4 ppniv -- 55
S02 ppmv 86 <1
H2$ ppmv — 9
particulate gr/DSCF <.0002
8-2
-------
8.2 REVISIONS TO MONITORING REQUIREMENTS
^•2.1 Total Sulfur Monitors
Although monitoring specifications have not been made for monitors
under the sulfur plant MSPS, several total sulfur monitors have been
reported on refinery sulfur plants.7,3 To date, performance of these
monitors has been less than satisfactory to the operators, although many
oroblems pertain to sanole collection and conditioning.3 since sample
collection problems can normally be solved, a further investigation of
total sulfur monitors seems warranted with the goal of developing performance
specifications to complement the monitoring requirements of the NSPS.
•q• -• 2 Hydrogen Sulfide Monitors
Monitors specifically for hydrogen sulfide are essentially the same
tyoe (lead acetate tape - linht dispersion) as observed during preparation
of the MSPS.-'. - Since the state-of-the-art for H,S monitors has apoarPntlv
iot advanced since the MSPS, it would seem expedient to investigate H2S "
•nonitorlna in combination with total sulfur monitoring with thP ooal of
simultaneous monitoring of reduced sulfur and H2S, just as both are
currently measured by £p(\ Method IS.
?l •2 • 3 Sulfur Dioxide and Oxygen Monitors
Sulfur dioxide monitors are found on many new NSPS facilities where
a final incinerator is used for H2S destruction.U.12 Most surveyed use
an in-stack S02 and oxygen monitors similar to that employed on coal-fired
utility boilers. The standard currently does not address the need for
oxygen monitors to convert S02 to an oxygen-free basis. It would apoear
that specifications can be applied to refinery sulfur plants It is
therefore recommended to amend the sulfur olants MSPS to include oxygen
monitoring.
3.3 REVISIONS TO COMPLIANCE TESTING REQUIREMENTS
At some sites, minor modifications to EPA Method 15 have been
instituted to alleviate problems in samnle collection such as moisture
and sulfur accumulation.13 Tnese pr0blems are generally recoonized and
approval of modifications by the enforcement authority has been granted.14
Method 6 for sulfur dioxide is considered a universally accepted
reference method and no change is indicated herein.
8-3
-------
8.4 SUMMARY OF RECOMMENDATIONS
Based on costs, cost-effectiveness, and other environmental impacts,
the current NSPS including the 20.32 Mg/D lower capacity exemption should
be retained. Oxygen monitoring requirements should be added to the NSPS,
and total sulfur monitors should be examined to see if specifications
based on a reliable system may be developed. Temperature monitoring for
incinerators should also be considered to ensure minimal non-S02 emissions
where only S02 emissions are regulated. No other changes to the NSPS
appear warranted, save a possible revision to EPA Test Method 15.
8.5 REFERENCES
1. Letter, W. T. Knowles, Shell Oil Company, to Charles B. Sedman, U.S.
EPA, dated August 24, 1982.
2. Reference 1.
3. Letter, H. J. Grimes, ARCO Petroleum Products Company, to Charles 3.
Sedman, U.S. EPA, dated October 5, 1982.
4. Letter, C. V. Rice, Amoco Oil Company, to Charles B. Sedman, U.S.
EPA, dated October 18, 1982.
5. Letter, R. M. Thompson, Shell Oil Company, to Charles 3. Sedman, U.S.
EPA, dated October 12, 1982.
6. Letter, L. C. Worley, Exxon Company, USA, to Charles B. Sedman, U.S.
EPA, dated October 14, 1982.
7. Sedman, C. 8., U.S. EPA - Trip Report - Phillips Petroleum Refinery,
Sweeny, Texas, dated September 27, 1982.
8. Sedman, C. B., U.S. EPA - Trip Report - Mobil Oil Refinery, Beaumont,
Texas, dated October 15, 1982.
9. Sedman, C. B., U.S. EPA - Trip Report - Beavon Sulfur Removal Units,
dated November 5, 1973.
10. Reference 8.
11. Reference 7.
12. Letter, C. M. Tyler, Standard Dil Company of Ohio, to C. B. Sedman,
U.S. EPA, dated July 15, 1982.
13. Confidential letter, R. J. Niederstadt, Mobil Oil Corporation, to
Don Goodwin, U.S. EPA, dated June 15, 1982.
14. Telephone conversation, B. Ferguson, Harmon Engineering and Testing,
Inc., to C. Sedman, U.S. EPA, dated November 18, 1982.
8-4
-------
APPENDIX A
COST ESTIMATING TECHNIQUES
AND RESULTS OF COST/ANALYSES
FOR SULFUR PLANTS
A.I. CAPITAL COST ESTIMATES
A.1.1 Glaus Plants
The most recent work involving capital cost estimates for Cl aus
Plants is the 1981 Ralph M. Parsons Company study prepared for The Onshore
Gas Production NSPS.l Although the study was directed primarily toward
lean «50% H2S) acid gas streams, the cost estimates allow for reasonable
extrapolation to the 80% H2S refinery case and direct comparison to other
data sources. Additional cost estimates were obtained from responses to
EPA inquiries via 114 letters and phone calls to facilities having Glaus
Plants subject to the NSPS. Though not directly used, previous cost
estimates from the original EPA study on refinery Cl aus plants (1975) and
the GPA Panel discussions in the Oil and Gas Journal were consulted for
comparison.2,3 Since all previous cost studies were performed in English
umts, English units are used in these appendices for consistency, then
converted to metric units in the main report body.
Table A-l presents the Cl aus capital cost estimates used to develop
model costs. These costs are all indexed to July 1982 dollars using the
process industry cost indices from Chemical Engineering-
1974 165.4 "
June 1975 182.4
1978 218.8
April 1980 257.3
1980 261.2
January 1981 276.6
July 1982 314.2
The Parsons capital estimates in Table A-l are for 2 or 3 stage
Claus plants with thermal oxidizer and stacks selected to give uniform
ground level S02 concentrations. Some cost estimates for the larger
Uaus plants also include oxidizer and stack, but with unknown design
basls. Flgure A-l is a logarithmic plot of cost data from Table A-l
A-l
-------
Table A-l. VARIOUS ESTIMATES OF CLAUS INVESTMENT COSTS
Source of
Estimate
Parsons
Study
EPA Background
Document
OGJ GPA
Panel Report
SOHIO
Parsons
Glaus Capacity
LT/D
10
10
10
10
10
10
100
100
100
555
1000
1000
5
10
100
100
100
10.3
Acid. Gas
H2S. %
50
50
20
20
12.5
12.5
50
20
12.5
20
80
50
80
80
80
80(?)
75
80
No. of
Stages
3
2
3
2
3
3
3
3
3
3
3
3
3
3
3
3
Estimated
Capital Cost $xlO&
(corrected to July 1982)
2.50
2.87
2.95
3.29
3.08
3.34
(2.84)
(3.26)
(3.35)
(3.74)
(3.50)
(3.79)
6.47 (7.35)
9.05 (10.28)
11.21 (12.73)
26.23 (29.80)
22.30 (25.33)
26.10 (29.65)
0.757 (1.30)
0.902 (1.55)
2.783 (4.79)
3,5 (5.03)
Year, Month
of Estimate
Jan. 1981
Comment
Installed cost,
no heat recoverv
Waste heat recovery
from thermal oxvlizer
(incinerator)
5.45
2.07
(655)
(2.53)
June 1975 No heat recovery
1978 Assumes typical
refinery installation
July 1980 With stack heat rec.
April 1980 No heat recovery
-------
IU|'UII'1M" < I ULIIS *I
Capital Investment
$xlO° (1982)
[Includes stack a oxidizer]
-------
based on relative size of the unit Dased on total gas flow. As shown,
the 100 LT./D case appears a good estimate as compared to data supplied by
SOHIO4; however, the 10 LT/D case does not correlate as well with Parson's
own estimate on a refinery case.
From Figure A-l, the extrapolated data for an 30 percent H2$ case
(plant size = 1.25) were plotted as a function of Glaus capacity in long
tons per day (LT/0), as shown in Figure A-2. The capital cost curve to
he used for modelling is based upon The Parson's estimates above 100 LT/D
and a fit to the Parsons 10.3 LT/D estimate. This curve is for a no heat
recovery assumption.
Above 100 LT/D, the cost curves are essentially a straight-line
relationship of the form
y = mx'^-S
Deferring back to Figure A-l, data for the 100 LT/D case also
anoroximate a straight-line relationship of the form
y = mx3-11
Therefore, for any sulfur plant of known capital cost (1982 dollars)
C[, of capacity rating LTDi, and ciH?0 in feed (H?S)-[, the cost of a second
Glaus plant C? with caoacity LTD2 and feed composition (H?S)2 niay be
found by:
Equation A-l - = - LTD^ °'6* (H9S), °'4
^ i
where 100 _<_ LTD]., LTD2 _<_ 1000
12.5 <_ (H2S)i, (H;?S)2 <_ 30
The above formula is obviously for rough estimates only and includes
the incinerator and stack. Should heat recovery or an unusual
incinerator/stack requirement be desired, adjustments to costs estimated
as above o^ from Figure A-2 should be considered as discussed below.
The estimated typical Claus stack and incinerator capital costs in
July 1982 dollars are plotted as a function of plant size (gas flow basis)
in Figure A-3. Figure A-4 shows a similar plot, but as a function of
* At 10-40 LT/D the exponent is 0.20, at ^0-80 LT/D 0.40.
A-4
-------
Figure A-2. Glaus Only Capital Cost vs.
Plant Sulfur Capacity @ 80S H2S Feed
A-5
-------
tt.
CO
O1
l/l
o
o
ao
rO O
X
•<=<=)-
m~=^:r^L
-: ; --^-jpy-_-
- - - (-.— J? — -..
~- " ~: ^T" --"-"^-
— 3^ r.~r:-:P^^
?M
'-^T-
-jr:.
i-a:_-
~t-:
3 -a
A-'?
nerator Caoita" Costs
-------
Figure A-4. Stack and Incinerator Costs vs
Claus Capacity (80% H-S)
A-7
-------
sulfur capacity. Figure H-5 plots capital costs of waste heat recovery
boilers for plants greater than IOC* LT/D. A plot similar to Figure A-4
is not necessary since waste 'neat recovery boilers are not considered
below 100 LT/D, the largest model to be examined.
A . 1.2. Tall Gas Treating Capital Costs
As discussed in Chapter 5, the purpose of this report is to assess
the imoact of MSPS upon Claus plant operation. Therefore, it is unnecessary
to evaluate all potential tail gas processes, rather, a representative
process will suffice. Further, the area of interest in determining cost
impacts is the snail (10-50 LT/D) sulfur plant which represent worst-case
impacts. Ultimately this analysis should answer the Questions, "What are
typical control costs?", and "Is tre current 20 LT/D capacity exemption
reasonable considering costs?"
To answer these questions, three model facilities at 10, 50, and 100
LT/D were chosen to span the area cf most interest and provide a 3-point
cost curve for possibly evaluating models within this range. Assuming
that control costs at 100 LT/D are reasonable, larger facility costs are
of minimal interest for the purposes of this study.
Because the amine tail gas process is dominant in the less than 100
LT/D size range (13 of 20 operataing tail gas treaters or 90 percent), it
is chosen as a representative model basis. It is important to note that
the amine system is not necessarily the lowest cost process in this size
range, rather the most common. One vendor of both amine and Stretford
processes indicates that the amine may be less costly for units of 30 LT/D
and smaller.^
Capital costs for actually installed amine tail gas units in the 10
to 100 LT/D range are presented in Table A-2 and adjusted to a July 1982
basis.
Table A-2. CAPITAL COSTS FOR AMINE TAIL GAS TREATERS
Parent Claus Capacity, LT/D Capital Installed Cost SxlO6 (1982)
10 2.31
20 2.84
60 2.50*
100 4.68
165 5.97
* Thought to be 1973 enuioment + 1978-82 construction.
A-8
-------
I'J
lUUKIIHMil , I I UCIIS M out 1,11
Waste Heat Boiler Capital Cost
D (1982)
-------
liX
:~p-
"1-P
I .ru.—
.znr.::f_._ -_r_._
r~"..~}.~: J". ~
I
._zt~
j^rfgr^:-
•ur_ t —j^
-^— .—r " -i r^r"
- -1 - f'
— I
"_:~t:.::tr-'
- I- - - -
s 9 100
Figure A-6. Capital Costs of Amine-Based
Tail Gas Treaters
A-10
-------
H
llll'HIIMMIL / » | [ttili AlOUt
Capital Investment
Amine Tail Gas Units
$xlO (1982)
-------
The costs in Table 4-2 represent a combination of retrofit and new
tail gas treaters. In the case of "etrofit units, costs have been adjusted
down to account for retrofit costs, while for new units, the costs were
disaggregated from total sulfur recovery costs.5 Therefore, a significant
degree of uncertainty is reflected in the above costs because no data
were available for a new tail gas unit with costs of the tail gas treater
separated from the Glaus plant and, in some cases, Glaus plant amine
treater and boilers. The $2.50 million estimate at 60 LT/0 is thought to
be the 1978 equipment cost + installation during 1978-82. A reasonable
1982 estimate would be S3.6 x 106.
A.1.3. Effects of Combined Glaus/Tail Gas Treater on Capital Costs
To estimate the combined cost of Glaus + tail gas treate^ is not
straightforward. First, if the tail gas unit recycles the removed material
to the Clans plant, the Glaus olant requires increased canacity to
^coommndatp the increased qas flow and sulfur recovery. This increased
capital expenditure is offset by the lower capital incurred by a smaller
stack renui^sd to disperse emissions.
In the "arsons study, the increase in Glaus olant expenditure due to
amine tail gas testing were S0.32xl06 at 100 LT/D, 50% H2S for a 7.06
percent increase in cost; S1.03xlO£ at 100 LT/0, 2Q% H2S for a 17.3"
increase in capital cost. In the model 100 LT/D plant chosen (80% HoS) ,
the average increase in Glaus capacity is 3.3 percent. Also, the gas
flow is increased by some 4.27 percent; hence, the percentage HjS drops
from 30 to 78.63 percent. Also, the engineering design allows for doubling
of anticipated recycle stream for safe design; therefore, the increased
capital cost based on the formula developed earlier is estimated at
[(1.066) ' (yn gn) ' - 1] or a 4,60 percent increase in capital cost.
These results are plotted in Figure A-8 and appear to correlate well with
the Parsons study.
Since stack size (height) is assumed to be proportional to the mass
emission rate, the capital expenditure for a stack is therefore a function
of the mass emission rate. From tie Parsons study, the data for stack
expenditure versus emission rate in Ibs/hr is plotted in Figure A-9 for
selected cases. Below 150 Ib/hr SDg, the stack cost is essentially fixed
at 530,670 (July 1982).
A-12
-------
1503
:>:! niniy UI.IAJ i »
-------
5 6
tmission Rate, Ib/hr
Figure A-9. Capital Cost of Stack vs.
SCL Emission Rate
A-14
-------
A.2. OPERATING COST ESTIMATES
In general, the operating costs were structured according to the
methodology presented in the January 1933 draft Background Information
Document for the Natural Gas Production Industry (EPA 450/3-82-023c) J
Operating costs are broken down into the following categories:
utility consumption and credits
chemical consumption and credits
labor-operating and supervisory
maintenance and repair
0 miscellaneous (supplies and laboratory changes)
° fixed costs - capital charges, taxes, and insurance
overhead, including administrative and marketing
In lieu of actual cost data for refinery sulfur plant operations
the following costs and/or assumptions were extracted directly from the
nas production document:
utility orices and credits (see Table A-20)
0 operating suoplies and laboratory charges at 10 percent each of
operating labor charges
taxes - 1 percent of fixed capital costs
0 insurance - 0.5 percent of fixed capital costs
0 overhead - 25 percent of operating labor and maintenance
0 administrative and marketing - 1 percent each of total annualfzed
costs
Other operating cost estimates require more detailed explanation as
in the following sections.
A-2-l- Utility Consumption and Credits
A.2.1.1 Glaus Plants
Steam, feedwater, and electric power figures for Glaus plants were
estimated using graphs prepared from the Parsons study cited earlier
Figures A-10 and A-ll graphically illustrate steam and condensate production
(consumotion for 600 Psig steam) in Ibs/hr per lona ton sulfur production
as a function of gas flow for 2-stage and 3-stage Glaus plants with no
heat recovery; Figure A-12 shows similar figures for a heat recovery
system as proposed by Parsons, based on incineration at 1200°F. Tables A-3
A-4, and A-5 show these data numerically for the three cases examined by
-------
#..
-------
46
KLUKf t i U, Lbi>fc H (. o
Steam Production
///hr/LT
Steam Production
-------
Table A-3. 2-STAGE CLAUS
NO HEAT RECOVERY
STEAM PRODUCTION (#/HR/l.T)
H2S/C02 ratio 50/50 20/80 12.5/87.5 80/20 (estimated)
600 Psi9 (23.9) (84.8) (130.2) (10.5)
250 Psig 140.9 127.7 123.9 154
50 Psi"9 0 12.2 22.1 0
15 psig 21.7 41.1 54.5 15.0
condensate 109.6 184.8 230.2 81 5
oo
-------
H2S/C02 ratio
50/50
condensate
120.2
Table A-4 3-STAGE CLAUS
NO HEAT RECOVERY
STEAM PRODUCTION W/HR/LT)
20/80
'
36.4
31.4
19
12.
(,615,
123.9
51.2
80/20 (estimated)
(175)
-------
Table A-5. CLAUS PLANT
HEAT RECOVERY
STEAM PRODUCTION (#/HR/LT)
H2S/C02 ratio 50/50 20/80 12 5/87.5 80/20 (estimated)
60° psig 89.2 185.0 278.6 63.5
250 Psig 10.7 22.14 33.46 7.4
50 psig 65.9 80.0 70.0 53.8
15 psig
condensate 65.9 80.0 70.0 53.8
I
ro
o
-------
Figure A-12
/ooo
o ,
3)
o =
-------
Parsons and also include the extrapolated figures at 30% l^S for a typical
refinery application. Tables A-6 and A-7 then combine these results for
model 2-stage and 3-stage Claus plants with heat recovery.
Using the total steam and condensate values, the boiler feed water
requirements may be estimated by assuming a 2.7-3.0 percent system loss
of steam and condensate; i.e., the total steam and condensate divided by
.9715 equals boiler feedwater requirements.
Electric power requirements nay be estimated by using either of two
curves shown in Figures A-13 and A-14. These show electric power
consumption as a function of gas flow and of sulfur in feed.
Fuel gas reouirements for incinerators were calculated for each case
based upon tail gas composition and temperatures according to principles
outlined in Chemical Engineering Thermodynamics by Smith and Van ^ess.3
T'ie calculations scheme is similar to that employed in Appendix C-II of
EP^ 450/2-78-012, Control of Emissions from Lurgi Coal Gasification
Plants; oanp C-19 of that report is reprinted here as "inure A-15.9 The
vnly difference here involves recalculation of the average specific heats
to correspond with the temperature ranges evaluated in this study--1200°F
combustion temperature. Also fuel was assumed to be fuel gas having a
composition of C1>15H^ 3 having a heating value of 3.85xl05 Btu/lb-mole
(995.6 Btu/scf). All exhaust streams are oxidized at 2'i percent excess
air, to be consistent with the Parsons study.
A.2.1.2 Amine Tail Gas Treaters
There are very little data available for actual steam, electric
power, and fuel gas consumption figures for amine tail gas treaters,
since most reported data are combined with the Claus and fuel gas ainine
data. Two estimates of amine treater utility consumption for a 100 LT/D
case are available along with one report of actual consumption figures
for two systems of 170 LT/D and 2^-0 LT/D.10*11^ Table A-8 shows these
figures with the actual data converted to a 100 LT/D figure for comparison.
As shown, the actual figures from ARCO and the Parsons estimates generally
agree except for fuel gas consumption, where the ARCO and SOHIO estimates
are similar. For purposes of model analyses, the ARCO data will be used,
along with the condensate generation estimate from Parsons. Fuel gas
consumption will also be calculated based on material and heat balances
for comparison.
A-22
-------
Table A-6. 2-STAGC CLAUS
WITH HEAT RECOVERY
STEAM PRODUCTION (#/HR/LT)
600 PSI9
50ps'9
15 PS1'9 21 7 /MI
1-1 54-5 150
condensate 43 7 in/1 „
4J'X 104-8 160.2 28 2
3=-
I
ro
CO
-------
Table A-7. 3-STAGE GLAUS
WITH HEAT RECOVERY
STEAM PRODUCTION (#/HR/LT)
H^S/C02 ratio 50/50 20/80 12.5/87.5 80/20 (estimated)
600 psig 54.9 79.3 117.1 46.0
250 psig 151.6 149.84 157.36 157.5
50 psig 78.8 116.4 121.2 60.4
15 psig 16.2 31.4 46.1 12.4
condensate 54.3 125.7 191.5 34.2
-------
Figure A-13
o»
a -
o =
r- -
o
+J
Q.
c
o
O
i.
0)
o
Q_
5.0 6 7 8 9
lo
-------
I
ro
en
JltAH) IHMIC
X I t /(.L I .
KLUJUL Ik L^
74OO
-f!
TT
T
:w*-
1
-I4-U-.-
ITT]
+4
CO
-s
CD
i
.P.
-------
C laus tai 1 (j()<, @ 284°F
Vent cxp. gases @ 77°F
Fuel:
N2 rf = To = 7rF
Reactions:
]- CO + 1/2 02 , C02
( = Em; CF; (T.-l
QJ
Oxidized _
Products T2al600°F or
30 , 2 CO
-
"•
H?S
H20 „,
* /2 0, -. 2 CO, t 3 ,,20 „,
3/2 0,
'• cos . VZ o
0 fnl
<;n
fll
298 '
'298 '
'298 = 233,097.4
•0 Btu/lb-mole
.4
,6
.6
.0
-
298 =1169,010.0
u. s o
SO,
Re
-------
Table A-8. UTILITY CONSUMPTION E!Y AMINE TAIL GAS UNITS
(100 LT/D Glaus basis)
Source Power, kw
SOHIO (estimate) 180
ARCO (actual) 175
PARSONS (estimate) 200
i
INJ
00
Cooling water,
GPM
- -
170
870
840
50 # steam
1 h/hr
8390
12472
13277
Fuel gas
106 Btu/hr
5.33
5.2
6.5
Condensate
Generated
Ib/hr
14,300
-------
A'2'?- Catalyst and Chemical Consumption
Catalyst consumption figures for Glaus plants are based on an
assumption that the first stage catalyst is replaced on a two-vear cycle
the second stage at four years, and the third at six years. The assumed'
catalyst is alumina at S17 Per cubic foot or $765 per short ton (3856 80
per long ton). Catalyst charge is estimated at 230 pounds per reactor
per long ton Glaus capacity for 80 percent H2S feed.13.14
For amine tail gas units, catalyst replacement (cobalt-molybdenum)
for the reduction reactor is assumed to be once every two years. The
catalyst charge is assumed to be about 1/2 of a Glaus stage, or
115 oounds oer reactor Per long ton Glaus capacity (30% H2s'feed) The
Burned catalyst cost is 10 tin.es that of the Glaus catalyst or S3568/long
*• n n - J J
torl.l
Tail gas chemical consumption is a more elusive subject as amine
tvoe, degree of fouling, and degree of enhanced recovery hy use of
Ipfoanina agents and organic contaminant removal varies from plant to
^ant. Actual ^igures provided for three systems show consumption of
3!'* at 0.56, 0.67, and 3.1 lb/hr per 100 LT/0 parent Haus canaritv
averaging 1.44 Ib/hr per 100 LT/D.16,1?,^
For model purposes, a figure of 0.70 lb/hr per 100 LT/D Glaus capacity
"/ill be used.
A • 2 • 3. Labor, Maintenance, and Repair Costs
In the Parsons study, labor costs were estimated as follows-riaus
Plants @ 1.25 operators per shift, Glaus + tail gas treater I? 2 25
operators Per shift. Supervision was assumed at 0.25 per shift for both
r a c a c
cases.
-rom two new operating plants having both Glaus plants and tail gas
treatment, the following data were obtained:19,20
Start-up
Plant Date
1 1980
2 1981
Glaus
Capacity
100
475
Type of Tail
Gas Treater
Amine
Amine
Manpower/shift
ClaUS Tail Ca*.
2/3
2/3
2/3
2/3
A-29
-------
Hence, for new plants the labor assumptions are 2/3 operator per
shift each for the Glaus and tail gas treater and I/A supervisor per
shift. Hourly rates per the Gas Production NSPS study are 314.50/hr for
ooerators and S13.30/hr for supervision.
Maintenance and repair costs also varied widely from plant to plant.
^or Glaus plants, labor and materials ranged from 2.3 to 6.1 percent of
estimated fixed capital costs for Glaus plants and 2.1 to 6.3 percent of
capital costs for tail gas units. Other studies have assumed 3 percent
of fixed capital (EPA 1975) and 3.5 of fixed capital: (Gas Production HSPS
Document, 1983).
Since the average labor and materials cost from the six new MSPS
and two dozen or so older units was about 3 percent for both systems,
t'n's figure is assumed cor model purposes.
A.3. MODEL PLANT LIME COSTS
*\.3.1. Gapital Cost and Operating Parameter Estimates
'Jsing the economic assumptions and cost curves presented in the
first h,!vn sections of this Apoendix, the following model plants were
eviluate-l:
Case Glaus Plant, LT/D Tail Gas Treater
1A 2 stage, 10 LT/D
13 2 stage, 10.48 LT/D Amine 0.96 LT/D design
?A 3 stage, 50 LT/D
2B 3 stage, 51.55 LT/D Amine 3.3 LT/0 design
3A* 3 stage, 100 LT/D
38* 3 stage, 103.3 LT/D Amine 6.6 LT/0 design
* Waste heat boiler included for incinerator
Case 1A
Key line item estimates for case 1A are:
Item Source Estimate (July 1982)
Capital Cost:
2-stage Glaus Figure A-2 SI.97 x 10°
Incinerator Fiqure A-A $0.2* x 10°
Stack Figure A-9 $0.31 x 106
S2.54 x 1C"6"
A-30
-------
jtem
Operating Cost (Credit):
600
250
osig
?sig
psig
steam
steam
steam
Source^
Figure A-10, Table A-3
Estimate (July 1982)
condensate
electric power
fuel nas
catalyst
105
(1540 4/hr)
( 150 */hr)
Figure A-13
Figure A-15 (calculated)
Section A.2.2
50.5 KWH/hr
0.60 10" Btu/hr
1725
Case IB 10.43 LT/D (78.07% H2S)
Key line item estimates for case 1 B are:
Item
Capital Tost:
2-stage Glaus
Amine Theater
Incinerator
Stack
Operating Cost:
600 psig
250 psig
50 psig steam
15 psig steam
15 condensate
electric power
fuel gas
cooling water
catalyst
chemicals:
DIPA
Soda
Source
Equation A-l
Figure A-2
Figure A-5
Figure A-4
Figure A-9
pigure A-10
M
Table A-3
Figure A-10
Figure A-10/
Figure A-13/
Table A-3
Figure A-15
(calculated)
Section A.2.2
Section A.2.2
Case 2A 50 LT/D (80% H2S)
I tern
Capital Cost:
3-stage Claus
Incinerator
Stack
Source
Figure A-2
Figure A-4
Figure A-9
Estimate
SO.26 x 106
SO.31 x 1Q6
S2.61 x 10*
$2.35 x 106
117 4/hr
(1603 *
(160 #
(367
51 KWH/H
1314 =Vhr
(2080 4/hr)
39.5 KWH/H
0.58 x 106 Btu/hr 0.76 x 1Q6 Btu/hr
1852 */yr
126.5 gprn
617.5
616 Ib/yr
1000 Ib/yr
Estimate
S3.50 x 106
0.33 x 106
0.51 x IQo
14 J3 x 10o
A-31
-------
Item
Operating Cost (Credit)
600 psig steam
250 psig steam
50 psig stean
15 psig steam
condensate
electric power
fuel gas
catalyst
Source
Table
Figure
rigure
A-14
A-15
(calculated)
Section A.2.2
Case 2B 51.65 LT/D (78.7%
Item
Capital Cost:
3-stage Claus
Amine Treater
Incinerator
Stack
Operating Cost:
600 psig steean
250 psig steam
50 psig steam
15 osig steam
condensate
electric power
fuel gas
cool ing water
catalyst
chemicals:
DIPA
Soda
Source
Equation A-l
rigure A-2
Figure A-6
Figure A-4
rigure A-9
rigure A-ll
it
" /Table A-8
it
" /Table A-8
rigure A-14/Table A-3
Figure A-15 (calculated
Table A-8
Section A.2.2
Case 3A 100 LT/D (80%
Item
Capital Cost:
3-stage Cl aus
Incinerator
Stack
Waste Heat
Recovery System
Operating Cost (Credit)
600 psig
250 psig
50 psig
15 psig
condensate
electric power
fuel gas
catalyst
Source
Figure A-2
rigure A-4
Figure A-9
Figure A-5
Table A-7
Figure A-14
Section A.2.2
Estimate
875 #/hr
(7500 =Vhr)
( 330 */hr)
( 620 #/hr)
(4400 #/hr)
125 KWH/H
3.0 x 106 Btu/hr
10,542 Ib/yr
Estimate
TJTaus Tall Gas
$3.60 x 106
SO.32 x 106
$0.31 x 106
$4.23 x 106"
919 #/hr
(7748 =Vhr)
( 346 */hr)
( 635 ?/hr)
(4597 =/hr)
128 KWH/H
209 x 106
S3.60 x 106
623 Ib/hr
(7150 Ib/hr)
99 KWH/H
Btu/hr 2.6 x 106
435 gpm
11,067 Ib/yr 3180 Ib/yr
Btu/hr
3037
4928
Ib/yr
Ib/yr
Estimate
S4.50 x 106
0.41 x 106
0.75 x 1C)6
0.56 x 1C)6
S6.26 x 10°
(4600 Ib/hr)
(15740 Ib/hr)
(6040 Ib/hr)
(1240 Ib/hr)
(3420 Ib/hr)
212 KWH/H
6.0 x 106 Btu/hr
21,084 Ib/yr
A-32
-------
Case 3B_ 103.3 LT/0 (78.7% H2S)
Item
Capital Costs:
3-stage Glaus
Amine Treater
Incinerator
Stack
Waste Heat
Recovery System
Operating Cost (Credit):
Source
Equation A-l
Figure A-6
Figure A-4
Figure A-9
Figure A-5
+ Fig. A-2
600
250 psig
50 psig
15 psig
condensate
electric power
fuel gas
cool ing water
catalyst
chemical s:
DIPA
Soda
Figure A-ll
ti
", Table A-8
Figure A-ll
", Table A-8
Figure A-14/Table A-8
calculated
Table A-8
Section A.2.2
Estimate
Glaus
S4.63 x 106 "
0.41 x 106
0.31 x 10°
0.57 x 1QQ
b.S^ x 10°
(4750 Ib/hr)
(16260 Ib/hr)
(6240 Ib/hr)
(1280 Ib/hr)
(3530 Ib/hr)
215 KWH/H
.15 x 106 Btu/hr
22,134 Ib/yr
Tail Gas
S4.68 x 105
12,472 Ib/hr
(14300 Ib/hr)
165 KWH/H
5.2 x 106 Btu/hr
370 gpm
6360 Ib/yr
6074 Ib/yr
9856 Ib/yr
Combining the above figures with the prices in Table A-9 results in
line item costs as presented in Table A-10. A significant portion of
annual operating costs is the capital recovery factor. For comparison
Table A-ll shows the annual costs, and costs per ton S02 controlled '
for interest rates of 10, 15, and 20 percent for a 15-vear lifetime
A-33
-------
Table A-9. ECONOMIC ASSUMPTIONS USED TO CALCULATE ANMUALIZED COSTS*
I. Utility prices:
1. 600 psig steam $15.98/Mg ($7.25/1,000 lb)
2. 250 psig steam $14.88/Mg ($6.75/1,000 Ib)
3. 50 psig steam $12.68/Mg (35.75/1,000 lb)
4. 15 psig steam $ 9.92/Mg ($4.50/1.000 lb)
5. boiler feedwater $ 3.31/Mg ($1.50/1,000 lb)
6. steam condensate $ 2.76/Mg ($1.25/1,000 Ib)
7. cooling water $13.21/103m3 ($ .05/1,000 gal)
8. catalyst:
a. alumina $352.64/Mg ($0.38/lb)c
b. cobalt-molybdenum (Co/Mo)
$3,5256/Mg ($3.80/lb)c
9. Chemicals:
a. diisopropanolamine $0.49/Kg ($1.07/lb)b
b. soda $:!30.6/Mg ($300/ton)b
10. fuel gas $3.64/lQ9/J ($3.50/106 Btu)d
11. electric power S0.05/KWH
12. sulfur $118,08 Mg ($120/LT)e
II. Labor (8,760 hours per year basis)
1. operators: ($14.50/hr)
2/3 per shift for Claus
2/3 per shift for tail gas treater
2. supervision: ($18.80/hr)
1/4 per shift for sulfur recovery facility
III. Maintenance and Repair
Labor and materials: 3.0 percent of fixed capital
Costs6
IV. Other Miscellaneous Costs
1. Operating supplies: 10 percent of operating labor
2. Laboratory charges: 10 percent of operating labor
V. Fixed Charges
1(1+1)"
1. Capital charges = fixed capital costs x TY+f)11-!
= a) .13148 for n = 15 years and 1 = 10%
b) .17106 for n = 15 years and i = 15%
c) .21382 for n = 15 years and i = 20%
1 Local taxes - 1 percent of fixed capital costs
3. Insurance - 0.6 percent of fixed capital costs
A-34
-------
Table A-9. ECONOMIC ASSUMPTIONS USED TO CALCULATE ANNUALIZED COSTS^ (continued;
VI. Overhead
1. plant overhead - 25 percent of operating labor + 25 percent of
maintenance and repair
2. administrative - 1 percent of annualized costs
3. distribution and marketing - 1 percent of annualized costs
a All assumptions and values assigned from Reference 1 unless otherwise
noted; actual consumption figures for model plants from EPA survey and
Reference 2.
b Chemical Market Reporter, October 4, 1982.
c Telephone conversation with Mr. R. E. Warner of Ralph M. Parsons Co.,
February 1, 1983.
d Memorandum: R. E. Jenkins to C- 3. Sedman, EPA, dated September 7,
1982.
e Average of EPA survey.
A-35
-------
Table A-10. LINE FEM COSTS FOR MODEL PLANTS
MODEL 1A (10.16 Mg/d)
Capital cost - $2.54 x
Direct operating cost
A. Utilities & Chemicals
1. 4,300 Kp steam
2. treated boiler feedwater
3. electric power
4. fuel gas
5. catalyst
B. Labor
1. Operators
2. Supervision
C. Maintenance and Repair
D. Supplies and laboratory charges
Fixed Charges:
A. Capital
B. Taxes
C. Insurance
Plant Overhead:
General Expenses
A. Administrative
B. Distribution and sales
Total Annual i zed Costs
Credits
1. 1,960 Kp steam
2. 106 Kp steam
3. steam condensate
4. sulfur
Total Credits
Net Annual Operating Cost for Case 1A
i = 15%
$ 6,395
21,615
21,210
17,640
655
$84,680
41,170
$76,200
$16,940
$434,490
25,400
15,240
$40,220
$ 8,020
$ 8,020
$817,895
$ 87,320
5,670
8,558
399,420
$499,265
$320,439
i = 10%
$ 6,395
21,615
21,210
17,640
655
$84,680
41,170
$76,200
$16,940
$333,960
25,400
15,240
$40,220
7,160
7,160
715,645
$ 87,320
5,670
8,558
399,420
$499,265
$218,189
i = 20%
$ 6,395
21,615
21,210
17,640
655
$84,680
41,170
$76,200
$16,940
$543,105
25,400
15,240
$40,220
9,100
9,100
928,670
$ 87,320
5,670
8,558
399,420
$499,265
$431,214
A-36
-------
Table A-10. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL IB (10.16 Mg/d)
Capital Cost - $4.96 x 106
Direct operating cost i - ic* ,- ,n.
A. Utilities ft Chemicals 1^-^- -1 = 10% 1 =
s: >
: eSS *-*" He :
5. fuel gasAdrogen %'°£ f|-°" 38,010
: : 1:S5
8. chemicals 3'?50 3,050
810 810 810
Labor
C. Maintenance S Repair $148>80o $148,8oo ,148.800
0. Suppl fes S Lab Charges , 33,870 533,870 $33,870
Fixed Charges
^2,140 1,060,545
C. Insurance ^9'76°° 94Q9'6C00 49,600
29,760 29,760 29,760
PI ant Overhead: c 7fl r.n
S 79,540 79,540 79,540
General Expenses
A. Administrative $ l
B. 01 stHbU«on and sales * {
Total Annual i zed Costs ,1>6go>065 Ii489j945
Credits
1. 1,960 Kp steam * qn oQn on onn
2. 106 Kp steam $ 92»g|° 90'^0 90,890
3. steam condensate
W47.465 S547,465 ,547,465
Net Annual Operating Cost for Case IB $1,142,600 $942,480 $1,359,085
A-37
-------
Table A-10. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 2A (50.8 Mg/D)
Capital Cost - $4.33 x 1Q6
Direct Operating Cost i =15% i = 10% 1 a
A. Utilities S, Chemicals
1. 4,300 Kp steam $ 53,290 $ 53,290 $ 53 290
2. treated boiler feedwater 155,310 155,310 155 310
3. electric power 52,500 52,500 52,' 500
4- fue1 9as 88,200 88,200 88,200
5. catalyst 4,005 4,005 4,005
B. Labor
1. Operators 84,680 84,680 84,680
2. Supervision 41,170 41,170 41,170
C. Maintenance and Repair 129,900 129,900 129,900
D. Supplies and Lab Charges 16,940 16,940 16,940
Fixed Charges
A- Capital 740,690 569,310 925,840
B- Taxes 43,300 43,300 43,300
C. Insurance 25,980 25,980 25,980
Plant Overhead 53,645 53,645 53,645
General Expenses
A. Administrative 15,000 13,300 16,850
B. Distribution and Sales 10,000 13,300 16,850
Total Annual ized Costs $1,519,610 1,344,830 1,708,460
Credits
1. 1,960 Kp steam $425,250 $425,250 $425,250
2. 352 Kp steam 15,940 15,940 15,940
3. 106 Kp steam 23,435 23,435 23,435
4. steam condensate 46,200 46,200 46,200
5- sulfur 2,028,600 2,028,600 2,028,600
Total Credits $2,539,425 2,539,425 2,539,425
Net Annual Operating Cost for Case 2A ($1,019,815) (1,194,595) (830,965)
A-38
-------
Table A-10. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 2B (50.8 ffc/D)
Capital Cost - $7.83 x 10^
Direct Operating Cost i - i«- • 1na,
A. Utilities & Chemicals 2^15% i = 10% T = 20%
1. 4,300 Kp steam
3. ch»ls
Labor
C. Maintenance S Repair
0. Suppl ies t Lab Charges
Fixed Charges
B.' T^es
C. Insurance
Plant Overhead
General Expenses
A. Administrative
B. Distribution & Sales
Total Annual ized Cost
Credits
2: lflSgSS
3. steam condensate
4. sulfur
Total Credits *o
52,684,560 $2,684,560 $2,684,560
Annual Operating Cost for Case 2B $ 157,590 ($158,520) $ 499,100
A-39
$ 55,965
284,485
172,770
95,340
161,000
10,960
16,290
3,990
"IIS
234,900
33,870
/ o • oUU
Af* QftO
101,065
":™
2,842,150
2,097^900
$ 55,965
284,485
172,770
95,340
161,000
10,960
16,290
3,990
82^340
234,900
33,870
1,029,490
46,' 980
101,065
24,650
24,650
2,526,040
439,310
24,005
123,345
2,097,900
* *— w nj
$ 55,965
284,485
172,770
16, '290
3,990
169,360
82,340
234,900
33,870
78 ,'300
46,980
101,065
31,100
31,100
3,183,660
439,310
24,005
123,345
2,097,900
-------
Table A-10. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 3A (101.6 Mg/D)
Capital cost - $6.26 x
Direct Operating Cost i = 15% i = 10% i = 20%
A. Utilities & Chemicals -- - -
1. treated boiler feedwater $402,575 $402,575 $402,575
2. electric power 89,040 89,040 89,040
3. fuel gas 176,400 176,400 176,400
4. catalyst 8,010 8,010 8,010
B. Labor
1. Operators 84,680 84,680 84,680
2. Supervision 41,170 41,170 41,170
C. Maintenance & Repair 187,800 187,800 187,800
D. Suppl ies & Lab Charges 16,940 16,940 16,940
Fixed Charges
A- Capital 1,070,835 823,065 1,338,515
B- Taxes 62,600 62,600 62,600
C. Insurance 37,560 37,560 37,560
Plant Overhead 68,120 68,120 68,120
General Expenses
A. Administrative 22,460 19,980 25,135
B. Distribution & Sales 22,460 19,980 25,135
Total Annual ized Costs $2,290,650 2,037,890 2,563,680
Credits
1. 4,300 Kp steam 280,140 280,140 280,140
2. 1,960 Kp steam 92,460 92,460 92,460
3. 352 Kp steam 291,730 291,730 291,730
4. 106 Kp steam 46,870 46,870 46,870
5. steam condensate 35,910 35,910 35,910
6. sulfur 4,057,200 4,057,200 4,057.200
Total Credits $5,604,310 5,604,310 5,604,310
Net Annual Operating Cost for Case 3A ($3,313,660) ($3,566,420) ($3,040,630)
A-40
-------
Table A-10. LINE ITEM COSTS FOR MODEL PLANTS (continued)
MODEL 3B (101.6 Mg/D)
Capital cost - $10.60 x 106
Direct Operating Cost * - i«
A. Utilities & Chemicals • " 1 = 10% 1 = 20%
-as
ass
catalyst fp1'925 21 925 21,925
7. chemicals 3H«n 3?'^ 32'580
7,980 7,980 7,980
B. Labor
1. Operators icQ ->cn ,^n
2. Supervision ^f^ "9.360 169,360
d^,340 82,340 82,340
C. H.1»t.n.nce * tep.,r 318,000 318,000 318,000
B. Supp, ,.s 4 Lab Charges 33,870 33,870 33,870
pixed Charges
B.' TaxPesa1 lf?n«fnnn 1'393'690 2,266,490
C. Insurance ^f'NW 106,000 106,000
63,600 63,600 63,600
Plant Overhead 101 0/ln
121,840 121,840 121,840
General Expenses
A. Administrative ->n
B. Distribution a Sales |H|° 35,550 44,280
jy,/bU 35,550 44,280
Total Annual ized Cost tA noo 7^r
$4,088,745 3,660,800 4,551,060
Credits
2- fl^gSS Jg.|75 289,275 289,275
3- 106 Kp steam 9g'|« 921»940 921,940
4. steam condensate S f 48'385
«.««.«5 5,642,615 5,642,615
Net Annua, Operatfng Cost for Case 38 («.S53.870) (11,981,815) ($i,09l 555,
A-41
-------
Table A-ll. COST & COST-EFFECTIVENESS OF MODEL CONTROLS
i = 10 percent
Base Case Annual Cost, $
Base Case S02 Removed, tons/yr
NSPS Case Annual Cost, $
NSPS Case SOe Removed, tons/yr
Cost-Effectiveness, $/ton
Plant Size, LT/D
10" ~
218,189
7,455.84
$942,480
7,832.16
51,929
50
($1,194,595)
37,867.2
($158,520)
39,160.8
$801
' 100
($3,566,420)
75,734.4
($1,981,815)
78,321.6
$612
t = 15 percent
Base Case Annual Cost, $
Base Case $03 Removed, tons/yr
NSPS Case Annual Cost, $
NSPS Case SOa Removed, tons/yr
Cost-Effectiveness, S/ton
320,439 ($1,019,815)
7,455.84 37,867.2
$1,142,600 $157,590
7,832.16 39,160.8
$2,190 $910
($3,313,660)
75,734.4
($1,553,870)
78,321.6
$680
i = 20 percent
Base Case Annual Cost, $
Base Case SOg Removed, tons/yr
NSPS Case Annual cost, $
NSPS Case S02 Removed, tons/yr
Cost-Effectiveness, S/ton
$431,214 ($830,965)
7,455.84 37,867.2
$1,359,085 $499,100
7,832.16 39,160.8
$2,471 $1,028
($3,040,630)
75,734.4
($1,091,555)
78,321.6
$753
A-42
-------
A. 4 REFERENCES
Jib MoUl6lU6SRr°o^yhSiUdX ' 0nSiTe S°Ur GaS P™duct'i°n Facilities,
Job MO. 6165-1, Ralph ?1. Parsons Company, July 1981.
noH S!iPU°rt and Env1™nmenta1 Impact Statement Volume 1-
Pla ? Wllw ll mfirFOrm?nCe f°r Pet™^m ^finery Sulfur Recovery
giants. tP<\ 450/2-76-016-a, September 1976.
op. 88-91?25 Rem°Val Pane1"5' The Oh1° and Gas Journal. September 11, 1976
Tyler' SOHI0' to Don R' Goodwi'n' u-s- EPA'
Br°C°fP> ™ "' ParS°n . EPA,
7. S"? Emissions in Matural Gas Production Industry -
^^
10. Reference 5
12. Reference 1
IS. Reference 13.
A-43
-------
17. Telephone Conversation - C. B. Sedman, U.S. EPA and L. Landrum,
ARCO, October 25, 1982.
13. Reference 5.
19. Reference 16.
20. Reference T.
A-44
-------
APPENDIX B
RESULTS OF COST ANALYSES
FOR INTERMEDIATE CONTROL SYSTEM
As a basis for comparison of an MSPS control svstem analyzed in
Appendix ^, s lower capital cost system with control efficiency somewher*
between that of a Claus and a Claus + reduction tail gas svstem is
evaluated in this Appendix. Currently, the only available svstem operating
in the United States and, hence a source of operating data, is the IFP-
1500 system. At present, it operates at four refineries of 100 180
250, and 400 LT/day capacities each.i From these sources, ooerating'data
-vere obtained to enable a rough cost estimate for a 100 LT/o'case as follows
B.I CAPITAL COST ESTIMATES
The 100 LT/D Claus plant from Figure A-2 is S4.50 x Ifl6. Th°
incinerator from Figure <\-4 is SO.41 x 1Q6. From Ffgure A.q' tne"stack
cost is estimated at 50.45 x 10" based on a 250 Ib/hr S02 emission rate
(93.66 percent sulfur recovery - see Reference 1). The heat recovery
system is identical to that of Case 3A at SO.55 x 10*.
The IFP-1500 at mo LT/D is reoorted to cost SI.234 x i06 for a
100 LT/n system and $2.35 x 10* for 180 LT/D, December 1975 basis 2
However, the 180 LT/D was a retrofit application. Therefore the'
SI.234 x 106 corrected to July 1982 is approximately $2.12 x lfl6 for *he
IFP portion of the Claus plant.
To make the system truly comparable to the cases examined in
Appendix A, a heat recovery boiler is also required, estimated at
$0.56 x 100. Therefore, the total investment is $8.04 x lfl6 for a 3_st
100 LT/D Claus plant with IFP-1500 tail gas treatment, incinerator with
waste heat recovery, and stack.
B.2 OPERATING COST ESTIMATES
All Claus operating costs will be taken by procedures in Appendix A
m most instances transformed directly from Case 3A. Fuel gas requirements
eor the incinerator, however, must be recalculated due to inlet gas
temperature differences. For simplicity, it is assumed that the steam
generation by Claus stages are identical to Case 3A, although in actual
practice, the first stage might be operated at higher temperatures
8-1
-------
(less net 250 # steam generation, more 50 * steam generation) than in
Glaus only operation, in order to minimize sulfide formation with carbon
dioxide (COS + CS?).
cor !FP operating costs, the following estimates for a 100 LT/D unit
are used based upon letters from operating facilities:^>4
utility requirements:
electricity 21 KWH/H
condensate 1.5 gpm
chemical/catalyst requirements (include routine make-up and periodic
inventory replacement)
solvent (PEG + salicyclic acid + sodium hydroxide): 124,000 Ib/yr
These figures are based on an assumed solvent inventory of 62 short
tons with 50 percent replacement annually and a complete inventory
replacement every two years; equivalent to a 52 short ton replacement
annually. Again, this is a simplification as the sodium hydroxide and
salicyclic ac^'d are replaced more -"requently than the polyethylene glycol
(PEG), but are minor ( 1 percent each) components of the overall solvent.
PEG costs in 1982 varied from 3.46/lb Gulf Coast to 3.53/lb West Coast,
so an average of S0.50/lb is used.^>^
Maintenance costs are assumed as an annual 3.55 percent of the IFP
capital cost. Two plants surveyed reported costs at 3.41 and 3.74 percent,
respectively.'' »3
All other costs are assumed similar to those in Appendix A and are
calculated as a function of capital and operating costs accordingly.
8-2
-------
B.3 LINE ITEM COSTS
Case -3C 100 LT/D (30% H2S)
Item
Capital Cost:
3-Stage Claus
IFP
Incinerator
Stack
Haste Heat Recovery
Operating Cost (credit)
500 psig
250 psig
50 psig
15 psig
condensate
electric power
fuel gas
catalyst
solvent
Source
Figure A-2
Section 8.1
Figure A-4
Figure A-9
Figure A-5
Table A-8
Table A-3/Section
Figure A-14/Section
calculated
Section A.2.2
Section B.2
Estimate
$4.50 x 106
$0.41 x 106
SO.45 x 106
SO.56 x 1Q6
"IFF
S2.12 x 106
The corresponding costs are tabulated
:iaus only case (3A) in Table 3-2.
(4600 Ib/hr)
(15740 Ib/hr)
(6040 Ib/hr)
(1240 Ib/hr)
B.2 (3420 Ib/hr)
8.2 212 KWH/H
6.15 x 1Q6 Btu/hr
2,084 Ib/yr
124,000 Ib/yr
750 Ib/hr
21 KWH/H
in Table B-l and compared to the
B-3
-------
Table B-l. LINE ITEM COST FOR CASE 3C
Capital Cost - $8.04 x 10s
Direct Operating Cost
1 = 15% i = 10% j = 20%
A. Utilities and Chemicals
1. treated boiler feedwater 392,350 392,350 392,850
2. electric power 97,860 97,860 97,860
3- fuel gas 180,810 180,810 180,810
4. catalyst 8,010 8,010 8,010
5. solvent 62,000 62,000 62,000
3. Labor
1. operators 169,360 169,360 169,360
2. supervision 82,340 82,340 82,340
C. Maintenance & Repair 252,860 252,860 252,860
0. Supplies and Lab Charges 33,370 33,870 33,870
cixed Charges
A. Capital 1,375,320 1,057,100 1,719,110
B. Taxes 80,^00 80,400 80,400
C. Insurance 48,240 48,240 48,240
Plant Overhead 105,550 105,550 105,550
General Expenses
A. Administrative 28,895 25,715 32,340
B. Distribution and Sales 28,895 27,715 32,340
Total Annualized Costs 2,947,260 2,622,680 3,297,940
Credits
1. 600 psig steam 280,140 280,140 280,140
2. 250 psig steam 892,460 892,460 892,460
3. 50 psig steam 291,730 291,730 291,730
4. 15 psig steam 46,870 46,870 46,870
5. stean condensate 28,035 28,035 28,035
6. sulfur 4,143,720 4,143,720 4,143,720
Total Credits 5,682,955 5,682,955 5,682,955
Nat Annual Operating Cost for Case 3C (52,735,695) ($3,060,275) ($2,385,015)
3-4
-------
Table 8-2. COST-EFFECTIVENESS OF IFP CONTROL
= 10 percent
Base Case Annual Cost, 5
Base Case SO? Removed, tons/yr
Claus + IFP Annual Cost, $
Claus + IFP S02 Removed, tons/yr
Cost Effectiveness, S/ton
(53,566,420)
75734.4
($3,060,275)
77349.44
5313
i = 15 percent
Base Case Annual Cost, $
Base Case SO? Removed, tons/yr
Claus + IFP A'nnual Cost, S
Claus + IFP SO? Removed, tons/yr
Cost-Effectiveness, S/ton
($3,313,660)
75734.4
(52,735,695)
77349.44
$353
i = 20 percent
Base Case Annual Cost, S
Base Case SOe Removed, tons/yr
Claus + IFP Annual Cost, S
Claus + IFP SO? Removed, tons/yr
Cost-Effectiveness, S/ton
($3,040,630)
75734.4
($2,385,015)
77349.44
$406
B-5
-------
B.4. REFERENCES
1. "Survey Report on SO? Control Systems for Mon-Utility Combustion and
Process Sources - May 1977", prepared by PEDCo Environmental, Inc.,
Contract Mo. 68-02-2603.
2. Telephone Conversation, C. 8. Sedman, U.S. EPA, and B. F. Ballard,
Phillips Petroleum Co., dated December 2, 1982.
3. Confidential Letter, C. Rice, Anoco, to C. Sedman, U.S. EPA, dated
October 18, 1982.
4. Confidential Letter, J. E. Hardaway, TOSCO, to C. Sedman, U.S. EPA,
dated January 14, 1983.
1. Reference 3.
'•>. Reference 4.
7. Reference 3.
•q. Reference 4.
B-6
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
|1 REPORT NO. I 2
EPA 450/3-33-014
TLE AND SUBTITLE
pftloT °f Jjew.. Source Performance Standards for
Petroluem Refinery Claus Sulfur Recovery Plants
|7 AUTHOR(S)
. PERFORMING ORGANIZATION NAME AND ADDRESS"
Office of Mr Quality Planning and Standards
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
AND ADDRESS
riffir* ^A-""":,1^ Plann1n9 and Standards
Office of_Air, Noise, and Radiation
U. b. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
is. SUPPLEMENTARY NOTES "
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
August 1983
6. PERFORMING ORGANIZATION CODE
. PERFORMING ORGANIZATION REPORT NC
. PROGRAM ELEMENT NO.
^ON I HACT/GRANT NO. ~
3. TYPE OF REPORT AND PERIOD COVERED
4. SPONSORING AGENCY CODE
EPA/200/04
(16. ABSTRACT
is document provides bsrknvnimH in-Po™*,-,.(-• -,
for claus sulfi DdCK9r°und information on sulfur emissions and
Federal emission regulation'sTre'su'mmarizeS ^cTu^r^185: ^ and
emphasis on factors which affect em ss inns Fm?!c Pr°cess is described with
with accompanying costs and performance data ntSIr" C°ntrols are also detailed
the New Source Performance Standard for M3?,±..^ !°,r ,a fo^-year review of
buo part J) as required by the
17.
DESCRIPTORS
' •
Air Pollution
Pollution Control
Standards of Performance
Claus Sulfur Plants
Tail Gas Treaters
Petroleum Refineries
13. DISTRIBUTION STATEMENT
Release unlimited
—•——i——__
EPA Form 2220-1 (9-73)
KEY WORDS AND DOCUMENT ANALYSIS
TERMS Ic^COSATI Field/G^T
Air Pollution Control
Sulfur Recovery Plants
13 B
19. SECURITY CLASS (THaRepon)
20. SECURITY CLASS (Thispage)
unclassified
21. NO. OF PAGES'
122
!2. PRICE
------- |