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

-------