./450/3-86/011
 ERA
             I States
             nmental Protection
             V
           Office of Air Quality
           Planning and Standards
           Research Triangle Park NC 27711
EPA-450/3-86-011
October 1986
           Air
Review of
New Source
Performance
Standards for
Petroleum Refinery
Fuel Gas

-------

                                   EPA-450/3-86-011
 I
   Review of New Source Performance
Standards for Petroleum Refinery Fuel Gas
 V
 ^
 I
 I
f^              Emission Standards and Engineering Division
X
f\
               U.S. ENVIRONMENTAL PROTECTION AGENCY
                    Office of Air and Radiation
               Office of Air Quality Planning and Standards
                 Research Triangle Park, NC 27711

                       October 1986

-------
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 endorsemnent 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



   1.1   CONTROL TECHNOLOGY                                         1-1



   1.2  ECONOMIC CONSIDERATIONS AFFECTING THE NSPS                 1-1






2.   CURRENT STANDARD                                               2-1



   2.1  NEW SOURCE PERFORMANCE STANDARDS                           2-1



   2.2   LEGISLATIVE HISTORY                                        2-12



   2.3  STATE REGULATIONS                                          2-15



   2.4  OTHER FEDERAL REGULATIONS                                  2-16



   2.5  REFERENCES                                                 2-18






3.   INDUSTRY DESCRIPTION



   3.1  BACKGROUND INFORMATION                                     3-1



   3.2  INDUSTRY CHARACTERIZATION                                  3-4



   3.3  EMISSIONS FROM COMBUSTION OF REFINERY FUEL GAS             3-8



   3.4  SELECTION OF REFINERY COMBUSTION  DEVICES FOR NSPS CONTROL  3-10



   3.5  REFERENCES                                                 3-10






4.  STATUS OF CONTROL  TECHNOLOGY                                   4-1



   4.1  ALKANOLAMINE PROCESS SYSTEM                                4-1



   4.2  THE LO-CAT® HYDROGEN SULFIDE OXIDATION PROCESS             4-8



   4.3  FACILITIES SUBJECT TO THE NSPS                             4-12



   4.4  COMPLIANCE TEST  RESULTS                                    4-13



   4.5  EMISSION MONITORING                                        4-15




   4.6  REFERENCES                                                 4-16

-------
                                                                   Page



5.  MODEL PLANTS                                                   5-1



   5.1  MODEL PLANTS AND CONTROL SYSTEMS                           5-1



   5.2  EMISSION REDUCTIONS                                        5-2






6.  COST ANALYSIS                                                  6-1



   6.1  INTRODUCTION                                               6-1



   6.2  AMINE TREATMENT WITH CLAUS SULFUR  RECOVERY                  6-2



   6.3  LO-CAT® PROCESS                                            6-6



   6.4  COST-EFFECTIVENESS                                         6-16



   6.5  COST COMPARISON                                            6-23



   6.6  REFERENCES                                                 6-24

-------
                               1.   SUMMARY

1.1  CONTROL TECHNOLOGY
     Currently, petroleum refineries are using two types of control
technologies, the alkanolamine or the LO-CAT® system, to comply with the
hydrogen sulfide (H2S) concentration limit (230 mg/dscm) of this new
source performance standard (NSPS).  Data for fifteen compliance tests
were obtained from nine different refineries.  All of the test data  are
from refineries that use an alkanolamine system.  The compliance test
results range from 2.97 mg/dscm to 119.43 mg/dscm.
     No compliance test data were obtained from a LO-CAT® system; however,
one refinery with such a system has installed a continuous H2$ monitor.
According to information submitted by the refinery, the f^S concentration
ranges from 20 ppmv to 50 ppmv and averages 30 ppmv {NSPS is 162 ppmv).
     The review did not find any demonstrated technologies for controlling
emissions that achieve more control than the alkanolamine or the LO-CAT®
systems.
1.2  ECONOMIC CONSIDERATIONS AFFECTING THE NSPS
     Another primary issue involving review of the NSPS is the cost  of
controls.  The cost effectiveness of controlling the H2$ concentration in
refinery fuel gas was estimated for four model plant sizes at three  H2S
concentrations for both types of control systems.  For most of the new NSPS
units, the costs of control per unit of sulfur dioxide (S02) removed will
be less than those discussed in this section.  The cost effectiveness
ranges from $201 per ton of S02 to a credit of $23.9 per ton of S02  for
the alkanolamine system with a Claus sulfur recovery unit.  The cost
effectiveness for the LO-CAT® system ranges from $89 per ton to $399 per
ton of S02 removed.

-------
                          2.  CURRENT STANDARDS

     This chapter presents and discusses  the  current  regulations  for  air
pollutant emissions from refinery fuel gas combustion devices.   Federal
regulations for new sources, other Federal  regulations,  and State regulations
(for exisiting and new sources) are all  addressed in  order to give an overall
picture of the regulatory structure for this  emission source category.   The
focus, however, is on the new source performance standards (NSPS) for sulfur
dioxide emissions from refinery fuel  gas  combustion devices.
     A summary of the NSPS is first presented,  followed  by detailed discussions
of the specific requirements, definitions,  and  specifications of  the  NSPS.
This is followed by a brief description of other Federal  and State regulations
that may also affect existing and new sources in this category.

2.1   NEW SOURCE PERFORMANCE STANDARDS
2.1.1  Background
     New source performance standards regulate  emissions of air  pollutants
from new, modified, and reconstructed facilities in various industrial categories.
The regulations establish emission limits and require emission performance
testing, continuous monitoring, and periodic  reporting.   The authority for
the NSPS regulations is granted to the U.S. Environmental  Protection  Agency
(EPA) under Section 111 of the Clean Air  Act.1
     The regulation for fuel  gas combustion devices in petroleum  refineries
is listed in Subpart J of 40 CFR 60, (Code  of Federal  Regulations;  Title 40 -
Protection of Environment; Part 60 - Standards  of Performance for New

-------
                                     2-2

Stationary Sources;  Subpart J - Standards of Performance for Petroleum Refineries)
Subpart J addresses  specific requirements for this source category,  but
Subpart J also incorporates the general  requirements for any NSPS.   These
general requirements are listed in Subpart A (General  Provisions)  of 40 CFR 60.
     Other sources of air pollution emissions from petroleum refineries are
also regulated under the new source performance standard regulatory  program.
Subpart J also regulates sulfur dioxide  emissions from Claus sulfur  recovery
plants, and sulfur dioxide, carbon monoxide, and particulate emissions from
fluid catalytic cracking unit catalyst regenerators.  Subpart H regulates
sulfuric acid mist and sulfur dioxide emissions from sulfuric acid plants.
Volatile organic compound (VOC) emissions from storage vessels for petroleum
liquids are regulated under Subparts K and Ka.  Subpart GGG regulates VOC
emissions due to leaks from process equipment.
2.1.2  Summary of the NSPS for Refinery  Fuel Gas Combustion Devices
     New source performance standards were promulgated by the EPA  on March  8,
1974, limiting emissions of sulfur dioxide ($02) from new, modified  and
reconstructed fuel gas combustion devices at petroleum refineries.   No signif-
icant changes have occurred since then.   These standards apply to  an affected
facility which commences construction or modification after June 11, 1973.
     The affected facility for this standard is any fuel gas combustion
device in a petroleum refinery.  These devices are defined as any  equipment
used to burn fuel gas, such as process heaters, boilers, and flares, but some
combustion sources in a refinery are specifically exempted in the  definition.
     The regulated air pollutant is S02-  Sulfur dioxide emissions from fuel
gas combustion devices can be controlled by reducing the hydrogen  sulfide

-------
                                     2-3

content of the fuel  gas prior to combustion or by flue gas desulfurization
(FGD) after combustion.  The standard was written with the intent to limit
the H2$ content of fuel gas, although the owner/operator has the option of
using FGD.  The standard prohibits the burning of fuel gas containing more
than 230 mg HgS/dscm (0.10 gr/dscf)  in any fuel  gas combustion device; however,
the standard does not apply to unusual situations, such as emergency gas
releases or process  upsets.  Compliance is demonstrated by an initial
performance test using EPA Method 11.  Subsequent continuous monitoring of
H2$ in the incoming  fuel  gas is required; however, since monitor performance
specifications have  not been established yet,  this monitoring requirement is
not in effect.
     An alternative  compliance option is included.  Instead of controlling the
H2S in the incoming  fuel  gas, the S02 emissions  may be controlled directly by
treating the effluent gases resulting from the combustion of fuel  gas.
However, it must be  shown that treating the effluent combustion gases will
control S02 emissions as effectively as controlling the h^S in the incoming
fuel gas.  Compliance for this option is demonstrated by an initial  performance
test using EPA Method 6,  and continuous monitoring of S02 in the effluent gas
is required.
     The regulation  also specifies a series of reporting and recordkeeping
requirements.  A refinery that has combustion  devices subject to the NSPS
is required to keep  records, submit  reports to EPA, and notify EPA of particular
plans and occurrences as described in section  2.1.8.

-------
                                     2-4

2.1.3  Applicability of Standards2
     2.1.3.1  Affected facilities.
     The NSPS is applicable to any new,  modified,  or reconstructed combustion
device which commenced construction after June 11, 1973,  and which burns fuel
gas in a petroleum refinery.
     Petroleum is defined as,
         "the crude oil removed from the earth and the oils derived from
          tar sands, shale, and coal."
     A petroleum refinery is defined as,
         "any facility engaged in producing gasoline, kerosene,  distillate
          fuel oils, residual fuel oils, lubricants, or other products
          through distillation of petroleum or through redistillation,
          cracking or reforming of unfinished petroleum derivatives."
The EPA's definition of a petroleum refinery is thus interpreted broadly to
encompass shale oil refineries, solvent refined coal plants, one-step topping
plants, and small re-refining operations.
     Fuel gas is defined as,
          "any gas which is generated at a petroleum refinery and which is
           combusted.  Fuel gas also includes natural gas when the natural
           gas is combined and combusted in any proportion with  a gas
           generated at a refinery.  Fuel gas does not include gases
           generated by catalytic cracking unit catalyst regenerators and
           fluid coking burners."
The specific exemption is included for gases generated by these  particular
processes because it is impractical to control the S02 emissions that would
result from burning the H2$ in these gases.  These off-gases contain relatively
low levels of H2$ and contain very high levels of carbon dioxide, making it
difficult to reduce the H2S concentration further in conventional amine
treating units.  However, if these exempted, off-gases are combined with fuel

-------
                                     2-5

gas from another part of a refinery, then the combined gas stream would be
subject to the NSPS limit if it is combusted in a new combustion device.
     Natural gas refers to pipeline standard natural  gas (meeting specifications
of < 0.25 grains H2S / 100 scf).   Thus,  if natural  gas is purchased and burned
exclusively in a combustion device, then H2$ content  of the gas will  necessarily
be below the NSPS limit.  If natural gas is mixed with refinery generated
fuel gas, then the combined stream is considered "fuel  gas" and its H2S
                                                                >
concentration must be under the NSPS limit prior to combustion in a new source.
     A fuel gas combustion device is defined as,
         "any equipment, such as  process heaters, boilers and flares used
          to combust fuel gas, except facilities in which gases are
          combusted to produce sulfur or sulfuric acid."
There are two reasons for including the  specific exemption for combustion
devices used to produce sulfur or sulfuric acid.  First, the combustion in
this case is a step in the chemical conversion process, and the resulting
post-combustion stream is considered a process stream that undergoes further
processing, not an effluent stream.  Second, there are separate NSPS's limiting
air emissions from these processes.  (Sulfuric acid plants are regulated
under Subpart G; Claus sulfur recovery plants under a different part of
Subpart J.)
     2.1.3.2  Applicability date.
     The NSPS applies only if the construction or modification commenced
after June 11, 1973, (the date of the original proposal of the regulation).
The term "commenced" is defined in the General Provisions to 40 CFR 60,
(Section 60.2),

-------
                                     2-6
         "Commenced means that an owner or operator has  undertaken  a
          continuous program of construction  or modification  or  that  an
          owner or operator has entered into  a  binding agreement or
          contractual  obligation to undertake and complete, within  a
          reasonable time, a continuous program of construction  or
          modification."
Thus, a fuel  gas combustion device that existed prior to the  proposal  on
June 11, 1973, and has not been significantly changed or altered since then
would not be regulated under the NSPS.
     2.1.3.3  Modification.
     While NSPS are intended primarily  for newly constructed  facilities,
existing sources can become subject to  an NSPS  through either "modification"
or "reconstruction."  These terms are defined in detail  in  the General Provi-
sions for Part 60, (40 CFR 60.14 and 40 CFR 60.15).
     An existing fuel  gas combustion device becomes subject to the  NSPS  under
the modification provision if there is  any physical or operational  change
that causes an increase in the emission rate.  A number  of  clarifications,
exemptions, and exceptions to the modification  provision are  listed.   The
following actions by themselves are not considered to be modifications:
          0 routine maintenance, repair, and replacement
          0 production increases achieved without any capital  expenditure
          0 production increases resulting from an increase in the  hours
            of operation
          0 use of an alternative fuel  if the existing  facility  was
            originally designed to accommodate  such an  alternative  use
          0 addition or replacement of  equipment for emission control  (as
            long as the replacement does not increase emissions)
          0 relocation or change of ownership of an existing  facility.

-------
                                     2-7

Also, the addition or modification of one facility at a source will  not cause
other unaltered facilities at that source to become subject to the NSPS.
Specifically, for petroleum refineries, if one fuel gas combustion device is
added or modified, then only that particular device must meet the NSPS; the
other existing combustion devices throughout the refinery are not affected.
     2.1.3.4  Reconstruction
     An existing facility becomes subject to the NSPS upon reconstruction
regardless of any change in the rate of emissions.  Reconstruction is defined
as the replacement of components of an existing facility to the extent that
the cumulative fixed capital cost of the new components exceeds 50 percent
of the cost that would be required to construct a comparable entirely new
facility.
2.1.4  Controlled Pollutant
     The NSPS limits emissions to the atmosphere of SOg from the exhaust of
refinery combustion devices which burn fuel gas.  Although the regulated air
pollutant is S02, the S02 emissions to the atmosphere are not regulated directly,
Instead SOg emissions are controlled indirectly by regulating the amount of
H2$ in the incoming fuel gas.
2.1.5  Exceptions to the Emission Standard
     The standard prohibits the burning of fuel gas containing more than 230
mg HgS/dscm (0.10 grain/dscf) in any new, modified, or reconstructed combustion
device.  The H2S content of the incoming fuel gas can be reduced in three
ways - by using sweeter crude with a lower sulfur content, by pretreatment of
the fuel gas before combustion in an acid gas treating unit, and by blending
natural gas with the fuel gas.  Although this mixing dilutes the H2S concentra-
tion without reducing overall S02 emissions, the blending of natural gas with

-------
                                     2-8

fuel gas Is not considered a circumvention of the standard, because this is
often normal  and necessary refinery operating practice.   Natural  gas is used
as auxiliary fuel since many refineries require more gas than they generate.
Also, natural gas is used to ensure a constant flow of fuel to processes,
while the amount of refinery generated fuel  gas may vary with operating
conditions or upsets in other process units.
     The standard does not apply to unusual  situations,  such as emergency gas
releases due to process upsets.  Process upset gas is defined as,
         "any gas generated by a petroleum refinery process unit
          as a result of start-up, shut-down, upset, or  malfunction."3
Start-up, shut-down, upset, and malfunction considerations are defined in the
General Provisions for 40 CFR Part 60.  The combustion of process upset gas
in a flare, or the combustion in a flare of process gas  or fuel gas which is
released to the flare as a result of relief valve leakage is specifically
exempted from this regulation.  However, flares which burn a continuous
process gas stream are subject to the NSPS regulations.
2.1.6  Testing Requirements
     The owner or operator of a fuel gas burning device  subject to NSPS
is required to conduct performance tests within a specified period after
start-up, and thereafter from time to time as may be specified by the EPA.
These performance tests are required in order to demonstrate that the standards
are being met by the new device.  General testing and reporting requirements
are listed in the General Provisions, (Section 60.7), while testing details
specific to this source category are found in Subpart J, (Section 60.106).
     The initial test of performance of a facility must be conducted within
60 days after the facility first achieves its maximum intended rate of operation.

-------
                                     2-9

However, If the Intended rate of operation is not achieved within 120 days of
initial start-up, the initial test must nevertheless be conducted within 180
days of start-up.  Thirty days must be allowed for prior notice to the EPA,
to allow the Agency to designate an observer to witness the test.
     To demonstrate compliance with the standard limiting the amount of ^S
in the fuel gas prior to combustion, EPA Reference Method 11  is used to
determine the concentration of H2S.  A performance test consists of 3 runs,
                       ,/
with each run consisting of 2 samples.  Samples are taken at approximately
one-hour intervals with a minimum sampling time of 10 minutes per sample.
The arithmetic mean of the three runs constitutes the value used to determine
whether the facility is in compliance.  (Necessary modifications in the
details of the test methods may be made, if approved in advance by the EPA.)
A written report of the test is to be furnished to the EPA.
2.1.7  Monitoring Requirements
     The regulation requires a continuous H2$ monitor on the fuel gas line at
the inlet to each combustion device that is subject to the NSPS.  When a
refinery has several fuel gas combustion devices having a common source of
fuel gas, monitoring may be done at one location instead of each combustion
device having a separate monitor.  This situation is common in many refineries
where a centralized acid gas treatment plant treats H2S rich gases from
several refinery process units and then routes the treated exit gases to
combustion devices throughout the refinery.  Excess emissions are defined  as
any 3 hour period when the integrated (or arithmetic) average concentration
of H2S in the fuel gas exceeds the standard of 230 mg H2S/dscm.

-------
                                     2-10

     The monitoring system must continuously monitor and record the H2S
concentration.  Under the General  Provisions (Section 60.13),  "continuous"  is
defined to mean that the monitoring system must complete at least one cycle
of operation (sampling, analysis,  and data recording) for each successive
15-minute period.   The owner or operator must install, calibrate, maintain,
and operate the continuous monitor according to the requirements which are
detailed in the subpart and the general  provisions.  Also, the continuous
monitoring system must satisfy the performance specifications  in Appendix B
of 40 CFR 60.
     The EPA has not yet developed instrument performance specifications for
H2S continuous monitoring systems.  Therefore, refinery combustion devices
subject to the NSPS are effectively exempt from the H2S monitoring requirements
until EPA establishes performance  specifications for an H2S monitor.
2.1.8  Recordkeeping and Reporting Requirements
     A refinery that has combustion devices subject to the NSPS, is required
to keep records, submit certain reports to EPA, and notify EPA of certain
plans and occurrences.
     One-time "notification" reports concerning the start of construction or
reconstruction, anticipated and actual startup dates, and physical or operational
changes to existing facilities are required so that the EPA will be able to
identify affected sources.  In addition, other records and reports are necessary
to enable the EPA to identify sources that may not be in compliance with the
standard.  These include initial performance test results, quarterly reports
of excess emissions, and retaining records of plant upsets and excess emissions
for 2 years.  Details of these requirements are listed in the  General Provisions
(40 CFR 60.7).

-------
                                     2-11
2.1.9  Option for an SO? Emission Limit
     Sulfur dioxide emissions from fuel gas combustion devices can be controlled
by reducing the H2S in the incoming fuel  gas or by reducing the 862 in the
exhaust gases.  The standard was written  with the intent to limit the ^S
content of fuel gas, and all of the refinery sources subject to the NSPS to
date are currently reducing S02 emissions by pretreatment of the fuel gases
to remove the ^S.
     However, the regulation includes a second, alternative provision which
allows the burning of fuel  gas with a higher H2S content provided the effluent
gases are treated to reduce the S02 emissions.  This flue gas desulfurization
would probably be accomplished with an add-on air pollution control  device.
In case this emission control option is selected, the regulation sets forth a
parallel set of emission standards and testing, monitoring, reporting, and
recordkeeping requirements.  Because no refinery has yet elected the FGD
options, the alternative provisions in the regulation will  not be discussed
in detail in this report.
     The NSPS does not set a specific S02 emission limit if this alternative
approach is followed.  Instead, the source must calculate and determine an
equivalent S02 emission level that would  control S02 emissions as effectively
as complying with the standard for H2$ concentration in the fuel  gas.  Because
the inlet fuel gas streams, operating conditions, and parameters may vary for
each combustion device, an equivalent S02 emission limit would probably need
to be calculated for each affected facility on a case-by-case basis.  The
information must be submitted to the EPA  for approval.

-------
                                     2-12

     Compliance for this option is demonstrated by a performance  test using
EPA Methods 1, 2, and 6 for determining the sampling site,  volumetric flowrate,
and S02 concentration.  Continuous monitoring of the $02 concentration in  the
effluent is also required in the regulation.   Since EPA has listed performance
specifications for continuous SOg instruments (in Performance Specification 2
of Appendix B to 40 CFR 60), these monitoring requirements  are in effect and
must be carried out.  Excess emissions are defined as a 3-hour period when
the average S02 concentration exceeds a predetermined level  that  was previously
calculated and approved for that particular combustion source. The testing,
monitoring, recordkeeping, and reporting requirements are similar to those
discussed for the H2$ concentration standard, and details are listed in
Subpart J and the General Provisions.
2.2  LEGISLATIVE HISTORY
     Standards of performance for air emissions from petroleum refineries
were first proposed on June 11, 1973 (38 FR 15406) and promulgated on
March 8, 1974 (39 FR 9308).  Since then, there have been 3  proposed and
8 final rulemakings which affected the standard for sulfur  dioxide emissions
from fuel gas combustion sources in refineries.  These rulemakings consisted
of minor clerical corrections, changes to the monitoring requirements, and
changes to the definition of fuel gas.  A listing and brief summary of these
is given in Table 2.1.
2.2.1  Changes to Definitions
     The definitions of "fuel gas" and "fuel  gas combustion device" were
changed to clarify the original intent of the regulation and to match the
conventional nomenclature used in the industry.  These formal changes to

-------
                                     2-13
                                TABLE 2.1

                  LIST OF RULEMAKINGS AFFECTING NSPS FOR
                   REFINERY FUEL GAS COMBUSTION DEVICES
  DATE
06-11-73
03-08-74
09-11-74
10-06-75
07-25-77



08-17-77


03-03-78
10-04-76
03-15-78
FEDERAL REGISTER
   CITATION
 38 FR 15406
 39 FR  9308
 39 FR 32852
 40 FR 46250
 42 FR 37936




 42 FR 41424


 43 FR  8800
 41 FR 43866
 43 FR 10866
TYPE OF
ACTION
Proposal
Final
Proposal
Final
        SUMMARY OF ACTION
Final



Final


Final
Proposal
Fi nal
0 Original  regulation limiting S02
  emissions and including testing,
  monitoring, and reporting requirements.

0 Added universal monitoring and reporting
  requirements to General Provisions,
  (40 CFR 60.7 & 60.13).
0 Eliminated monitoring requirements
  for H2$ in fuel gas; revised
  monitoring requirements for S02 in
  exhaust gas.

0 Clarification, adding applicability
  date to the definition of the
  affected facility.

0 Clerical  change to revise statuatory
  authority citations for clarity.

0 Clerical  change to amend statuatory
  authority citations per Clean Air
  Act Amendments of 1977.

0 Re-added monitoring requirements for
  H2S in fuel gas.
03-12-79
03-03-80
12-01-80
 44 FR 13480
 45 FR 13991
 45 FR 79452
Final       ° Change definition of "fuel gas" and
              "fuel gas combustion device" to
              clarify when an incinerator-waste
              heat boiler is affected by the NSPS.

Proposal    ° Change definition of "fuel gas" to
Final         clarify which gaseous fuels are
              covered by NSPS, particularly when
              "natural gas" is a "fuel gas".

-------
                                     2-14

the regulation were Initiated in part as a response to several  questions
which had arisen concerning applicability determinations.   Since the changes
were merely clarifying the original  intent of the regulation, these changes
did not have a significant impact on emissions.
2.2.2  Monitoring
     The original standard included  requirements for the monitoring of the
H2$ at the inlet or S02 at the outlet of a fuel  gas combustion  device.  At
the time, however, no guidance or instructions were provided on how to conduct
the monitoring, and so sources were  temporarily exempted from monitoring.
(Incidentially, this approach was taken for all  of the early NSPS regulations
covering several industries because  no monitoring guidelines had been set by
EPA at the time for any pollutant.)
     Major changes and additions to  monitoring requirements were issued a
short time later (proposed 9/11/74 and promulgated 10/6/75).  These changes
added overall monitoring and reporting requirements to the General Provisions
for all NSPS's and set forth detailed performance specifications for SOg
monitoring instruments.  Concurrently, there were coordinating  revisions
to the S02 monitoring requirements in Subpart J for the exhaust gases from
refinery combustion devices.  At the same time, all monitoring  requirements
for H2S levels in the inlet gases were eliminated because specifications for
those instruments had not yet been set.  Then, this approach was reversed in
a later rulemaking (proposed on 10/4/76 and promulgated on 3/15/78), when
monitoring requirements for H2S in the fuel gas were reinstated; however,

-------
                                     2-15

since detailed performance specifications for H2S monitors have not been
promulgated, the H2S monitoring requirements still do not have to be carried
out.  The overall  result of these actions and current status of requirements
for H2$ monitoring is that universal, general monitoring requirements and
guidelines are in  place but specific, detailed requirements for H2S monitors
have not yet been  determined.  Thus, no H2S monitoring is required at this
time.
2.3  STATE REGULATIONS
2.3.1  State Regulations for Existing Sources
     A review of State regulations has shown a wide variation in the types of
regulations, degree of stringency, and methods of enforcing the limitations.
Many States have several forms of regulations, each applying to a different
fuel or type of source, and do not necessarily conform to or parallel the
NSPS format.  For  example, with regard to source category, the NSPS specifies
fuel gas combustion devices in petroleum refineries, whereas a State regulation
may specify a different source category:  petroleum refinery combustion
sources; a combustion source in any industrial facility; combustion sources
that burn gaseous  fuels.
     State regulations may limit emissions of S02 by limiting the H2S in the
fuel gas (as does  the NSPS), the total S02 emissions from the combustion
device, or the total S02 emissions from the petroleum refinery.  In general,
S02 emissions are  limited by a regulation restricting the quantity of S02
emitted per unit quantity of heat input or by limiting the sulfur content of
the fuels.  In some States, the regulation specifies the maximum allowable
ground level S02 concentration resulting from the emissions.

-------
                                     2-16

     Sulfur dioxide regulations fall  Into  one of the following regulatory
formats:
      1.  ppm S02,  by volume,  in the effluent
      2.  Pounds of S02 per million  Btu's of heat Input
      3.  Requirements on the  sulfur content of the fuel,  such  as  ppm HSS In
         the fuel  gas or weight percent sulfur In fuel  oil
      4.  Ambient air quality  levels similar or the same as  the National
         Ambient Air Quality  Standard (NAAQS) for S02
2.3.2  State Regulations for  New Sources
     For new sources, most States have been delegated enforcement authority
for the NSPS for the petroleum refining industry, and most  have adopted  the
NSPS as written.  No State has adopted an  emission limit more  stringent  than
the NSPS limit of 0.10 gr H2S/dscf;  although some States or local  air pollu-
tion control agencies have adopted more stringent monitoring requirements.
The Puget Sound agency (in Washington State) and three counties in southern
California have required continuous emission monitoring with a continuous
automated instrument even though this requirement is officially not in effect
under the NSPS until Instrument performance specifications  are promulgated.
2.4  OTHER FEDERAL REGULATIONS
     There has been additional regulatory  activity since the promulgation of
the current NSPS which affects the emissions from refinery  fuel gas combustion
sources.   In addition to State regulations (for existing and new sources) and
the NSPS, some petroleum refineries may be required to achieve more stringent
emission levels from fuel gas combustion devices under regulations for the
Prevention of Significant Deterioration (PSD)4 or under State  Implementation
Plans (SIP) which are subject to nonattainment review by EPA.5
2.4.1  PSD Regulations
     If a new facility is built in an area which is attaining  the NAAQS  for

-------
                                     2-17
S02, then it falls under the PSD regulations and must use the best available
control technology (BACT).   In general,  BACT determination is applied on a
case-by-case basis and must be at least  as stringent as the NSPS level.   For
this source category, the BACT emission  level  has been defined as equal  to
the NSPS level.6  Some State and local enforcing agencies have also used the
PSD regulations as a means  of requiring  some form of emission monitoring and
corresponding emission monitoring reports.
2.4.2  Nonattainment Area Regulations
     If a new source is located in a nonattainment area for NAAQS for S02,
then emission control technology capable of the lowest achievable emission
reduction (LAER) is required.  In general, LAER is at least as stringent
as the NSPS, but for this source category, LAER has also been defined as
equal to the NSPS.7
2.4.3  Other NSPS Regulations
     The NSPS under review in this study limits-the S0£ emissions from all
new, modified, or reconstructed fuel gas combustion devices, which include
process heaters, boilers, and flares used to combust fuel  gas.  The air pollution
emissions from these combustion sources  may also be regulated by other
NSPS's (either current or under development).   The additional requirements
and restrictions imposed by these other  NSPS's do not affect or conflict
with the emission reductions required by the fuel  gas NSPS, but they may
affect the planning and design of new combustion sources by refinery owners.
     A new or modified boiler which has  a heat input greater than 250 million
BTU per hour may also be subject to the  provisions of Subpart D or Subpart Da
of 40 CFR 60; the applicable subpart is  based on the date of construction
of the source and whether the source is  classified as an industrial boiler

-------
                                     2-18

or a utility boiler.   These subparts limit opacity,  participate  emissions,
and NOX emissions,  as well  as S02 emissions.   Similarly,  a  medium-sized,  new
or modified boiler  (100 to  250 million  BTU per hour)  may  be subject  to  the
provisions of Subpart Db, which is currently  under  development.
2.5  References
1.   Clean Air Act  As Amended, August 1977.   42 U.S.C.  Title  I--Air
     Pollution Prevention  and Control.   Part  A--Air Quality and  Emission
     Limitations; Section  111—Standards of Performance for New  Stationary
     Sources.  Washington,  D.C.
2.   U.S. Environmental Protection Agency. Code of Federal  Regulations.
     Title 40, Part 60.  Sections 60.101.   Office of the  Federal  Register.
     Washington, D.C.  July 1, 1985.
3.   Same as reference #2.
4.   Reference 1.   Part C--Prevention of Significant Deterioration of Air
     Qua!i ty.
5.   Reference 1.   Part D—Plan Requirements  for Nonattainment Areas.
6.   U.S. Environmental Protection Agency. BACT/LAER Clearinghouse  -
     A Compilation  of Control Technology Determinations.   U.S. EPA,
     Research Triangle Park, N.C.  EPA 450/3-85-016 (a-d).   June 1985.
7.   Same as reference #6.

-------
                         3.  INDUSTRY DESCRIPTION

     This chapter describes a typical petroleum refinery,  its various
production processes and the range of products produced by a refinery.
Also, the current number of operating refineries and their geographical
location are discussed as well  as the industry growth rate.  Finally,
emissions from combustion of refinery fuel  gas and the rational  for
choosing refinery fuel gas for new source performance standard control
are discussed.
3.1  BACKGROUND INFORMATION
     A petroleum refinery transforms crude oil into a variety of useful
products.  The petroleum refining industry produces more than 2500 products
that can be categorized into the following classes:   fuel  gas, gasoline,
kerosene, fuel oil, lubricating oil, grease,  wax, asphalt, coke, chemicals,
and solvents.  There is no "typical" refinery, since the number of products
and the product mix varies widely within a refinery as well as between
refineries.  The manufacturing processes also vary depending on refinery
age, type of technology, capacity, location,  and type of crude processed.
     Petroleum refinery operations involve physical  separation of components
of the crude oil  (e.g., crude distillation) and chemical conversion
processes which transform some of the less useful components of the oil
into more useful  products (e.g., cracking of  high molecular weight oils
into lower molecular weight products such as  gasoline).
     The processing sequence of a refinery is illustrated  in Figure 3-1.
The crude oil is heated and charged to an atmospheric distillation

-------
              3-2
ATMOSPHESIC CHUDE OIL SEPAHAIIOH UNIT
                                                                      -o
                                                                      72
                                                                      O
                                                                      00
                                                                      oo
                                                                      o    m

                                                                            GO
                                                                      3»     I

-------
                                   3-3
tower where it is separated into several light, intermediate, and heavy
fractions.  The bottoms from the tower are sent to a vacuum distillation
unit for further separation.  The bottoms from the vacuum still are
thermally cracked in a coker to produce a wet gas, coker gasoline,
and coke.  A portion of the bottoms from the vacuum still may be processed
into asphalt.   Gas oils from the atmospheric and vacuum distillation
units are used as feedstocks for the catalytic cracking and hydrocracking
units.  These units convert the gas oils to gasoline and distillate fuel.
The gasoline from these units is fed to a catalytic reformer to improve
the octane number and then blended with other refinery streams to make
gasolines for marketing.
     The wet gas streams from the distillation, coker, and cracking units
are combined and fractionated into fuel gas, liquified petroleum gas, and
unsaturated and saturated branched chain and straight chain, light
hydrocarbons containing from three to five carbon atoms.  The fuel gas is
used as fuel in the refinery furnaces.  The straight chain saturated
hydrocarbons are blended into gasoline.  The unsaturated hydrocarbons and
the branched chain hydrocarbons, primarily isobutane, are processed in an
alkylation unit.  In the alkylation unit the unsaturated hydrocarbons
react with isobutane to form isoparaffins which are blended into gasoline
to increase the octane.
     The middle distillates from the crude unit, the coker unit, and the
cracking unit are blended into diesel and jet fuels and furnace oil.
Heavy vacuum gas oils and reduced crude oil  from some crudes can be
processed into lubricating oils, waxes, and grease.

-------
                                   3-4

3.2  INDUSTRY CHARACTERIZATION
     As of January 1,  1985,  there were  191  operable  refineries  in the
United States with total  crude oil  distillation  capacity  of  15.1 million
barrels per calendar-day^/ and 15.9 million barrels  per stream  day,]V
Table 3.1 lists the number of operating refineries along  with the total
crude capacity located in each State.   These refineries are  distributed
among 35 states, with 79 refineries (41  percent) being concentrated  in
the three major refining States of Texas,  California,  and Louisiana.
These three States account for an even  higher percentage  (57 percent) of
the total U.S. crude oil  refining capacity.  Texas alone  accounts for
approximately 27 percent of the total  crude oil  refining  capacity, while
Louisiana and California account for about 15 percent  each.
     In addition, there are four operating refineries  in  Puerto Rico, Virgin
Islands, Guam, and the Hawaiian Foreign Trade Zone with a combined crude
capacity of 771,000 barrels per calendar day.2  Because these are territories
and not formally part of the United States, these refineries are usually
not included in industry studies and figures for the U.S.; however,  these
refineries are regulated under federal  EPA new source  performance standards.
     £/ Barrels per calendar day (b/cd):   the average volume a refinery unit
processes each day including downtime used for turnarounds.  This is actual
total volume for the year divided by 365.
     ]V Barrels per stream day (b/sd):   the amount a unit can process
running at full capacity under optimal  crude and product slate conditions
for short periods.

-------
                                   3-5
                                TABLE 3.1



         OPERATING REFINERIES IN THE U.S. (AS OF JANUARY 1, 1985)3
State
Al abama
Alaska
Ari zona
Arkansas
California
Colorado
Delaware
Georgia
Hawaii
Illinois
Indiana
Kansas
Kentucky
Louisiana
Mary! and
Michigan
Minnesota
Mississippi
Montana
Nevada
New Jersey
New Mexico
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Tennessee
Texas
Utah
Virginia
Washington
West Virginia
Wisconsin
Wyoming
No.
Plants
1
4
1
4
30
3
1
2
2
8
5
7
2
16
1
4
2
5
6
1
5
3
2
5
5
1
8
1
33
6
1
7
2
1
6

b/cd
80,000
138,930
5,000
69,170
2,265,098
94,700
140,000
28,800
1'09,500
946,000
431,300
338,000
218,900
2,188,793
14,200
119,400
204,143
362,400
147,500
4,500
503,000
63,050
62,800
515,700
374,000
15,000
658,700
57,000
4,145,900
154,950
51,000
410,550
16,500
39,000
162,778
Crude Capacity
b/sd
81,300
142,910
5,263
70,950
2,381,417
98,500
150,000
32,000
118,426
1,003,550
445,500
352,383
226,000
2,280,958
14,947
•126,094
211,220
383,104
154,147
4,700
533,210
66,000
65,400
540,000
390,394
15,789
691,300
60,000
4,385,273
161,868
53,000
427,543
17,000
40,000
168,052
Total
191
15,136,262
15,898,198

-------
                                   3-6
     Since January 1,  1981,  a net total  of 101  refineries  have  been  shut
down with a total  capacity of 2.5 million barrels  per calendar  day.
(Table 3-2 and Table 3-3).  During 1984,  a total of  26 refineries  were
shut down and 2 refineries were started  up resulting in a  net decrease  of
24 plants with an  associated loss in crude distillation capacity of
0.5 million barrels per calendar day. The majority  of these  closings
occurred at refineries with crude distillation  capacity of 30,000  barrels
per calendar day or less.   These closings accounted  for a  net reduction
of 19 facilities.   Refineries with crude distillation capacity  greater
than 30,000 barrels per calendar day showed a net  reduction of  5
facilities during  1984.
     Refinery utilization  (actual production vs. production capacity)
peaked at 78.2 percent in  August 1984;  the average rate for the year rose
to 76.2 percent, from the  previous year's average  of 71.7  percent.4
     Total downstream charge capacity on January 1,  1985,  stood at 28.3
million barrels per stream day, a net decrease  of  0.4 million barrels  per
stream day.  Downstream charging capacity includes the following processes:
vacuum distillation, thermal operation,  catalytic  cracking (fresh  and
recycled), catalytic reforming, catalytic hydrocracking, and  catalytic
hydrotreating.  New construction at existing refineries and the start-up
of previously closed refineries was more than offset by the nearly 0.7
million barrels per stream day capacity  closed  during 1984.  The most
significant declines were  in vacuum distillation and catalytic  reforming,
dropping 2 percent and 4 percent, respectively. However,  during  the
year, downstream charge capacity increased 3 percent for catalytic cracking
(recycled) and nearly 11 percent for catalytic  hydrocracking.5

-------
                                   3-7
                                TABLE 3-2
           Number of Operable Refineries in the U.S.6>7«8»9>10
Year
1981
1982
1983
1984
1985
Total
303
273
225
220
191
                                TABLE 3-3
                  Crude Oil  Disti11ationll,12,13,14,15,16
                 (Thousands  of Barrels per Calendar Day)
U.S. Capacity
Puerto Rico
Virgin Islands and Guam
Hawaiian Foreign Trade Zone
1981
18,465
228
714
68
1982
17,669
255
744
68
1983
16,157
244
559
60
1984
15,862
121
558
60
1985
15,13'
121
588
62

-------
                                   3-8
     Refiners project that  total  downstream charge  capacity  may  drop  to
28.1  million barrels per stream day by 1986.  However,  the downstream
processing mix is not projected to change appreciably  from January  1,
1985, levels.  All  major downstream refinery processes, except catalytic
hydrotreating, are expected to fall  below current year  levels.^
     Refinery receipts of crude oil  averaged 12.2 million barrels per day
during 1984, up 0.5 million barrels per day from 1983.   This increase
represented a reversal  of the downward trend that started in 1979.
Receipts of domestic crude oil averaged 8.8 million barrels  per  day and
foreign averaged 3.3 million barrels per day.  Most of  the increase
in refinery receipts of crude oil  was from domestic sources, rising from
8.6 million barrels per day during 1983 to 8.8 million  barrels per  day
during 1984.  Foreign crude receipts also rose during  1984,  reaching 3.3
million barrels per day from 3.2 million barrels per day during  1983.18

3.3  EMISSIONS FROM COMBUSTION OF REFINERY FUEL GAS
     An integrated refinery uses energy equivalent  to  about  10 percent
of the total energy content in the crude oil it processes.^ The fuel
requirements of any one refinery depends on the nature of the feed, the
final product yield and the level  of the individual product  quality.  All
of the refinery's energy needs could be derived from its own crude  oil
feed, usually refinery fuel gas and residual oil,  but  most  refineries are
designed to use available supplemental fuels such  as natural gas.
     Figure 3-1 illustrates a modern refinery, the  various  processes and
their respective products.  As indicated in Figure  3-1, several  refinery
processes produce refinery fuel gas as a by-product.  After  removing the

-------
                                   3-9
H2S, the refinery fuel  gas is burned in various combustion devices (boilers
and heaters) located throughout the refinery.  Based on five plant surveys
and trip reports, there do not appear to be any particular processes  or
combustion devices in which refineries utilize refinery fuel  gas as a fuel.
The number of combustion devices and the various processes in which refinery
fuel gas is burned varies greatly from one refinery to another.
     Since all  crude oil contains some amount of sulfur,  the refinery
fuel gas produced by the various processes will also contain sulfur.   The
sulfur content of crude oil ranges from less than 0.1  percent to greater
        »
than 5 percent sulfur by weight.  Sulfur in the refinery  fuel gas will  be
in the form of hydrogen sulfide (H2S), carbonyl sulfide,  mecaptan sulfur plus
GI and C2 sulfides.  As the sulfur concentration of the crude oil increases,
so does the concentration of H.2$ and other sulfur compounds in the refinery
fuel gas.  Combustion of refinery fuel gas containing H2$ produces sulfur
dioxide (S0£).  Thus, when untreated refinery fuel  gas derived from crude
oil with a high sulfur content is burned in the various process  combustion
sources, substantial quantities of S02 will  be emitted to the atmosphere.
     As discussed in Chapter 2, the new source performance standard
prohibits the burning of refinery fuel gas containing more than  230 mg
H2S/dscm (0.10 gr/dscf) in any combustion device that burns refinery  fuel
gas.  The combustion of process upset gas in a flare, and process gas or
refinery fuel  gas released to a flare from relief valve leakage  is exempt
from this standard.
     The alternative to the 230 mg H2S/dscm refinery fuel gas standard is
that an owner or operator may elect to treat the gases resulting from the

-------
                                   3-10

combustion of refinery  fuel  gas  so  as  to  limit the  release of S02 to the

atmosphere.  The EPA Administrator  must be  satisfied that treatment of

the combustion gases controls  S02 emissions as effectively as compliance

with the H2S standard.

     The standard is equivalent  to  a S02  content  of approximately

20 gr/100 scf of refinery fuel  gas  burned.   Burning such fuel will  result

in a concentration of 15 to 20 parts per  million  S02 in the combustion

products.20

3.4  SELECTION OF REFINERY COMBUSTION  DEVICES FOR NSPS CONTROL

     Combustion devices that burn refinery  fuel gas were originally selected

for NSPS development because of  their  potential to  emit sulfur  dioxide  (S02)

in significant quantities.  At the  time of  the NSPS development,  (in the

early 1970's) the nationwide emissions of S02 were  estimated to be  28 million

tons per year.  It was  estimated that  in  1970 approximately 0.8 million tons

of sulfur dioxide were emitted from petroleum  refineries.^1  The  background

study for the original  NSPS predicted  an  overall  emissions reduction for

controlled sources of 95 percent.

3.5  REFERENCES

1.   Laster L.L., 1973.  Atmospheric Emissions  from the Petroleum Industry.
     National Environmental Research Center, Research Triangle  Park, N.C. 27711.
     NTIS, Springfield, Virginia.   PB-224-040.

2.   National Petroleum Refiners Association.   U.S. Refining Capacity.
     Washington, D.C.  July 1, 1985.

3.   Cantrell, A.  Annual Refining  Survey.   Oil  and Gas Journal.  Page  122.
     March 18, 1985.

4.   Same as reference #2.  Page iii.

5.   Same as reference #2.  Page iii.

6.   Cantrell, A.  Annual Refining  Survey.   Oil  and Gas Journal.  Page  112.
     March 30, 1981.

-------
                                   3-11
7.   Cantrell,  A.   Annual  Refining Survey.   Oil  and Gas Journal.  Page 130.
     March 22,  1982.

8.   Cantrell,  A.   Annual  Refining Survey.   Oil  and Gas Journal.  Page 130.
     March 21,  1983.

9.   Cantrell,  A.   Annual  Refining Survey.   Oil  and Gas Journal.  Page 112.
     March 26,  1984.

10.  Same as reference #3.   Page 123.

11.  Same as reference #6.

12.  Same as reference #7.

13.  Same as reference #8.

14.  Same as reference #9.

15.  Same as reference #3.   Page 123.

16.  Same as reference #2.

17.  Same as reference #2.   Page V.

18.  Same as reference #2.   Page V.

19.  1985 HPI Market  Data.   Hydrocarbon  Processing,  page 26.

20.  Background Information  for Proposed New Source Performance Standards:
     Asphalt Concrete Plants,  Petroleum  Refineries, Storage Vessels,
     Secondary  Lead Smelters and Refineries, Brass or Bronze Ingot
     Production Plants,  Iron and Steel Plants,  and Sewage Treatment
     Plants. U.S. Environmental  Protection  Agency.  Research Triangle
     Park, N.C. Publication  Number APTD-1352a.   June 1973.  Page 27.

21.  National Air Pollutant  Emission Estimates,  1940-1983.  U.S. EPA,
     RTP, N.C.   EPA 450/4-84-028.   December  1984.

-------

-------
                     4.  STATUS OF CONTROL TECHNOLOGY

     This chapter discusses the control  techniques  being  used to  meet the
new source performance standards (NSPS).  In order  to  comply with the
hydrogen sulfide (H2$) emission requirements (230 mg/dscm,  0.10 gr/dscf)
of this NSPS, the owner/operator of the affected facility has the option
of either reducing the H2S concentration prior to burning the refinery
fuel gas in the affected facility or treating the sulfur  dioxide  (S02)
emissions from the affected facility.   At this time, all  known affected
facilities control S0£ emissions by reducing the H2$ concentration in the
fuel gas prior to being burned.
     A review of the literature, the Environmental  Protection Agency
(EPA) compliance data system,  discussions with refinery personnel,  trade
organizations, local, State, and EPA regional  agencies reveals that two
processes are being used to comply with the NSPS.  These  two systems are
the alkanolamine process system and the LO-CAT® system.   These processes
are discussed below.
     4.1.  ALKANOLAMINE PROCESS SYSTEM
     The first commercially available alkanolamine  was triethanolamine (TEA)
which was used in early natural gas treating plants.   As  other members of
the alkanolamine family were introduced into the market,  they were evaluated
as possible acid gas absorbents.l  Alkanol amines are categorized  as being
primary, secondary, or tertiary, depending upon the degree of substitution
of the central nitrogen atom by organic groups.  Structural formulas for
the various alkanolamines are presented in Figure 4-1. Two commercially
utilized primary amines are monoethanol amine (MEA)  and diglycol amine (DGA),

-------
                                   4-2
each shows single substitution of organic groups at the central  nitrogen
atom.  Two commercially available secondary amines are diethanolamine (DEA)
and diisopropanolamine (DIPA).  The two secondary amines show double
substitution of organic groups at the central  nitrogen atom.   A triple
substitution of organic groups at the central  nitrogen atom is possible,
hence the tertiary amines:  triethanolamine (TEA) and methyldiethanolamine
(MDEA).
     Alkanolamines are weak organic bases with each one of the amines
illustrated in Figure 4-1 having at least one hydroxyl group (OH) and one
amino group.  In general, it can be considered that the hydroxyl  group
serves to reduce the vapor pressure and increase the water solubility,
while the amino group provides the necessary alkalinity in water solutions
to cause the absorption of acidic gases.2  As crude oil is processed the
following acid gases are formed and can be found in untreated refinery
fuel gas:  hydrogen sulfide (H2S), carbon dioxide (003), and carbonyl
sulfide (COS).  These compounds are considered to be acid gases because
when dissolved in an aqueous medium, they dissociate to form weak acids.
The acid gas and amine base will combine chemically to form an acid-base
complex called a salt, thus removing the acid gas from the process stream.3
     The principal reactions of the alkanolamines with H2$ are listed in
Table 4-1. The reactions shown in Table 4-1 proceed to the right at low
temperatures and to the left at higher temperatures.  This is the reason
that H2S can be absorbed by alkanolamine solutions at ambient temperatures.
At elevated temperatures (as exist in the stripper column), the reactions
are reversed with the sulfide and carbamate salts being decomposed and the
acid gases released in the stripper column.*

-------
                                   4-3
                                FIGURE 4-1

           MOLECULAR STRUCTURES OF COMMON GAS TREATING AMINES5
Primary Amines:  single substitution of organic group at the nitrogen atom

      Monoethanolamine (MEA)                Diglycolamine (DGA)
         MW =61                               MW = 105
                                       H

        ;N-CH2-CH2-OH                       N-CH2-CH2-0-CH2-CH2-OH
Secondary Amines:   double substitution of organic group at the nitrogen atom

      Diethanolamine (DEA)                  Diisopropanolamine (DIPA)
         MW = 105                              MW = 133
      HO-CH2-CH2-N-CH2-CH2-OH               HO-CH-CH2-N-CH2-CH-OH
                 I                              III
                 H                             CH3    H     CH3
Tertiary Amines:  triple substitution of organic group at the nitrogen atom

      Triethanolomine (TEA)                  Methyldiethanolamine (MDEA)
         MW = 149                              MW = 119
      HO-CH2-CH2-N-CH2-CH2-OH               HO-CH2-CH2-N-CH2-CH2-OH

                 CH2-CH2-OH                            CH3

-------
                                   4-4
                                TABLE 4-1

       PRINCIPAL REACTIONS OF ALKANOLAMINES WITH HYDROGEN SULFIDE6
Primary Amines
           + H2s <::::::::::> (RNH3)2s
     (RNH3)2s + H2s <:::::::::
Secondary Amines
      2R2NH + H2s <:::::::::> (R2NH2)2s
      (R2NH2)2s + H2s <:::::::::> 2R2NH
Tertiary Amines
      2R3N + H2s <:::::::::> (R3NH)2s
      (R3NH)2s + H2s <:::::::::> 2R3NHHs
R = C2H4OH

-------
                                   4-5
     Various literature sources Indicate that as alkanolamines were
developed and field tested, the primary and secondary amines displaced
TEA.  Triethanolamine was displaced largely because of its low ^B capacity
(resulting from higher equivalent weight), its low reactivity (as a
tertiary amine), and its relatively poor stability.7  Information obtained
via plant visits and contacts with petroleum refineries indicate that
primary and secondary amines are indeed the principle types of amines
used in alkanolamine process systems.  However, one refinery responded
that it used MDEA, a tertiary amine.  Also, two of the five refineries
that were visited are considering switching from DEA to MDEA because
of its selectivity for removing only HgS and the associated reduction in
energy needed to regenerate the MDEA.8,9
     The general process flow for an alkanolamine processing plant is
shown in Figure 4-2.  The process flow scheme varies little, regardless
of the aqueous amine solution used to remove H2S.  The primary pieces of
equipment of concern are the absorber column and stripper column, together
with the associated piping, heat exchange, and separation equipment.
     The sour refinery fuel gas containing H2$ will nearly always enter
the alkanolamine process plant through a separator (not shown) to remove
any free petroleum liquids and/or entrained solids.  The sour gas then
enters the bottom of the absorber column and flows upward through the
absorber in intimate counter-current contact with the aqueous amine
solution.  The H2S is removed and sweetened gas leaves the top of the
absorber and flows to another separator (not shown) to remove any amine
solution and/or entrained solids that are carried over with the exiting
refinery gas.

-------
                                   4-6
                               FIGURE 4-2

                      ALKANOLAMINE TREATING  UNIT
     10
                                         ^•TREATED PROCESS GAS
                      ,REGENERATED
                          AMINE-,    FOUL AMINE
SOUR PROCESS GAS
JU^ACIO - WATER

      * ^ /
               ABSORBER
                                             STRIPPER
                                                                 HYDROGEN
                                                                 SULFIDE,
                                                                     TO SULFUR
                                                                     RECOVERY
                                                                    WATER
                              AMINE

-------
                                  4-7
     Lean amine solution from the bottom of the stripper column is pumped
through an amine-amine heat exchanger and then through a water or air-
cooled exchanger before being introduced to the top tray of the absorber
column.  The amine moves downward through the absorber counter-current to
the sour gas, and absorbs HgS from the gas stream.  Rich amine solution
flows from the bottom of the absorber through the amine-amine heat
exchanger and then to the top of the stripper column.
     The amine-amine heat exchanger serves as a heat conservation device
and lowers total heat requirements for the process.  The rich solution
flows downward through the stripper in counter-current contact with vapor
generated .in the reboiler.  The reboiler vapor (primarily steam) strips
the acid gases from the rich solution.  The acid gases and steam leave
the top of the stripper and pass through a condenser, where the major
portion of the steam is condensed and cooled.  The acid gases are separated
in a separator and sent to the sulfur recovery unit.  The condensed steam
is returned to the top of the stripper column as reflux or a slip stream
is bled off to control the ammonia concentration in the top of the stripper
column to help maintain the amine system water balance.
     Rich amine solution leaves the bottom of the absorber at an elevated
temperature due to the heat of reaction released when acid gases react
with the amine.  The amine cooler serves to lower the lean amine temperature
to the 100°F range.  Higher temperature on the lean amine solution will
result in excessive amine losses through vaporization and also lower acid
gas carrying capacity in the solution because of temperature effects.

-------
                                   4-8

4.2 THE LO-CAT® HYDROGEN  SULFIDE  OXIDATION  PROCESS

     The LO-CAT® process, a  process based on  a  liquid-phase-oxidation

technique originated by Humphreys and Glasgow (London), was developed to

provide an isothermal  process  for carrying  out  the modified Claus  reaction.H

                H2S + 1/2 02	>  H20 + S°

The LO-CAT® process has two  basic designs:  aerobic, which absorbs H2S

from air laden streams, and  anaerobic, used when the gas  stream has

little or no oxygen present.  The anaerobic application is used in treating

refinery fuel  gas.  Figure 4-3 is a flow diagram of an anaerobic LO-CAT®
              *
H2S oxidation process used to  treat refinery  fuel gas.

     The LO-CAT® process removes  sulfur  by  using a proprietary catalyst

that consists of an aqueous  solution of  chelated iron, buffered with soda

ash (Na2C03), potassium hydroxide (KOH)  or  any  other common alkali to a

pH of about 8.  The LO-CAT®  catalyst solution is circulated in a closed loop

between the absorbers and the  oxidizer vessel.  Sour gas  passes through

an inlet gas scrubber to remove entrained liquids and  enters  the absorber

at line pressure through a specially designed venturi  prescrubber.  The

gas passes upward through a  low liquid-filled vessel,  an  absorber, in

which the sour gas contacts  the LO-CAT®  catalyst solution flowing  downward.

In the absorbers the H2S gas is absorbed very rapidly  into the catalyst

solution where it is immediately oxidized to  precipitate  elemental sulfur.

The following chemical  reactions  take  place in  the absorber vessel:12

     (1)  H2S Absorption

              H2S + H20	> H2S  (aqueous) + H20

     (2)  First lonization

               H2S (aqueous)	>  H+  +  HS"

-------
                                          4-9
                                     FIGURE  4-3

                        FLOW DIAGRAM  OF A  LO-CAT SYSTEM
13
SOUR GAS
SCRUBBER
VEiNTURI LIQUID FILLED
PRE-CCINTACTOR ABSORBER
OXIDIZER AIR
COMPRESSOR
SEPARATOR
SULFUR
LOADOUT
STATION

MOLTEN
SULFUR
STORAGE

                •ntr •« our
                                                     SOLUTION
                                                     CNtCUiniON
CMCUICAL UBITION tOUmKMT MOT »MOWW.

-------
                                   4-10
     (3)  Second lonization
               HS"	> H+ + S=
     (4)  Oxidation of Sulfide
               2Fe+++ + S= -	> 2Fe++ + S°
                                                                            f,
     Sweet gas with less than 100 ppm H2S leaves  the absorber vessel  and
passes through a sweet gas scrubber (not shown in Figure 4-3) to guard
against mist carryover before entering the fuel  system.
     Active LO-CAT® catalyst solution enters the  absorber column at the top
and flows downward through the vessel counter-current to the gas flow.  The
spent solution leaves at the bottom of the vessel.  Partially reduced
solution is regenerated in the oxidizer vessel by direct contact with
compressed air.
     Regeneration of the LO-CAT® solution in the  oxidizer vessel involves the
following reactions:I4
     (5)  Oxygen Absorption
             02 + H20	-> 2(0) (aqueous)  + H20
     (6)  Regeneration of Iron
             2Fe++  + 0 (aqueous) + H20 —	-> 2 Fe+++ + 2 (OH)~
Overall, the reaction is
     (7)  H2S + 0.5 02	> H20 + S°

     The air used to generate the catalyst solution is supplied by an air
blower to an air sparger grid near the bottom of the oxidizer.  The spent
air, with a small fraction of the oxygen removed, leaves the top of the
oxidizer vessel and discharges to the atmosphere.

-------
                                   4-11
     The circulating catalyst solution, which contains elemental  sulfur
in a slurried form, flows from the absorber back to the oxidizer  through
a solution cooler which removes heat generated by the exothermic  reactions
in the absorber.  Sulfur formed in the absorber vessel circulates with
the solution and the particles grow to the 10-20 micrometer range.  The
larger particles settle out of the bulk solution, flow in the bottom of
the oxidizer vessel, and are flushed out the oxidizer vessel  as a slurry
of 10 to 20 percent by weight of sulfur.  The sulfur slurry is pumped at
approximately 100 psig through a special non-plugging heat exchanger or
melter, where it is heated to 270°F by 50 psig steam.  The aqueous catalyst
solution/molten sulfur mixture formed in the melter is fed through steam
jacketed piping to a steam jacketed separator vessel.  Molten sulfur is
withdrawn from the separator vessel to a molten sulfur storage tank or a
sulfur pit where it is stored for ultimate shipment by truck  or rail  car.
     The clarified catalyst solution leaves the sulfur settler through a
back pressure control  valve set at about 75 psig.  The bulk of the solution is
returned to the settler vessel and then to the absorber/oxidizer  circuit.
     In the oxidation of H2S to sulfur, some side reaction takes  place which
may be represented by the equation:15
     (8)  2HS~ + 1.5 02	> H2S203
     This reaction reduces the pH of the scrubbing solution,  and  NagCOs,
KOH, or other alkaline salt must be added to maintain the pH  of the solution
in the 8-8.5 pH range.  This leads to a gradual buildup of sodium thiosulfate
(Na2S203) or similar water-soluble sulfur-containing salts in the solution.
These have no deleterious effect at concentrations below 30 percent by
weight.  When the thiosulfate concentration exceeds 30 percent by weight

-------
                                   4-12
some of the spent catalyst solution is withdrawn  to  a  holding tank  for
disposal.  The thiosulfate has  a commercial  value; thus,  this solution
can be sold and the thiosulfate reclaimed.   Another  means of disposal is
to treat this solution in the refinery's  biological  wastewater  treatment
system.
4.3  FACILITIES SUBJECT TO THE  NSPS
    In Chapter 2 the term "affected facility"  was discussed  and defined.
As noted, the affected facility is  defined  as  the combustion device that
is capable of burning refinery  fuel gas  and not  the  H2S control  system.
Information obtained from plant trips  and surveys indicates  that refinery
fuel gas is burned in combustion devices  that  are used in a  wide variety
of production processes within  a refinery.   There does not appear to  be a
tendency for refineries to use  refinery  fuel  gas  as  an energy source  in
any particular production process.   Thus  the utilization  of  refinery  fuel
gas will vary from one refinery to  another.
    According to information supplied  by  the EPA's Stationary Source
Compliance Division (SSCO), fifty-four (54) petroleum  refineries have become
subject to the H£S portion of this  NSPS  since 1975.  Unfortunately, an
accurate number of affected facilities within each refinery  could not be
determined.  The information supplied  by  SSCD is  considered  to  underestimate
the number of refineries and affected  facilities  subject to  this NSPS.
This low estimate is attributed to  the fact that  some  State  agencies  that
have authority to enforce the Clean Air  Act do not report all the refineries
and affected facilities to SSCD.
    Trade journals were also reviewed  for information.  These sources
report new refinery construction projects plus modification  and reconstruction

-------
                                   4-13
projects that are occurring at  petroleum refineries.  Unfortunately, the
manner in which the modification and reconstruction  projects  are  reported
does not indicate whether these projects will  be  subject to the ^S portion of
this NSPS nor how many combustion sources are  involved.  A review of various
trade journals from 1980 to 1985 indicates there  have been construction
projects at 42 petroleum refineries that involve  either an amine treater
unit, a LO-CAT® unit,  or unspecified refinery  fuel gas recovery.
4.4  COMPLIANCE TEST RESULTS
     The Environmental Protection Agency regional  offices, State  agencies,
and petroleum refineries were contacted to obtain compliance  test informa-
tion for new, modified, or reconstructed facilities.
     All of the test data that  were submitted  were from refineries that use
an alkanolamine process system to remove the hydrogen sulfide (H2S) from
the refinery fuel gas.  No compliance test data were obtained from a
refinery that uses the LO-CAT® system to treat sour  refinery  fuel gas.
     4.4.1.  Alkanolamine Process System Compliance  Data16»17»18»19'20»21»22»23
     Data for fifteen (15) compliance tests were  obtained from nine (9)
different refineries.   The results of these tests are listed  in Table 4-2.
The new source performance standard (NSPS) for H£S is 230 mg/dscm
(O.lOgr/dscf).  Compliance test results for the H2S  concentration in
refinery fuel gas ranges from 2.97 mg/dscm to  119.43 mg/dscm. The data
indicate that refineries are able to meet the  NSPS.  Also, none of the
refineries that were contacted indicated that  they were experiencing any
problems complying with the NSPS.

-------
                                   4-14
                                TABLE 4-2



         COMPLIANCE  TEST  RESULTS FOR ALKANOLAMINE  PROCESS  SYSTEMS
Ref i nery
A
B
C
D
E
F
G
H
I
Date of
Test
2/81
4/82
10/82
3/84
7/84
7/76
8/84
8/84
11/80
11/81
1/84
--
10/84
8/81
6/85
Type of
Alkanolamine Used
MEA/primary
ti H
n H
ii n
n n
DEA/secondary
MEA/primary
n H
DEA/secondary
DEA/secondary
n n
MEA/primary
MEA/primary
DEA/secondary
DGA/primary
Avg.H2S Concentration
(mg/dscm)
101
32
12
57
77
119.43
2.97
8.09
12.5
89.2
63.3
105.6
18.1
27.7
3.1
NSPS = 230 mg/dscm

-------
                                   4-15
4.4.2  LO-CAT® Process System24
     Currently, the LO-CAT® treatment system is used in only two (2)
refineries within the United States to remove H2S from refinery fuel gas.
Only one of these refineries was operating as of January 1985.  Both of these
LO-CAT® systems are small  units, less than 20 LT/D.  Compliance tests have
not been performed for either of the two LO-CAT® systems.  Thus, there are
no data for this type of system.  However, one refinery has installed a
continuous H2$ monitor.  According to information submitted by the refinery,
the \\2$ concentration of the refinery gas treated by the LO-CAT® system
ranges from 20 ppmv to 50 ppmv and averages 30 ppmv.  The refinery reported
no excess emissions have occurred since the LO-CAT® system became operational.
4.5  Emission Monitoring
     The NSPS requires an H2S continuous monitoring device to be installed
in order to determine compliance with the H2S standard of 230 mg/dscm.
However, the EPA has not yet developed performance specifications for H2S
continuous monitoring systems.
     In April 1979, the EPA initiated work to establish specifications
and also to determine the durability, maintenance requirements, and data
validity of commercially available H2S continuous emission monitors.25
Five (5) H2$ monitors were selected for evaluation.  Selection criteria
included operating principles, engineering judgment about suitability for
use at petroleum refineries, and total cost.  The performance of the five
(5) H2$ monitors tested was disappointing.  The absolute agreement between
reference Method 11 and all monitors was poor and variable in eight out
of ten relative accuracy tests.  Thus, a conclusion of the test program

-------
                                   4-16

was that the use of H2S monitors  for  compliance purposes could not be
recommended at that time nor could  performance specifications for H2S
monitors be written.  Even though performance specifications for continuous
H2$ monitoring devices have not  been  promulgated  by  the EPA, some refineries
have been required by State agencies  to  install an \\2$ continuous monitoring
device.
     Refineries subject to the I^S  portion  of this NSPS that do not have  a
continuous H2S monitor are using manually collected  grab samples (Draeger
Tubes) to test the refinery fuel  gas  for H2S concentration.  The frequency
with which these samples are collected varies from one refinery to another
and can range from 1 manual sample  per shift to 3 manual samples per shift.
4.6  REFERENCES
1.   Kohl, Arthur L. and Fred C. Riesenfeld.  Gas Purification (Third
     Edition).  Houston, Gulf Publishing Company. 1979. Page 28.
2.   Same as reference #1.  Page 29.
3.   Butwell, K. F., D. J. Kubek, and P. W. Sigmund. Alkanolamine Treating.
     Hydrocarbon Processing.  Page  108.  March 1982.
4.   Maddox, Dr. R. N.  Gas and  Liquid Sweetening.   Norman, Oklahoma.
     Campbell Publishing Series.  1974.  Page 47.
5.   Same as reference #3.
6.   Same as reference #4.  Pages 45-47.
7.   Same as reference #1.

-------
                                   4-17

8.   Memo from El more, William Larry, EPA, to Kenneth R.  Durkee,  EPA.
     January 27, 1986.  Trip report for plant visit to Alliance Refinery,
     B.P. Oil Incorporated, Belle Chasse, Louisiana.

9.   Memo from Elmore, William Larry, EPA, to Kenneth R.  Durkee,  EPA.
     January 22, 1986.  Trip report for plant visit to Tenneco Oil  Refinery,
     Chalmette, Louisiana.

10.  Background Information for Proposed New Source Performance Standards:
     Asphalt Concrete Plants, Petroleum Refineries, Storage Vessels,
    ,Secondary Lead Smelters and Refineries, Brass or Bronze Ingot
     Production Plants, Iron and Steel  Plants, and Sewage Treatment
     Plants.  U.S.  Environmental Protection Agency.  Research Triangle
     Park, N.C. Publication Number APTD-1352a.  June 1973.  Page  27.

11.  Hardison, L.  C.  Treating Hydrogen Sulfide:  An Alternative  to Claus.
     Chemical Engineering.  January 21, 1985.  Page 62.

12.  Letter and attachments from L. C.  Hardison, ARI Technologies,  Inc.  to
     Elmore, Larry, EPA.  October 18, 1985.

13.  Same as reference #12.

14.  Same as reference #12.

15.  Same as reference #12.

16.  Letter and attachments from Thompson, J. E., Citgo Petroleum Corporation
     to Farmer, J.  R., EPA.  March 15,  1985.  Response to Section 114
     letter on petroleum refineries.

17.  Letter and attachments from Ballard, B. F., Phillips Petroleum Company
     to Farmer, J.  R., EPA.  March 21,  1985.  Response to Section 114
     letter on petroleum refineries.

18.  Letter and attachments from Livermore, A. R., La Gloria Oil  and Gas
     Company, to Farmer, J. R., EPA.  April 9, 1985.  Response to
     Section 114 letter on petroleum refineries.

19.  Letter and attachments from Mull ins, J. A., Shell  Oil Company, to
     Farmer, J. R., EPA.  April 9, 1985.  Response to Section 114 letter
     on petroleum refineries.

20.  Letter and attachments from Cox, R. A., Texaco USA,  to Farmer, J. R.,
     EPA.  March 29, 1985.  Response to Section 114 letter on petroleum
     refineries.

21.  Letter and attachments from Owing, E. C., National Cooperative Refinery
     Association,  to Farmer, J. R., EPA.  February 21, 1985.  Response to
     Section 114 letter on petroleum refineries.

-------
                                  4-18

22.  Memo and attachments  from Mclaughlin, N. 0., EPA, to Durkee, K. R., EPA,
     February 5,  1985.   Trip  report  on plant visit to Murphy Oil Company,
     Meraux, Louisiana.

23.  Same as reference #9.

24.  Letter and attachments from Hullinger, T. T., U.S. Oil and Refining
     Company, to Farmer, J. R.,  EPA.  May 15, 1985.  Response to
     Section 114 letter  on  petroleum refineries.

25.  A Study To Evaluate Carbon  Monoxide and Hydrogen Sulfide Continuous
     Emission Monitors  at  an  Oil  Refinery.  U.S. Environmental Protection
     Agency.  Research Triangle  Park, N.C.  EPA-600/4-82-054.  August 1982.

-------
                   5.  MODEL PLANTS AND CONTROL SYSTEMS

5.1  MODEL PLANTS AND CONTROL SYSTEMS
     Model plants and model  plant parameters are selected to represent
the range of facilities that have become subject to the  NSPS since  its
promulgation in 1974 or that are likely to be constructed, modified,  or
reconstructed in the future.  The control  systems applied to the  model
plants are selected to represent those control  techniques that have been
commonly installed to meet the NSPS.  The model  plants  are defined  in
this chapter and a cost analysis of these model  plants  is presented in
Chapter 6.
     As discussed in Chapter 4, two types of control  systems are  commonly
used by refineries to control H£S in refinery fuel  gas.   These two  systems
are amine treater/Claus sulfur recovery units and LO-CAT® units.
     The amine treater/Claus sulfur recovery unit combination is  the  most
widely used process.  The LO-CAT® system is a relatively new process  that
has been introduced since the last review of this standard and currently
only two LO-CAT® units (sulfur capacity 7 and 17.35 LT/D) have been
installed in the U.S. for HgS removal from refinery fuel  gas.
5.1.1  Model Control Systems and H2S Concentrations
     The selection and sizing of model ^S control  systems is based on
the total sulfur loading in the untreated refinery  fuel  gas.  Six sizes
of control units were evaluated for this review:  1,  2,  5.1, 10.2,  50.8,
and 101.6 Mg/D (1, 2, 5, 10, 50, and 100 LT/D).  Three  H2S concentrations
in the untreated refinery fuel gas (1, 5, and 10 percent H£S by volume)
were analyzed for each model plant.

-------
                                   5-2
     Tables 5-1 and 5-2 list  the model  plants.  The  amine treater/Claus
sulfur recovery unit was modeled for  the  following sizes:   5.1, 10.2,
50.8, and 101.6 Mg/D (5, 10,  50, and  100  LT/D).  The LO-CAT®  system was
modeled for 1, 2, 5.1,  and 10.2  Mg/D  (1,  2,  5,  and 10 LT/D).
5.2 EMISSION REDUCTIONS
     This section presents the sulfur dioxide (S02)  emission  reductions
that are achieved by each model  control system.  The emission  reduction
for each model control  system is determined  by  comparing the  emissions
from a control system meeting the new source performance standard  (NSPS)
versus the emissions from the baseline  situation of  no control.
5.2.1  Amine Treater/Cl aus Sulfur Recovery Unit
     The S02 emission reductions for  the  model  amine/Claus  sulfur  recovery
units are presented in Table  5-3.  All  incoming sulfur (in  the form of
H2S) in the untreated fuel gas is either  routed to fuel gas combustion
devices, converted in the Claus  sulfur recovery unit to elemental  sulfur,
or emitted from the Claus unit to the atmosphere.  The model  amine treaters
were designed to, achieve the  NSPS limit of 230  mg of H2S/dscm in the fuel
gas stream to combustion sources, while the  remainder of the  H2S is sent
to the Claus unit.  The model Claus units were  designed to  achieve a 96.6
percent sulfur recovery efficiency.  For  calculating emission reduction,
it was assumed that the amine treater/Claus  system was operating at the
capacity of the Claus unit for 350 operating days per year.

-------
                   5-3
                TABLE 5-1



MODEL PLANTS:  AMINE TREATER AND CLAUS SRU
Model
Sulfur
(Mg/D)
5.08


10.16


50.8


101.6


Plant Size
Processed
(LT/D)
(5)


(10)


(50)


(100)


Fuel Gas
H2S Cone.
(% by Volume)
1
5
10
1
5
10
1
5
10
1
5
10
Amine
Efficiency
(%)
98.38
99.68
99.84
98.38
99.68
99.84
98.38
99.68
99.84
98.38
99.68
99.84
Glaus SRU
Efficiency
(%)
96.6


96.6


96.6


96.6



-------
                               5-4
                            TABLE  5-2

                      MODEL  PLANTS:  LO-CAT®
  Model  Plant Size                  RFG                    Unit
  Sulfur Processed               H2$ Cone.              Efficiency
(Mg/D)         (LT/D)            (% by Volume)
1.016


(1)


1
5
10
98.38
99.68
99.84
 2.032          (2)                   1                    98.38
                                     5                    99.68
                                    10                    99.84

 5.08           (5)                   1                    98.38
                                     5                    99.68
                                    10                    99.84

10.16          (10)                   1                    98.38
                                     5                    99.68
                                    10                    99.84

-------
                                       5-5
                                    TABLE 5-3

                          S02 EMISSION REDUCTION USING A

                     AMINE TREATER/CLAUS SULFUR RECOVERY UNIT
  Model     Control            Uncontrolled S02               S02 Emission
  System     Size3             Emissions b>c                   Reductions^
  (Mg/D)    (LT/D)               (Mg/yr)                         (Mg/yr)
5.08
10.16
50.8
101.6
(5)
(10)
(50)
(100)
3556
7112
35560
71120
3435
6870
34350
68700
a Model control system size (i.e., amount of sulfur processed) is based on the
  sulfur loading into the amine treater/Claus sulfur recovery unit.

b 1 year = 350 operating days

c 1 Mg of sulfur (S2) converts to 2 Mg of S02 emissions

d Efficiency of the sulfur recovery unit was assumed to be 96.6%

-------
                                   5-6
5.2.2  LO-CAT® Unit
     The S02 emission reductions for the model  LO-CAT® units are presented
in Table 5.4.  As indicated in Figure 4-3,  the  LO-CAT® process has no
emission stream to the atmosphere.   All  the incoming sulfur (in the form
of H2$) is either converted to elemental sulfur or routed to the fuel gas
combustion devices.  The model LO-CAT® units were designed to achieve the
NSPS limit of 230 mg of H2S/dscm in the outlet  fuel  gas stream.  For
calculating emission reductions, it was assumed that the LO-CAT® unit was
operating at capacity for 350 operating days per year.

-------
                                       5-7
                                    TABLE 5-4

                   S02 EMISSION REDUCTION USING A LO-CAT® UNIT
Model
System
(Mg/D)
1.016


2.032


5.08


10.16


Control
Size9
(LT/D)
(1)


(2)


(5)


(10)


H2S
Cone
(% by Vol.)
1
5
10
1
5
10
1
5
10
1
5
10
Uncontrolled S02
Emissions" »c
(Mg/yr)
712
712
712
1422
1422
1422
3556
3556
3556
7112
7112
7112
Control System
SO? Emissions
(Mg/yr)
11.5
2.3
1.1
2.3
4.6
2.3
57.6
11.4
5.7
115.2
22.8
11.. 4
S02 Emission
Reductions
(Mg/yr)
701
710
711
1399
1417
1420
3498
3545
3550
6997
7089
7101
a Model  control  system size (i.e.  amount of sulfur processed)  is  based  on  the
  sulfur loading into the LO-CAT®  unit.

b 1 year = 350 operating days

c 1 Mg of sulfur (S2) converts to  2 Mg of S02 emissions

-------
                             6.  COST ANALYSIS
6.1  INTRODUCTION
     This section presents costs (in January 1985 dollars)  of model  plant
control systems necessary to meet the hydrogen sulfide (and consequently
the sulfur dioxide) provisions of the current NSPS for refinery fuel  gas.
Two control  systems are analyzed: (1) for large sources of  refinery  fuel gas,
an amine treating section coupled with a Claus sulfur recovery plant;  (2)  for
smaller sources, a LO-CAT® process which absorbs the hydrogen sulfide  and
oxidizes it to elemental sulfur with the aid of an iron-based catalyst.
(A more detailed discussion of the control  systems appears  in Chapter  4.)
     Capital and annualized costs are estimated for the following model  plant
sizes, all  given on a nominal  sulfur basis:   LO-CAT® — 1.016 Mg per day
(1 long ton per day), 2.032 Mg/D (2 LT/D),  5.08 Mg/D (5 LT/D), and 10.16 Mg/D
(10 LT/D);  amine-Claus -- 5.08  Mg/D (5 LT/D), 10.16 Mg/D  (10 LT/D), 50.8
Mg/D (50 LT/D), 101.6 Mg/D (100 LT/D), 203.2 Mg/D (200 LT/D), 508 Mg/D (500
LT/D) and a plant expansion from 10.16 Mg/D to 10.668 Mg/D  (10 to 10.5 LT/D)*.
(A more detailed discussion of model plants is presented in Chapter  5.)
Cost-effectiveness is calculated for all model plants and  is discussed
for the two model plant sizes that are common to both control systems.
Except for the plant expansion, all estimated costs apply  to new control
systems installed at new sources of refinery fuel gas.  Note that the
cost of Claus tail gas treating is not considered in the analysis.
     The costs presented for the amine-Claus process are based on raw  cost,
data provided in References 1 and 2.  Costs for the LO-CAT® process  are
based on information from References 3, 4,  and 5.  Capital  costs are on  a

*Note:  1 Mg/D = 1.1025 short tons/D = 0.9844 LT/D

-------
                                    6-2
turnkey basis and thus include the purchase cost of equipment  and  auxiliaries,
taxes, freight, and all  necessary installation costs,  as well  as  indirect
costs such as engineering and supervision,  construction and  field  expense,
contractor fee, and contingency.  Annualized costs include direct  operating
costs such as operating  labor, maintenance  labor, utilities, and materials,
as well as ^indirect costs such as capital  charges, overhead, property taxes,
insurance, and administration.  Net annualized cost is also  presented,
representing total annualized cost less the credit for recovered  sulfur
where applicable.  The costs thus obtained  are "order-of-magnitude"--i.e.,
nominally j^ 50 percent in accuracy.  Because most of the operating and
maintenance (O&M) costs  have been calculated directly, their accuracy should
approach that of a study estimate (+_ 30 percent).  Finally,  because the
annualized costs are comprised of O&M costs and capital cost components,
their accuracy should fall between +_ 30 and +_ 50 percent.
6.2     AMINE TREATMENT  WITH CLAUS SULFUR  RECOVERY
6.2.1   Capital Costs
     The capital costs for this combination process are estimated  for
five model plants—four new and one expansion of an existing plant.
(Detailed model plant parameters are given in Chapter 5.)  The capital
costs for the amine treatment section were calculated from cost factors
given in Reference 1 for gas sweetening operations.  The process  employs
diethanolamine (DEA) as the absorbent.
     The capital costs for the Claus section were developed by updating the
costs in Reference 2 via the Chemical Engineering (CE) Plant Cost  Index.
The capital cost for the 5.08  Mg/0 Claus  plant was extrapolated  logarithmically
from the 10.16 Mg/D model plant cost in Reference 1 (the smallest  size)
using a 0.2 exponent, in accordance with cost vs. capacity formulas presented

-------
                                    6-3
in Appendix A of Reference 1.   The 0.2  exponent  was  found  to  hold  in  the  low
size range, based on data gathered for  Claus plants  in  Reference 1.   Costs  for
the 50 and 100 LT/D plants were updated from costs  in Reference 1  for these sizes.
     Capital costs for the two sections were added together for each  model
plant and are presented in Table 6-1.   The capital cost of the plant
expansion was calculated as the incremental  cost of  a 10.668  Mg/D  plant
over a 10.16 Mg/D plant and does not include any retrofit  costs.   This
approach was taken as an expedient to arriving at an order-of-magnitude
cost for the expansion; it must be recognized that such an expansion  would
be a site-specific case, the cost of which could vary significantly.
     The capital costs for 200 and 500  LT/D  model plants can  be estimated
by multiplying the cost of the 100 LT/D plant by a factor  of  two or five,
respectively.  This results from the fact  that plants above 100 LT/D  capacity
are normally constructed as trains, so  that  their costs vary  linearly with
size.
6.2.2   Annualized Costs
     The annualized costs associated with  owning and operating the amine-
Claus plants are estimated for each model  plant. The direct  operating cost
includes operating and supervisory labor,  maintenance,  supplies, utilities,
chemicals, and catalysts.  For the amine section, direct operating costs
were calculated by means of the cost factors given in Appendix A.  For the
Claus section, the direct operating costs  were developed by combining
consumption data with unit cost data, both from  Reference  2.  (The latter
are listed in Table 6-2.)  The unit cost data were updated slightly by use
of Bureau of Labor Statistics factors in most cases, such  as  the Producer
Price Index.

-------
                   6-4
                TABLE  6-1

           CAPITAL COST SUMMARY

AMINE TREATMENT WITH CLAUS  SULFUR  RECOVERY
          (January 1985 Dollars)
       Cost in Thousands of Dollars
   Plant, Capacity,  Mg/D (LT/D)  SuTfur
5.08 Mg/D (5 LT/D)
     1% H2S
     5% H2S
    10% H2S

10.16 Mg/D (10 LT/D)
     1% H2S
     5% H2S
    10% H2S
Ami nea
110
 88.8
 84.6
220
178
169
C1 ausb
2290
2290
2290
2630
2630
2630
                                                 Total c
                                                 2390
                                                 2370
                                                 2370
                                                 2840
                                                 2800
                                                 2790
50.8 Mg/D (50 LT/D)
1% H2S
5% H2S
10% H2S
101.6 Mg/D (100 LT/D)
1% H2S
5% H2S
10% H2S
Expansion of 0.508 Mg/D
(0.5 LT/D)d
1% H2S
5% H2S
10% H2S
a Cost developed from i
b Cost developed from i

1100
888
846

2200
1780
1690


10.7
9.2
8.5
nformation in Reference 1
nformation in Reference 2

4480
4480
4480

6470
6470
6470


25.8
25.8
25.8

.

5570
5360
5320

8670
8240
8160


36.5
35.0
34.3


c Totals may not add due to rounding.
d These costs represent
a 10.16 Mg/D plant.
the incremental costs of

a 10.688 Mg/D

plant over


-------
                                    6-5


                                 TABLE  6-2



                           ASSUMED UNIT COSTS  FOR

                         OPERATING COST ESTIMATION

                       CLAUS SULFUR RECOVERY PLANTS^,c
Operating Cost Item

Operating 1 abor

Supervision  labor

4300 Kp steam

1760 Kp steam

352 Kp steam

106 Kp steam

Steam condensate

Boiler feedwater

Electricity

Fuel gas

Sulfur
Unit Cost (Credit)

    $15.00/h

    $20.00/h

    $17.50/Mgd

   ($16.25/Mg)

   ($14.00/Mg)

   ($11.00/Mg)

   ($ 3.00/Mg)

    $ 3.60/Mg

    $15.28/Gj

    $ 4.00/Gj

   ($110/Mg)e
English Equivalent
   $  7.94/103lb

  ($  7.37/103lb).

  ($  6.35/103lb)

  ($  4.99/103lb)

.  ($  1.36/103lb)

   $  1.63/103lb

   $  0.055/KWh

   $  4.22/MM BTU

  ($111.8/LT)
a  For Claus plant operating cost estimation,  unit  costs  were  multiplied  by
   consumption factors given in Reference 2.
b  Unit costs were not used for operating cost estimation for  the  amine
   process or LO-CAT® process;  operating costs for  these  processes were given
   in References 1 and 3,  respectively,  as functions  of gas  flowrate  or plant
   capacity.
c  The total costs for catalysts given  in Reference 2 were multiplied by  1.1
   to update them to January 1985 dollars.
d  Cost or credit depending upon model  plant.
e  Based on the cost given for  recovered (i.e., Claus plant) sulfur,  Houston
   terminal, in Chemical Marketing Reporter, January  21,  1985.

-------
                                    6-6
     Most indirect costs were factored  from capital  costs  or  direct  operating
costs using appropriate factors  from References  6  and  7.   Capital  recovery
was calculated from the capital  cost with  a 10 percent rate of  return  and a
15-year equipment life (Reference 2).  There is  no waste disposal  cost,
because spent catalysts are regenerated.   The annualized cost includes an
estimate of the credit for recovered sulfur at prices  applicable  in  January
1985, a period of relative stability for  recovered sulfur  prices.  Note
also that credit for different pressures  of steam  varies in that,  in
general, the larger the Glaus plant, the  greater the heat  recovery practiced
and the higher the pressure of steam that  is recovered.
     The annualized costs for both sections were added together for  each
model plant and are presented in Tables 6-3 through  6-9.   Again,  costs for
expanding the existing 10.16 Mg/D plant to 10.668  Mg/D were calculated as
incremental costs.  All annualized costs  are based on  350  operating  days
per year.  See Appendix B for an example  of annualized cost estimation for
the individual processes that make up the control  system  (amine treating
and Claus sulfur recovery).
6.3    LO-CAT® PROCESS
6.3.1  Capital Costs
     Capital costs for LO-CAT® model plants are  also estimated.  Although the
model plants envision incoming hydrogen sulfide  concentrations  of 1, 5, and 10
percent, according to Reference 5 the differences  in capital  cost among the three
would not be significant.  Therefore, the costs  are  developed independent of
hydrogen sulfide concentration.  The purchased  equipment  cost was taken directly
from Reference 3 and includes all primary and auxiliary equipment and
instrumentation.  Taxes and freight costs were  obtained from  Reference 6.

-------
                                    6-7



                                Table 6-3

                        ANNUALIZED COST SUMMARY*
                AMINE TREATMENT WITH GLAUS  SULFUR RECOVERY
                         5.08 Mg/D (5 LT/D) PTMT
                          (January 1985 Dollars)

                       Cost in Thousands of Dollars
                                                     H2S      5% H2S      10%H2S
Direct Operating Cost
Operating labor
Supervision
Maintenance, repairs
Supplies, laboratory
Utilities
4300 Kp steam
Boiler feedwater
Electricity
Fuel Gas
Chemicals, catalyst
Indirect Operating Cost
Overhead (50 percent of all labor costs)
Capital Recovery (10 percent rate of return,
15 years equipment life)
Taxes (one percent of capital cost)
Insurance (one percent of capital cost)
Administration (two percent of capital cost)
Total Annual i zed Cost Without Credit
Credits
4300 Kp steam
1760 Kp steam
352 Kp steam
106 Kp steam
Steam Condensate
Sulfur
Total Credits
Net Annual ized Cost

210
61
70
17

. 4
12
57
10
3

153

315
24
24
48
1,007

0
48
0
3
5
189
244
762

210
61
70
17

4
12
57
10
3

153

312
24
24
47
1,003

0
48
0
3
5
189
244
759

210
61
70
17

4
12
57
10
3

153

312
24
24
47
1,002

0
48
0
3
5
189
244
758
Totals may not add due to rounding

-------
                                 6-8
                              Table  6-4

                      ANNUALIZED  COST  SUMMARYa
              AMINE  TREATMENT  WITH GLAUS  SULFUR RECOVERY
                      10.16  Mg/D  (10 LT/D)  PLANT
                        (January  1985  Dollars)
                     Cost  in  Thousands  of  Dollars
                                                   H2S
5%
10%H2S
Direct Operating Cost
Operating labor
Supervision
Maintenance, repairs
Supplies, laboratory
Utilities
4300 Kp steam
Boiler feedwater
Electricity
Fuel Gas
Chemicals, catalyst
Indirect Operating Cost
Overhead (50 percent of all labor costs)
Capital Recovery (10 percent rate of return,
15 years equipment life)
Taxes (one percent of capital cost)
Insurance (one percent of capital cost)
Administration (two percent of capital cost)
Total Annual i zed Cost Without Credit
Credits
4300 Kp steam
1760 Kp steam
352 Kp steam
1-06 Kp steam
Steam Condensate
Sulfur
Total Credits
Net Annual i zed Cost

210
61
82
17

7
24
114
19
7

155

- 374
28
28
57
1,184

0
95
0
6
9
378
489
695

210
61
82
17

7
24
114
19
7

155

369
28
28
56
1,177

0
95
0
6
9
378
489
688

210
61
82
17

7
24
114
19
7

155

367
28
28
56
1,175

0
95
0
6
9
378
489
686
Totals may not add due to rounding

-------
                                 6-9
                              Table 6-5

                      ANNUALIZED COST  SUMMARYa
              AMINE  TREATMENT WITH  GLAUS  SULFUR  RECOVERY
                       50.8 Mg/D (50 LT/D)  PLANT
                        (January  1985  Dollars)

                     Cost  in  Thousands of  Dollars
H2S
                                                           5%  H2S      10%H2S
Direct Operating Cost
Operating labor
Supervision
Maintenance, repairs
Supplies, laboratory
Utilities
4300 Kp steam
Boiler feedwater
Electricity
Fuel Gas
Chemicals, catalyst
Indirect Operating Cost
Overhead (50 percent of all labor costs)
Capital Recovery {10 percent rate of return,
15 years equipment life)
Taxes (one percent of capital cost)
Insurance (one percent of capital cost)
Administration (two percent of capital cost)
Total Annual i zed Cost Without Credit
Credits
4300 Kp steam
1760 Kp steam
352 Kp steam
106 Kp steam
Steam Condensate
Sulfur
Total Credits
Net Annual i zed Cost

210
61
151
17

58
170
512
97
35

169

733
56
56
111
2,436

0
464
18
26
50
1,889
2,448
(11)

210
61
151
17

58
170.
512
97
35

169

705
54
54
107
2,400

0
464
18
26
50
1,889
2,448
(47)

210
61
151
17

58
170
512
97
35

169

700
53
53
106
2,393

0
464
18
26
50
1,889
2,448
(55)
Totals may not add due to rounding

-------
                                    6-10

                                  Table 6-6

                          ANNUALIZED COST SUMMARYa»b
                  AMINE TREATMENT WITH CLAUS SULFUR RECOVERY
                                 Mg/D I 100 LT/D) PLANT
                             (January 1985 Dollars)

                         Cost in Thousands of Dollars
H2S     5% H2S
                                                                          10%H2S
Direct Operating Cost
Operating 1 abor
Supervision
Maintenance, repairs
Supplies, laboratory
Utilities
4300 Kp steam
Boiler feedwater
Electricity
Fuel Gas
ChemicaTs, catalyst
Indirect Operating Cost
. Overhead (50 percent of all labor costs)
Capital Recovery- (10 percent rate of return,
15 years equipment life)
Taxes (one percent of capital cost)
Insurance (one percent of capital cost)
Administration (two percent of capital cost)
Total Annual ized Cost Without Credit
Credits
4300 Kp steam
1760 Kp steam
352 Kp steam
106 Kp steam
Steam Condensate
Sulfur
Total Credits
Net Annual ized Cost

210
61
227
17

0
441
1,007
194
71

184

1,140
87
87
173
3,897

307
101
322
52
39
3,779
4,600
(702)

210
61
227
17

0
441
1,007
194
71
•
184.

1,084
82
82
165
3,825

307
101
322
52
39
3,779
4,600
(775)

210
61
227
17

0
441
1,007
194
71

184

1,073
82
82
163
3,810

307
101
322
52
39
3,779
4,600
(789)
a Totals may not add due  to  rounding.
b Note:   For estimating the  cost  of  plants with capacities in excess of 100 LT/D,
  the costs in this table would be used  as costs of individual trains.  For
  example, a 200 LT/D plant  would consist of two 100 LT/D trains, and the costs
  for the plant would be  twice the costs in this table.

-------
          6-11
       Table 6-7
ANNUALIZED COST SUMMARY
AMINE TREATMENT WITH GLAUS
EXPANSION OF 10.16 Mg/D (10 LT/D) PLANT
(January 1985 Do!
1% H2S


Direct Operating Cost
Operating labor
Supervision
Maintenance, repairs
Supplies, laboratory
Utilities
Chemicals, catalyst
Indirect Operating Cost
Overhead (50 percent of all labor costs)
Capital Recovery (10 percent rate of return,
15 years equipment life).
Taxes (one percent of capital cost)
Insurance (one percent of capital cost)
Administration (two percent of capital cost)
Total Annual i zed Cost Without Credit
Credits
4300 Kp steam
1760 Kp steam
352 Kp steam
106 Kp steam
Steam Condensate
Sulfur
Total Credits
Net Annual i zed Cost
SULFUR RECOVERY
TO 10.668 Mg/D
lars)

10.66 Mg/D


210,000
60,900
83,000
16,800
172,300
7,268

155,300

379,600
28,870
28,870
57,740
1,200,600
-
0
100,100
0
6,601
9,767
396,800
513,268
687,330

(10.5 LT/D)


10.16 Mg/D


210,000
60,900
82,070
16,800
164,100
6,922

155,300

374,800
28,500
28,500
57,000
1,184,900

0
95,360
0
6,287
9,302
377,900
488,849
696,050




Incre-
mental

0
0
930
0
8,200
346

0

4,775
370 '
370
740
15,700

0
4,740
0
314
465
18,900
24,419
(8,720)

-------
                               6-12
                            Table 6-8

                    ANNUALIZED COST SUMMARY
            AMINE TREATMENT WITH GLAUS "SDUFUR RECOVERY
EXPANSION OF 10.16 Mg/D (10 LT/D) PLANT  TO 10.668 Mg/D (10.5 LT/D)
                      (January 1985 Dollars)

                              5% H2S
10.66 Mg/D 10.16 Mg/D Incre-
mental
Direct Operating Cost
Operating labor
Supervision
Mai ntenance , repai rs
Supplies, laboratory
Utilities
Chemicals, catalyst
Indirect Operating Cost
Overhead (50 percent of all labor costs)
Capital Recovery (10 percent rate of return,
15 years equipment life)
Taxes (one percent of capital cost)
Insurance (one percent of capital cost)
Administration (two percent of capital cost)
Total Annual i zed Cost Without Credit
Credits
4300 Kp steam
1760 Kp steam
352 Kp steam
106 Kp steam
Steam Condensate
Sulfur
Total Credits
Net Annuali zed Cost

210,000
60,900
83,000
16,800
172,300
7,268

155,300

373,850
28,430
28,430
56,860
1,193,140

0
100,100
0
6,601
9,767
396,800
513,268
679,870

210,000
60,900
82,070
16,800
164,100
6,922

155,300

369,250
28,080
28,080
56,160
. 1,177,660

0
95,360
0
6,287
9,302
377,900
488,849
688,810

0
0
930
0
8,200
346

0

4,600
350
350
700
15,480

0
4,740
0
314
465
18,900
24,419
(8,940)

-------
                               6-13
                            Table 6-9

                    ANNUALIZED COST SUMMARY
            AMINE TREATMENT NITH CLAPS  STUFUR  RECOVERY
EXPANSION OF 10.16 Mg/D (10 LT/D) PLANT TO 10.668 Mg/D  (10.5 LT/D)
                      (January 1985 Dollars)

                             10% H2S
10.66 Mg/D 10.16 Mg/D Incre-
mental
Direct Operating Cost
Operating labor
Supervision
Maintenance, repairs
Supplies, laboratory
Utilities
Chemicals, catalyst
Indirect Operating Cost
Overhead (50 percent of all labor costs)
Capital Recovery (10 percent rate of return,
15 years equipment life)
Taxes (one percent of capital cost)
Insurance (one percent of capital cost)
Administration (two percent of capital cost)
Total Annual i zed Cost Without Credit
Credits
4300 Kp steam
1760 Kp steam
352 Kp steam
106 Kp steam
Steam Condensate
Sulfur
Total Credits
Net Annual ized Cost

210,000
60,900
83,000
16,800
172,300
7,268

155,300

372,580
28,330
28,330
56,660
1,191,470

0
100,100
0
6,601
9,767
396,800
513,268
678,200

210,000
60,900
82,070
16,800
164,100
6,922

155,300

368,070
27,990
27,990
55,980
1,176,120

0
95,360
0
6,287
9,302
377,900
488,849
687,270

0
0
930
0
8,200
346

0

. 4,510
340
340
680
15,350

0
4,740
0
314
465
18,900
24,419
(9,070)

-------
                                     6-14

Direct and indirect installation  factors from Reference 7 were then applied
to each model  plant purchased equipment  cost to  obtain total  installed cost.
     The capital  costs for the four model  plants are  presented in Table 6-10.

6.3.2  Annualized Costs
     The annualized costs for the LO-CAT®  model  plants have  also been estimated.
Most of the direct operating costs were  taken directly from  Reference 3,
interpolating values from tabular data therein.  Operating labor was calculated
based on one-third of a man per shift for  all model plants (Reference 5), 350
days per year, and a labor rate of $15 per hour  (see  Table 6-2).  Supervision
was figured at 15 percent of labor and maintenance  labor was assumed to equal
operating labor (Reference 6).  Catalyst costs,  obtained from Reference 7,
were found to vary according to the method used  to  separate  the recovered
sulfur from the catalyst solution.
     Indirect costs were factored from capital costs  or direct operating costs
using factors from References 6 and 8.  As with  the amine-Claus control system,
capital recovery is based on a 10 percent  rate of  return and a 15-year equip-
ment life.  There are no waste disposal  costs for  spent catalyst, in accordance
with Reference 3.  Product recovery credits for  sulfur have  also been calculated.
Note that because the LO-CAT® process directly produces the  desulfurized refinery
fuel gas, these model plants are designed  to meet  (or better) the limit of 162
parts per million (ppm) H2$ in the gas.  The design sulfur removal efficiencies
vary with H2S inlet concentration, with  a  consequent  variation in sulfur
recovery.  Note also that there is some  question about the quality of sulfur
recovered from the LO-CAT® process.  To  be conservative, therefore, each
annualized cost table includes a cost for  sulfur disposal, calculated at a
unit cost of $24.60 per long ton ($24.22 per Mg).   The total  annualized

-------
                                      6-15
                                   Table 6-10

                              CAPITAL COST SUMMARY
                                IP-CAT® PROCESS
                             (January 1985 Dollars)
                                         Cost in Thousands  of Dollars3
                                      Plant Capacity.  Mg/D  (LT/D)  Sulfur
Purchased equipmentb(A)

Taxes and freight(O.OSCXA)

      Base Cost (B)

Direct/indirect installation

  (0.50dXB)

Total Installed Cost
1.016(1)

  520

   £2

  562

  281
  843
2.032(2)  5.08(5)    10.16(10)

   810     1460       2380

    65.      117.        190

   875     1577       2570

   438      789       1285
  1313
2366
3855
3  Cost for each capacity applies for inlet hydrogen sulfide concentrations
   of 1, 5, or 10 percent.
b  Costs were developed from information in Reference 3  and include  instru-
   mentation.
c  Reference 6 was the source of this factor.
d  Overall installation factor suggested by Reference 7  for refinery installations,

-------
                                    6-16



cost for the process is therefore bracketed  between  a  low  figure  that


includes full credit and a high figure that  includes the full  cost  of


disposal.  In reality,  the annualized cost would  probably  fall  in between,


depending on the quality of the sulfur produced and  the available markets


for sulfur of that quality.


     The annualized costs for the model  plants  are presented  in Tables  6-11


through 6-14.
                                    >


6.4  COST-EFFECTIVENESS


     The cost-effectiveness values are determined from the costs  developed


in Sections 6.2 and 6.3 and the emission  reductions  presented in  Chapter 5.


The cost-effectiveness values represent the  estimated  cost (in January  1985


dollars) that would be incurred by a refinery for each ton of S02 emissions


that is controlled.


     Cost-effectiveness values for both control systems and all model  plants


are shown in Tables 6-15 and 6-16, for the  amine-Claus and LO-CAT®  processes,


respectively.  All cost-effectiveness values are  based on  a baseline of


zero control.  For both processes, the C/E  ratio  increases as the model


plant size decreases.


     The amine-Claus system's cost-effectiveness  improves  to a credit  of


$10 to $12/Mg at the 101.6 Mg/D model plant  size. The model  plant  size at


where the transition from cost to credit occurs is approximately  30 Mg/D.


(Note that because costs for the 203.2 and  508 Mg/D  plants are multiples of


the cost for the 101.6 Mg/D plant, the cost-effectiveness  for all plants


101.6 Mg/D and larger is exactly equal in this analysis.)   Cost-effectiveness


is not a strong function of inlet H2S concentration  in the range  of concen-


trations examined in this analysis.

-------
                                    6-17
                                 Table 6-11

                         ANNUALI ZED COST SUMMARY a.
                              LO-CAT® PROCESS
                         1.016 Mg/D (1 LT/D) PLANT
                           (January 1985 Dollars)
                                                                Cost  in
                                                               Thousands
                                                               of  Dollars
Direct Operating Cost
  Operating labor                                                 42
  Supervision                                                     6
  Maintenance labor                                               42
  Utilities (electricity)                                          2
  Catalysts and chemicals                                          64

Indirect Operating Cost
  Overhead (50 percent of  all  labor costs)                         45
  Capital Recovery (10 percent rate of return,  15 years
                   equipment life)                                111
  Taxes (one percent of capital  cost)         .                     8
  Insurance (one percent of capital cost)                          8
  Administration (two percent  of capital  cost)                     17

Total Annualized Cost Without  Credit                             346

Sulfur Credit0                                                    38

Net Annualized Cost (With  Credit)                                308

Sulfur Disposal0                                                  8

Total Annualized Cost with Disposal                              355
aWith the exception of the sulfur credit and sulfur disposal  cost,  the  costs
 shown apply for inlet H2S concentration of 1,5,  and 10  percent.  See Note  c.
^Totals may not add due to rounding.
cThe sulfur credit and disposal  cost  shown are for 1 percent  I^S.   For  other
 inlet H2S concentrations, the sulfur credit would be as follows:   5 percent
 H2S — 39; 10 percent H2S -- 39.
 H2S -- 9; 10 percent H2S -- 9.
The disposal cost would be:   5 percent

-------
                                    6-18
                                 Table  6-12

                         ANNUALIZED  COST  SUMMARY a,b
                              LO-CAT® /PROCESS
                         2.032 Mg/D  (2  LT/D)  PLANT
                           (January  1985  Dollars)
                                                               Cost  in
                                                              Thousands
                                                              of Dollars
Direct Operating Cost
  Operating labor                                                42
  Supervision                                                     6
  Maintenance labor                                              42
  Utilities (electricity)                                          3
  Catalysts and chemicals                                         129

Indirect Operating Cost
  Overhead (50 percent of  all  labor costs)                        45
  Capital Recovery (10 percent rate of return,  15 years
                   equipment life)                                173
  Taxes (one percent of capital  cost)                             13
  Insurance (one percent of capital.cost)                         13
  Administration (two percent  of capital  cost)                    26

Total Annualized Cost Without  Credit                             493

Sulfur Credit0                                                   77

Net Annualized Cost (With  Credit)                                416

Sulfur Disposal0                                                 17

Total Annualized Cost with Disposal                               509
aWith the exception of the sulfur credit and sulfur disposal  cost,  the  costs
 shown apply for inlet H2S concentration of 1,5,  and 10  percent.  See Note  c.
^Totals may not add due to rounding.
cThe sulfur credit and disposal  cost  shown are for 1 percent  H2S.   For  other
 inlet H2S concentrations, the sulfur credit would be as follows:   5 percent
 H2S -- 78; 10 percent H2S -- 78.
 H2S -- 17; 10 percent H2S -- 17.
The disposal cost would be:   5 percent

-------
                                    6-19
                                 Table 6-13

                         ANNUALI ZED COST SUMMARY a
                              LO-CAT® PROCESS
                         5.08 Mg/D (5 LT/D)  PLANT
                           (January 1985 Dollars)
                                                                Cost  in
                                                               Thousands
                                                               of  Dollars
Direct Operating Cost
  Operating labor
  Supervision
  Maintenance labor
  Utilities (electricity)
  Catalysts and chemicals

Indirect Operating Cost
  Overhead (50 percent of  all  labor costs)
  Capital Recovery (10 percent rate of return,
                   equipment life)
  Taxes (one percent of capital  cost)
  Insurance (one percent of capital cost)
  Administration (two percent  of capital  cost)

Total Annualized Cost Without  Credit

Sulfur Credit0

Net Annualized Cost (With  Credit)

Sulfur Disposalc

Total Annualized Cost with Disposal
15 years
                   42
                    6
                   42
                    5
                  142
 45

311
 24
 24
 47

688

192

496

 42

731
aWith the exception of the sulfur credit and sulfur disposal  cost,  the  costs
 shown apply for inlet H2S concentration of 1,5,  and 10  percent.  See Note  c.
^Totals may not add due to rounding.
cThe sulfur credit and disposal  cost  shown  are for 1 percent  f^S.   For  other
 inlet H2S concentrations, the sulfur credit would be as follows:   5 percent
 H2S -- 195; 10 percent t^S -- 195.   The disposal  cost would  be:  5 percent
 H2S — 43; 10 percent H2S -- 43.

-------
                                   6-20
                                 Table  6-14

                         ANNUALIZED  COST  SUMMARY a.b
                              LO-CAT® PROCESS
                         10.16 Mg/D  (10 LT7DT  PLANT
                           (January  1985  Dollars)
                                                                Cost  in
                                                              Thousands
                                                              of  Dollars
Direct Operating Cost
  Operating labor                                                 42
  Supervision                                                     6
  Maintenance labor                                               42
  Utilities (electricity)                                          7
  Catalysts and chemicals                                "         284

Indirect Operating Cost
  Overhead (50 percent of  all  labor costs)                         45
  Capital  Recovery (10 percent rate of return,  15 years
                   equipment life)                                507
  Taxes (one percent of capital  cost)                              39
  Insurance (one percent of capital cost)                          39
  Administration (two percent  of capital  cost)                     77

Total Annualized Cost Without  Credit                            1088

Sulfur Credit0                           .                        385

Net Annualized Cost (With  Credit)                                703

Sulfur Disposal0                                                  85

Total Annualized Cost with Disposal                             1172
aWith the exception of the sulfur credit and sulfur disposal  cost,  the  costs
 shown apply for inlet H2S concentration of 1,5,  and 10 percent.  See Note  c.
bTotals may not add due to rounding.
cThe sulfur credit and disposal cost  shown are for 1 percent  l^S.   For  other
 inlet H2S concentrations, the sulfur credit would be as follows:   5  percent
 H2S — 390; 10 percent F^S -- 391.  The disposal  cost would  be:  5 percent
     -- 86; 10 percent f^S -- 86.

-------
                                        6-21



                                     Table  6-15

                                 COST EFFECTIVENESS
                               SULFUR DIOXIDE  CONTROL
                    AHINE  TREATMENT  WITH GLAUS SULFUR RECOVERY^

                               (January  1985 Dollars)
Annualized
Plant Size Cost (Credit)
et Concentration Mg/day (LT/D) SlO^/yr Mg/yr
H2S




H2S


j

% H2S




5.08 (5)
10.16 (10)
50.8 (50)
101.6 (100)
0.51 (0.50)b
5.08 (5)
10.16 (10)
50.8 (50)
101.6 (100)
0.51 (0.50)b
5.08 (5.)
10.16 (10)
50.8 (50)
101.6 (100)
0.51 (0.50)b
762
695
(11)
(702)
(8.7)
759
688
(47)
(7.75)
(8.9)
758
686
(55)
(789)
(9.1)
3,435
6,870
34,350
68,700
344
3,435
6,870
34,350
68,700
344
3,435
6,870
34,350
68,700
344
S02 Removed
tons/yr $/Mg
3,787
7,573
37,867
75,734
379
3,787
7,573
37,867
75,734
379
3,787
7,573
37,867
75,734
379
222
101
(0.32)
(10.2)
(25.3)
221
100
(1.37)
(11.3)
(26.0)
221
100
(1.60)
(11.5)
(26.4)
C/E
$/ton
201
92
(0.29)
(9.3)
(23.0)
200
91
(1.24)
(10.2.)
(23.6)
200
91
(1.45)
(10.4)
(23.9)
The cost-effectiveness ratios  for  200  LT/D  and  500 LT/D plants would be the same as
shown for 100 LT/D,  in that  the annualized  costs  for these plants are multiples of the
100 LT/D costs in this analysis.   However,  if large plants were to be custom-designed,
their C/E ratios would be expected to  be  lower,  reflecting economies of scale in the
capital  costs.
Expansion case (10.16 to 10.668  Mg/D).   All
between the base and expansion capacities.
values shown on these lines  are incremental

-------
                                           6-22
                                      Table 6-16

                                   COST EFFECTIVENESS
                                 SULFUR DIOXIDE CONTROL
                                   LO-CAT® PROCESS

                                 (January  1985 Dollars)
Inlet Concentration
1% H2S







5% H2S







10% H2S







Plant Size
Mg/D (LT/D)
1.016 (1)

2.032 (2)

5.08 (5)

10.16 (10)

1.016 (1)

2.032 (2)

5.08 (5)

10.16 (10)

1.016 (1)

2.032 (2)

5.08 (5)

10.16 (10)

Sulfur
Status
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
recovered
disposed
*For the special case where neither a sulfur
Annual i zed
Cost (Credit)
$lo3/yr*
308
355
416
509
496
731
703
1,172
308
355
415
509
493
732
698
1,173
308
355
415
509
493
732
697
1,173
recovery credi
S02
Mg/yr
700
700
1,399
1,399
3,499
3,499
6,997
6,997
709
709
1,418
1,418
3,545
3,545
7,090
7,090
710
710
1,420
1,420
3,551
3,551
7,101
7,101
t nor a
Removed
tons/yr
771
771
1,543
1,543
3,856
3,856
7,713
7,713
781
781
1,563
1,563
3,907
3,907
7,815
7,815
783
783
1,565
1,565
3,914
3,914
7,827
7,827
disposal cost
$/Mg
440
507
297
364
142
209
100
168
434
501
293
359
139
206
98
165
434
500
292
358
139
206
98
165
were
C/E
$/t
39
46
27
33
12
18
9
15
39
45
26
32
12
18
8
15
39
45
26
32
12
18
8
15
taken
the annualized cost would be found in Table  6-11,  -12, -13, or -14.  The cost-effective-
ness would then be the quotient of this  cost (the  "Total Annualized Cost Without Credit")
and the "S02 Removed" value in column 5.

-------
                                      6-23

     The most interesting aspect of Tables 6-15 and 6-16  is  the comparison  of
cost-effectiveness between the two control systems at the common model  plant
sizes--5.08 and 10.16 Mg/D.  At the 10.16 Mg/D size, the  cost-effectiveness
ratios of the amine-Claus and LO-CAT® systems (sulfur recovery  case)  are
essentially equal.  But as the model plant size decreases to 5.08 Mg/D,
the amine-Claus system becomes less cost-effective than the  LO-CAT® system
(sulfur recovery case).  This indicates that for these lower sulfur loadings,
the LO-CAT® would be the preferred control system to use.

6.5  COST COMPARISON
     To determine their representativeness, the capital costs plus the  operating
and the maintenance (O&M) costs in Sections 6.2 and 6.3 were compared to  cost
data obtained from refineries that responded to Section 114  letter requests.
Of the 13 plants covered in these 114 responses, seven contained enough cost/
process data to allow meaningful comparisons.  Because some  of  the respondents
requested their cost data to be held confidential, no specific  results
are shown herein.  However, some general  information about the  cost comparisons
can be provided:
     o    After escalating all costs to January 1985 dollars, costs were
          compared for seven amine treating units (ATU's), one  LO-CAT®
          system, and one combination ATU-CLAUS system.   (Note:   some
          refineries provided data on more than one system.)
     o    All but one of the capital cost sets differed by less than  +_  50%,
          the nominal accuracy of the total installed costs  in  the chapter.
     o    Larger discrepancies were seen in the O&M costs, due  to differences
          in unit prices, cost allocations, accounting methods,  and other
          factors.  In all but one case, the O&M costs in the chapter were
          higher than the respondents'.

-------
                                   6-24
     o    Details  of  the  cost comparison have been placed in the confidential

          portion  of  the  project docket.



6.6  REFERENCES FOR CHAPTER 6
1.  Shumaker,  Jeffrey  L.,  Memorandum to Docket A-80-20A, dated September 11,
    1984, concerning sweetening plant cost factors.

2.  Review of  New Source Performance Standards for Petroleum Refinery  Glaus
    Sulfur Recovery Plants, U.S. Environmental Protection Agency, EPA-450/
    3-83-014,  August 1983.
3.  Hardison,  L.C.,  "Minimizing Gas Treating Costs with LO-CAT® for
    Removal",  for presentation at the Panhandle Plains Regional Meeting of
    the Gas Processors  Association, Amarillo, TX, October 11, 1984.
4.  Hardison,  L.C.,  "Go  from H2$ to S in One Unit", Hydrocarbon Processing,
    April  1985,  pp.  70-71.

5.  Telecons—Thomas Beggs  (JACA Corporation, Fort Washington, PA)  with
    Mr. L.C.  Hardison (President, ARI Technologies, Inc., Palatine, IL):
    April  1,  1985; May 21,  1985; and June 10, 1985.

6.  Vatavuk,  William M.  and Neveril, Robert B., "Estimating' Ai r-Pollution  Control
    Costs—Part  II:  Estimating Capital and Operating Costs," Chemical  Engineering,
    November 3,  1980, pp.   157-162.

7.  Letter from  L. C. Hardison (ARI Technologies, Inc., Palatine, IL)  to W.
    L. Elmore (U.S.  Environmental Protection Agency, Research Triangle Park,  NC),
    February 26, 1986.

8.  Peters, M.S. and Timmerhaus, K.D., Plant Design and Economics for  Chemical
    Engineers, Third Edition.  McGraw-Hill, New York, NY, 1980, pp. 203-209.

-------
                                 APPENDIX  A



     Listed below are the  cost  factors  that were  obtained  or derived from

Reference 1 and used in estimating capital and  annual i zed  costs for the

amine treating process.

            Component                              Cost Factor

Capital  cost (10% H2S)                    $0.51 per standard c,ubic feet

                                          per day (SCFD) of acid  gas*
Capital  cost (1% H2S or 5% H2S)
Operating labor (including some
  maintenance
 0.9 (cap. cost at 10%)

+0.1 (cap. cost of 10%)j~   10   ~l°'6
                             conc
 $126,000 per year
Maintenance
 $0.01 per year per SCFO acid gas
Utilities (heat and electricity)
 $0.274 per year per SCFD acid  gas
Chemicals
 $0.0187 per year per SCFD acid gas
   The acid gas flowrate is  related to the  sulfur  loading,  as  follows:

           Flowrate (SCFD)  = 33,160 x  S
               where:     S  = sulfur fed to  amine treater  (LT/D)

  Via this equation,  the above factors may  be  rewritten in  terms of the sulfur
  loading.

-------
                                 APPENDIX  B





     This section presents  three examples  of  line  item  annualized costs  for



the control  systems analyzed.   The model plant  used  for illustration  is  the



5.08 Mg/D (5 LT/D) plant with  an inlet  concentration of one  percent



The examples are shown in Tables B-l, B-2, and  B-3.

-------
                                    B-2
                                 Table B-l
                     LINE  ITEM ANNUALIZED COST  EXAMPLES
                      AMINE  TREATMENT,  5 LT/D,  1% H?S
                           (January  1985 -Dollars)
Di rect
   Operating labor&
   Supervision
   Maintenance
   Utilities
   Chemicals

Indi rect
   Overhead
   Capital  recovery

   Taxes

   Insurance

   Administration  '
(direct from Reference 1)               $126,000
15% of above:   $126,000 x  0.15           18,900
$0.01 per SCFD:   165,800 x 0.01           1,658
$0.274 per SCFD:   165,800  x 0.274        45,429
$0.0187 per SCFD:   165,800 x 0.0187       3,100
50% of labor:  (126,000 + 18,900) x 0.50  72,450
CRF (10 percent, 15 years) = 0.1315:
  $109,925 x 0.1315                      14,455
1 percent of capital  cost:
  $109,925 x 0.01                     .    1,099
1 percent of capital  cost:
  $109,925 x 0.01                         1,099
2 percent of capital  cost:
  $109,925 x 0.02                         2,199

                                       $286,389
a  Direct factors from Reference 1.  (See Appendix A).

b  Includes an unspecified amount of maintenance labor.

-------
               B-3
            Table B-2
LINE ITEM ANNUALIZED COST EXAMPLES
CLAUS

Di rect
'Operating Labor
/
Supervision
Maintenance, repai
Supplies, lab
4300 Kp steam
Boiler feedwater
Electricity
Fuel gas
Catalyst
Indi rect
Overhead
Capital recovery
Taxes
Insurance
SULFUR RECOVERY PLANT 5 LT/D, 1% H?S
(January 1985 Dollars)
5600 hours x $15
year hour
2100 hours x $20
year hour
r13 3 percent of capital cost:
$2,285,000 X 0.03
20 percent of operating labor:
$84,000 X 0.20
200 Mq x $17.50
Year • Mg
3267 Mg_ X $3-60
Year Mg
764 Gj x $15.28
Year Gj
' 2423 Gj x $4
Year Gj
$655 * 5 LT/D * ,.,
10 LT/D
50 percent of labor:
0.50 X (84,000 + 42,000 + 0.5 X
CRF (10 percent, 15 years) =
0.1315: $2,285,000 X 0.1315
1 percent of capital cost:
$2,285,000 X 0.01
1 percent of capital cost:

$84,000
42,000
68,550
16,800
3,500
11,760
11,670.
9,690
361
80,138
68,550)
300,478
22,850
22,850
        $2,285,000 X 0.01

-------
                                    B-4
                             Table  B-2  (cont'd)
Indirect (cont'd)

     Administration          2  percent  of  capital  cost:         45,700
                             $2,285,000 X  0.02

Total  Annualized Cost w/o Credit                               720,347
Credits

     1760 Kp steam           2934 Mg   x  $16.25                 47,680
                              year         Mg

     106 Kp steam            286 Hg   x  $11.00                  3,144
                              year         Mg

     Steam condensate        1550 Mg   x  $3.00                   4,651
                              year         Mg

     Sulfur •'    5.08 Mg  x 0.966 recovery x 350  days  y  $110    188,930
                   day                         year      Mg

Total Credits                                                  244,405
NET ANNUALIZED COST                                      .     $475,942
a  Consumption figures derived from cost data in Reference 2.

b  Assumed to be equal parts labor and materials.

-------
                                    B-5

                                 Table B-3

                     LINE ITEM ANNUALIZED COST EXAMPLES
                       LO-CAT® PROCESS. 5 LT/D, 1% HgS
Direct

     Operating Labor


     Supervision

     Maintenance labor

     Utilities6


     Catalysts and
      Chemicals'3

Indirect

     Overhead


     Capital recovery


     Taxes


     Insurance


     Administration
               8 hours   x  350 days x  $15
                day           year      hour

             15% of above:  $43,000  X 0.15

             Same as operating
             $2.70
                                        5  LT
                                        day
              $81.20  x  5 LT
               LT        day
 350  days
  year

( 350  days
     year
            50 percent of labor:
            (42,000 + 6,300 + 42,000)  X  0.50

            CRF (10 percent, 15 years)  =
            0.1315:  $2,366,000 X 0.1315

            1 percent of capital  cost:
            $2,366,000 X 0.01

            1 percent of capital  cost:
            $2,366,000 x 0.01

            2 percent of capital  cost:
            $2,366,000 X 0.02
Total Annual ized Cost w/o Credit

Sulfur Credit
5.08 Mg  x 0-9838 recovery x 350 days
  day                          year
                                                         $110
                                                          Mg
NET ANNUALIZED COST (WITH CREDIT)
Sulfur Disposal 5.08 Mg x 0.9838 recovery x 350 days  x  $24.22
                  day                         year       Mg

TOTAL ANNUALIZED COST WITH DISPOSAL
 $42,000


   6,300

  42,000

   4,725


 142,100
  45,200


 311,100


  23,700



  23,700

  47,300


 688,100

 192,400


$495,700

  42,400


 $730,500
   In most cases, unit consumption data were taken from References  3,  4,
   and 5.
   Unit costs for utilities and catalysts are not the same for all  plant  sizes,
   as the quantities required of steam, electricity,  catalyst, etc.,  vary
   nonlinearly with capacity.

-------
                                 TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
 . REPORT NO.
 EPA-450/3-86-011
                            2.
                                                        3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
 Review of  New  Source  Performance Standards
 for  Petroleum  Refinery Fuel Gas
           5. REPORT DATE
               October 1986
           6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                        8. PERFORMING ORGANIZATION REPORT NO.
 . PERFORMING ORGANIZATION NAME AND ADDRESS
 Office of  Air  Quality Planning  and Standards
 Environmental  Protection Agency
 Research Triangle Park, NC  27711
                                                         10. PROGRAM ELEMENT NO.
           11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
 DAA for  Air Quality  Planning &  Standards
 Office of  Air and Radiation
 U.S.  Environmental Protection Agency
 Research Triangle Park, NC  27711
           13. TYPE OF REPORT AND PERIOD COVERED
               Final
           14. SPONSORING AGENCY CODE


             EPA 200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
      As required by Section lll(b)  of the Clean Air Act,  as amended,  a four
 year review of the new source performance standards for petroleum refineries
  (40 CFR Subpart J) was conducted.  This review was limited to the sulfur dioxide
 standard as applied to refinery fuel gas.  The report presents  a summary of the
 current standard, the status of current applicable control technology,  and the
 ability of plants to meet the standard.  No revision to the standard  is recommended;
 however, EPA should investigate an alternative method of  continuously measuring
 the sulfur concentration of refinery fuel gas.
17.
                              KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                            b.lDENTIFIERS/OPEN ENDED TERMS
                        c. COSATI Field/Group
 Air Pollution
 Petroleum Industry
 Hydrogen Sulfide
 Standards of Performance
 Pollution Control
Air Pollution Control
13B
18. DISTRIBUTION STATEMENT
 Unlimited
                                            19. SECURITY CLASS (ThisReport)

                                              Unclassified
                                                                      21. NO. OF PAGES
                                            20. SECURITY CLASS (Thispage)
                                              Unclassified
                             -Bl.
                                                                      22. PRICE
 EPA Form 2220-1 (R«v. 4-77)   PREVIOUS EDITION is OBSOLETE

-------