United States                   EPA-600/R-0 1-109
           Environmental Protection
           Agency                     April 2002
oEPA    Research and
           Development
           CONTROL OF MERCURY EMISSIONS
           FROM COAL-FIRED ELECTRIC
           UTILITY BOILERS:
           INTERIM REPORT
           INCLUDING ERRATA DATED 3-21-02
           Prepared for
           Office of Air Quality Planning and Standards
           Prepared by
           National Risk Management
           Research Laboratory
           Research Triangle Park, NC 27711

-------
                                 Foreword
      The U.S. Environmental  Protection  Agency is  charged by  Congress  with
protecting the Nation's land, air,  and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and  implement actions leading to
a compatible balance  between human activities and the ability of natural systems to
support and nurture life. To meet this mandate, EPA's research program is providing
data and technical support for solving environmental problems today and building a
science knowledge base  necessary to manage  our ecological resources wisely,
understand how pollutants affect our health, and prevent or reduce environmental risks
in the future.

      The National Risk Management Research Laboratory (NRMRL) is the Agency's
center for investigation of technological and management approaches for preventing
and reducing risks from pollution that threaten human health and the environment.  The
focus of the Laboratory's research program is on methods and their cost-effectiveness
for prevention and control  of pollution to air, land,  water, and subsurface resources,
protection of water quality in public water systems; remediation of contaminated sites,
sediments and  ground  water; prevention and control  of  indoor air  pollution;  and
restoration of ecosystems.  NRMRL collaborates with both public and private sector
partners to foster technologies that reduce the cost of compliance and to anticipate
emerging problems. NRMRL's research provides solutions to environmental problems
by: developing and promoting technologies that protect and  improve the environment;
advancing scientific and engineering  information  to support regulatory and policy
decisions; and  providing the technical support  and information transfer to ensure
implementation  of environmental  regulations and strategies  at the national, state, and
community levels.

      This publication has  been  produced as part of the Laboratory's  strategic
long-term  research plan.  It  is published and made available  by  EPA's Office of
Research and Development to assist the user community and to link researchers with
their clients.
                                 E. Timothy Oppelt, Director
                                 National Risk Management Research Laboratory

                           EPA REVIEW NOTICE

     This report has been peer and administratively reviewed by the U.S. Environmental
     Protection Agency, and approved for  publication.  Mention of trade names or
     commercial products does not constitute endorsement or recommendation for use.

     This document is available to the public through the National Technical Information
     Service, Springfield, Virginia 22161.

-------
                                      EPA-600/R-01-109
                                      December 2001
Control of Mercury Emissions from Coal-
        Fired Electric Utility Boilers:
                Interim Report

James D. Kilgroe, Charles B. Sedman, Ravi K. Srivastava,
   Jeffrey V. Ryan, C. W. Lee, and Susan A. Thorneloe
           U.S. Environmental Protection Agency
           Office of Research and Development
       National Risk Management Research Laboratory
        Air Pollution Prevention and Control Division
            Research Triangle Park, NC 27711

-------
                 Errata Pages
                       for
                 EPA-600/R-01-109

Control of Mercury Emissions from Coal-Fired
     Electric Utility Boilers: Interim Report
                  December 2001

                   Errata Pages
                       xvi
                       xxii
                       ES-10
                       6-3
                       6-4
                       6-19
                       6-21
                       6-43a
                       6-43b
                       6-43c
                       6-48
                       6-49
                       6-51
                       6-52

                  March 21, 2002
                       i-a

-------
                                       Abstract

In December 2000, the U.S. Environmental Protection Agency (USEPA) announced its intent to
regulate mercury emissions from coal-fired electric utility steam generating plants. This report,
produced by EPA's Office of Research and Development (ORD), National Risk Management
Research Laboratory (NRMRL), provides additional information on mercury emissions control,
following the release of " Study of Hazardous Air Pollutant Emissions from Electric Utility
Steam Generating Plants - Final Report to Congress," in February 1998.  The first three chapters
describe EPA's December 2000 decision to regulate mercury under the National Emission
Standards for Hazardous Air Pollutants (NESHAP) provisions of the Clean Air Act, coal use in
electric power generation, and mercury behavior in coal combustion. Chapters 4-9 report: new
information on current electric utility fuels, boilers, and emission control technologies; mercury
emissions associated with these diverse technology combinations; results and implications of
tests to evaluate the performance of mercury control technologies and strategies; retrofit control
cost modeling; and mercury behavior in solid residues from coal combustion.  The final chapter
summarizes current research and identifies future efforts needed to ensure cost-effective control
of mercury emissions.  References are provided at the conclusion of each chapter.
                                        Preface

    This is an interim report, based on data available as of mid-2001, which in some cases are
limited. As more data are collected and evaluated, some of the conclusions reached in this report
                                    may be modified.
                                           11

-------
                                        Contents
                                                                                  Page
Abstract 	ii
Preface	ii
Figures	x
Tables	xiv
Acronyms	xix
Acknowledgements	xxii

Executive Summary	ES-1

Chapter 1.  Report Background
    1.1   Introduction	1-1
    1.2   Report Purpose	1-2
    1.3   NESHAP Statutory Background	1-2
    1.4   Major Findings of EPAReports to Congress	1-4
         1.4.1  Study of HAP Emissions from Electric Utility Steam Generating Units	1-4
         1.4.2 Mercury Study Report	1-5
         1.4.3 Great Waters Reports	1-6
    1.5   Information Collection Request to Electric Utility Industry	1-6
    1.6   Regulatory  Finding on HAP Emissions from Electric Utility Steam Generating
         Units	1-8
    1.7   Mercury Emissions Research Programs  	1-9
    1.8   Relationship to Mercury Emission Control Research for Municipal Waste
         Combustors	1-9
    1.9   Report Organization	1-12
    1.10  References	1-15

Chapter 2.  Coal-fired Electric Utility Boilers
    2.1 Introduction	2-1
    2.2 Coal	2-1
         2.2.1 Coal Property Tests	2-1
              2.2.1.1 Coal Heating Value	2-2
              2.2.1.2 Coal Proximate Analysis	2-2
              2.2.1.3 Coal Ultimate Analysis	2-2
              2.2.1.4 Coal Mercury Analysis	2-3
         2.2.2 Coal  Classification	2-3
         2.2.3 United States Coal Resources	2-4
         2.2.4 Mercury Content  in Coals	2-7
    2.3 Coal Cleaning	  2-9
         2.3.1 Coal  Cleaning Processes	2-9
         2.3.2 Mercury Removal by Coal Cleaning	2-10
    2.4 Coal-fired Electric Utility Boilers	2-11
         2.4.1 Conventional Coal-fired Electric Utility Power Plants	2-11
         2.4.2 Coal-fired Cogeneration Facilities	2-13

                                            iii

-------
                                     Contents (continued)
                                                                                  Page
        2.4.3 Integrated Coal Gasification Combined Cycle Power Plants	2-14
   2.5 Coal-firing Configurations for Electric Utility Boilers	2-14
        2.5.1 Pulverized-coal-fired Furnace	2-18
        2.5.2 Cyclone Furnace	2-18
        2.5.3 Fluidized-bed Combustor	2-18
        2.5.4 Stoker-fired Furnace	2-19
        2.5.5 Gasified-coal-fired Combustor	2-19
   2.6 Ash from Coal Combustion	2-20
   2.7 Coals Burned by Electric Utilities in 1999	2-22
   2.8 References	2-26

Chapter 3.  Criteria Air Pollutant Emission Controls for Coal-fired Electric Utility Boilers
   3.1 Introduction	3-1
   3.2 Criteria Air Pollutants of Concern from Coal Combustion	3-2
        3.2.1 Particulate Matter	3-2
        3.2.2 Sulfur Dioxide	3-3
        3.2.3 Nitrogen Oxides	3-3
   3.3 Existing Control Strategies Used for Coal-fired Electric Utility Boilers	3-4
   3.4 Particulate Matter Emission Controls	3-5
        3.4.1 Electrostatic Precipitators	3-5
        3.4.2 Fabric Filters	3-8
        3.4.3 Particle Scrubbers	3-11
        3.4.4 Mechanical  Collectors	3-11
   3.5  SC>2 Emission Controls 	3-12
        3.5.1 Low-sulfur Coal 	3-12
        3.5.2 Fluidized-bed Combustion with Limestone	3-14
        3.5.3 Wet FGD Systems	3-14
        3.5.4 Spray Dryer Adsorber	3-15
        3.5.5 Dry Injection	3-15
        3.5.6 Circulating Fluidized-bed Adsorber	3-15
   3.6 NOXEmission Controls	3-16
        3.6.1 Combustion Controls	3-16
        3.6.2 Selective Catalytic Reduction	3-18
        3.6.3 SelectiveNoncatalytic Reduction	3-18
   3.7 Emission Control Configurations for Coal-fired Electric Utility Boilers	3-19
   3.8 References	3-21

Chapter 4.  Measurement of Mercury
   4.1 Introduction	4-1
   4.2 Manual Methods for Hg Measurements 	4-2
   4.3 Continuous Emission Monitors for Hg Measurements	4-8
   4.4 Summary, Conclusions, and Recommendations	4-17

                                            iv

-------
                                  Contents (continued)
                                                                                  Page
   4.5 References	4-19

Chapter 5. Mercury Speciation and Capture
   5.1 Introduction	5-1
   5.2 General Behavior of Mercury in Coal-fired Electric Utility Boilers	5-1
   5.3 Speciation of Mercury	5-3
        5.3.1 Gas-phase Oxidation	5-4
        5.3.2 Fly Ash Mediated Oxidation	5-9
        5.3.3 Oxidation by Post-combustion NOXControls	5-16
        5.3.4 Potential Role of Deposits, Fly Ash, and Sorbents on Mercury Speciation .. 5-16
   5.4 Capture of Mercury by Sorbent Injection	5-17
        5.4.1 Sorbent Characterization	5-17
        5.4.2 Experimental Methods Used in Sorbent Evaluation	5-18
              5.4.2.1  Bench-scale Reactors	5-18
              5.4.2.2  Pilot-scale Systems	5-25
              5.4.2.3  Full-scale Tests	5-25
        5.4.3 Research on Sorbent Evaluation	5-26
              5.4.3.1  Sorbent Evaluation Using Enhanced-flow Reactors	5-26
              5.4.3.2  Sorbent Evaluation Using Packed-bed Reactors	5-26
              5.4.3.3  Sorbent Evaluation Using Fluidized-bed Reactors	5-31
   5.5 Sorbent Development	5-32
        5.5.1 Powdered Activated Carbons	5-32
              5.5.1.1  Effects of Temperature, Mercury Concentration, and Acid Gases.... 5-32
              5.5.1.2  Role of Surface Functional Groups	5-33
              5.5.1.3  In-flight Capture of Mercury by a Chlorine-impregnated Activated
                    Carbon	5-33
        5.5.2 Calcium-based Sorbents	5-34
              5.5.2.1  Capture  of Low Concentrations of Mercury Using Calcium-based
                    Sorbents	5-34
              5.5.2.2  Simultaneous Control of Hg, SO2, andNOxby Oxidized-calcium-
                    based Sorbents	5-36
        5.5.3 Development of Low-cost Sorbents	5-37
        5.5.4 Modeling of Sorbent Performance	5-38
   5.6 Capture of Mercury in WetFGD Scrubbers	5-39
        5.6.1 Wet Scrubbing	5-39
        5.6.2 Oxidation	5-40
        5.6.3 Gas and Liquid  Oxidation Reagents	5-45
   5.7 Observations and Conclusions	5-45
   5.8 References	5-47

-------
                                 Contents (continued)
Chapter 6. Mercury Capture by Existing Control Systems Used by Coal-fired Electric Utility
          Boilers
   6.1 Introduction	6-1
   6.2 EPA ICR Part III Data	6-1
   6.3 Mercury Content of Utility Coals Burned in 1999	6-4
   6.4 Potential Mercury Capture in Existing Units	6-6
        6.4.1  Units with an ESP or FF	6-7
        6.4.2  Units with SDA Systems	6-8
        6.4.3  Units with Wet FGD Systems	6-8
        6.4.4  Units with Other Control Devices	6-9
   6.5 EPA's Part III ICR Data Evaluation Approach	6-9
        6.5.1  Evaluation Method	6-9
        6.5.2  Measure of Performance	6-10
        6.5.3  Comparison of Hgx (Inlet) Using OH Measurement and Coal Hg Data	6-14
        6.5.4  Development of Data Sets for Coal-boiler-control Classes	6-15
        6.5.5  Questionable Nature of OH Speciation Measurements Upstream of
               PM Controls	6-15
   6.6 Fuel, Boiler, and Control Technology Effects	6-17
        6.6.1  Coal Effects	6-19
        6.6.2  Control Technology Effects	6-21
        6.6.3   Post-combustion PM Controls	6-24
              6.6.3.1 Cold-side ESPs	6-24
              6.6.3.2 Hot-side ESPs	6-30
              6.6.3.3 FFBaghouses	6-33
              6.6.3.4 Comparison of ESPs andFFs	6-33
              6.6.3.5 Other PM Controls	6-36
        6.6.4  Hg Capture in Units with Dry FGD Scrubbers	6-37
        6.6.5  Hg Capture in Units with Wet FGD Scrubbers	6-41
        6.6.6  Potential  Effects of Post-combustion NOX Controls 	6-49
   6.7 Combustion System Effects	6-50
        6.7.1  Cyclone-fired Boilers	6-52
        6.7.2  Fluidized-bed Combustors	6-54
        6.7.3  IGCC Facilities	6-56
   6.8 National and Regional Emission Estimates	6-57
   6.9 Summary and Conclusions	6-59
   6.10 References	6-62

Chapter 7. Research and Development Status of Potential Retrofit Mercury Control Technologies
   7.1 Introduction	7-1
   7.2 Technology-based Mercury Control Strategies for Existing Coal-fired Electric Utility
        Boilers	7-2
        7.2.1  Remove Mercury Prior to Burning by Coal Cleaning	7-2

                                           vi

-------
                               Contents (continued)
                                                                                Page
     7.2.2 Retrofit Mercury Controls to Existing Air Pollution Control Systems	7-3
     7.2.3 Integrate Mercury Control Under a Multipollutant Control Strategy	7-4
7.3 Post-combustion Mercury Control Retrofit	7-4
     7.3.1 Cold-side ESP Retrofit Options	7-4
     7.3.2 Hot-side ESP Retrofit Options	7-5
     7.3.3 Fabric Filter Retrofit Options	7-5
     7.3.4 Spray Dryer Adsorber Retrofit Options	7-6
     7.3.5 WetFGD Scrubber Retrofit Options	7-6
     7.3.6 Particle Scrubber Retrofit Options	7-6
7.4 Retrofit Control Technology Research and Development Programs	7-7
     7.4.1 MWC Mercury Control Technology	7-7
     7.4.2 Pilot-scale Coal-fired Test Facilities	7-9
7.5 Mercury Control Retrofits for Existing Coal-fired Electric Utility Boilers Using
     ESP or FF Only	7-13
     7.5.1 Sorbent Injection Configurations	7-13
     7.5.2 Sorbent Adsorption Theory	7-14
     7.5.3 Pilot-scale and Full-scale Research and Development Status	7-15
          7.5.3.1  Coal Fly AshReinjection	7-15
          7.5.3.2 Activated Carbon Sorbent Injection	7-19
          7.5.3.3  Calcium-based Sorbent Injection	7-25
          7.5.3.4 Multipollutant Sorbent Injection	7-28
          7.5.3.5  Noble-metal-based Sorbent in Fixed-bed Configuration	7-30
7.6 Mercury Control Retrofits for Existing Coal-fired Electric Utility Boilers Using Semi-
     dry Adsorbers	 7-31
     7.6.1 Retrofit Options	7-31
     7.6.2 Pilot-scale and Full-scale Research and Development Status	7-31
7.7 Mercury Control Retrofits for Existing Coal-fired Electric Utility Boilers Using Wet
     FGD Scrubbers	7-31
     7.7.1 Retrofit Options	7-31
     7.7.2 Mercury Absorption  Theory	7-32
     7.7.3 Pilot-scale and Full-scale Research and Development Status	7-32
          7.7.3.1  Oxidation Additives	7-32
          7.7.3.2 Mercury Oxidation Catalysts	7-33
          7.7.3.3  WetFGD Scrubber Design and Operating Modifications	7-37
7.8 Multipollutant Control Technologies	7-43
     7.8.1 Corona Discharge	7-43
     7.8.2 Electron Beam Irradiation	7-44
     7.8.3 Oxidant Injection in Flue Gas	7-44
     7.8.4 Catalytic Oxidation	7-45
     7.8.5 Oxidant Addition to Scrubber	7-45
     7.8.6 Catalytic Fabric Filters	7-45
     7.8.7 Carbon-fiber FFs and ESPs	7-45
                                        vn

-------
                                 Contents (continued)
                                                                                Page
   7.9  Summary	7-46
   7.10 References	7-47

Chapter 8. Cost Evaluation of Retrofit Mercury Controls for Coal-fired Electric Utility Boilers
   8.1 Introduction	8-1
   8.2 Cost Estimate Methodology	8-2
        8.2.1  Mercury Control Technologies Evaluated	8-3
        8.2.2  Model Plant Descriptions	8-5
        8.2.3  Computer Cost Model	8-7
        8.2.4  PAC Injection Rate Algorithms	8-7
        8.2.5  Cost Estimate Assumptions	8-9
   8.3 Estimated Costs of Reducing Mercury Emissions	8-10
        8.3.1  Bituminous-coal-fired Boiler Using CS-ESP	8-11
        8.3.2  Subbituminous-coal-fired Boiler Using CS-ESP	8-14
        8.3.3  Subbituminous-coal-fired Boilers Using FF	8-14
        8.3.4  Coal-fired Boilers Using SCR for NOX Control	8-17
   8.4 Impacts of Selected Variables on Mercury Control Costs	8-17
        8.4.1  Acid Dew Point Approach Setting	8-17
        8.4.2  PAC Recycle	8-18
        8.4.3  Increased Flue Gas Residence Time	8-18
        8.4.4  Use of Composite PAC and Lime Sorbent	8-23
   8.5 Cost Indications for Other Model Plant Scenarios	8-23
   8.6 Projection of Future Mercury Control Costs	8-25
   8.7 Comparison of Mercury and NOX Control Costs	8-27
   8.8 Summary	8-29
   8.9 References	8-30

Chapter 9. Coal Combustion Residues and Mercury Control
   9.1 Introduction	9-1
   9.2 CCR Types	9-1
   9.3 CCR Mercury Concentrations	9-2
   9.4 Nationwide Management Practices	9-2
        9.4.1  Reuse and Recycling of CCRs	9-6
        9.4.2  Land-disposal of CCRs	9-6
   9.5 Current Status of CCR Research Activities	9-8
   9.6 Future  CCR Research Activities and Needs	9-9
   9.7 References	9-9

Chapter 10.  Conclusions and Recommendations
   10.1 Electric Utility Coal Combustion and Air Pollution Control Technologies	10-1
   10.2 Mercury Measurement Methods	10-2
   10.3 Mercury Speciation and Capture	10-3
                                          Vlll

-------
                              Contents (concluded)
                                                                             Page
     10.3.1  Mercury Speciation	10-3
     10.3.2  Development and Evaluation of Sorbents	10-4
10.4 Evaluation of EPA ICR Mercury Emission Test Data	10-4
10.5 Potential Retrofit Mercury Control Technologies	10-6
     10.5.1  Cold-side ESP, Hot-side ESP, and FF Systems	10-6
     10.5.2  Semi-dry FGD Scrubbers	10-7
     10.5.3  Wet FGD Scrubbers	10-7
10.6 Costs of Retrofit Mercury Control Technologies	10-7
10.7 Coal Combustion Residues and Mercury Control	10-8
10.8 Current and Future Research	10-9
                                  Appendices

A. Summary of Part II EPA ICR Data  Mercury Content and Selected Fuel Properties
   of As-fired Coals and Supplemental Fuels Burned in Coal-fired Electric Utility
   Boilers Nationwide in 1999 	A-l

B. Background Material of Methodology Used to Estimate 1999 Nationwide Mercury
   Emissions from Coal-fired Electric Utility Boilers	B-l

C. Summary of Part II EPA ICR Data  Mercury Capture Efficiencies of Existing
   Post-combustion Controls Used for Coal-fired Electric Utility Boilers	C-l

D. Assessment of Mercury Control Options for Coal-fired Power Plants	D-l
                                       IX

-------
                                        Figures

Figure                                                                        Page No.

2-1.    Distribution of coal deposits in the continental United States by USGS coal resource
       region  	2-6

2-2.    Simplified schematic of coal-fired electric utility boiler burning pulverized,
       low-sulfur coal	2-12

4-1.    Diagram of sampling train for Ontario-Hydro Method	4-3

4-2.    Comparison of Hg speciation measured by manual test methods from UND/EERC
       pilot-scale evaluation tests firing Blacksville bituminous coal and sampling and
       spiking Hg at FF inlet	4-6

4-3.    Comparison of gaseous Hg speciation measured by manual test methods from
       UND/EERC pilot-scale evaluation tests firing Blacksville bituminous coal and
       sampling and spiking Hg atFF outlet	4-7

4-4.    Comparison of total Hg results for CEMs at lowHg levels	4-15

4-5.    Comparison of Hg speciation results for CEMs at lowHg levels	4-16

5-1.    Mercury species distribution in coal-fired electric utility boiler flue gas	5-2

5-2.    Predicted distribution of Hg species at equilibrium, as a function of temperature for a
       starting composition corresponding to combustion of a bituminous coal (Pittsburgh) in
       air at a stoichiometric ratio  of 1.2	5-5

5-3.    Predicted distribution of Hg species at equilibrium, as a function of temperature for a
       starting composition corresponding to combustion of a subbituminous coal (Powder
       River Basin) in air at a stoichiometric ratio of 1.2	5-6

5-4.    Effects of SC>2 and water vapor on the gas-phase oxidation of Hg at 754 C and at three
       different Cl concentrations	
5-5.    Hg oxidation in the presence of the three- and four-component model fly ashes
       containing iron at abed temperature of 250 C	5-11

5-6.    Hg oxidation in the presence of the three- and four-component model fly ashes
       containing copper at abed temperature of 250 C	5-12

5-7.    Schematic of bench-scale fixed-bed reactor	5-21

-------
                                  Figures (continued)

Figure                                                                        Page No.

5-8.    Schematic of the bench-scale flow reactor with methane burner	5-22

5-9.    Schematic of bench-scale fluidized-bed reactor system	5-24

5-10.   Effect of particle size on adsorption for Darco FGD at 100 C, 86 ppb
       Hgconcentration, and 8.4 s contact time	5-27

5-11.   Example of the sampling and measurements taken during testing of the baseline test
       gas with HC1, NO2, and SO2	5-30

5-12.   Adsorption and subsequent oxidation of gaseous Hg in a simulated flue gas at 149 C
       (300 F)	5-42

5-13.   Adsorption and oxidation of gaseous Hg by various catalysts at 149 C (300 F)
       and371C(700F)	5-43

5-14.   Adsorption and oxidation of gaseous Hg by various coal fly ashes at 149 C (300 F)
       and371C(700F)	5-44

6-1.    1999 ICR data analyses - mercury in fuels	6-5

6-2.    Inlet versus outlet mercury concentration for all tests 	6-12

6-3.    Inlet mercury concentration versus  percent reduction for all tests  	6-12

6-4.    Effect of OH sample filter solids onHg speciation	6-17

6-5.    Inlet and outlet mercury concentrations for bituminous PC-fired boilers with
       CS-ESP	6-27

6-6.    Mercury emissions from bituminous coal-fired PC boilers with CS-ESP	6-28

6-7.    Mercury emissions for subbituminous- and lignite-fired PC boilers with
       CS-ESP	6-28

6-8.    Hypothetical effect of inlet and outlet Hgx concentration changes  on run-to-run Hgx
       capture	6-30

6-9.    Mercury emissions from bituminous-fired PC boilers with HS-ESP	6-32
                                           XI

-------
                                  Figures (continued)

Figure                                                                       Page No.

6-10.   Mercury emissions from subbituminous- and lignite-fired PC boilers with
       HS-ESP 	6-32

6-11.   Mercury emission reductions for PC-fired boilers withESPs andFFs	6-35

6-12.   Mercury speciation for PC-fired boilers with ESPs and FFs	6-35

6-13.   Relative mercury speciation for PC-fired boilers withESPs and FFs	6-36

6-14.   Mercury control for dry FGD scrubbers	6-40

6-15.   Mercury speciation for PC boilers with  SDA	6-40

6-16.   Relative mercury speciation for PC boilers with SDA	6-41

6-17.   Mercury speciation for PC boilers with wet FGD	6-45

6-18.   Mercury emissions for PC boilers with wet FGD	6-45

6-19.   Relative mercury speciation for PC boilers with wet FGD	6-46

6-20.   Mercury speciation for cyclone-fired boilers	6-53

6-21.   Relative mercury speciation for cyclone-fired boilers	6-53

6-22.   Mercury speciation for FBCs	6-55

6-23.   Relative mercury speciation for FBCs	6-55

7-1.    Schematic of 10-MWe coal-fired Babcock & Wilcox (B&W) Clean Environment
       Development Facility (CEDF) as used for Advanced Emissions Control
       Development Program (AECDP)	7-10

7-2.    Schematic of Particulate Control Module (PCM) at Public Service Company of
       Colorado (PSCO) Comanche Station	7-11

7-3.    Schematic of DOE/NETL in-house 500-lb/hr coal combustion test facility	7-12

7-4.    Hg removal by activated carbon injection measured at AECDP test facility burning
       Ohio bituminous coal and using ESP	7-22
                                         xn

-------
                                 Figures (concluded)

Figure                                                                     Page No.

7-5.   Hg removal by limestone injection measured in Combustion 2000 furnace using
      mechanical cyclone separator	7-26

7-6.   Hg removal by limestone injection measured at AECDP test facility burning Ohio
      bituminous coal and using ESP	7-27

7-7.   Effect of using H^S as an oxidation additive on wet FGC scrubber Hg removal
      measured at AECDP test facility burning Ohio bituminous coal	7-34

7-8.   Effect of using EDTA as an oxidation additive on wet FGD scrubber Hg removal
      measured at AECDP test facility burning Ohio bituminous coal	7-3 5

7-9.   Effect of oxidation air on wet FGD scrubber Hg removal as measured at AECDP
      test facility burning Ohio bituminous coal	7-39

7-10.  Effect of oxidation air on Hg in wet FGD scrubber flue gas as measured at AECDP
      test facility burning Ohio bituminous coal	7-40

7-11.  Effect of ESP operating voltage on wet FGD scrubber Hg removal as measured at
      AECDP test facility burning Ohio bituminous coal	7-41

7-12.  Effect of ESP operating voltage on Hg in wet FGD scrubber flue gas as measured at
      AECDP test facility burning Ohio bituminous coal	 7-42

8-1.   Change in total annual cost resulting from addition of ductwork to provide
      additional residence time	 8-22

8-2.   Change in total annual cost resulting from use of a composite PAC-lime sorbent
      instead of PAC	8-24

9-1.   Nationwide CCR management practices in the year 1999  	9-5
                                        Xlll

-------
                                         Tables

Table                                                                           Page No.

1-1.    Current research areas related to controlling Hg emissions from coal-fired electric
       utility power plants	1-10

2-1.    Demonstrated reserve base of major coal ranks in the United States estimated by
       DOE/EIA	2-5

2-2.    Mercury content of selected as-mined coal samples by coal rank and USGS coal
       resource region	 2-8

2-3.    Characteristics of coal-firing configurations used for electric utility power plants ... 2-15

2-4.    Nationwide distribution of electric utility units by coal-firing configuration for the
       year 1999 as reported in the Part II EPA ICR data  	2-17

2-5.    Nationwide quantities of coals and supplemental fuels burned in coal-fired electric
       utility boilers for the year 1999 as reported in the Part II EPA ICR data 	2-23

2-6.    Mercury content of as-fired coals and supplemental fuels burned in coal-fired electric
       utility boilers for the year 1999 as reported in the Part II EPA ICR data 	2-25

3-1.    Criteria air pollutant emission control strategies as applied to coal-fired electric utility
       boilers in the United States for the year 1999 as reported in the Part II EPA ICR data 3-6

3-2.    Nationwide distribution of existing PM emission controls used for coal-fired electric
       utility boilers for the year 1999 as reported in the Part II EPA ICR data 	3-7

3-3.    Comparison of PM collection efficiencies for different PM control device types 	3-9

3-4.    Nationwide distribution of existing SC>2 emissions controls used for coal-fired electric
       utility boilers for the year 1999 as reported in the Part II EPA ICR data	3-13

3-5.    Nationwide distribution of existing NOX emissions controls used for coal-fired
       electric utility boilers for the year 1999 as reported in the Part II EPA ICR data	3-17

3-6.    Nationwide distribution of post-combustion emission control configurations used for
       coal-fired electric utility boilers for the year 1999 as reported in the Part II EPA ICR
       data	3-20

4-1.    Summary of selected manual test methods evaluated for measurement of Hg  in
       combustion gases	4-4
                                           xiv

-------
                                  Tables (continued)

Table                                                                          Page No.

4-2.    Results from EPA Method 301 evaluation tests for the Ontario-Hydro Method  	4-9

5-1.    Percent oxidation of Hg by simulated and actual coal-fired electric utility boiler fly
       ash	 5-13

5-2.    Comparison of bench-scale fixed-bed with entrained-flow reactors	5-20

5-3.    Composition of test gases to simulate coal combustion flue gas used for UND/EERC
       bench-scale study	5-29

5-4.    Mercury removal by lime sorbent injection as measured by EPA bench-scale tests.. 5-35

5-5.    Simulated flue gas conditions with the most active catalysts and fly ashes indicated
       for oxidation of gaseous Hg to gaseous Hg2+	5-41

6-1.    Distribution of ICR mercury emission test data by boiler-coal type configurations.... 6-3

6-2.    Distribution of ICR mercury emission test data for pulverized-coal-fired boilers by
       post-combustion emission control device configuration	6-4

6-3.    Comparison of mercury content normalized by heating value in as-fired coals and
       supplemental fuels for electric utility boilers in 1999	6-6

6-4.    ICR mercury emission test allocations by coal-boiler-control class	6-16

6-5.    Average mercury capture by existing post-combustion control configurations used for
       PC-fired boilers	6-19

6-6.    Effects of coal and control technology inlet and outlet SPF and capture for
       PC-fired boilers	6-21

6-7.    Average mercury emission factors and percent reduction for coal-boiler-control
       classes	6-22

6-8.    Number of coal-fired utility boilers equipped with particulate matter controls only . 6-24

6-9.    Type of fuel used in PC-fired units equipped with CS-ESP	6-24

6-10.   Post-combustion controls: cold-side ESPs	6-25

6-11.   Post-combustion controls: hot-side ESPs	6-31

                                          xv

-------
Errata Page xvi, dated 3-21-02
                                   Tables (continued)

  Table                                                                       Page
  No.

  6-12.  Mercury (HgT) reduction at PC-fired units with FF baghouses	6-33

  6-13.  Post-combustion controls:  FF baghouses	6-34

  6-14.  Post-combustion controls:  miscellaneous PM controls	6-37

  6-15.  Post-combustion controls:  dry FGD scrubbers	6-38

  6-16.  PC-fired boiler PM controls for wet FGD systems	6-42

  6-17.  Post-combustion controls:  wetFGD scrubbers	6-43a

  6-18.  WetFGD scrubbers burning bituminous coal	6-47

  6-19.  WetFGD scrubbers burning subbituminous coal	 6-48

  6-20.  Wet FGD scrubbers burning TX lignite	6-49

  6-21.  Potential effects of post-combustion NOX control technologies on mercury capture
         in PC-fired boilers burning bituminous coal	6-49

  6-22.  Cyclone-fired boilers	6-51

  6-23.  Fluidized-bed combustors	6-52

  6-24.  Comparison of class average Hgx reductions for PC- and cyclone-fired boilers	6-54

  6-25.  Calculated mercury removal in IGCC power plants using bituminous coal	6-56

  6-26   Nationwide coal burned and mercury emitted from electric utility coal-fired power
         plants in 1999	6-59

  7-1.    Comparisons of typical uncontrolled flue gas parameters for coal-fired utility
         boiler versus municipal waste combustor (MWC)	7-8

  7-2.    Hg removal by native fly ashes measured across PM control devices at PSCO power
         plants burning selected western coals	7-17

  7-3.    Hg removals by fly ash reinjection measured across PCM at PSCO Comanche power
         plant for selected western coals	7-18
                                       xvi

-------
                                  Tables (continued)

Table                                                                         Page No.

7-4.    Effect of flue gas temperature on fly ash Hg adsorption measured across PCM at
       PSCO Comanche power plant burning PRB subbituminous coal 	7-20

7-5.    Hg removal by activated carbon injection measured at PSE&G Hudson Station
       burning low-sulfur bituminous coal and using ESP	 7-21

7-6.    Hg removal by activated carbon injection measured at DOE/NETL in-house test
       facility burning low-sulfur bituminous coal and using FF	7-24

7-7.    Comparison of Hg removals for activated carbon injection versus limestone
       injection measured at AECDP test facility burning  Ohio bituminous coal and using
       ESP	 7-29

7-8.    Comparison of field test results using flue gas from electric utility boiler firing Texas
       lignite versus bench-scale results using simulated flue gas for selected candidate Hg
       oxidation catalysts	7-36

8-1.    Mercury control technologies	8-4

8-2.    Matrix of model plant scenarios	8-6

8-3.    Estimated total annual mercury control costs for bituminous-coal-fired boiler with
       existing CS-ESP	8-12

8-4.    Estimated total annual mercury control costs for subbituminous-coal-fired boiler
       with existing CS-ESP	8-15

8-5.    Estimated total annual mercury control costs for subbituminous-coal-fired boiler
       with existing FF	 8-16

8-6.    Impact of acid dew point setting on annual mercury control costs for a 500-MWe
       electric utility boiler burning bituminous coal	8-19

8-7.    Impact of acid dew point setting on annual mercury control costs for a 500-MWe
       electric utility boiler burning subbituminous coal	8-20

8-8.    Effect of PAC recycle on annual mercury control costs for a 500-MWe electric utility
       boiler burning bituminous coal	8-21

8-9.    Projected future mercury control costs	8-26
                                          xvn

-------
                                 Tables (concluded)




Table                                                                       Page No.




8-10.  Comparison of mercury control costs withNOx control costs	8-28




9-1.   Coal combustion residues	9-3




9-2.   Calculated Hg concentrations in CCRs using EPA ICR data	9-3




9-3.   Summary of available test data on Hg concentrations in major types of CCRs	9-4




9-4.   Commercial uses for CCRs generated in 1999	9-7
                                        xvin

-------
                               Acronyms
 ADP
AES
AHC
ASTM
CAA
CCR
CEM
CFBA
COHPAC
CS-ESP
CuCl
CVAAS
CVAFS
DI
DOE
EPA
EPRI
ESP
ETV
FBC
FF
FGD
HAP
Hg
Hg
HgO
Hg2+
HgP
HgT
HgCl2
HgS04
HS-ESP
Acid Dew Point
Atomic Emission Spectrometry
Advanced Hybrid Collector
American Society for Testing and Materials
Clean Air Act
Coal Combustion Residues
Continuous Emission Monitors
Circulating Fluidized-bed Adsorber
Compact Hybrid Particulate Collector
Cold-side Electrostatic Precipitator
Cuprous Chloride
Cold-vapor Atomic Absorption Spectrometry
Cold-vapor Atomic Fluorescence Spectrometry
Dry Injection
United States Department of Energy
United States Environmental Protection Agency
Electric Power Research Institute
Electrostatic Precipitator
Environmental Technology Verification
Fluidized-bed Combustion
Fabric Filter
Flue Gas Desulfurization
Hazardous Air Pollutant
Mercury
Elemental Mercury
Mercuric Oxide
Oxidized or Ionic Mercury
Particle-bound Mercury
Total Mercury
Mercuric Chloride
Mercuric Sulfate
Hot-side Electrostatic Precipitator
                                    xix

-------
                           Acronyms (continued)
IGCC
ICR
kWh
LNB
MC
MESA
MWC
MWe
MWFA
NESCAUM
NETL
NOX
OAR
OH Method
O&M
PAC
PFF
PM
PRB
PS
PTFE
QA/QC
RfD
SC
SCR
SDA
SEM
SNCR
Integrated Gasification Combined Cycle
Information Collection Request
Kilowatt Hour
Low NOX Burner
Mechanical Collector
Mercury Speciation Adsorption
Municipal Waste Combustor
Megawatt Electric
Mixed Waste Focus Area
Northeast States for Coordinated Air Use Management
National Energy Technology Laboratory (DOE)
Nitrogen Oxides
EPA's Office of Air and Radiation
Ontario-Hydro Method
Operation and Maintenance
Powdered Activated Carbon
Polishing Fabric Filter
Particulate Matter
Powder River Basin
Particle Scrubber
Polytetrafluoroethyl ene
Quality Assurance/Quality Control
Reference Dose
Spray Cooling
Selective Catalytic Reduction
Spray Dryer Adsorber
Scanning Electron Microscope
Selective Noncatalytic Reduction
                                    xx

-------
                            Acronyms (concluded)
TGM                Total Gaseous Mercury
T TNTT-V rcT^n /->          University of North Dakota/Energy and Environmental
UJNJJ/iiiiKLx                 ,   _
                     Research Center
UVDOAS            Ultraviolet Differential Optical Absorption Spectroscopy
Wet FGD            Flue Gas Desulfurization by Liquid Scrubbing

-------
Errata Page xxii, dated 4-23-02
                             Acknowledgements

      This report summarizes and interprets research sponsored by the U.S. Environmental
Protection Agency (EPA), the U.S. Department of Energy, the Electric Power Research
Institute, and other organizations. The Research Triangle Institute (RTI) was instrumental in
summarizing the Phase I and Phase II Mercury Information Collection Requests.  The RTI
summaries provided: data on equipment used at coal-fired utility electrical generating plants
in 1999, data on coal characteristics and usage, and estimates of the 1999 annual mercury
emissions. RTI staff members who made significant contributions to this effort included
Jeffrey Cole, Paul Peterson, James Turner, and Robert Zerbonia. Many thanks to William H.
Maxwell of EPA's Office of Air Quality Planning and Standards, Research Triangle Park, NC,
who reviewed the draft version of this report and provided many helpful suggestions.

      The report was reviewed  by an external peer review panel chaired by Constance L.
Senior of Reaction Engineering International, Salt Lake City, UT. The panel also included
Praveen Amar, Director of Science and Policy at NESCAUM, Boston, MA, and Massoud
Rostam-Abadi, Senior Chemical  Engineer at Illinois State Geological Survey and Adjunct
Professor of Environmental Engineering at the University of Illinois at Urbana-Champaign,
Champaign, IL. They provided many excellent comments and advice that resulted in
substantial improvements in the document.
                                       xxn

-------
                                 Executive Summary
 Overview
        This report documents current knowledge on the emission and control of mercury (Hg)
 from coal-fired electric utility plants. The purpose of the report is to provide information on the
 status of government and industry efforts in developing improved technologies for the control of
 Hg emissions.

        This is an interim report, which contains information available in the public domain prior
 to June 2001.  Since then, the results of additional research have been published. This additional
 information can be found in DOE, EPA,  and EPRI reports, in journal articles, and in the
 proceedings of conferences. Two recent conferences provided significant new information on
 the control of Hg emissions  the A&WMA 2001 Annual Conference (Orlando, FL, June 2001),
 and the A&WMA Specialty Conference  on Mercury (Chicago, IL, August 2001).

        The first part of the report (Chapters 1 through 3) is directed to readers outside the
 research community who are interested in Hg emission and Hg control issues. Information is
 provided on:

          Legislative and regulatory background of EPA's December 2000 decision to regulate
           Hg emissions from coal-fired electric utility generating stations,

          Studies made in support of EPA's regulatory determination,

          Fuels, combustion technologies, and pollution control technologies used for coal-fired
           steam electric generating units, and

          Research results from an official  Information Collection Request (ICR) on the fuels
           and technologies used by the  utility industry in 1999 at coal-fired steam electric
           generating stations.

       The second part of the report (Chapters 4 through 10) is directed to all readers. It focuses
on the review and evaluation of information that has been gathered since the publication of:
EPA's Mercury Study Report to Congress: EPA's Study of Hazardous Air Pollutant Emissions
from Electric Utility Steam Generating UnitsFinal Report to Congress:  and the  A&WMA
                                          ES-1

-------
Critical Review:  Mercury Measurement and Its Control. The second part of the report contains
information on:

            Hg measurement methods,

            Forms of Hg (speciation) and the capture of Hg in flue gas from combustion of coal,

            Evaluation of the ICR flue gas data on Hg concentrations upstream and downstream
            of air pollution control devices (APCDs),

           Summary of retrofit control technologies that can be used to limit Hg emissions at
            coal-fired plants currently equipped with particulate matter (PM) control devices,
            and dry or wet flue gas desulfurization (FGD) scrubbing systems,

          Estimates of the costs of controlling Hg emissions by the use of powdered activated
            carbon (PAC),

           Overview of the current coal combustion residue (CCR) management practices and
            the identification of environmental issues requiring additional research, and

           Conclusions, overview of current research, and research recommendations.

       Detailed supporting information is provided in Appendices.


 Background

       The 1990 Clean Air Act Amendments required EPA to study the health and
 environmental impacts of hazardous air pollutants (HAPs) emitted from electric utility boilers.
 The Agency was also required to conduct a study of the potential health and environmental
 impacts of Hg emitted from anthropogenic sources in the United States. The EPA  subsequently
 published an 8-volume Mercury Study Report to Congress in December 1997 and a Study of
 Hazardous Air Pollutant Emissions from Electric Utility Steam  Generating UnitsFinal Report
 to Congress in February 1998.  The Hg report to Congress identified coal-fired utility boilers as
 the largest single anthropogenic source of Hg emissions in the United States. The utility HAP
 report indicated that there was a plausible link between Hg emissions from coal-fired boilers and
 health risks posed by indirect exposure to methylmercury.

       In December 2000, EPA announced its intent to regulate HAP emissions from coal- and
 oil-fired electrical generating stations. The decision to regulate HAP emissions from coal-fired
 units was based on:
                                          ES-2

-------
         A National Academy of Science study on the health effects of methylmercury,

         The collection and analysis of coal- and flue-gas Hg data under an official Information
         Collection Request (ICR), and

         Studies concerning the status of Hg emission control technologies.

       Three important milestones are incorporated in EPA's decision to regulate HAP
emissions from coal-fired electric generating units:

         The proposal of regulations by December 2003,

         The promulgation of regulations by December 2004, and

         Compliance with the regulations by December 2007.


Electric Utility Coal Combustion and Air Pollution Control Technologies

       The EPA ICR data collection effort was conducted in three phases. In Phase I,
information was collected on the fuels, boiler types, and air pollution control devices (APCDs)
used at all coal-fired utility boilers in the United States. In Phase n, coal data were collected and
analyzed by the utility industry for 1,140 coal-fired and three integrated gasification, combined
cycle (IGCC) electric power generating units.  Each coal sample was analyzed for Hg content,
chlorine (Cl) content, sulfur content, moisture content, ash content, and calorific value. In Phase
IE, flue gas Hg measurements were  made using the modified Ontario-Hydro (OH) Method for
total and speciated Hg.  Additional coal samples were collected and analyzed in conjunction with
the OH Method measurements.

       The EPA ICR data indicated that, in 1999, coal-fired steam electric generating units in the
U.S. burned 786 million tons of coal of which about 52 percent was bituminous and 37 percent
was subbituminous. Other fuels included lignite, anthracite coal, reclaimed waste coal, mixtures
of coal and petroleum coke (pet-coke), and mixtures of coal and tire-derived fuel (TDF).
Pulverized coal-fired (PC) boilers represent approximately 86 percent of the total number and 90
percent of total utility boiler capacity. Based on capacity,  other types of boilers include cyclone-
fired boilers (7.6 percent), fluidized-bed combustors (1.3 percent), and stoker-fired boilers (1.0
percent).

       The 1999 EPA ICR responses indicate that a variety of emission control technologies are
employed to meet requirements for sulfur dioxide (802), nitrogen oxides (NOX), and particulate
matter (PM).  Most utilities control NOX by combustion modification techniques and SO2 by the
use of compliance coal.  For post-combustion controls, 77.4 percent of the units have PM control
only, 18.6 percent have both PM and SO2 controls, 2.5 percent have PM and NOXcontrols, and 1.3
percent have three post-combustion control devices.
                                          ES-:

-------
       The different types of post-combustion control devices are listed below:

             Particulate matter (PM) control technologies include electrostatic
             precipitators (ESPs), fabric filters (FFs) (also called "baghouses"), and
             particulate scrubbers (PS). ESPs and FFs may be classified as either cold-
             side (CS) devices [installed upstream of the air heater where flue gas
             temperatures range from 284 to 320 F (140 to 160 C)] or hot-side
             [installed downstream of the air heater and operate at temperatures ranging
             from 662 to 842 F (350 to 450 C)]. Based on current information, it
             appears that little Hg can be captured in HS-ESPs.

             SO2post-combustion control technologies are systems that are classified as
             wet flue gas desulfurization (FGD) scrubbers, semi-dry scrubbers, or dry
             injection. Wet FGD scrubber controls remove 862 by dissolving it in a
             solution.  A PM control device is always located upstream  of a wet
             scrubber. PM devices that may be used with wet FGD scrubbers include a
             PS, CS-ESP, HS-ESP, or FF baghouse.  Semi-dry scrubbers include spray
             dryer absorption (SDA). Dry injection involves injecting dry powdered
             lime or other suitable sorbent directly into the flue gas. A PM control
             device (ESP or FF) is always installed downstream of a semi-dry scrubber
             or dry injection point to remove the sorbent from the flue gas.

             NOxpost-combustion control technologies include selective non-catalytic
             reduction (SNCR) and selective catalytic reduction (SCR) processes. With
             both of these methods,  a reducing agent such as ammonia or urea is
             injected into the duct to reduce NOX to N2.  SCR operates at lower
             temperatures than SNCR and is more effective at reducing  NOX, but it is
             more expensive.

       For PM control, ESPs are used on 84 percent of the existing electric utility coal-fired
boiler units, and FF baghouses are used on 14 percent of the utility units.  Post-combustion SC>2
controls are less common. Wet flue gas desulfurization (FGD) systems are used on 15.1 percent
of the units; and, dry scrubbers, predominantly spray dryer absorbers (SDA), are used on 4.6
percent of units that were surveyed. While the application of post-combustion NOX controls is
becoming more prevalent, only 3.8 percent of units used either selective non-catalytic reduction
(SNCR) or selective catalytic reduction (SCR) systems in 1999.

Mercury Measurement Methods

       When the coal is burned in an electric utility boiler, the resulting high combustion
temperatures vaporize the Hg in the coal to form gaseous elemental mercury (Hg).  Subsequent
cooling of the combustion gases and interaction  of the gaseous Hg with other combustion
products result in a portion of the Hg being converted to gaseous oxidized forms of mercury
(Hg2+) and particle-bound mercury (Hgp). The term speciation is used to denote the relative
                                          ES-4

-------
amounts of these three forms of Hg in the flue gas. The total Hg in flue gas (Hgx) is the sum of
Hgp, Hg2+, and Hg. It is the ability to measure these forms of Hg, either collectively or
individually, which distinguishes the capabilities of available measurement methodologies.

       The Hg in flue gas can be measured by either manual sampling methods or by the use of a
continuous emission monitor (CEM). Manual methods are available for the measurement  of HgT
and the speciation of Hg, including Hgp. CEMs are now available to measure gas-phase Hgx.

Manual Test Methods

       Manual sampling methods for measuring Hgx from combustion processes are well
established.  EPA Methods 101A and 29 are routinely used to measure HgT in flue gas from
incineration and coal combustion. While a validated reference method for the measurement of
the speciated forms of Hg does not exist, the Ontario-Hydro (OH) method is the de facto method
of choice.

       Generally, sampling trains used to collect flue gas samples for Hg analysis consist of the
same components: a nozzle and probe operated to extract a representative sample from a duct or
stack; a filter to collect PM; and a series of impingers with liquid reagents to capture gas-phase
Hg.  Sampling trains used for speciation measurements sequentially capture Hg2+ and Hg in
different impingers. After sampling, the filter and sorption media are prepared and analyzed for
Hg in a laboratory.

       While several research methods exist for performing speciated Hg measurements, the OH
Method is presently the method of choice for measuring Hg species in the flue gas from coal-
fired utility plants.  The OH method has been shown to provide valid Hg speciation
measurements when samples are taken downstream of an efficient PM control device. However,
the OH Method can give erroneous speciation measurements for locations upstream of PM
control devices because of sampling artifacts.

       Fly ash captured by the sampling train filter can absorb Hg2+ and Hg. Catalytic
properties of the fly ash can also oxidize Hg, resulting in physical and chemical transformations
within the sampling train. Transformations caused by the sampling process are called artifacts,
and the resulting measurements do not accurately reflect critical properties of Hg at the locations
where the samples were taken. Sampling methods have not yet been developed to overcome
measurement artifacts associated with high flue gas concentrations of fly ash.

Continuous Emission Monitors (CEMs)

       Continuous emission monitors (CEMs) are in some respects superior to manual
measurement methods. CEMs provide a rapid real-time or near real-time response, which  can be
used to characterize temporal process variations that cannot be measured with manual
measurement methodologies. Mercury CEMs are similar to most combustion process CEMs in
that a flue gas sample must be extracted from the stack and then transferred to the analyzer for
                                         ES-5

-------
detection. However, Hg monitoring is complicated by the fact that Hg exists in different forms
and that quantitative transport of all forms is difficult.

       The CEMs designed to measure total gas-phase Hg (Hg2+ and Hg) are now routinely
used in Europe and Japan to measure Hg emissions from incinerators. The Hg concentrations in
the stack gas from well-controlled emission sources contain negligible amounts of Hgp, and the
measurement of gas-phase Hg downstream of the emission control devices can be considered to
be equivalent to the measurement of HgT.

       The detectors in Hg CEMs typically measure Hg by the use of cold vapor atomic
absorption spectroscopy (CVAAS) or cold vapor atomic fluorescence spectroscopy (CVAFS).
HgT concentrations are measured by converting (reducing) all of the Hg2+ in the sample to Hg
before it enters the detector. Various conversion techniques exist, including thermal, catalytic,
and wet chemical methods. The wet chemical technique is currently used in commercial
monitors that are capable of speciation measurement. The use of wet chemical reagents results in
high operating costs, which are the primary limitation to the Hg CEM's use as a compliance tool.

       Speciating Hg CEMs are highly valuable as research tools. Several commercially
available Hgx CEMs have been modified to indirectly measure Hg2+ by determining the
difference between gas-phase HgT and Hg.  Hg CEMs are susceptible to the same PM-related
measurement artifacts associated with manual measurements, and users of Hg CEMs in high dust
conditions must consider this problem.

       Regardless of the sampling method, the key to reliable and accurate Hg sampling and
continuous monitoring is maintaining sample integrity.  Flue gases may contain particles that
change the  species of Hg within the sampling train or CEM system. While this does not change
the total Hg measurement, it may bias the determination of Hg vapor species, which may be used
to estimate the potential for Hg capture, as well as to assess the performance of control devices.
Similarly, common flue gas constituents, such as SC>2, HC1 and NOX, may affect quantitative
measurement performance.

       Additional research is needed to investigate and overcome measurement obstacles so that
speciating CEMs can serve as process monitors and as a research tool for evaluating the
effectiveness of emission controls.  Such research can also provide a better understanding of the
factors that affect Hg speciation.

Speciation and Capture of Mercury

Mercury Speciation

       The capture of Hg by flue gas cleaning devices is dependent on Hg speciation.  Both Hg
and Hg2+ are in vapor-phase at flue gas cleaning temperatures. Hg is insoluble in water and
cannot be captured in wet scrubbers.  The predominant Hg2+compounds in coal flue gas are
weakly to strongly soluble, and the more-soluble species can be generally captured in wet FGD
                                         ES-6

-------
                  n       9+
scrubbers. Both Hg and Hg  are adsorbed onto porous solids such as fly ash, powdered
activated carbons (PAC), or calcium-based acid gas sorbents for subsequent collection in a PM
                 94-                                              n
control device.  Hg  is generally easier to capture by adsorption than Hg . Hgp is attached to
solids that can be readily captured in ESPs and FFs.

       Flue gas cleaning technologies that are applied on combustion sources employ three basic
methods to capture Hg:

            Capture of Hgp in PM control devices;
            Adsorption of Hg and Hg2+ onto entrained sorbents for subsequent capture in PM
             control devices; and
            Solvation of Hg2+ in wet scrubbers.

       The factors that affect the speciation and capture of Hg in coal-fired combustion systems
include the type and properties of coal, the combustion conditions, the types of flue gas cleaning
technologies employed, and the temperatures at which the flue gas cleaning systems operate.

       Oxidation reactions that affect the speciation of Hg include homogeneous, gas-phase
reactions and heterogeneous gas-solid reactions associated with entrained particles and surface
deposits. Suspected flue gas oxidants involved in Hg oxidation include oxygen (62), ozone (63),
hydrochloric acid (HC1), chlorine (Cl), nitrogen dioxide (NO2) and sulfur trioxide (SOs). Many
of these oxidants are also acid species, which may be significantly impaired by the presence of
alkaline species in fly ash, such as sodium, calcium and potassium. Heterogeneous oxidation
reactions may be catalyzed by metals such as iron, copper, nickel, vanadium, and cobalt.
Conversion of Hg to Hg2+ may be followed by adsorption to form Hgp.

       The determination of which mechanisms, oxidants, and catalysts are dominant is crucial
in developing and implementing Hg control strategies. For example, the impaired oxidation of
Hg in subbituminous coals and lignites is probably related to lower concentrations of HC1 in flue
gas and high alkalinity of the fly ash. PM collectors and scrubbers reflect this in the low
removals of Hg in the ICR database.

Fundamentals ofSorption

       Sorbents used for the capture of Hg can be classified as Hg sorbents or multipollutant
sorbents. Sorbents evaluated for Hg capture have been manufactured from a number of different
materials such as lignite, bituminous coal, zeolites, waste biomass, and waste tires. The
manufacturing process typically involves some type of thermal treatment. Additives are often
used to produce impregnated sorbents.

       For coal-fired electric utility boiler applications,  the use of sorbents to capture gas-phase
Hg (or gas-phase Hg and acid gases) is limited to the use of finely ground powdered sorbents.
These sorbents can be injected upstream of PM control devices to collect the sorbent and
adsorbed Hg. The development of improved sorbents is needed because of poor sorbent
                                          ES-7

-------
utilization that results from low flue gas concentrations of Hg and short sorbent exposure times
in units equipped with CS-ESPs. The performance of a sorbent is related to its physical and
chemical characteristics.  The best performing sorbents must be carefully matched to
performance requirements as defined by the application for which it is to be used. For example,
properties and performance requirements of sorbents used for capture of SC>2 and Hg are quite
different.  In a similar fashion, the performance criteria for sorbents used with flue gas from
bituminous coal will probably be different from the sorbents used with sub-bituminous coals.

       Sorbents are porous materials. The most common physical properties related to sorbent
performance are surface area, pore size distribution, and particle size distribution. The capacity
for Hg capture generally increases with increasing surface area and pore volume. The  ability of
Hg and other sorbates to penetrate into the interior of a particle is related to pore size distribution.
The pores of the sorbent must be large enough to provide free access to internal surface area by
Hg and Hg2+while avoiding excessive blockage by previously adsorbed reactants.  As particle
size decreases, access to the internal surface area of the particle increases, along with potential
adsorption rates. Powdered activated carbons used for Hg control typically have diameters of 44
|im or smaller.

       Mercury can be either physically or chemically adsorbed. Physical adsorption
(physisorption)  typically results from van der Waals and Coulombic (electrostatic) interactions
between the sorbent and the sorbate.  The resulting bonds are weak (typically < 10-15 kcal/mole)
and are easily reversed.

       Chemical adsorption (chemisorption) involves the establishment of a chemical bond (as
the result of a chemical reaction, electron transfer). Chemisorption results in stronger bonds than
physisorption and is not necessarily reversible.  Chemical adsorption is also dependent on the
presence of chemically active sites where the sorbate is chemically bound.  Some of the chemical
constituents of activated carbons influencing Hg capture include: sulfur content, iodine content,
and chlorine content.  Impregnation of carbons with sulfur, iodine, or chlorine can increase the
reactivity and capacity of sorbents.  Hg is likely oxidized and sorbed in a rapid two step reaction,
either chemically by reaction with strong ionic groups such as Cl", I", or S~ or physically through
interaction with functional groups in  sorbent pores.

       The HgCb is readily adsorbed onto both carbon and calcium based sorbents, probably
by acid-base reactions. Section 5.5 details the fundamental research to develop carbon and
calcium sorbents for Hg vapor capture.

Evaluation of Sorbents

       Sorbents may be evaluated by bench-, pilot-scale, or full-scale experiments.  The initial
screening of sorbents has typically been conducted using bench-scale, packed-bed experimental
reactors. These reactors are used to evaluate the adsorption  capacity of sorbents exposed to Hg
in a synthetic flue gas made from compressed bottled gases. The reactor is held at a
predetermined temperature, and either Hg or HgCb is fed into the synthetic flue gas upstream of
                                           ES-8

-------
the reactor. An on-line Hg analyzer is used to continuously monitor the Hg content of the inlet
flue gas and of that after exposure to the sorbent fixed bed. These reactors are used to determine
the effects of temperature and flue gas composition on the performance of sorbents. These
reactors provide results that are primarily applicable to the capture of Hg in FF baghouses.

       Flow reactors that expose sorbents to flue gas during short residence experiments can be
used to simulate conditions associated with ESPs. These reactors can be used to explore the rates
of Hg adsorption and determine the effects of temperature and flue gas composition.  The most
effective screening tests are conducted with reactors that are installed on a slip stream from  a
pilot- or full-scale coal combustion system.  Large pilot- or full-scale tests must be used to
assess the effects of mass transfer limitations (i.e., mixing and diffusion of flue gas constituents)
and long-term equipment operability.

 Wet FGD Scrubbers

       Oxidized mercury compounds such as HgCb are soluble in water and alkaline scrubbing
solutions. Thus, the oxidized fraction of Hg vapors in flue gas is effectively captured when a
power plant is operated with wet or semi-dry scrubbers for removing 862.  The elemental
fraction, on the  other hand, is insoluble and is not removed to any significant degree. The
challenge to Hg removal in wet FGD scrubbers, then, is to find some way to oxidize the
elemental Hg vapor before it reaches the scrubber or to modify the liquid phase of the scrubber to
cause oxidation to occur.

Evaluation of EPA ICR Mercury Emissions  Data

       The methods used to evaluate the ICR data were based on two interrelated objectives.  The
first method was to estimate the speciated amount and the geographical distribution of national Hg
emissions from  coal-fired power plants in 1999. The second method was to characterize the
effects of coal properties, combustion conditions, and flue gas cleaning methods on the speciation
and capture ofHg.

Mercury Capture by Existing Air Pollution Control Devices

       The air pollution control technologies now used  on pulverized-coal-fired utility boilers
exhibit average  levels of Hg control that range from 0 to 98 percent, as shown in Table ES-1.  The
best levels of control are generally obtained by emission control systems that use FFs. The
amount of Hg captured by a given control technology is better for bituminous coal than for either
subbituminous coal or lignite.

       The lower levels of Hg capture in plants firing subbituminous coal and lignite are
attributed to low fly ash carbon content and the higher relative amounts of Hg in the flue gas from
combustion of these fuels. The average capture of Hg based on OH Method inlet measurements
in PC fired plants equipped with a cold-side ESP is 35 percent for bituminous coal, 3 percent for
sub-bituminous coal and near zero for lignite.
                                          ES-9

-------
Errata Page ES-10, dated 3-21-02

Table ES-1.   Mean mercury emission reduction for pulverized-coal-
fired boilers.
Post-combustion Emission
Controls
Used for PC Boiler
PM Control
Only
PM Control
and
Spray Dryer
Adsorber
PM Control
and
Wet FGD
System
CS-ESP
HS-ESP
FF
PS
SDA + ESP
SDA + FF
SDA + FF +
SCR
PS + FGD
CS-ESP + FGD
HS-ESP + FGD
FF + FGD
Average Mercury Emission Reduction (%) a
Bituminous-coal-
fired
36%
9%
90%
not tested
not tested
98%
98%
12%
75%
49%
98%
Subbituminous-
coal-fired
3%
6%
72%
9%
35%
24%
not tested
-8%
29%
29%
not tested
Lignite-
fired
-4%
not tested
not tested
not tested
not tested
0%
not tested
33%
44%
not tested
not tested
  a) Mean reduction from test 3-run averages for each PC boiler unit in Phase III EPA ICR data base.

      Plants that employ only post-combustion PM controls display average Hg emission
reductions ranging from 0 to 89 percent. The highest levels of control were observed for units
with FFs. Decreasing levels of control were shown for units with ESPs, PS, and mechanical
collectors.

      Units equipped with lime spray dryer absorber scrubbers (SDA/ESP or SDA/FF
systems) exhibited average Hg captures ranging from 98 percent for units burning bituminous
coals to 3 percent for units burning subbituminous coal. The predominance of Hg in stack gas
units that are fired with subbituminous coal and lignite results from low levels of Hg
oxidization.

      The capture of Hg in units equipped with wet FGD scrubbers is dependent on the
relative amount of Hg + in the  inlet flue gas and on the PM control technology used. Average
Hg captures in wet FGD scrubbers ranged from 23 percent for one PC-fired HS-ESP + FGD
unit burning subbituminous coal to 97 percent in a PC-fired FF + FGD unit burning
bituminous coal.  The high Hg capture in the FF + FGD unit is attributed to increased
oxidization and capture of Hg  in the FF.

      Mercury captures in PC-fired units equipped with spray dry scrubbers and wet limestone
scrubbers appear to provide similar levels of control on a percentage reduction basis. However,
this observation is based on a small number of short-term tests at a limited number of facilities.
                                      ES-10

-------
Additional testing will be required to characterize the effects of fuel, combustion conditions, and
APCD conditions on the speciation and capture of Hg.

National Emission Estimates

       The data used for estimating the national Hg emissions were:  (1) the mean Hg content of
coal burned in any given unit during 1999, (2) the amount of coal burned in that unit during 1999,
and (3) best match coal-boiler-control device emission factor for the unit.  The results of these
estimates indicated that:

         Coal and related fuels burned in coal-fired utility boilers in 1999 contained 75 tons of
          Hg, and

         Forty-eight tons of Hg was emitted to the atmosphere in 1999 from coal-fired utility
          power plants.

Multipollutant Controls

       The EPA ICR data indicate that technologies currently in place for control of criteria
pollutants achieve reductions in Hg emissions that range from 0 to  > 90 percent. Current levels
of Hg control can be increased by application of retrofit technologies or methods designed to
increase capture of more than one pollutant. This multipollutant approach can utilize the
synergisms that accrue through the simultaneous  application of technologies for NOX and Hg
control, SC>2 and Hg control, or 862, NOX, and Hg control.

       Bench- and pilot-scale tests have shown that Hg capture in PM control devices generally
increases as the carbon content of fly ash increases. Increased use of combustion modification
techniques that increase ash carbon content will generally increase  the amount and capture of
Hgp.

       The EPA ICR data indicate that SCR systems may enhance the oxidation and capture of
Hg. Recent pilot- and full-scale tests on bituminous coal-fired units equipped with SNCR + CS-
ESP and SCR + SDA/FF systems have confirmed these results. However, improvement in Hg
capture appears to be highly dependent on the type of coal burned and the design and operating
conditions of SCR systems. The potential in increased Hg capture  associated with the NOX
control system cannot now be quantified.  It is believed, however, that the use of combustion
modification techniques  and post combustion NOX control technologies on NOX state
implementation plan (SIP) units will also increase the capture of Hg in these units.

       The retrofit of coal-fired electric utility boiler units to control emission of 862 and fine
PM is also expected to provide co-benefits in the control of Hg.  This is apparent from the
increased control of Hg on units equipped with FFs, dry FGD scrubbers, and wet FGD scrubbers.
Mercury or multipollutant sorbents will add minimal capital costs to units that are retrofitted with
FFs or SDA/FF for control of other pollutants.  The use of multipollutant sorbents would be more
                                         ES-11

-------
costly, but the incremental costs of Hg control would be modest.  Technologies designed for use
on existing wet FGD units could also be used for new scrubbers that are intended to control 862
and the precursors to secondary fine PM.

       Generally, the control of Hg emissions via multipollutant control technologies can
provide a cost-effective method for collectively controlling the various pollutants of concern.

Potential Retrofit Mercury Control Technologies

       A practical approach to controlling Hg emissions at existing utility plants is to minimize
capital costs by adapting or retrofitting the existing equipment to capture Hg. Potential retrofit
options for control of Hg were investigated for units that currently use any of the following post
combustion emission control methods: (1) ESPs or FFs for control of PM, (2) dry FGD
scrubbers for control of PM and 862, and (3) wet FGD scrubbers for the control of PM and 862.

ESP and FF Systems

       Least costly retrofit options for the control of Hg emissions from units with ESP or FF are
believed  to include:

      Injection of a sorbent upstream of the ESP or FF.  Cooling of the stack gas or
       modifications to the ducting may be needed to keep sorbent requirements at acceptable
       levels.

      Injection of a sorbent between the ESP and a pulsejet FF retrofitted downstream of the
       ESP.  This approach will increase capital costs but reduce sorbent costs.

      Installation of a semi-dry circulating fluidized-bed absorber (CFA) upstream of an
       existing ESP used in conjunction with sorbent injection. The CFA recirculates both fly
       ash and sorbent to create an entrained bed with a large number of reaction sites. This
       leads to higher sorbent utilization and enhanced fly ash  capture of Hg and other
       pollutants.

       Units equipped with a FF require less sorbent than units equipped with an ESP. ESP
systems depend on in-flight adsorption of Hg by entrained fly ash or sorbent particles. FFs
obtain the same in-flight Hg adsorption as ESPs and additional adsorption as the flue gas passes
through the FF cake.

       In general, the successful application of cost-effective sorbent injection technologies for
ESP and  FF units will depend on: (1) the development of lower cost and/or higher performing
sorbents, and (2) appropriate modifications to the operating conditions of equipment being
currently used to control emission of PM, NOX,  and 862.
                                         ES-12

-------
Semi-Dry FGD Systems

       SDA systems that use calcium-based sorbents are the most common dry FGD systems
used in the electric utility industry. An aqueous slurry containing the sorbent is sprayed into an
absorber vessel where the flue gas reacts with the drying slurry droplets.  The resulting, particle-
laden, dry flue gas then flows to an ESP or a FF where fly ash and 862 reaction products are
collected.

       CFAs are "vertical duct absorbers" that allow simultaneous gas cooling, sorbent injection
and recycle, and gas absorption by flash drying of wet lime reagents. It is believed that CFAs can
potentially control Hg emissions at costs lower than those associated with use of spray dryers.

       Dry FGD systems are already equipped to control emissions of SC>2 and PM.  The
modification of these units by the use of appropriate  sorbents for the capture of Hg and other air
toxics is considered to be the easiest retrofit problem to  solve.

Wet FGD Systems

       Wet FGD systems are typically installed downstream of an ESP or FF. Wet limestone
FGD scrubbers are the most commonly used scrubbers on coal-fired utility boilers. These FGD
units are expected to capture more than 90 percent of the Hg2+in the flue gas  entering the
scrubber. Consequently, existing wet FGD scrubbers may lower Hg emissions between 20 and
80 percent, depending on the speciation of Hg in the inlet flue gas.

       Improvements in wet scrubber performance in capturing Hg depend primarily on the
oxidation of Hg to Hg2+. This may be accomplished by (1) the injection of appropriate
oxidizing agents or (2) the installation of fixed oxidizing catalysts upstream of the scrubber to
promote oxidization of Hg to soluble species.

       An alternative strategy for controlling  Hg emissions from wet FGD scrubbing systems is
to inject sorbents upstream of the PM control  device. In wet FGD systems equipped with ESPs,
performance gains are limited by the in-flight  oxidization of Hg and the in-flight capture of Hg2+
and Hg. In systems equipped with FFs, increased oxidization and  capture of Hg can be achieved
as the flue gas flows through the FF.  Increased oxidization of Hg in the FF will result in
increased Hg removal in the downstream scrubber.

Multipollutant Control Methods

       From a long-term perspective, the most cost-effective Hg controls will be those
implemented with a multipollutant emission control  scheme, wherein Hg sorbents also remove
other pollutants, and catalysts and absorbers are employed to remove bulk contaminants such as
NO and 862.  Mercury is also removed as a consequence of using particular bulk gas sorbents,
catalysts, particle collectors, and absorbers. Therefore, while sorbents injected upstream of PM
collectors may be readily employed for Hg control, the best long-term schemes will result from
                                         ES-13

-------
modifying or adding control systems for other pollutants that also control Hg emissions. Chapter
9 discusses several applications under development.

Costs of Retrofit Mercury Control Technologies

       Preliminary annualized costs of Hg controls using powdered activated carbon (PAC)
injection have been estimated based on recent pilot-scale evaluations with commercially
available adsorbents (see Table ES-2).  These control costs range from 0.305 to 3.783 mills/kWh,
with the highest costs associated with plants having hot-side electrostatic precipitators (HS-
ESPs). For plants representing 89 percent of current capacity and using controls other than HS-
ESPs, the costs range from 0.305 to 1.915 mills/kWh.  Assuming a 40 percent reduction in
sorbent costs by use of a composite lime-PAC sorbent for Hg removal, cost projections range
from 0.18 to 2.27 mills/kWh with higher costs again being associated with plants using HS-
ESPs.
Table ES-2.  Estimates of current and projected annualized operating costs for
retrofit mercury emission control technologies.
Coal Type
(sulfur content)
Bituminous
(3% S)
Bituminous
(0.6% S)
Subbituminous
(0.5% S)
Existing
APCDa
CS-ESP+FGD
FF+FGD
HS-ESP+FGD
CS-ESP
FF
HESP
CS-ESP
FF
HESP
Retrofit
Mercury Control b
PAC
PAC
PAC+PFF
SC+PAC
SC+PAC
SC+ PAC+PFF
SC+PAC
SC+PAC
SC+PAC+PFF
Current Cost
(mills/kWh)
0.727-1.197
0.305-0.502
1.501 -NAC
1.017-1.793
0.427-0.753
1.817-3.783
1.150-1.915
0.423-1.120
1.419-2.723
Projected Cost
(mills/kWh)
0.436-0.718
0.183-0.301
0.901 - NAC
0.610-1.076
0.256-0.452
1.090-2.270
0.69-1.149
0.254-0.672
0.851 -1.634
    a)  CS-ESP = cold-side electrostatic precipitator; HS-ESP = hot-side electrostatic precipitator; FF= fabric filter;
       FGD = flue gas desulfurization
    b)  PAC=powdered activated carbon; SC=spray cooling; PFF=polishing fabric filter
    c)  NA = not available


       In comparison, the estimated annual costs of Hg controls, as a function of plant size, lie
mostly between the costs for low-NOx burners (LNBs) and selective catalytic reduction (SCR)
systems.  The costs of Hg control will dramatically diminish if retrofit hardware and sorbents are
employed for control of other pollutants such as NOX, 862, or fine PM.

       The performance and cost estimates of PAC injection-based Hg control technologies
presented in this document are based on relatively few data points from pilot-scale tests and are
considered to be preliminary. However, based on pilot-scale tests and the results of ICR data
                                         ES-14

-------
evaluations, better sorbents and technologies now being developed will reduce the costs of Hg
controls beyond current estimates.

       Within the next 2 to 3 years, the evaluation of retrofit technologies at plants where co-
control is being practiced will lead to a more thorough characterization of the performance and
costs of Hg control.  Future cost studies will focus on the development of performance and cost
information needed to refine cost estimates for sorbent injection based controls, will develop cost
estimates for wet scrubbing systems that employ methods for oxidizing Hg, and will determine
the costs of various multipollutant  control options.

       The issue of Hg in residues will also be examined to address concerns related to the
release of captured Hg species into the environment. These evaluations will be conducted in
conjunction with the development  and evaluation of air pollution emission control technologies.

Coal Combustion Residues  and Mercury Control

       Operation of power plants results in solid discharges including fly ash, bottom ash, boiler
slag, and FGD residues. These residues already contain Hg, presumably bound Hg that is
relatively insoluble and non-leachable. In 1998, approximately 108 million  tons of coal
combustion residues (CCRs)  were  generated. Of this amount, about 77 million tons were
landfilled and about 31 million tons were utilized for beneficial uses.

       Increased control of Hg emissions from coal-fired power plants may change the amount
and composition of CCRs. Such changes may increase the potential for release of Hg to the
environment from either landfilling or uses of CCRs. Mercury volatilization from CCRs  in
landfills and/or surface impoundments is expected to be low due to the low temperatures
involved and the existence of relatively small surface area per unit volume of residue. For Hg
control retrofits involving dry or wet FGD scrubbers, the residues are typically alkaline and the
acid leaching potential of Hg from  these residues is expected to be minimal.

       There are several commercial uses of CCR where available data on which to characterize
the Hg emission potential are lacking. The following CCR uses are given a  priority for
developing additional data in order to characterize the ultimate fate of Hg:

      The use of fly ash in cement production,
      The volatilization and leaching of residues used for structural fills,
      Leaching of residues exposed to the acidic conditions during mining applications,
      Volatilization of Hg during the production of wallboard from gypsum in wet scrubber
       residues,
      Mercury volatilization during the production and application of asphalt with fly ash
       fillers, and
      Leaching or plant uptake of Hg from fly ash, bottom ash, and FGD sludge that are used as
       soil amendments.
                                         ES-15

-------
Current and Planned Research

       DOE, EPA, EPRI, the utility industry, and the control technology industry are funding
research on the control of Hg emissions from coal-fired boilers.  A major portion of this research
is being funded under cooperative agreements with DOE. These agreements include cost sharing
by EPRI and other industrial partners.  Research on these projects is being jointly coordinated
under DOE's, EPA's, and EPRI's Hg control technology programs. These research efforts will be
used to:

          Develop hazardous air pollution Maximum Achievable Control Technology (MACT)
          requirements for electric utility generating units,

          Optimize control of Hg emissions from units that must comply with more stringent
          NOX emission requirements under the NOX SIP, and

          Develop technologies that can be used to control emissions under multipollutant
          control legislation options that are currently being considered.

       Current research efforts include three full-scale test projects, six pilot-scale test projects
on coal -fired units, the evaluation of Hg CEMs, supporting research on the speciation and
capture of Hg, and research on CCRs and CCBs.  This research includes:

         One full-scale ESP sorbent injection project with tests at four sites,

         One full-scale wet FGD scrubber project at two sites,
         One full-scale project on the effects of SNRC, SCR, and SOs conditioning
          systems at five sites,

         On-going research on the development and use of Hg CEMs,

         On-going speciation, capture, and sorbent development research, and

         Small Business Administration projects on development of sorbents, and
          measurement methods.

Six new pilot-scale DOE projects have been announced in FY2001 . These are:

         Advance parti culate collector with sorbent injection (North Dakota-EERC)

         Evaluation of Hg oxidization catalysts (URS Radian Group)

         Spray cooling and multipollutant sorbents (CONSOL)

         Evaluation of multipollutant sorbents and CFBA (SRI)

         Electrical discharge multi-pollution control system (Power Span)

         Evaluation of advanced sorbents (Apogee Scientific)
                                         ES-16

-------
Conclusions and Recommendations

       Additional efforts are planned to characterize the behavior of Hg in coal combustion
systems.  Further research is needed on the speciation and capture of Hg and on the stability of
Hg in CCRs and residue by-products. Studies on the control capabilities and costs of potential
Hg retrofit technologies currently under pilot-scale development are being continued and
appropriate control technologies are to be evaluated on full-scale units. Additionally, an
evaluation of the co-control of Hg with available PM, 862, and NOX controls is needed.

       Mercury measurement and monitoring capabilities must be consistent with the regulatory
approaches being considered; e.g., speciated vs. total Hg emissions.  Field activities need to be
coordinated to (1) improve the emissions data base, (2) develop the technologies most
appropriate for Agency goals (e.g., Hg-specific vs. multipollutant), and (3) refine cost data and
cost-performance models based on actual field experience.

       Finally, EPA must continue to work closely with DOE, EPRI and the utility industry to
develop Hg and multipollutant control technologies.   Collaboration will help ensure that all of
the scientific knowledge, engineering skills, and financial  resources  needed to develop  control
technologies and establish the most cost-effective regulatory requirements are available.

       Current and future research should focus on:

         Control of emissions for units with ESPs,

         Control of Hg emissions from subbituminous coals and lignite,

         Evaluation of CFA systems,

         Demonstration of Hg control for units with SD A/ESP and SDA/FF systems,

         Development of Hg oxidizing methods for wet FGD systems,

         Evaluation additives for the oxidization of Hg and the sequestration
              2+
         of Hg  in wet scrubbers,

         Enhancement of fly ash capture by combustion modification techniques,

         Optimization of NOX controls for Hg control,

         Control of Hg and other air toxic emissions from units equipped with SCR
         and wet FGD scrubbers,

         Use and evaluation of Hg CEMs,

         Tests with CEMs to study the variability of Hg emissions,

         Effects of coal blending on Hg capture, and

         Effects of cyclone-, stoker-, and fluidized-bed combustion on Hg control.


                                         ES-17

-------
                                      Chapter 1
                                 Report Background
1.1 Introduction
      Mercury (Hg) is a metallic element that can be released into the atmosphere from both
anthropogenic (i.e., made by humans) and natural sources. Ambient Hg concentrations in the air
are typically very low. Human exposure by direct inhalation of Hg in the air is not the
predominant public health concern for this metal. However, the Hg in ambient air eventually can
be re-deposited on land surfaces or directly into rivers, lakes, and oceans.  Mercury that enters
bodies of water by direct deposition from the air or runoff from land surfaces ultimately is
transformed by biological processes into a highly toxic form of Hg (methylmercury [MeHg]) that
concentrates in fish and other organisms living in these waters.  A study by the National
Academy of Sciences (NAS) concluded that human exposure to MeHg from eating contaminated
fish and seafood is associated with adverse health effects related to neurological and
developmental damage varying in severity depending on the Hg concentrations in the ingested
food.1 An extreme example of these health effects cited by this study is the high-dosage
exposure from the consumption of MeHg-contaminated fish by the residents living near
Minamata Bay in Japan in the 1950s that resulted in fatalities and severe neurological damage.2

      The largest anthropogenic source of Hg emissions in the United States is the Hg released
from burning coal to produce steam for generating electricity. Mercury naturally occurs in trace
amounts in all coal  deposits.  When coal is burned in a steam boiler or a furnace, most of the Hg
bound in the coal is released during the combustion process as gaseous elemental mercury (Hg).
Subsequent cooling of the combustion gases and interaction of the gaseous Hg with other
combustion products  result in a portion of the Hg being converted to gaseous oxidized forms of
mercury (Hg2+) and particle-bound mercury (Hgp).

      Coal-fired electric utility power plants currently do not use air pollution controls
specifically designed  to reduce Hg emissions to the atmosphere. However, certain control
technologies now used at coal-fired electric utility power plants to reduce other air pollutant
emissions (particulate matter [PM], sulfur dioxide [802], nitrogen oxides [NOX]) also reduce Hg
emissions with varying levels of effectiveness.  Methods for enhancing Hg removal by these
existing controls are being studied.  New control technologies to specifically control Hg
emissions from coal combustion are being developed. Multipollutant control technologies that
will achieve both high Hg removal and effective control of PM, SC>2, and NOX are being
investigated.

                                         1-1

-------
      The Clean Air Act (CAA) directs the U.S. Environmental Protection Agency (EPA) to
regulate emissions of air toxics from stationary sources by establishing national air emission
standards for hazardous air pollutants (NESHAP).  Mercury is one of the compounds listed under
CAA Section 112 as a hazardous air pollutant (HAP). The EPA Administrator has found that it
is appropriate and necessary to establish a NESHAP regulating HAP emissions, including Hg,
from coal-fired electric utility power plants.
1.2 Report Purpose

      The EPA Office of Research and Development (ORD) National Risk Management
Research Laboratory (NRMRL) has prepared this Hg emission control technology report. The
overall purpose of the report is to review and evaluate recent scientific data and new knowledge
about control technologies that potentially can be used to reduce Hg emissions from coal-fired
electric utility power plants. The first part of the report is directed to readers outside the research
community involved in Hg emission control issues by providing background information
regarding EPA's NESHAP decision, the use of coal for electrical  power generation, and Hg
behavior in coal combustion gases. The second part of the report  is directed to all readers and
focuses on a review and evaluation of new information that has been gathered by the EPA since
the Agency's reports to Congress related to the control of Hg emissions from electric utility
power plants. Also included in this report are summaries of the results to date from companion
NRMRL studies investigating the costs of retrofitting potential Hg control technologies to
existing coal-fired electric utility power plants in the United States and Hg behavior in the ash
and other solid residues from coal combustion.

      The remainder of Chapter 1 provides a summary of the statutory authority and past major
studies completed by the EPA that led to the Agency's regulatory  finding on the HAP  emissions
from electric utility power plants.  Background on major research  programs investigating Hg
emissions from coal combustion is presented. This chapter concludes with a description of
topics presented in Chapters 2 through 10 of this report.
1.3 NESHAP Statutory Background

      Title HI of the CAA regulates stationary sources that emit HAPs. Section 112 in Title HI
was comprehensively amended in 1990.  Under the amended CAA Section 112(b), Congress
listed specific chemicals, compounds, and groups of chemicals as HAPs.  Mercury is one of the
chemicals included on this HAP list. The EPA is directed by Section 112 to regulate the HAP
emissions from stationary sources by establishing "national emission standards for hazardous air
pollutants or ANESHAP.@ The EPA develops and promulgates individual NESHAPs for specific
categories of stationary sources. The NESHAP for a given source category is codified under its
own subpart in the Code of Federal Regulations under part 63  to title 40.
                                        1-2

-------
      Section 112 of the CAA established specific directives as to how the EPA must develop
NESHAPs. The statute requires that each NESHAP must require the maximum degree of HAP
emission reduction that is achievable, taking into consideration the cost of achieving such an
emission reduction and any non-air quality health and environmental impacts and energy
requirements.  The control technology that achieves this level of HAP emission control is called
Amaximum achievable control technology or AMACT.@

      The 1990 CAA Amendments include several provisions in Section 112 that specifically
address the regulation of HAP emissions from electric utility steam generating units.  First, CAA
Section 112(a) defines the term Aelectric utility steam generating unit@ to mean

      A. . . any fossil fuel fired combustion unit of more than 25 megawatts that serves a
      generator that produces electricity for sale.  A unit that cogenerates steam and
      electricity and supplies more than one-third of its potential electric output capacity
      and more than 25 megawatts electrical output to any utility power distribution
      system for sale shall be considered an electric utility steam generating unit. @

      Section 112(n)(l)(A) directs the EPA to perform a study and report to Congress about the
hazards to public heath reasonably anticipated to occur as result of exposure to HAP emissions
from electric utility steam generating units.  After considering the result of this study, the EPA
must determine whether regulation of electric utility steam generating units under Section 112 is
appropriate and necessary. In July 1995, the EPA submitted its draft version of the report for
peer review and, concurrently, released that version of the report for public review and comment.
The EPA completed the final report and submitted  to it Congress in February 1998.3

      A related directive in Section 112(n)(l)(B) requires the EPA to perform a second study
and report to Congress about Hg emissions from electric utility steam generating units, municipal
waste combustion units, and other sources including area sources.  This section directs the EPA=s
study to consider the rate and mass of the Hg emissions from these sources, the health and
environmental effects of such emissions, the technologies that are available to control such
emissions, and the cost of these technologies.  The  EPA completed this study and submitted its
final report to Congress in December  1997.4

      The 1990 CAA amendments to Section 112  also direct the EPA to perform additional
studies that include analyses of Hg emissions from  electric utility steam generating units.
Included among these studies is the requirement under CAA Section 112(m) for the EPA to study
the atmospheric deposition of HAPs to the Great Lakes, Chesapeake Bay, Lake Champlain, and
coastal waters.  This group of surface water bodies  collectively is referred to as the AGreat
Waters. Section 112(m) directs the EPA to investigate the contribution of atmospheric
deposition to pollutant loadings in the Great Waters; environmental and public health effects of
atmospheric pollution deposited to these waters; and the sources of the pollutants deposited to
these waters.  Three reports to Congress on the atmospheric deposition of pollutants to the Great
Waters have been prepared to date (May  1994, June 1997, and June 2000).5'6'7
                                         1-2

-------
      In addition to requiring the EPA to prepare the above cited reports, Congress directed the
EPA to fund an independent evaluation conducted by the NAS of the available data related to the
health impacts of MeHg and provide recommendations for the reference dose (RfD) to be used
for health impact analyses.  The RfD is the amount of a chemical which, when ingested daily
over a lifetime, is anticipated to be without adverse health effects to humans, including sensitive
subpopulations.  The NAS conducted an 18-month study of the available data on the health
effects of MeHg and published a report of its findings in 2000.l  On the basis of its evaluation,
the NAS committee's consensus is that the value of EPA's current RfD for MeHg is a
scientifically appropriate level for the protection of public health.
1.4 Major Findings of EPA Reports to Congress

1.4.1  Study of HAP Emissions from Electric Utility Steam Generating Units

      The findings of the EPA=s study of the hazards to public heath reasonably anticipated to
occur as result of exposure to HAP emissions from electric utility steam generating units are
presented in the two-volume report titled Study of Hazardous Air Pollutant Emissions from
Electric Utility Steam Generating UnitsFinal Report to Congress.^ The assessment for Hg in
the report includes a description of Hg emissions, deposition estimates, control technologies,  and
a dispersion and fate modeling assessment that includes predicted levels of Hg in various media
(including soil, water, and freshwater fish) based on modeling from four representative utility
plants using hypothetical scenarios. The EPA did not evaluate human or wildlife exposures to
Hg emissions from utilities in that report. With regard to non-inhalation exposures (e.g.,
ingestion) to other HAPs, the report presents a limited qualitative discussion of arsenic,
cadmium, dioxins, and lead.

      Based on information  and analyses available at the time the report was prepared, electric
utility steam generating units can emit a significant number of the HAPs listed in CAA Section
112(b).  However, except for Hg, electric utility steam generating units are responsible for a very
small percentage of the total  nationwide emissions of these particular HAPs. The EPA
concluded that Hg emitted from coal-fired steam generating units is the HAP of greatest potential
concern for electric utility steam generating units. For two other HAPs (arsenic and dioxin), the
EPA=s analysis concluded that further evaluations and review are needed to better characterize
the impacts of these HAP emissions from coal-fired steam generating units.

      Nickel emissions are the only HAP emissions of potential concern from oil-fired electric
utility steam generating units. The EPA acknowledged that there are significant uncertainties
concerning the chemical forms of nickel emitted from these units and the health effects of those
various nickel compounds. At the time the study was prepared, the EPA projected that future
nationwide nickel emissions  from oil-fired steam generating units would decrease because of
anticipated declining use of oil by utilities for electric power generation.
                                         1-4

-------
       The impacts due to HAP emissions from natural-gas-fired steam generating units are
negligible based on the results of the study. The EPA concluded that no further evaluation is
needed of HAP emissions from natural-gas-fired electric utility steam generating units.

      The EPA identified uncertainties that make it difficult to quantify the magnitude of the
risks due to Hg emissions from coal-fired electric utility steam generating units, and identified
the research areas where more information is needed to gain a better understanding of the risks
and impacts of these Hg emissions. Included among the research areas that the EPA
recommended for further evaluation were:  1) collection and assessment of additional data on the
Hg content of various types of coals; 2) collection and assessment of additional data on Hg
emissions from coal-fired steam generating units; 3) collection and assessment of additional
information on control technologies or pollution prevention options; and 4) further review of the
available data on the health impacts associated with exposure to Hg. Following completion of
the report, the EPA initiated studies addressing the identified research needs.

1.4.2  Mercury Study Report

      The findings of the EPA=s assessment of the magnitude of Hg emissions from sources in
the United States, the health and environmental implication of those emissions, and the
availability and costs of control technologies are presented in the eight-volume report titled
Mercury Study Report to Congress4 The report provides an extensive analysis of the public
health impacts and environmental impacts resulting from Hg emissions to the atmosphere and
deposition on surface waters and land. The findings of the report related to Hg emissions from
electric utility steam generating units and other anthropogenic sources in the United States (as
discussed in Volume n of the report) are  summarized below.

      Mercury cycles in the environment occur as a result of both natural processes and human
activities (anthropogenic sources).  The EPA prepared a nationwide inventory of annual Hg
emissions from anthropogenic sources in the United States.  This inventory was based on the
period 1994-1995 and estimated the total annual nationwide emissions of Hg to be 144
megagrams (158 tons).  Most of these emissions (approximately 87  percent) are produced when
waste or fuels containing Hg are burned.  Four specific source categories account for
approximately 80 percent of the total nationwide anthropogenic emissions:  coal-fired electric
utility boilers (33 percent), municipal waste combustors (19 percent), industrial and commercial
boilers (18 percent), and medical waste incinerators (10 percent).  Another 10 percent of the Hg
emissions were estimated to be from manufacturing sources that use Hg as a processing agent,
product ingredient, or where Hg is present as a trace constituent in a process raw material.  The
largest manufacturing sources are chloro-alkali plants and Portland cement manufacturing plants.
The remaining 3 percent of the emissions were estimated to be released from area and
miscellaneous sources.

      In the report, the EPA also assessed future trends in Hg emissions. Emissions from two of
the significant combustion sources identified in the 1994-95 nationwide inventory are predicted
to decline significantly when the national emission standards for municipal waste combustors
                                         1-5

-------
(MWCs) and medical waste incinerators are fully implemented.  Industrial use of Hg was found
to be declining in those manufacturing sectors where acceptable substitute materials can be used
(e.g., use of electronic thermometers in place of Hg thermometers, elimination of Hg additives in
paints and pesticides, reduced use of Hg in batteries).

1.4.3  Great Waters Reports

      The findings of the EPA=s study of the atmospheric HAP deposition to the Great Waters
are presented in a series of three reports to Congress; the first report dated May 1994, the second
report dated June 1997, and the third report dated June 2000. The HAPs of concern emitted from
electric utilities addressed by the Great Waters study include lead, cadmium, dioxins, and, in
particular, Hg.

      The first Great Waters report to Congress noted that the water bodies are polluted by
HAPs that originate from both local and distant sources; however, more data are needed to
identify the specific sources of the pollutants.  The report recommendations were the following:
1) the EPA should strive to reduce emissions of the pollutants of concern through
implementation of the CAA; 2) a comprehensive approach should be taken, both within the EPA
and with other agencies, to reduce and preferably prevent pollution in air, water, and soil; and 3)
the EPA  should continue to support research for emissions inventories, risk assessment, and
regulatory benefits assessment.

      The second Great Waters report to Congress confirmed, and provided additional support
for, the findings of the first report that persistent and bioaccumulative toxic pollutants and
excessive nitrogen can adversely affect the environmental conditions of the Great Waters.
Electric utilities and mobile sources are identified by the report based on air modeling studies and
emissions data as major contributors of nitrogen oxides to the Chesapeake Bay and its watershed.

      The most recent Great Waters report to Congress presents updated scientific and
programmatic information to support and build upon the broad conclusions presented in the first
two reports. Specific to Hg, fate and transport modeling and exposure assessments presented in
the report predict that the anthropogenic contribution of the total amount of MeHg in fish is, in
part, the result of Hg releases from combustion and industrial sources. Furthermore,
consumption offish is the dominant pathway of exposure to MeHg for fish-consuming humans
and wildlife.
1.5 Information Collection Request to Electric Utility Industry

      The EPA=s 1998 report to Congress on HAP emissions from electric utility steam
generating units identified additional information needed to gain a better understanding of the
risks, impacts, and control of Hg emissions from coal-fired steam generating units.  As part of the
Agency=s effort to gather this information, the EPA conducted an information collection project
beginning in late 1998 to survey all coal-fired steam generating units meeting the CAA Section
                                         1-6

-------
112(a) definition that were operating in the United States. This information collection provided
the EPA with data on the Hg content and amount of coal burned by these units during the 1999
calendar year.  As part of the information request, the EPA also selected a subset of the coal-fired
electric utility steam generating units at which field-source testing was performed to obtain Hg
emission data for the air pollutant control devices now being used for these units.

      There were three parts to the EPA information collection effort. Part I of this effort
consisted of gathering the information to first identify the location of each coal-fired steam
generating unit meeting the CAA Section 112(a) definition that was operating in the United
States. The EPA sent information collection requests (ICRs) to the owners and operators of
approximately 1,100 facilities that potentially could be operating coal-fired steam generating
units. Information requested in the Part I questionnaire sent to each of these facilities included
the type of coal burned, the method of firing the coal, and the methods used for control  of air
pollutants.  Based on the ICR responses, 1,143 coal-fired steam generating units that meet the
CAA Section 112(a) definition were identified at 461 facilities. These coal-fired steam
generating units were located across the entire nation in 47 of the 50 states, with the exceptions
being Idaho, Rhode Island, and Vermont.

      Part n of the information collection effort,  during calendar year 1999, consisted of
gathering information on the quantities, Hg  content, and other selected properties of coal burned
by each of the  1,143 coal-fired steam generating units. The owner or operator of each coal-fired
steam generating unit provided to the EPA,  on a quarterly basis, analysis results for samples of
the coal fired in the steam generating unit.  These analyses were performed according to a
demonstrably acceptable protocol and reported the Hg content of the coal burned and other
important coal properties (e.g., coal heating value and the sulfur, ash, moisture, and chlorine
contents). Each owner or operator also reported data on the  total amount of coal burned on a
monthly basis during 1999.

      Part HI of the information  collection effort consisted of conducting Hg emission source
testing at selected electric utility  power plants operating coal-fired steam generating units. The
test locations were selected by the EPA to approximate the nationwide distribution of coal-fired
steam generating units by type of boiler, coal burned, and air emission controls used.  The testing
at each location was performed by the facility owner or operator (or a source testing contractor
hired by the facility). At each of the selected test locations, measurements were made of the Hg
content in the inlet and outlet gas stream for the farthest downstream control device used on the
unit. The testing followed an EPA-approved sampling protocol and included three sample runs
at each sampling location. Samples of the coal burned during the source test were also  collected.
Each test was completed and a final test report was provided to the EPA. The EPA review of the
test reports ultimately found acceptable test results for 80 coal-fired steam generating units.

      All of the nationwide industry survey data (information collected for Part I of the survey),
coal analysis data (information collected for Part n of the survey), and Hg emission testing (data
collected for Part in of the survey) are available to the public on the EPA web site,
. Selected information from the ICR
                                          1-7

-------
data base are also summarized in chapters of this report as related to characterizing the coal
properties, control configurations, and Hg emissions from existing coal-fired electric utility
steam generating units. In this report, the term "EPA ICR data" is used to refer to the
compilation of coal-fired electric utility power plant, coal property, and Hg emissions data
gathered by this nationwide information collection project.
1.6 Regulatory Finding on HAP Emissions from Electric Utility Steam Generating Units

      On December 20, 2000, the EPA published in the Federal Register a notice (65 FR 79825)
presenting the EPA Administrator's finding as to whether regulation of emissions of HAP from
fossil-fuel-fired electric utility steam generating units is appropriate and necessary. This finding
is based on the results of EPA's reports to Congress, the EPA's analysis of the ICR responses,
and other information the Agency subsequently collected concerning HAP emissions from
electric utility steam generating units.

      Based on the available information, the Administrator concluded that Hg is both a public
health concern and a concern in the environment. The EPA's analysis shows that coal-fired
electric utility steam generating units are the largest source of Hg emissions to the atmosphere in
the United States. Further, the Administrator concluded that there is a plausible link between
MeHg concentrations  in fish and Hg emissions from these coal-fired steam generating units.
Therefore, the Administrator found that it is appropriate and necessary to regulate HAP
emissions, including Hg, from coal-fired electric utility steam generating units under CAA
Section 112 (i.e., establish a NESHAP), because the implementation of other requirements under
the CAA will not adequately address the serious public health and environmental hazards arising
from these emissions.  As a result, the EPA added coal-fired electric utility steam generating
units to the list of source categories under CAA Section 112(c).

      In its  1998 report to Congress, the EPA found that nickel emissions are the only HAP of
potential concern from oil-fired electric utility steam generating units.  The Administrator found
that there remained uncertainties regarding the extent of the public health impact from nickel
emissions from oil-fired electric utility steam generating units. Therefore, the EPA also added
oil-fired  electric utility steam generating units to the CAA Section 112(c) source category list.

      The Administrator found that regulation of HAP emissions from natural-gas-fired electric
utility steam generating units is not appropriate or necessary. Because the EPA believes that the
CAA Section 112(a)(8) definition of electric utility steam generating units excludes stationary
combustion turbines, the Administrator's finding for natural-gas-fired electric utility steam
generating units  does not apply to stationary combustion turbines.

      In response to the regulatory finding, the EPA has begun development of a NESHAP to
specifically control HAP emissions from coal-fired electric utility steam generating units. The
current schedule for this rule is to propose a NESHAP for the source category by December 15,
2003, and take final action on the rule by December 15, 2004.

-------
1.7 Mercury Emissions Research Programs

      Mercury emissions from combustion sources including coal-fired electric utility power
plants have been the subject of extensive research and study throughout the 1990s by government
agencies, the electric utility industry, and university researchers. Researchers at the U.S.
Department of Energy's (DOE) National Energy Technology Laboratory (NETL) (previously
known as the Federal Energy Technology Center) have prepared a comprehensive literature
search and review summarizing the data and findings of many of these studies published in
1999.9

      Currently, the EPA, the DOE/NETL, and the Electric Power Research Institute (EPRI) are
funding major on-going research work on Hg emissions from coal combustion.  Each
organization conducts these projects "in-house" as well as through contracts with university
researchers and private companies. In addition, the EPA, the DOE/NETL, and EPRI are
collaborating on a number of joint projects. The on-going projects range from fundamental
studies based on bench-scale laboratory experiments and computer modeling to field test
programs at coal-fired electric utility power plants.  Table  1-1 presents a summary overview of
the research topics being investigated. Major objectives of these research efforts include:

        $ Improving the test methods for measuring Hg emissions from coal-fired electric
          utility boilers and other coal combustion systems.  The current focus of this effort is
          development of continuous emission monitors (CEMs) to measure Hg.

        $ Understanding the chemical,  physical, and operating factors that affect Hg behavior in
          combustion gases and residues from burning coal.

        $ Developing cost-effective techniques for controlling Hg that can be readily retrofitted
          to existing coal-fired electric utility power plants.

        $ Developing Hg control technologies for application to new coal-fired electric utility
          power plants.

        $ Developing multipollutant control technologies that will control Hg emissions
          together with SO2 or NOX emissions.
 1.8  Relationship to Mercury Emission Control Research for Municipal Waste Combustors

      The EPA has identified MWCs as the second largest source category of Hg emissions in
the United States after coal-fired electric utility steam generating units.4  The control of Hg
emissions from MWCs has been, and continues to be, the subject of research in both the United
States and Europe.
                                         1-9

-------
Table 1-1. Current research areas related to controlling Hg emissions from coal-fired electric utility
power plants.
Research Area
Hg content of coals
Coal cleaning methods for Hg removal
Hg behavior and speciation in coal combustion gases
Hg measurement and monitoring in coal combustion gases
Hg adsorption on fly ash in coal combustion gases
Improving Hg capture by conventional wet scrubber systems
Improving Hg capture by conventional particulate control devices
Hg capture using activated carbon sorbent injection
Hg capture using other Hg-specific sorbents
Hg capture using multipollutant sorbents
Hg behavior in solid residues from Hg control systems
Hg control costs and economics
Major Research Sponsor
DOE

y
^
y
^
y
^
y
^

^
y
EPA
y

^
y

y

y

y
^
y
EPRI

y
^
y
^
y

y


^
y
USGS
y
y











-------
      An MWC is an enclosed combustion unit used to burn municipal solid waste for the
purpose of reducing the volume of waste that must be disposed in a landfill. Many people also
refer to these combustion units as waste incinerators. Although an MWC may function as a
simple incinerator, more commonly these combustion units are equipped with heat recovery
equipment that is used for producing steam. The steam is used in a variety of different ways
depending on the facility location including generating electrical power, industrial process steam,
or district heating systems.  Other terms sometimes used to refer to this type of MWC facility
include Aresource recovery facility and Awaste-to-energy plant.

      The EPA and some states have established regulations to reduce the level of Hg emissions
from MWC facilities operating in the United States. To comply with these regulations, a
combination of control strategies, including the application of add-on control devices, are now in
use for new and existing MWC facilities. Direct transfer to coal-fired electric utility steam
generating units of all of the specific control strategies that are used to meet the Hg emission
regulations for MWC facilities is not feasible, effective, or practical because of the distinct
differences between the two categories of combustion sources (e.g., properties of the fuel burned;
the design, operation, and scale of the combustion unit; and the characteristics of the post-
combustion gases).  Nevertheless,  understanding how Hg emissions are controlled in an MWC
does provide useful information to help identify potential Hg control technologies for coal-fired
electric utility steam generating units and to assess the performance and costs of using these
controls.

      In the United States, the municipal solid waste that can be burned in MWCs is primarily
composed of household, commercial, and institutional refuse. These wastes cannot include any
hazardous wastes regulated under  subtitle C of the Resource Conservation and Recovery Act
(RCRA). However, small amounts of Hg may be in certain discarded consumer products that are
not RCRA hazardous wastes and are burned in MWCs (e.g., batteries, some fluorescent bulbs,
electrical switches, thermometers). Most of this Hg is released during the combustion process
and remains in combustion gases vented from the MWCs.

      Mercury emissions from MWC facilities in the United States are decreasing for three
major reasons. First, Section 129 of the CAA requires the EPA to develop national emission
standards for Hg (and a number of other pollutants) being emitted from MWC facilities. The
EPA finalized the standards as new source  performance standards (NSPS) and Emission
Guidelines (EG) under 40 CFR part 60 in October 1995.  The NSPS (subpart Eb) applies to those
MWCs constructed after September 20, 1994 (i.e., Anew sources); the EG (subpart Cb) applies
to those MWCs built before this date (i.e., Aexisting sources).  For Hg, the same emission limit
of 0.08 milligram per dry standard cubic meter (mg/dscm) applies to both new and existing
MWC facilities.

      In addition to the Federal standards and emission guidelines, individual states with
significant numbers of MWC facilities operating within their jurisdiction have enacted legislation
controlling Hg emissions from these MWC facilities. Several states (e.g., Florida and New
Jersey) have established Hg emission limits for MWCs, effectively requiring these units to use a
                                         1-11

-------
specific control technology.  Some states have enacted regulations limiting or banning the sale of
certain Hg-containing products that, when discarded, would have been mixed in refuse burned in
an MWC. These regulations differ from state to state, with Minnesota having the most extensive
set of restrictions on the disposal of Hg-containing products.

      The third reason for the decline in Hg emissions from MWC facilities is the trend by
manufacturers to limit or discontinue the use of Hg in many products that ultimately are mixed in
the waste burned in MWCs.  These products include household alkaline batteries and interior and
exterior paints. Other products that traditionally have used Hg (e.g., Hg thermometers and
thermostats) are increasingly being replaced by digital, electronic versions that do not require Hg
components.

       Despite the reductions in the Hg content of the waste burned, MWCs still need to use
add-on emission controls to capture Hg in the combustion gases exhausted from the combustor.
Mercury removal from the combustion gases using these control systems can vary depending on
the combination of controls used and the site-specific conditions.  The injection of powdered
activated carbon into the gas upstream of a particulate matter control device is a common method
currently used in the United States to control Hg emissions from MWCs.  In Europe, wet
scrubbing systems are commonly used to control MWC  Hg emissions.  Because of factors such
as the differences in flue gas characteristics and duct configurations (discussed further in
Chapter 7), the Hg control technologies now used for MWCs cannot be directly transferred to
coal-fired utility boilers. However, the commercial experience with MWC Hg emission controls
does point to potential control technologies that should be investigated further for application to
coal-fired electric utility power plants.
1.9 Report Organization

      The remainder of this report consists of nine chapters (Chapters 2 through 10) presenting
background information, recent research findings,  and the current status of research studies
related to Hg emission behavior and control in coal-fired electric utility power plants. Each
chapter addresses specific topics related to the application of Hg emission control technologies to
coal-fired steam generating units.  Appendices are presented at the end of the report to support
and supplement information presented in the chapters.
                                        Chapter 2
                             Coal-fired Electric Utility Boilers
       Chapter 2 presents an overview of the coals burned and combustion technologies
       used for electric power generation.  The design and operating characteristics of the
       different types of coal-fired boilers used by electric utilities in the United States
       are presented.  The properties of the coal burned by electric utilities in the year
       1999 are summarized using information compiled from the EPA ICR database.
                                         1-12

-------
                                Chapter 3
                Criteria Air Pollutant Emission Controls for
                     Coal-fired Electric Utility Boilers
Chapter 3 presents a summary review of the different air pollution control devices
(APCDs) currently used at coal-fired electric utility power plants to meet criteria
air pollutant emissions standards. The nationwide distribution of APCD
configurations used at these power plants to comply with the air standards is
presented using information from the EPA ICR database.
                                Chapter 4
                         Measurement of Mercury
Chapter 4 discusses the principles, applications, and limitations of Hg
measurement methodologies, particularly with respect to understanding and
interpreting the ICR data. The chapter discusses the Ontario-Hydro method and
other manual test methods available for measuring Hg in coal combustion flue
gas.  This chapter introduces the principles and issues related to Hg continuous
emission monitors (CEMs) and their use as a valuable research tool.
                                Chapter 5
                     Mercury Speciation and Capture
Chapter 5 provides an introduction to Hg chemistry and behavior of Hg as it
leaves the combustion zone of the furnace and passes in the flue gas through the
downstream boiler sections, air heater, and air pollution control devices. Recent
laboratory research on Hg chemistry in coal combustion flue gas is summarized.
Mercury speciation is discussed as related to coal properties, combustion
conditions,  flue gas composition, fly ash properties, time/temperature profile
between the boiler and air pollution control devices, and post-combustion flue gas
cleaning methods.  Results from recent studies on the mechanisms for capturing
Hg by adsorption of gaseous Hg, by solid particles in the flue gas, and by
absorption capture of Hg by alkaline solutes/slurries are analyzed.
                                Chapter 6
           Mercury Capture by Existing Control Systems Used by
                     Coal-fired Electric Utility boilers
Chapter 6 discusses the level of Hg capture achieved by the air emission control
devices now in use at coal-fired electric utility power plants to meet Federal and
state air emission standards for particulate matter, sulfur oxides, and nitrogen
oxides.  The results of the Hg emission source testing compiled in the Part HI
EPA ICR data are presented and analyzed.  The methods used to evaluate these
                                  1-13

-------
Hg emissions data are described to meet two interrelated objectives.  First, an
analysis of the EPA ICR data is presented as used for EPA's estimate of
nationwide Hg emissions from coal-fired electric utility power plants in 1999.
Second, the EPA ICR data are analyzed to characterize the effects of coal
properties, combustion conditions, and flue gas cleaning methods on the
speciation and capture of Hg.
                                 Chapter 7
                    Research and Development Status of
              Potential Retrofit Mercury Control Technologies
Chapter 7 discusses potential retrofit control technologies for increasing Hg
emission capture levels in the air pollutant control systems now in use at existing
coal-fired electric utility power plants.  The use of activated carbon and other dry
sorbents for Hg emission control is discussed.   Current knowledge is summarized
regarding the enhancement of Hg capture by existing particulate matter control
devices and wet scrubbing systems. Recent pilot-scale and full-scale test data for
Hg capture by potential retrofit control technologies are presented. This chapter
also summarizes the status of emerging Hg and multipollutant control
technologies that are being developed for the control of Hg emissions from coal
combustion.
                                 Chapter 8
              Cost Evaluation of Retrofit Mercury Controls for
                     Coal-fired Electric Utility Boilers
Chapter 8 presents a preliminary evaluation of total annual costs to apply potential
activated carbon injection-based control technologies to existing coal-fired
electric utility power plants.  The evaluation is based on estimating the control
costs using a computer model for a series of model plant scenarios.  The cost
estimate methodology and assumptions are described.  The cost estimates are
presented and discussed.
                                 Chapter 9
              Coal Combustion Residues and Mercury Control
The EPA/NRMRL presently is conducting a life-cycle analysis project to help
evaluate any potential environmental trade-offs and to ensure that there is not an
increased environmental risk from the management of coal combustion residues
(CCRs) resulting from the implementation of Hg control technologies at coal-fired
electric utility power plants. In support of this evaluation, the NRMRL is
gathering data and information to assess future increases in Hg concentrations in
                                  1-14

-------
       CCRs resulting from application of Hg emissions control requirements to coal-
       fired electric utility boilers.  Chapter 9 summarizes some of the CCR information
       gathered by NRMRL to date and identifies the major data gaps and priorities of
       EPA's research to ensure that Hg controlled at the coal-fired electric utility power
       plant stack is not later released from CCRs in an amount that is problematic for
       the environment.
                                      Chapter 10
                          Conclusions and Recommendations
       Chapter 10 summarizes the major findings of this report and presents
       recommendations for further work, which would benefit the understanding of Hg
       behavior in the coal combustion processes at electric utility power plants.
1.10 References
1.  National Research Council.  lexicological Effects of Methylmercury. Committee on the
   Toxicological Effects of Methylmercury Board on Environmental Studies and Toxicology,
   Commission on Life Sciences. National Academy Press, Washington, DC, 2000. Available
   at: < http://www.nap.edu/books/0309071402/html/ >.

2.  Mishima, Akio. Bitter Sea: The Human Cost of Minamata Disease.  Kosei Publishing Co.,
   Tokyo, Japan, 1992.

3.  French, C.L., W.H. Maxwell, W.D. Peters, G.E. Rice, O.R. Bullock, A.B. Vasu, R. Hetes,
   A. Colli, C. Nelson, and B.F. Lyons.  Study of Hazardous Air Pollutant Emissions from
   Electric Utility Steam Generating Units  Final Report to Congress,  Volumes 1-2. EPA-
   453/R-98-004a and b.  Office of Air Quality Planning and Standards, Research Triangle Park,
   NC. February 1998. Available at:
   < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

4.  Keating, M.H., K.R. Mahaffey, R. Schoeny, G.E. Rice, O.R. Bullock, R.B. Ambrose, Jr.,
   J. Swartout, and J.W. Nichols. Mercury Study Report to Congress, Volumes 1-V111.  EPA-
   452/R-97-003 through 010.  Office of Air Quality Planning and Standards and Office of
   Research and Development, Research Triangle Park, NC. December 1997. Available at:
   < http://www.epa.gov/airprogm/oar/mercury.html >.

5.  U.S. Environmental Protection Agency. Deposition of Air Pollutants to the Great Waters:
   First Report to Congress. EPA-453/R-93-055. Office of Air Quality Planning and
   Standards, Research Triangle Park, NC. May  1994.
                                        1-15

-------
6.  Ackermann, J., M. McCullough, E. Ginsburg, D. Byrne, and L. Driver. Deposition of Air
   Pollutants to the Great Waters: Second Report to Congress. EPA-453/R-97-011. Office of
   Air Quality Planning and Standards, Research Triangle Park, NC. June 30, 1997. Available
   at: < http://www.epa.gov/oar/oaqps/gr8water/2ndrpt >.

7.  Lacy, G. and D. Evarts. Deposition of Air Pollutants to the Great Waters: Third Report to
   Congress. EPA-453/R-00-005. Office of Air Quality Planning and Standards, Research
   Triangle Park, NC. June 2000. Available at:
   < http://www.epa.gov/oar/oaqps/gr8water/3rdrpt >.

8.  U.S. Environmental Protection Agency. Database of information collected in the Electric
   Utility Steam Generating Unit Mercury Emissions Information Collection Effort. OMB
   Control No. 2060-0396. Office of Air  Quality Planning and Standards, Research Triangle
   Park, NC. April 2001.  Available at:
   < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

9.  Brown, T. D., D.N. Smith, R.A. Hargis, Jr., and W.J. O'Dowd. "1999 Critical Review:
   Mercury Measurement and Its Control:  What We Know, Have Learned, and Need to Further
   Investigate," Journal of the Air & Waste Management Association, June 1999. pp. 1-97.
   Available at: < http://www.lanl.gov/projects/cctc/resources/pdfsmisc/haps/CRIT991 .pdf >.
                                        1-16

-------
                                       Chapter 2
                          Coal-fired Electric Utility Boilers
2.1 Introduction
       The steam produced in a boiler can be used to drive a steam turbine that, in turn, spins an
electric generator.  In a conventional steam boiler used for electrical power generation, water is
heated under pressure to form high-temperature, high-pressure steam. The heat required to
produce steam can be supplied by burning a fossil fuel inside an enclosed space in the boiler.
Electricity generation in the Unities States relies extensively on burning coal in steam boilers.

      This chapter presents an overview of the use of coal by electric utilities for power
generation. An introduction to the properties of coal and coal resources in the United States is
presented. The major components and general operation of a conventional coal-fired electric
utility boiler are described.  A profile of the different coal-firing configurations used by electric
utility power plants in the United States is presented based on analysis of the Part II EPA ICR
data. Ash produced by coal combustion is described.  The chapter concludes with a summary of
the Part II EPA ICR data for the mercury content of the coals burned by electric utility power
plants in  1999.  Air pollutant emissions and the control strategies currently used for these
coal-fired electric utility power plants are discussed in Chapter 3.
2.2 Coal

       Coal is a combustible "rock" composed of organic and mineral materials that have
formed over time by vegetative decay and mineral deposition. The principal chemical
constituents of coal are carbon, hydrogen, oxygen, nitrogen, and sulfur. Coal also contains
incombustible mineral matter and trace amounts of metallic elements, oxides, and rare gases. The
properties of a given coal deposit vary depending on a variety of site-specific factors including
the type of vegetative matter from which the coal formed, the age of the deposit, and the
conditions under which the coal formed.

2.2.1  Coal Property Tests

       Standardized tests for determining the properties of coal have been adopted by the
American Society for Testing and Materials (ASTM).1  These ASTM methods are widely used in
                                             2-1

-------
the United States by coal producers, electric utility companies, and government agencies to
obtain coal property data for many purposes including classifying coal resources, designing coal
combustion equipment, pricing coal, and monitoring coal shipment quality.  Standardized
procedures for collecting coal samples for analysis also have been established by ASTM
methods.

2.2.1.1 Coal Heating Value

       One of the key properties of coal is the quantity of heat that can be released when the coal
is burned. The heating value of coal is determined using one of several ASTM test methods
(e.g., ASTM D2015 or D3286). These tests involve burning a coal sample in a bomb calorimeter
and measuring the temperature rise following the procedure specified in the method.  As used in
the United States, heating value is most commonly expressed in units of British thermal units per
pound of coal (Btu/lb). Heating value can also be expressed in units of joules per kilogram,
kilojoules or kilocalories per kilogram, or calories per gram. Also, heating value may be reported
as higher heating value (HHV) or lower heating value (LHV).  The HHV is the value measured
by the actual test.  The LHV is calculated by subtracting the heat of water vaporization from the
value measured in the bomb calorimeter.

2.2.1.2 Coal Proximate Analysis

      The proximate analysis is a widely used test procedure for determining for a given coal the
total moisture, volatile matter, fixed carbon, and ash contents expressed on a weight-percent
basis. The protocol for performing a proximate analysis for coal is established by ASTM D3172
that specifies the overall procedure to be followed and the other specific ASTM test methods to
be used.  The analysis involves performing a series of tests in a specific order on a given coal
sample. First, the total moisture of the coal is determined by drying the sample in an oven
according to ASTM test method 3173. The difference in weight before and after drying is the
amount of moisture in the coal.

      Volatile matter is not naturally present in coal. However, combustible gases (e.g.,
hydrogen, methane, and other hydrocarbons) are formed by thermal decomposition when the coal
sample is heated under controlled temperature and time conditions. The conditions are specified
in ASTM test method 3175.  The difference in weight before and after heating the coal  sample
for a second time in a furnace is the amount of volatile matter contained in the coal. The coal
sample is then completely burned under conditions specified in ASTM test method 3174.  The
weight of the noncombustible matter remaining after combustion is the ash content in the coal.
The percentage of fixed carbon is obtained by subtracting from 100 percent the sum of the
percentages of total moisture, volatile matter, and ash.

2.2.1.3 Coal Ultimate Analysis

      The second analysis procedure commonly performed is the ultimate analysis. This
analysis determines the composition of the coal based on elemental constituents. The protocol

                                            2-2

-------
for performing a coal ultimate analysis is established by ASTM D3176 which specifies the
overall procedure to be followed and the specific ASTM test methods to be used.  As defined in
ASTM D3176, the elements determined are total carbon, total hydrogen, total sulfur, total
nitrogen, and total oxygen. Determination of ash is included in the analysis. The quantity of
chlorine present in the coal is also commonly included as part of the ultimate analysis. However,
the contents of mercury and other trace constituents in the coal are not included in the results
from a coal ultimate analysis.

2.2.1.4 Coal Mercury Analysis

       A separate analysis must be conducted to determine the Hg content of coal. Several
ASTM test methods are available for measuring the total Hg concentration in a coal sample.
Two methods are established by ASTM D6414 "Standard Test Method for Total Mercury in Coal
and Coal Combustion Residues by Acid Extraction or Wet Oxidation/Cold Vapor Atomic
Absorption."  The lower quantitative limits for these methods are, respectively, 0.02 ppm and
0.03  ppm. A third, commonly used method is ASTM D3684 "Standard Test Method for Total
Mercury in Coal by the Oxygen Bomb Combustion/Atomic Absorption Method" with a lower
quantitative limit of 0.06 ppm.  An interlaboratory study conducted by EPRI evaluated the use of
these three analytical methods to measure coal Hg content for submitting data to the EPA ICR.2
The study indicated that all three methods had certain limitations, especially when used to
analyze very low Hg content coals and coal ashes.  However, the study concluded that the
uncertainty in these methods should not have a significant impact on the use of the data collected
by the EPA ICR for nationwide Hg emission estimates.

2.2.2 Coal Classification

      Over the years, a number of coal classification systems have been developed by the United
States Geological Survey (USGS) and others. These coal classification systems allow
assessments of coal resources and provide  data for designing coal combustion equipment.3 In the
United States, coals are classified using a hierarchy ranking coals relative to other coals based on
the degree of metamorphism (effectively, the geological age of the coal and the conditions under
which the coal formed). These classification criteria have been standardized by ASTM method
D-388. Under the ASTM method, coals are divided into four major categories called "ranks."
Each rank is further subdivided into groups.  The basic ranking criteria are coal heating value,
volatile matter content, fixed carbon content, and agglomerating behavior. The coal ranks are
summarized below.

       Anthracite coal. The highest rank class of coal that is defined to be a nonagglomerating
       coal having more than 86 percent fixed carbon and less than 14 percent volatile matter on
       a dry, mineral-matter-free basis.  This coal rank is subdivided into three groups based on
       decreasing fixed carbon and increasing volatile matter content: meta-anthracite,
       anthracite, and semianthracite.
                                            2-3

-------
       Bituminous coal. The second highest rank of coal defined to be high in carbonaceous
       matter, having less than 86 percent fixed carbon, and a 14 percent volatile matter on a
       dry, mineral-matter-free basis, and a heating value of more than 10,500 Btu/lb on a moist,
       mineral-matter-free basis. This coal can be either agglomerating or nonagglomerating.
       The rank is subdivided into five bituminous coal groups on the basis of decreasing heat
       content and fixed carbon and increasing volatile matter:  low-volatile bituminous coal,
       medium-volatile bituminous coal, and high-volatile bituminous coals A, B, and C.

       Subbituminous coal.  The third-highest rank of coal defined to be nonagglomerating coals
       having a heating value of more than 8,300 Btu/lb but less than  11,500 Btu/lb on a moist,
       mineral-matter-free basis. This rank of coal is subdivided on the basis of decreasing heat
       value into three groups: subbituminous A coal (10,500 to 11,500 Btu/lb),
       subbituminous B coal (9,500 to 10,500 Btu/lb),  and subbituminous C coal (8,300 to 9,500
       Btu/lb). Note that the heating value range for the upper-end subbituminous A coals
       overlaps with the heating value range for the lower-end high-volatile bituminous C coals.
       Lignite. The lowest rank of coal defined to consist of brownish-black coal having heating
       values less than 8,300 Btu/lb on a moist, mineral-matter-free basis. This rank of coal is
       subdivided into two groups: lignite A (6,300 to 8,300 Btu/lb) and lignite B (less than
       6,300 Btu/lb).

2.2.3  United States Coal Resources
       Coal is the most abundant fossil fuel in the United States. The  DOE Energy Information
Administration (EIA), the Federal government agency responsible for estimating coal resources
in the United States, estimates that the demonstrated reserve base of coal in the United States is
approximately 508 billion tons.4 The distribution of this coal by major coal rank is presented in
Table 2-1. Over half of the coal reserve base is classified as bituminous coal.  Another third of
the reserves  are classified as subbituminous coal.

       Not all of the coal identified in the demonstrated reserve base can be extracted from the
ground for a variety of reasons. Of the estimated 508 billion tons of demonstrated coal reserves,
the DOE EIA estimates that approximately 275 billion tons of coal can be recovered by standard
mining technologies, assuming that a market and an adequate selling price exist for this coal.

      In the United States, coal deposits have been found in 36 states.  Figure 2-1 shows the
distribution of coal resources in the United States by coal region as designated by the USGS.
Coal resources in the Eastern United States are concentrated primarily  along the Appalachian
Mountains and are estimated by the DOE EIA to contain 108 billion tons. The major deposits of
bituminous coals are concentrated in the Central Appalachian region comprised of eastern
Kentucky, western Virginia, and southern West Virginia. Most of the anthracite coal resources
in the United States are located in eastern Pennsylvania (Pennsylvania  Anthracite  and Northern
Appalachian regions).
                                             2-4

-------
Table 2-1.  Demonstrated reserve base of major coal ranks in the United States
estimated by DOE/EIA (source: Reference 4).
Coal Rank
Anthracite
Bituminous
Subbituminous
Lignite
Total
Estimated
U.S. Demonstrated
Coal Reserves
(billion tons)
8
271
185
44
508
Percentage of
U.S Demonstrated
Coal Reserves
2%
53%
36%
9%
100%
                                     2-5

-------
to
            J 5. DE?'Rr; tuT OF THE ITBKCf<
                                                                                                            ?;pf|4 !l :4(*:-H *F".'V .aV-v* ;'=!
                     [ ^                             ^fe UnjQr, mm
                      jj        ; '.     fj Q;;f<^	~^n	TT-^	
                    / L~        /
                 r
                 L.

                               ,           i J' *"*' !>->:-:-'s f      Rhr
                               ""Mama Pork  f.m ts3w. P;':">: "L	          '	  \
                                                    '" i	        injifto*" '
                                                    ,   :
                                        ''TfTTp     A	~Sn^~	-

                            Swittniwbst Uttjl tt^'liji    1%    i   	
 p'San Juan PBVsfT'Jf	^
'(        -	^'2V?'
r'           I Si           i
                                                        Raton


                    ^   ^ .^a
                    V  /   \

'*1        4"
T	:tti
                                                                                   .
                                                                                   I
                                                                                   ;; \ y"

                                                                       >1i8rtfwm
                                                       as
                                                       S5SSSSS, ,-

                                                       "


                                                                                      is-^
                                                                    *
                                                                                    --      

                                                                                 +t3*$
                                                                                                      ^
               Figure 2-1.  Distribution of coal deposits in the continental United States by USGS coal resource
               region (source:  Reference 5)

-------
      The coal regions in the Central United States (Eastern Interior, Western Interior, Texas,
and Mississippi regions) are estimated by the DOE EIA to contain 160 billion tons of coal. Most
of the coal deposits in these regions are bituminous coal (largest deposits in the Eastern Interior
region). A band of lignite deposits occur along the Gulf Coast (Texas and Mississippi regions)
with the largest deposits in eastern Texas.

      The coal reserves in the Western United States coal regions are estimated by the DOE EIA
to be 240 billion tons. Subbituminous coal is the most prevalent coal type with the major
deposits located throughout Montana and Wyoming (Powder River, Bighorn Basin, Wind River,
and Green River - Hams Fork regions) and in northwestern New Mexico (San Juan River
region). Large deposits of lignite are found in eastern Montana and North Dakota (Fort Union
region). Bituminous coal is found mostly in the coal regions in Colorado and Utah (Uinta, Raton
Mesa, and Southwest Utah regions).

2.2.4 Mercury Content in Coals

      Mercury is a naturally occurring impurity contained in coal in trace amounts.  It can occur
in coal in several forms. Most of the Hg is believed to be present in  combination with sulfide
minerals, particularly pyrite.  The mercury-pyrite association accounts for as much as 65 to 70
percent of the Hg in some coals. Mercury is also associated with other ash-forming minerals and
with the organic fraction in coal.  On the order of 25 to 35 percent of the Hg in coal is typically
associated with the organic material.

       Data on the Hg content of "in-the-ground" coals are available in the USGS COALQUAL
database.6 One study evaluated the Hg content of coals using this database and selecting coal
types representing major coal producing regions in the United States.7 The data from the study
are summarized in Table 2-2. The average concentration  of Hg in the coal samples ranged from
0.08 to 0.22 |ig/g. These data show that the Hg content of coals is not constant but varies
depending on the coal deposit. The data also show that Hg  content is not a function of coal rank
(i.e., one coal type does not have inherently lower Hg concentrations than another coal type).

       A comparison of the Hg concentrations in the different coals cannot be directly related to
the amount of Hg emissions emitted  from boilers burning these coals. Other coal  properties and
how the coal is prepared prior to firing in a boiler affect the theoretical potential level of Hg
emissions that would occur in the absence of applying any Hg emissions controls. In other
words, one cannot conclude that burning a coal with higher as-mined Hg concentration will
necessarily result in higher Hg emissions from a coal-fired electric utility boiler.

      Coals with higher heating values require less coal to be burned in a boiler on a mass basis
to produce a given electricity output.  For two coals with the same Hg content but different
heating values, burning the coal with the higher heating value in a given boiler will result in less
Hg being emitted in boiler combustion gases per unit of electricity output. On an equal energy
basis, the Hg content of the bituminous and subbituminous coals listed in Table 2-2 span the

                                           2-7

-------
            Table 2-2. Mercury content of selected as-mined coal samples by coal rank and USGS coal resource
            region (source: table prepared by summarizing and interpreting coal data presented in Reference 7).
to
oo
Coal Rank
Anthracite
Bituminous
Subbituminous
Lignite
USGS Coal Resource Region
Pennsylvania Anthracite
Uinta
Raton Mesa
Eastern Interior
Western Interior
Appalachian
San Juan River
Hams River
Green River
Powder River
Southwest Utah
Wind River
Fort Union
Texas and Mississippi
Hg
Concentration
(ppm)
0.18
0.08
0.09
0.10
0.18
0.20
0.08
0.09
0.09
0.10
0.10
0.19
0.13
0.22
Heating Value
(Btu/lb)
12,440
10,800
12,320
11,400
10,970
12,790
9,610
10,570
9,580
8,090
9,290
9,560
6,360
6,490
Hg Content By
Heating Value
(lbHgper1012Btu)
15.4
7.3
6.6
8.2
16,1
15.4
7.7
4.8
6.6
12.6
11.0
18.7
21.8
36.4

-------
same general range of values. No trend is apparent from these data; both bituminous and
subbituminous coals are found at the lower and upper ends of the range. For example, a
bituminous coal from the Raton Mesa region and a subbituminous coal from the Green River
region each have an average Hg content of 6.6 Ib per 1012 Btu. At the other end of the range, a
bituminous coal from the Western Interior region has an average Hg content of 16.1 Ib per
1012 Btu and a subbituminous coal from the Wind River region has an average Hg content of
18.7 Ib per 1012 Btu.  On the other hand, the Hg contents reported for the two lignite coals listed
in Table 2-2 are significantly higher than any of the bituminous and subbituminous coals (an
average
lignite).
average of 21.8 Ib per 1012 Btu for Fort Union lignite and 36.4 Ib per 1012 Btu for Gulf Coast
       Another key reason why the Hg content of as-mined coals cannot be related to Hg
emissions is the as-mined coal frequently is not burned in an electric utility boiler as it comes
directly from the mine. The as-mined, or raw, coal often is first processed at a coal preparation
plant to remove non-coal impurities in order to provide the coal purchaser with a uniform coal
that meets a predetermined, contractual set of specifications. These processes commonly are
collectively referred to as "coal cleaning."  Depending on the properties of the coal and the type
of process used, coal cleaning can reduce the Hg content of the coal that is ultimately fired in the
electric utility boiler.
2.3 Coal Cleaning

2.3.1  Coal Cleaning Processes

       Raw coal from a mine contains separate rock, clay, and other minerals.  After the coal is
mined, it may first pass through a series of processes known as coal preparation or coal cleaning
before it is shipped to an electric utility power plant. The coal is processed for three main
reasons: 1) to reduce the ash content; 2) to increase the heating value; and 3) to reduce the sulfur
content to ultimately lower emissions of sulfur dioxide when the coal is burned in the utility
boiler. The removal of impurities from the coal also helps to reduce power plant maintenance
costs and to extend the service life of the boiler system.

       Coal cleaning processes currently in use separate the organic fraction of the as-mined coal
from the mineral materials according to the differences in either the density-based or surface-
based characteristics of the different materials. Physical coal cleaning typically involves a series
of process steps including: 1) size reduction and screening, 2) gravity separation of coal from
sulfur-bearing mineral impurities, and 3) dewatering and drying.

       Bituminous coals from mines in the Eastern and Midwestern United States frequently are
cleaned to meet the electric utility customer's specifications  for heating value, ash content, and
sulfur. It is estimated that about three-fourths (77 percent) of these coals are cleaned prior to
shipment to an electric utility power plant.8  The subbituminous and lignite coals from mines in
the Western United  States routinely are not cleaned before shipment to an electric utility power
                                           2-9

-------
plant, but in special cases these types of coals can be cleaned. For example, some of
subbituminous coal from mines in the Powder River coal region (a major source of coal for many
electric utilities) is cleaned for shipment to electric utility customers.

2.3.2  Mercury Removal by Coal Cleaning

      Conventional coal cleaning methods will also remove a portion of the Hg associated with
the incombustible mineral materials but not the Hg associated with the organic carbon structure
of the coal. Any reduction in Hg content of the coal shipped to an electric utility power plant
obtained from the Hg removed by coal cleaning processes transfers the removed Hg to the coal
cleaning wastes. Limited data have been gathered on the level of Hg removed by conventional
coal cleaning methods.

       A review of test data for 26 bituminous coal samples from coal seams in four states
(Illinois, Pennsylvania, Kentucky, and Alabama) prepared for EPA's Mercury Study Report to
Congress indicates a wide range in the amount of Hg removed by coal cleaning.8 In some cases,
analysis of coal samples from the same coal seam also showed considerable variability. Analysis
of five of the coal samples  showed no Hg removal associated with conventional coal cleaning
while the remaining 21 coal samples had Hg reductions ranging from approximately 3 to 64
percent. The average Hg reduction for all of the data was approximately 21 percent.

       Other studies have reported higher average Hg reductions for Eastern and Midwestern
bituminous coals. One study tested 24 samples of cleaned coal.7  These data also showed a wide
range in Hg reduction rates. The average decrease in Hg reduction on an energy basis was 37
percent, with values ranging from 12 to 78 percent.  On a mass basis, the average Hg reduction
from coal cleaning was 30  percent. A higher Hg reduction was reported on an energy basis than
on a mass basis because the coal cleaning raises the heating value per unit mass of the coal, as
well as removing Hg. A second study of the effects of coal cleaning on Hg content for three
Ohio coals reported reductions in Hg content of the coals ranging from 36 to 47 percent.9

      The variation in Hg reductions observed from the test data might be a function of the type
of process used to clean a given coal and the proportion of Hg in  the coal that is present in
combination with pyrite (iron disulfide). Coal-cleaning processes that make separations
according to the density differential of particles are generally more effective in removing Hg
associated with pyrite than are surface-based processes.  The heavier pyrite is easily removed by
density-based processes, but not by surface-based processes where the similar surface
characteristics of pyrite and the organic matter make separation of the two components difficult.
For coals that have larger portions of Hg associated with pyrite, density-based cleaning processes
are expected to have higher Hg removals. However, some coals may contain large portions of
Hg associated with the organic fraction of the coal; Hg removal in these cases would be expected
to be substantially lower since the organic fraction of coal is not removed during cleaning.
Additional reductions in Hg can probably be achieved by using more intensive coal cleaning
methods. Several advanced coal cleaning techniques being investigated to improve Hg removal
are discussed in Chapter 7.
                                          2-10

-------
2.4 Coal-fired Electric Utility Boilers

       The large steam boilers used by electric utilities are also referred to as "steam generators,"
"steam generating units," or simply "boilers."  As discussed in Chapter 1, CAA Section 112(a)
defines the term Aelectric utility steam generating unit@ to include those units that cogenerate
steam and electricity and supply more than one-third of its potential electric output capacity and
more than 25 megawatts electrical output to any utility power distribution system for sale. For
simplicity in the remainder of this report, the term "electric utility boiler" is used to mean
"electric utility steam generating unit" as defined in CAA Section 112(a)(8).

       A total of 1,143 coal-fired units meeting the CAA definition of an "electric utility steam-
generating unit" were reported in the Part II EPA ICR data to be in the United States in 1999.w
More than one boiler unit is often operated at an electric utility power plant.  The 1,143 units
were located at a total of 461 facilities. These facilities can be categorized in three facility types:
conventional  coal-fired electric utility power plants, coal-fired cogeneration facilities, and
integrated coal gasification and combined cycle (IGCC) power plants.

2.4.1  Conventional Coal-fired Electric Utility Power Plants n'n

       A conventional electric utility power plant burns coal in a boiler unit solely for the
purpose of generating steam for electrical power production.  A total of 1,122 coal-fired electric
utility boilers were reported in the Part II EPA ICR data to be operating at conventional electric
utility power plants.  Each of these boilers was designed to meet plant load and performance
specifications by burning coals within a specific range of coal properties (e.g., heating value, ash
content and characteristics, and sulfur content). While the specific  equipment and design of a
coal-fired electric utility boiler will vary from plant to plant, the same basic process is used to
generate electricity.  Figure 2-2 presents a simplified schematic of the major components of a
coal-fired electric utility boiler operated at a conventional electric utility power plant.

       Coal typically is delivered to a power plant by  railcars, trucks, or barges. At some power
plants located near the mine supplying the coal, coal is delivered by a slurry pipeline or an
extended conveyor system.  Also, a few power plants burn imported coal that is delivered to the
facility by ship.  The delivered coal is unloaded and stored in outdoor storage piles or covered
storage structures such as silos or bins. Depending on how the coal is burned in the boiler (e.g.,
in a bed or burned in suspension), the coal is crushed or pulverized  before being fed to the boiler.

      A conventional coal-fired electric utility boiler consists of multiple sections, each of which
serves a specific purpose.  The coal is ignited and burned in the section of the boiler called the
"furnace chamber."  Blowing ambient air into the furnace chamber  provides the oxygen required
for combustion. The carbon and hydrogen comprising the coal are  oxidized at the high
temperatures produced by combustion to form the primary combustion products of carbon
dioxide (CO2) and water (H2O).  Sulfur in the coal is oxidized to form SO2.  Molecular nitrogen
in the combustion air and nitrogen bound in the coal react with oxygen in certain sections of the
combustion zone in the furnace chamber to form NOX.  Small amounts of other gaseous
                                           2-11

-------
                                                           high-pressure steam
to
i
to
                                                                                          2=^
                                                                                      Forced Draft Fan
                                                                                                            Prepared for EPA/NRMRL
                                                                                                            by P. Peterson
                                                                                                            12/5/2001
               Figure 2-2. Simplified schematic of coal-fired electric utility boiler burning pulverized, low-sulfur coal.

-------
combustion products form from other impurities in the coal. These hot combustion products are
vented from the furnace in a gas stream called collectively Aflue gas.@ Additionally, most but not
all the carbon in the coal is burned in the furnace.  Unburned or partially burned solid carbon
particles are entrained and vented from the furnace in the flue gas.

       The walls of the furnace chamber are lined with vertical tubes containing water. Heat
transfer from the hot combustion gases in the furnace boils the water in the tubes to produce
high-temperature, high-pressure steam. This steam flows from the boiler to a steam turbine. In
the turbine, the thermal energy in the steam is  converted to mechanical energy to drive a shaft
that spins a generator, which produces electricity.  After the steam exits the turbine, it is
condensed  and the water is pumped back to the boiler.

       To improve overall energy conversion  efficiency, modern coal-fired electric utility boilers
contain a series of heat recovery sections.  These heat recovery sections are located downstream
of the furnace  chamber and are used to extract additional heat from the flue gas. The first heat
recovery section contains a "superheater," which is used to increase the steam temperature. The
second heat recovery section contains a "reheater," which reheats the steam exhausted from the
first stage of the turbine.  This steam is then returned for another pass thorough a second stage of
the turbine. The reheater is followed by an "economizer," which preheats feed water to the boiler
tubes in the furnace.  The final heat recovery section is the "air heater," which preheats ambient
air used for combustion of the coal.

       A portion of all coals is composed of mineral matter that is noncombustible.  This matter
forms the ash that continuously must be removed from the operating utility boiler.  The ash
collection points and removal systems used for a given boiler unit are dependent on the ash
properties and content in the coal-fired, the boiler design, and the air pollution control devices
used.  The removal and handling of the coal ash is discussed further in Section 2.6.

       The flue gas exhausted from the boiler passes through air pollution control equipment and
is vented to the atmosphere through  a tall stack. The types and configurations of air pollution
controls currently used for coal-fired electric utility boilers are discussed in Chapter 3.

2.4.2  Coal-fired Cogeneration Facilities

       Approximately six percent of the boiler units are at cogeneration facilities,  which are
owned and operated by independent power producers or industrial companies. Of the 1,143  total
coal-fired electric utility boilers reported in the EPA Part IIICR data, 68 are classified as
cogeneration units.  The total generating capacity of these cogeneration units is 867 MWe. There
are more coal-fired boilers in the United States operating as cogeneration units; however, these
units do not meet the criteria specified in the CAA definition of a steam-generating unit (i.e., the
cogeneration unit is rated below 25 MWe or less than one-third of the unit's electrical output is
sold). These units were not surveyed for the EPA ICR database.
                                           2-13

-------
       Operation of a cogeneration facility differs from the operating configuration of the
conventional electric utility power plant shown in Figure 2-2.  Two basic cogeneration unit
configurations are used: the Atopping@ mode or the Abottoming@ mode. In the topping
cogeneration configuration, steam produced by the coal-fired electric utility boiler is used first to
generate electricity and then all or part of the exhaust heat is subsequently used for an industrial
process. The bottoming cogeneration configuration reverses this sequence using waste heat
generated  by an industrial process to produce steam in a heat recovery boiler for driving a steam
turbine and generating electricity. All of the cogeneration boiler units listed in the EPA ICR data
operate using the topping mode configuration.

2.4.3 Integrated Coal Gasification Combined Cycle Power Plants

       The IGCC power plants represent a new technology and are different from conventional
electric utility power plants in two major characteristics.  First, the IGCC power plants do not
burn the coal in its solid form. Instead, the coal is first converted to a combustible gas using a
coal gasification process at the facility site. Second, the IGCC power plants generate electricity
using two separate thermal cycles and associated turbines referred to as a "combined cycle"
operation. The coal-derived gas from the gasification process is first burned in a gas turbine that
drives an electrical generator.  The exhaust gases from this gas turbine pass through a  heat
recovery boiler to generate steam to power a steam turbine that drives a second electrical
generator. Three IGCC power plants have been built in the United  States.  The operation of these
power plants is discussed further in Section 2.5.5.
2.5 Coal-firing Configurations for Electric Utility Boilers

       Coal can be burned in a boiler using one of three basic techniques: burning coal particles
in suspension, burning large coal chunks in a fuel bed, or in a two-step process in which the coal
is first converted to a synthetic gas which is then fired in the boiler. Five basic firing
configurations are used to burn coal for electric power generation: pulverized-coal-fired furnace,
cyclone furnace, fluidized-bed combustor, stoker-fired furnace, and gasified-coal-fired
combustor. A general comparison of the different coal-firing configurations used for electric
utility power plants is presented in Table 2-3.

      Table 2-4 shows the distribution of the  1,143 coal-fired electric utility boilers listed in the
EPA ICR data by coal-firing configuration. Pulverized-coal-fired designs account for the vast
majority of the coal-fired electric utility boilers both in terms of total number of units
(approximately 86 percent) and nationwide generating capacity. Cyclone furnaces are used to
burn coal in approximately eight percent of the units. Fluidized-bed combustors are used for
about four percent of the coal-fired electric utility boilers.  Stoker-fired furnaces account for
about three percent of the total number of coal-fired electric utility boilers but provide less than
one percent of the total coal-fired megawatts.  Only three IGCC units  have been built in the
United States.
                                           2-14

-------
            Table 2-3.  Characteristics of coal-firing configurations used for electric utility power plants.
                 Coal-firing
               Configuration
       Coal Combustion Process Description
          Distinctive Design/Operating Characteristics
               Pulverized-coal-
                fired furnace
Coal is ground to a fine powder that is pneumatically
fed to a burner where it is mixed with combustion air
and then  blown into the furnace.  The pulverized-coal
particles burn in suspension in the furnace.  Unburned
and partially burned coal particles are carried off with
the flue gas.
                                                                                            Wall-fired
                                                                                         Tangential-fired
                                                                                          (Corner-fired)
                        An array of burners fire into the furnace
                        horizontally, and can be positioned on
                        one wall or opposing walls depending on
                        the furnace design.
                        Multiple burners are positioned in
                        opposite corners of the furnace producing
                        a fireball that moves in a cyclonic motion
                        and expands to fill the furnace.
                  Cyclone
                  furnace
to
Coal is crushed into small pieces and fed through a
burner into the cyclone furnace. A portion of the
combustion air enters the burner tangentially creating a
whirling motion to the incoming coal.
Designed to burn low-ash fusion coals and retain most of the ash in
the form of a molten slag.
                Fluidized-bed
                 combustor
Coal is crushed to fine particles.  The coal particles are
suspended in a fluidized bed by upward-blowing jets of
air. The result is a turbulent mixing of combustion air
with the coal particles. Typically, the coal is mixed with
an inert material (e.g., sand, silica, alumina) and a
sorbent such as limestone (for SO2 emission control).
The unit can be designed for combustion within the bed
to occur at atmospheric or elevated pressures.
Operating temperatures for FBC are in the range of 850
to 900 C.
                                                                                      Bubbling fluidized bed
                                                                                              (BFB)
                                                                                           Circulating
                                                                                          fluidized bed
                                                                                             (CFB)
                        Operates at relatively low gas stream
                        velocities and with coarse-bed size
                        particles. Air in excess of that required to
                        fluidize the  bed passes through the bed in
                        the form of  bubbles.
                        Operates at higher gas stream velocities
                        and with finer-bed size particles. No
                        defined bed surface.  Must use high-
                        volume, hot cyclone separators to
                        recirculate entrained solid particles in flue
                        gas to maintain the bed and achieve high
                        combustion efficiency.

                                                    (continued)

-------
             Table 2-3.  (continued).
                  Coal-firing
                Configuration
       Coal Combustion Process Description
          Distinctive Design/Operating Characteristics
                  Stoker-fired
                    furnace
to
                                                                                          Spreader-stoker
Coal is crushed into large lumps and burned in a fuel
bed on a moving, vibrating, or stationary grate.  Coal is
pushed, dropped, or thrown onto the grate by a
mechanical device called a "stoker."
                                                                                            Underfeed
                                                                                          Traveling grate
                                                                            A flipping mechanism throws the coal into
                                                                            the furnace above the grate. The fine
                                                                            coal particles burn in suspension while
                                                                            heavier coal lumps fall to the grate and
                                                                            burn in a fuel bed.
                        Coal fed by pushing coal lumps along in a
                        feed through underneath the grate.
                                                                            Coal is fed by gravity onto a moving grate
                                                                            and leveled by a stationary bar at the
                                                                            furnace entrance.
                 Gasified-coal-
                fired combustor
Synthetic combustible gas derived from an on-site coal
gasification process is burned in a gas turbine
combustor. The hot combustion gases turn the gas
turbine blades mounted on a shaft that drives an
electric generator. The  hot exhaust gases from the gas
turbine pass through a waste heat boiler to produce
steam for driving a steam turbine/generator unit.
Gasified-coal-fired combustors are unique from the other coal-firing
configurations because a gaseous fuel is burned instead of solid
coal.

-------
Table 2-4. Nationwide distribution of electric utility units by coal-firing
configuration for the year 1999 as reported in the Part II EPA ICR data (source:
Reference 10).
Coal-firing
Configuration
Pulverized-coal-fired furnace
Cyclone furnace
Fluidized-bed combustor
Stoker-fired furnace
Gasified-coal-fired combustor
Nationwide Total
I Nationwide
Total Number of
| Units
979
87
42
32
3
1,143
Percent of
Nationwide Total
85.6 %
7.6 %
3.7 %
2.8 %
0.3 %
100%
Percent of
Nationwide
Electricity
Generating
Capacity
90.1 %
7.6 %
1.3%
1.0%
<0.1 %
100%
                                     2-17

-------
2.5.1  Pulverized-coal-fired Furnace

        To burn in a pulverized-coal-fired furnace, the coal must first be pulverized in a mill to
 the consistency of talcum powder (i.e.; at least 70 percent of the particles will pass through a
 200-mesh sieve).  The pulverized coal is generally entrained in primary air before being fed
 through the burners to the combustion chamber, where it is fired in suspension. Pulverized-coal
 furnaces are classified as either dry or wet bottom, depending on the ash removal technique.  Dry
 bottom furnaces fire coals with high ash fusion temperatures, and dry ash removal techniques are
 used. In wet bottom (slag tap) furnaces, coal with a low ash fusion temperature is fired, and
 molten ash is drained from the bottom of the furnace.

        Pulverized-coal-fired furnaces are further classified by the firing position of the burners.
 Wall-fired boilers are characterized by rows of burners on one or more walls of the furnace.  The
 two basic forms of wall-fired furnaces are single-wall (having burners on one wall) or opposed
 (having burners on walls that face each other). Circular register burners and cell burners are
 types of burner configurations used in both single-wall and opposed-wall-fired units. A circular
 register burner is a single burner mounted in the furnace wall, separated from other burners so
 that it has a separate, distinct flame zone. Cell burners are several circular register burners
 grouped closely together to concentrate their distinct flame zones.

        Tangential-fired boilers are based on the concept of a single flame envelope and project
 both fuel and combustion air from the corners of the furnace.  The flames are directed on a line
 tangent to a small circle lying in a horizontal plane at the center of the furnace.  This action
 produces a fireball that moves in a cyclonic motion and expands to fill  the furnace.

 2.5.2 Cyclone Furnace

        Cyclone furnaces use burner design and placement (i.e., several water-cooled horizontal
 burners) to produce high-temperature flames that circulate in a cyclonic pattern. The coal is not
 pulverized but instead crushed to a 4-mesh size.  The crushed coal is fed tangentially, with
 primary air, to a horizontal cylindrical combustion chamber. In this chamber, small coal particles
 are burned in suspension, while the larger particles are forced against the outer wall.  The high
 temperatures developed in the relatively small furnace volume, combined with the low fusion
 temperature of the coal ash, causes the ash to  form a molten slag, which is drained from the
 bottom of the furnace through a slag tap opening.

 2.5.3 Fluidized-bed Combustor

        Fluidized-bed combustion increasingly is being used for coal-fired electric utility power
 plants. A variety of coals, including those with high concentrations of ash, sulfur, and nitrogen,
 can be burned in a fluidized-bed combustor (FBC). The term "fluidized" refers to the state of the
 bed materials (fuel or fuel and inert material [or sorbent]) as gas passes through the bed.  In a
 typical FBC, combustion occurs when coal, with inert material (e.g., sand, silica, alumina, or ash)
 and a sorbent such as limestone, is suspended through the action of primary combustion air

                                           2-18

-------
which is distributed below the combustor floor.  The gas cushion between the solids allows the
particles to move freely, giving the bed a liquid-like characteristic (i.e., fluidized). In an FBC,
crushed coal (between 1A and 3/8 inches in diameter) is injected into a bed above a grate-like air
distributor. Air is injected upward through the grate, lifting and suspending the solid particles.
Inert materials such as sand or alumina are often mixed with the coal to maintain the bed in a
fluidized state.  Limestone particles can also be added to the bed to adsorb sulfur dioxide
produced during combustion (discussed in Chapter 3).

2.5.4 Stoker-fired Furnace

       Stoker-firing of coal is used for the oldest furnace designs in the electric utility industry,
being first introduced to the industry in the late 1800s.  Today, this design is used by only a few
of the operating power plants. New power plants are not expected to adopt this design.  In stoker
furnaces, coal is burned on a bed at the bottom of the furnace.  The bed of coal burns on a grate.
Heated air passes upward through openings in the grate. Stokers are classified according to the
way coal is fed to the grate; the three general classes in use today are underfeed stokers, overfeed
stokers, and spreader stokers.  Underfeed stokers feed coal by pushing it upward through the
bottom of the grate. In overfeed stokers, the coal is deposited directly on the grate from a
gravity-fed bin. In spreader stokers, a flipping mechanism throws the coal into the furnace above
the grate; in this method, fine coal particles burn in suspension while heavier particles fall to the
grate and burn. Additional combustion air is added above the grate to support suspension
burning. Overfeed stokers can burn every type of coal except caking bituminous coal; spreader
stokers can burn all types of coal except anthracite.

2.5.5 Gasified-coal-fired Combustor

       Unlike the four coal-firing configurations discussed above, IGCC power plants do not
burn solid coal. In place of the coal-fired boiler used at a conventional coal-fired electric utility
power plant, at an IGCC power plant a coal gasification unit is used coupled with a gas turbine
combustor and heat recovery boiler. The solid coal is gasified by a process in which a coal/water
slurry is reacted at high temperature and pressure with oxygen (or air) and steam in a vessel (the
gasifier) to produce a combustible gas. This combustible gas is composed of a mixture of carbon
dioxide and hydrogen and is often referred to as a synthetic gas or Asyngas.@ Molten ash flows
out of the bottom of the gasifier into a water-filled sump where it forms a solid slag. The syngas
is cleaned and conditioned before being burned in a gas turbine that drives an electrical
generator. The hot combustion gases from the gas turbine are exhausted directly through a heat
recovery boiler (i.e., no combustion takes place in the boiler) to produce steam that is then
expanded through a steam turbine that drives a second generator to produce more electrical
power.

       The generation of electricity using the IGCC process offers a number of advantages
compared to using  conventional coal-fired boilers including higher thermal conversion
efficiencies (e.g., more kilowatt-hours  of electricity generated per kilogram of coal burned),
greater fuel flexibility (e.g., capability to use a wider variety of coal grades), and improved

                                           2-19

-------
control of particulate matter and SC>2 emissions without the need for post-combustion control
devices (e.g., almost all of the sulfur and ash in the coal is removed during the gasification
process). Three IGCC power plant projects have been constructed in the United States as part of
the DOE=s Clean Coal Technology Program, a joint government-industry cost-share technology
development program.  These facilities are the 250 MWe Tampa Electric Company Polk Power
Project, the 307 MWe Wabash River Coal Gasification Repowering Project, and the 107 MWe
Sierra Pacific Pinon Pine IGCC Power Project. Two of the facilities currently are operating (the
Polk and Wabash River IGCC facilities). The Pinon Pine IGCC facility presently is shut down
because of recurring problems with parti culate matter in the syngas causing premature gas
turbine blade erosion.13

       In IGCC applications, the syngas from the gasifier is cleaned and conditioned before it is
burned in the gas turbine using several different techniques. For example,  at the Wabash River
IGCC facility, the syngas from the coal gasifier passes through a series of gas cleaning and
conditioning steps including  a barrier filter for particulate removal, a water scrubber for gas
cooling, and an amine scrubber for removal of reduced-sulfur species.  In contrast, at the Polk
IGCC facility, a hot-gas cleaning process is used  and the syngas from the coal gasifier is not
cooled before it is burned in the gas turbine.
2.6 Ash from Coal Combustion

       Coal contains inorganic matter that does not burn including oxides of silicon, aluminum,
iron, and calcium. This noncombustible matter forms ash when the coal is burned. Burning of
coal in electric utility boilers generates large quantities of ash that must be removed and disposed
of.  The finer, lighter ash particles are entrained in the combustion gases and vented from the
furnace section with the flue gas. This portion of the coal ash is referred to as "fly ash." The
coarser, heavier ash particles fall to the bottom of the furnace section in the boiler unit.  This
portion of the coal ash is referred to as "bottom ash."  The proportion of fly ash to bottom ash
generated in a coal combustion unit varies depending on how the coal is burned.

       In general, the fly ash is collected as a dry material at several points downstream of the
furnace section. These points include collection hoppers beneath the boiler economizer, air
heater, and the particulate matter control devices (other than wet scrubbers).  From the collection
hopper, the fly ash is conveyed using a mechanical system, vacuum system, pneumatic system, or
combination of these systems to a storage silo. If a wet scrubbing system is used for air pollutant
control, fly ash is captured and removed in the scrubber wastewaters.

       For most boiler designs, the bottom ash is collected in a pit or hopper at the bottom of the
boiler furnace. The ash is collected in the form of either a dry material or a molten slag
depending on whether the furnace operating temperature is above the ash fusion temperature (i.e.,
the temperature at which the mineral compounds composing the ash melt). The ash is
continuously removed from the ash pit using a mechanical, pneumatic, or hydraulic conveyance
system.

                                          2-20

-------
       When coal is burned in a pulverized-coal furnace, on the order of 60 to 80 percent of the
total ash generated is fly ash.  The high amount of fly ash results because the coal enters the
furnace in a fine powder form that burns rapidly in suspension resulting in many tiny, lightweight
ash particles that can easily be carried out of the furnace section with the flue gas. The heavier
ash particles fall to the bottom of the furnace where they are removed.  Two pulverized-coal
boiler design approaches are used to collect bottom ash. The more frequently used design
approach, commonly referred to as a "dry-bottom" furnace, collects the ash as essentially a dry
material. For the typical dry-bottom furnace, the ash and slag particles fall into a water-filled
hopper. The water serves several purposes including providing an air seal to prevent the
infiltration of ambient air into the furnace, solidifying molten slag particles, and facilitating ash
handling.  The ash is then continuously removed from the ash pit using either a mechanical or an
hydraulic conveyance system. The other design approach, referred to as a "wet-bottom" furnace,
positions the coal burners on the furnace wall to maintain the ash that collects on the furnace
floor in a molten state.  The slag is drained through a slag tap opening into a slag tank.

       The cyclone  furnace is specifically designed to burn low-ash fusion coals and retains most
of the ash in the form of a molten slag. The molten slag collects in a trough on the bottom of
furnace and is continually drained through a slag tap opening into a slag tank. Water in the slag
tank solidifies the ash for disposal.  Only 20 to 30 percent of the ash produced by burning coal in
a cyclone furnace is  entrained as fly ash.

       By nature of the fluidized-bed combustion process, most of the ash in the coal leaves the
fluidized-bed combustor as fly ash.  Because the temperatures in the FBC remain below the ash
fusion temperature, formation of slag is avoided. Bottom ash is removed as a dry material to
maintain the fluidized bed  at a constant level. The ash removal system can be either a
mechanical or pneumatic system.

       In  stoker-fired furnaces where the coal is burned in a fuel  bed, most of the ash remains on
the grate and is removed as bottom ash.  Some smaller ash particles are entrained in the upward
flow of combustion  air through the grate and exit the furnace section as fly ash. The spreader
stoker has a greater proportion of the ash entrained as fly ash (up  to 50  percent of the ash) than
the other stoker types (on the order of 20 percent fly ash).  This occurs because the spreader
stoker mechanically throws the crushed coal across the top of the grate. This allows the smaller
coal fines in the incoming coal to burn in suspension before falling to the grate.  This produces
the small, lightweight ash particles that are carried out of the furnace section with the flue gas.

       No ash is produced when burning syngas derived from coal in an IGCC power plant. The
ash contained in the coal is removed by the gasification process that is used to produce the
syngas.  Before the syngas can be burned in the gas turbine, the gas  must be precleaned to
remove all types of particulate matter in order to prevent premature wear and destruction of the
turbine blades.
                                          2-21

-------
2.7 Coals Burned by Electric Utilities In 1999

       The EPA ICR Part n survey collected data on the coal, coal wastes, and some
supplemental fuels burned in each coal-fired electric utility boiler operating in the United States
during the entire calendar year 1999.  Coal samples were analyzed for, at a minimum, the higher
heating value (HHV) and the coal sulfur, ash, Hg, moisture, and chlorine content. Samples were
collected every third to twelfth fuel shipment in each month of 1999, depending on the statistical
characteristics of initial analysis results for each boiler unit. Either the coal shipper or the power
plant operator could take the sample if the samples were collected at a point after any coal
cleaning had been completed. Thus, Aas-shipped@ or Aas-received@ coals are considered to be
equivalent to Aas-fired@ coals, and Hg analyses from such samples are assumed to represent the
quantity of Hg entering the boiler.

       In 1999, a nationwide total of approximately 786 million tons of coal and supplemental
fuels were burned in coal-fired electric utility boilers that met the CAA Section 112(a) definition
of an electric utility steam generating unit (i.e., boiler units of more than 25 megawatts that serve
a generator that produces electricity for sale). Table 2-5 shows the nationwide distribution of the
coal burned by rank as reported by the respondents to the EPA ICR (i.e., the power plant owners
and operators).

       Most electric utility power plants burn either bituminous or subbituminous coals. Half of
the coals burned by the electric utility industry in 1999 were bituminous coal (52 percent of the
total nationwide tonnage). Approximately one-third of the coals burned were subbituminous
coals (36.5 percent of the total nationwide tonnage).  Some power plants reported burning both
bituminous and subbituminous coals. At most of these facilities, the two coal types are blended
together before firing in the boiler unit. A few of the facilities switch between the two coal types
for firing in the boiler unit to address site-specific circumstances.  The vast majority of the
bituminous or subbituminous coals were supplied from mines in the United States. However,
imported coals were burned in 1999 at a few power plant locations. Ten plants, located near Gulf
of Mexico or Atlantic Ocean seaports, imported bituminous coal from South America and three
plants located in Hawaii and Florida imported subbituminous coal from Indonesia.

       In general, the burning of lignite or anthracite coals by electric utilities is limited to those
power plants that are located near the mines supplying the  coal.  Lignite accounted for
approximately 6.5 percent of the total coal tonnage burned at electric utility power plants in
1999. A total of 17 electric utility power plants reported burning lignite. All of these facilities
are located near the coal deposits from which the lignite is mined in Texas, Louisiana, Montana,
or North Dakota.  Similarly, burning of anthracite coal in 1999 was limited to a few power plants
located close to the anthracite coal mines in eastern Pennsylvania. The  coal-fired electric utility
boilers at these facilities burned  either newly mined anthracite coal or waste anthracite coal
reclaimed from mine waste piles.

       Table 2-5 also shows that small amounts of supplemental fuels  (e.g., petroleum coke or
tire derived fuel [TDF] chips) also were co-fired with coal in some coal-fired electric utility

                                          2-22

-------
Table 2-5. Nationwide quantities of coals and supplemental fuels burned in
coal-fired electric utility boilers for the year 1999 as reported in the Part II EPA
ICR data (source: Reference 10).
Fuel Type
Bituminous coal
Subbituminous coal
Lignite
Bituminous/subbituminous coal mixture
Bituminous coal/petroleum coke mixture
Waste anthracite coal
Waste bituminous coal
Petroleum coke
Other (a)
Total
Total Tonnage
Burned
(million tons)
406
287
51
24
6
5
4
2
1
786
Percentage
by Weight
51.7%
36.5%
6.5%
3.0%
0.7%
0.6%
0.5%
0.3%
< 0.2%
100%
      (a) Mixes of anthracite, bituminous, and waste bituminous fuel, tires, subbituminous coal and petroleum
         coke, or waste subbituminous coal.
                                        2-23

-------
boilers. At these facilities, the supplemental fuels are mixed with coal before firing in the boiler
unit. These supplemental fuels typically have heating values higher than that of coal and serve to
boost the overall heating value of the fuel mix burned in the boiler unit.  Less than 0.5 percent of
the total fuel tonnage burned in 1999 consisted of supplemental fuels.

       Selected properties of the coal and supplemental fuel burned nationwide in coal-fired
electric utility boilers in 1999, as reported in the EPA ICR Part II data, are summarized by fuel
type in Appendix A.  Table 2-6 presents a summary of the Hg content data reported for the coals
and supplemental fuels as fired in the boiler units.  The EPA ICR data do not identify the coal
resource regions from which the coal burned in a given boiler unit was mined. However,
consistent with the Hg content data for as-mined coals presented in Table 2-2, the data presented
in Table 2-6 indicate that there is no  general relationship between coal rank and Hg content of the
coal. For bituminous, subbituminous, and lignite coals, the Hg concentrations reported in the
EPA ICR data ranged from trace amounts to upper levels of approximately 1 ppm.

       A review of the EPA ICR data suggests that there is no direct correlation between the
sulfur content of a coal and its Hg content.  In other words, Ahigh@ sulfur coals are not necessarily
Ahigh@ Hg coals.  Trace concentrations of Hg were reported for coals with high-sulfur contents.
Conversely, Hg concentrations at the upper end of the concentration ranges also were reported
for high sulfur-content coals.  This observation is consistent with previous studies of the Hg
content in coal based on a much smaller database. For example, an earlier study comparing the
sulfur and Hg concentrations in  153  samples of coal shipments found no relationship between the
sulfur and Hg concentrations in these coals.14
                                          2-24

-------
                 Table 2-6.  Mercury content of as-fired coals and supplemental fuels burned in coal-fired electric
                 utility boilers for the year 1999 as reported in the Part II EPA ICR data (source: Reference 10).
to
Fuel Type
Bituminous coal
South American bituminous coal (a)
Subbituminous coal
Indonesian subbituminous coal (b)
Lignite coal
Anthracite coal
Waste anthracite coal
Waste bituminous coal
Waste subbituminous coal
Petroleum coke
Tire-derived fuel
Number of
Samples
Analyzed
27,793
270
8,180
78
1,047
65
426
572
53
1,150
149
Hg Concentration
(ppm dry)
Range | Mean
0.0-1.3 0.11
0.01-0.42 0.08
0.008-0.9 0.07
0.02-0.1 0.03
0.02-0.75 0.11
0.06-0.23 0.14
0.04-0.54 0.19
0.03-1.18 0.46
0.07-0.35 0.12
0.0009-0.5 0.05
0.01-0.33 0.056
Hg Content by
Fuel Heating Value
(lbHgper1012Btu)
Range i Mean
0.07-103.81 8.61
0.70-31.06 5.52
0.66-71.02 5.75
1.54-9.23 2.70
1.84-75.06 10.80
5.02-17.49 11.36
8.39-4.73 28.55
2.47-172.92 60.79
5.81-30.35 11.42
0.06-32.16 3.30
0.60-19.89 3.72
                  (a) Bituminous coal imported from South America and burned at 10 power plants.
                  (b) Subbituminous coal imported from Indonesia and burned at three power plants in Hawaii and Florida.

-------
2.8 References
1.  American Society for Testing and Materials. 2000 Annual Book of ASTMStandards. West
   Conshohocken, PA. December 2000.

2.  Electric Power Research Institute. Evaluation of Methods for Analysis of Mercury and
   Chlorine in Coal. EPRI Report 1000287, Palo Alto, CA. September 2000.

3.  Wood, G. H., Jr.,  T.M. Kehn, M.D. Carter, and W.C. Culbertson.  Coal Resource
   Classification System of the U.S. Geological Survey. U.S. Geological Survey Circular 891,
   1983.  Available at: < http://energy.er.usgs.gov/products/papers/C891/index.htm >.

4.  U.S. Department of Energy, Energy Information Administration.  U.S. Coal Reserves: 1997
   Update. DOE/EIA-0529(97). Office of Coal, Nuclear, Electric and Alternate Fuels, Office of
   Integrated Analysis and Forecasting, Washington, DC. February 1999.  Available at:
   < http://www.eia.doe.gov/cneaf/coal/reserves/front-l.html >.

5.  J.  Tully.  Coal Resource Regions of the Conterminous United States. U.S. Geological Survey
   Open-File Report 96-279. July 6,  1996. Available at:
   < http://energy.er.usgs.gov/products/openfile/OF96-279/ >.

6.  Bragg, L.J., J.K. Oman, S.J. Tewalt, C.J. Oman, N.H. Rega, P.M. Washington, and R.B.
   Finkelman.  U.S. Geological Survey Coal Quality (COALQUAL) Database:  Version 2.0. U.S.
   Geological Survey Open-File Report 97-134. June 15, 2001. Available at:
   < http://energy.er.usgs.gov/products/databases/CoalOual/index.htm >.

7.  Toole-O'Neil, B., S.J. Tewalt, R.B. Finkleman, and R. Akers. "Mercury  Concentration in
   Coal-Unraveling the Puzzle." Fuel, 78, 47-54 (1999).

8.  Keating, M.H., K.R. Mahaffey, R. Schoeny, G.E. Rice, O.R. Bullock, R.B. Ambrose, Jr.,
   J. Swartout,  and J.W. Nichols. Mercury Study Report to Congress, Volume 11. EPA-452/R-
   97-004b.  Office of Air Quality Planning and Standards and Office of Research and
   Development, Research Triangle Park, NC. December 1997.  Available at:
   < http://www.epa.gov/airprogm/oar/mercury.html >.

9.  McDermott Technology, Inc.. Mercury Emission ResultsCoal Content, Emissions and
   Control. Alliance, OH. Available at:
   <
   http://www.mtiresearch.com/aecdp/mercury.html#Coal%20Composition%20and%20Coal%2
   OCleaning >(accessed July 2001).
                                        2-26

-------
10.  U.S. Environmental Protection Agency. Database of information collected in the Electric
   Utility Steam Generating Unit Mercury Emissions Information Collection Effort. OMB
   Control No. 2060-0396.  Office of Air Quality Planning and Standards. Research Triangle
   Park, NC. April 2001. Available at:
   < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

11.  Singer, J.G. (Ed.). Combustion Fossil Power.  Fourth Edition. Combustion Engineering,
   Inc., Windsor, CT. 1991.

12. French, C.L., W.H. Maxwell, W.D. Peters, G.E. Rice, O.R. Bullock, A.B Vasu, R. Hetes,
   A. Colli,  C. Nelson, and B.F. Lyons. Study of Hazardous Air Pollutant Emissions from
   Electric Utility Steam Generating Units  Final Report to Congress, Volume 1.  EPA-453/R-
   98-004a.  Office of Air Quality Planning and Standards, Research Triangle Park, NC.
   February 1998.  Available at: < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

13. Cargill, P., G. DeJonghe, T. Howsley, B. Lawson, L. Leighton, and M. Woodward. Pinon
   Pine IGCC Project: Final Technical Report to the Department of Energy. DOE Award No.
   DE-FC21-92MC29309, Sierra Pacific Resources, Sparks, NV, January 2001. Available at:
   .

14. Baker, S.S. EPR1Mercury in Coal Study: A Summary Report for Utilities That Submitted
   Samples  Update.  Prepared for EPRI Utility Air Regulatory Group by  Systems Applications
   International Corporation, San Diego, CA.  June 1994.  pp. D-ltoD-4.
                                       2-27

-------
                                        Chapter 3
                    Criteria Air Pollutant Emission Controls for
                           Coal-fired Electric Utility Boilers
3.1 Introduction
        The EPA uses "criteria pollutants" as indicators of ambient air quality.  For each criteria
 air pollutant, the EPA has established maximum concentrations for specific exposure periods
 above which adverse effects on human health may occur. Under authority of the CAA, these
 threshold concentrations for the criteria air pollutants are codified as the national ambient air
 quality standards (NAAQS).  The EPA has set NAAQS for six criteria air pollutants: carbon
 monoxide (CO), lead (Pb), nitrogen dioxide (NC^), ozone (Os), particulate matter (PM), and
 sulfur dioxide (SO2).

        Estimates of national emissions for criteria air pollutants prepared by the EPA show that
 electric utility power plants that burn coal are significant emission sources of SC>2, nitrogen
 oxides (NOX), and PM.1  Electric utility power plants are the Nation's largest source of SO2
 emissions, contributing approximately 68 percent of the estimated total national SC>2 emissions in
 1998 (most recent year for which national estimates are available). Over 90 percent of these SO2
 emissions are coal-fired electric utility boilers. Electric utilities contributed 25 percent of total
 national NOX emissions in 1998. Again coal combustion is the predominant source of NOX
 emissions from the electric utilities (almost 90 percent of the estimated NOX emissions). Coal-
 fired electric utility power plants also are one of the largest industrial sources of PM emissions.
 In general, the high combustion efficiencies achieved by coal-fired electric utility boilers result in
 low emissions of CO and volatile organic compounds (a precursor for the photochemical
 formation of ozone in the atmosphere). Lead is listed as a HAP in addition to being listed as a
 criteria air pollutant. Lead emissions from electric utility boilers were evaluated as part of EPA's
 report to Congress on HAP emissions from electric utility power plants (discussed in Section
 1.4.1). 2 The EPA found that electric utility boilers contribute a very small percentage of the
 nationwide Pb emissions.

        All coal-fired electric utility power plants in the United States use control devices to
 reduce PM emissions.  Many coal-fired electric utility boilers also are required to use controls for
 SO2 and NOX emissions depending on site-specific factors such as the properties of the coal
 burned, when the power plant was built, and the area where the power plant is located.  As
 discussed in Chapter 6, certain control technologies used to reduce criteria air pollutant
                                          5-1

-------
emissions from coal-fired electric utility boilers also remove some of the mercury (Hg) from the
flue gas. In addition, the existing control configuration used for a given coal-fired electric utility
boiler to meet criteria air pollutant emissions standards directly can affect the applicability,
performance, and costs of retrofitting additional Hg controls to the unit.

       The purpose of this chapter is to present a summary review of the different control
technologies currently used by coal-fired electric utility boilers to meet the applicable criteria air
pollutant emissions standards.  The nationwide distribution of control configurations used at
coal-fired electric utility power plants to comply with these standards is presented using
information from the EPA ICR database.  The impact or influence of these control configurations
on control of Hg emissions is discussed in the Chapter 6.
3.2 Criteria Air Pollutants of Concern from Coal Combustion

3.2.1  Particulate Matter3 4

       Dust, dirt, soot, smoke, and liquid droplets are directly emitted into the air from
anthropogenic sources as well as natural sources such as forest fires and windblown dust. This
type of PM sometimes is called "primary particulate matter." In addition, gaseous air pollutants
(e.g., sulfur dioxide, nitrogen oxides, and volatile organic compounds) are considered to be PM
precursors causing "secondary particulate matter" through complex transformations that occur in
the ambient environment. Human exposure to concentrations of PM at various levels results in
effects on breathing and respiratory symptoms,  aggravation of existing respiratory and
cardiovascular disease, alterations in the body's defense systems against foreign materials,
damage to lung tissue, carcinogenesis, and premature death. The people most sensitive to the
effects of PM include individuals with chronic obstructive pulmonary or cardiovascular disease
or influenza, asthmatics, the elderly, and children.  Particulate matter also contributes to visibility
impairment in the United States.

       Primary PM emissions from coal-fired electric utility boilers consist primarily of fly ash.
Ash is the unburned carbon char  and the mineral portion of combusted coal.  The amount of ash
in the coal, which ultimately exits the boiler unit as fly ash, is a complex function of the coal
properties, furnace-firing configuration, and boiler operation. For the dry-bottom, pulverized-
coal-fired boilers, approximately 80 percent of the total ash in the as-fired coal will exit the boiler
as fly ash. Wet-bottom, pulverized-coal-fired boilers emit significantly less fly ash: on the order
of 50 percent of the total  ash exits the boiler as  fly ash.  In a cyclone furnace boiler, most of the
ash is retained as liquid slag; thus, the quantity  of fly ash exiting the boiler is typically 20 to 30
percent of the total ash.  However, the high operating temperatures unique to these designs may
also promote ash vaporization and larger fractions of submicron fly ash compared to dry bottom
designs.  Fluidized-bed combustors emit high levels of fly ash since the coal is fired in
suspension and the ash is present in dry form. Spreader-stoker-fired boilers can also emit high
levels of fly ash. However, overfeed and underfeed stokers emit less fly ash than spreader
stokers, since combustion takes place in a relatively quiescent fuel bed.
                                          5-2

-------
       In addition to the fly ash, PM emissions from a coal-fired electric utility power plant
result from reactions of the SO2 and NOX compounds as well as unburned carbon particles carried
in the flue gas from the boiler. The SC>2 and NOX compounds are initially in the vapor phase
following coal combustion in the furnace chamber but can partially chemically transform in the
stack, or near plume, to form fine PM in the form of nitrates, sulfur trioxide (SOs), and sulfates.
Firing configuration and boiler operation can affect the fraction of carbon (from unburned coal)
contained in the fly ash. In general, the high combustion efficiencies achieved by pulverized-
coal-fired boilers and cyclone-fired boilers result in relatively small amounts of unburned carbon
particles in the exiting combustion gases. Those pulverized-coal-fired electric utility boilers that
use special burners for NOX control (discussed in Section 3.7) tend to burn coal less completely;
consequently, these furnaces tend to emit a higher fraction of unburned carbon in the combustion
gases exiting the furnace.

       Another potential source of PM in the flue gas from a coal-fired electric utility boiler is
the use of a dry sorbent-based control technology.  Solid sorbent particles are injected into the
combustion gases to react with the air pollutants and then recaptured by a downstream control
device.  Sorbent particles that escape capture by the control device are emitted as PM to the
atmosphere. Control technologies using sorbent injection are discussed in Chapter 7.
3.2.2 Sulfur Dioxide
                    3,4
       Exposure of people to SC>2 concentrations above threshold levels affects their breathing
and may aggravate existing respiratory and cardiovascular disease. Sensitive populations include
asthmatics, individuals with bronchitis or emphysema, children, and the elderly.  Sulfur dioxide
is also a primary contributor to acid deposition, or acid rain, which causes acidification of lakes
and streams and can damage trees,  crops, historic buildings, and statues.  In addition, SOX
compounds in the air contribute to visibility impairment. In the United States, SO2 is primarily
emitted from the combustion of fossil fuels and by metallurgical processes.

       Coal deposits contain sulfur in amounts ranging from trace quantities to as high as
eight percent or more.  Most of this sulfur is present as either pyritic sulfur (sulfur combined with
iron in the form of a mineral that occurs in the coal deposit) or organic sulfur (sulfur combined
directly in the coal structure). During combustion, sulfur compounds in coal are oxidized to
gaseous SC>2 or SOs. When firing bituminous coal, almost all of the sulfur present in coal will be
emitted as gaseous sulfur oxides (on average 98 percent). The more alkaline nature of ash in
some subbituminous coals causes a portion of the sulfur in the coal to react to form various
sulfate salts; these salts are emitted as fly ash or retained in the boiler bottom ash. Generally, the
percentage of sulfur in the as-fired  coal that is converted to sulfur oxides  during combustion does
not vary with the utility boiler design or operation.
3.2.3  Nitrogen Oxides
                      4,5
       Nitrogen dioxide (NO2) is a highly reactive gas.  The major mechanism for the formation
of NO2 in the atmosphere is the oxidation of nitric oxide (NO) when exposed to solar radiation.

-------
These two chemical species are collectively referred to as nitrogen oxides (NOX).  Exposure of
people to NO2 can irritate the lungs, cause bronchitis and pneumonia, and lower resistance to
respiratory infections.  Nitrogen oxides are an important precursor together with volatile organic
compounds in the photochemical formation of ozone in the atmosphere.  Ozone is a criteria
pollutant and the major component of smog. Nitrogen dioxide is also a primary contributor to
acid rain.  The major NOX emissions sources are transportation vehicles and stationary
combustion units.

       Both NO and NO2 are formed during coal combustion by oxidation of molecular nitrogen
that is present in the combustion air or nitrogen compounds contained in the coal. Overall, total
NOX formed during combustion is composed predominantly of NO mixed with small quantities
of NO2 (typically less than 10 percent  of the total NOX formed).  However, once NO formed
during coal combustion is emitted to the atmosphere, the NO is oxidized to NO2.

       The NOX formed during coal combustion by oxidation of molecular nitrogen (N2) in the
combustion air is referred to as "thermal NOX." The oxidation reactions  converting N2 to NO and
NO2 become very rapid once gas temperatures rise above 1,700 C (3,100 F). Formation of
thermal NOX in a coal-fired electric  utility boiler is dependent on two conditions occurring
simultaneously in the combustion zone:  high temperature and an excess of combustion air.  A
boiler design feature or operating practice that increases the gas temperature above 1,700 C, the
gas residence time at these temperatures, and the quantity of excess combustion air will affect
thermal NOX formation.  The formation of NOX by oxidation of nitrogen  compounds contained in
the coal is referred to as "fuel NOX." The nitrogen content in most coals ranges from
approximately 0.5 to 2 percent. The amount of nitrogen available in the  coal is relatively small
compared with the amount of nitrogen available in the combustion air. However, depending on
the combustion conditions, significant quantities of fuel NOX can be formed during coal
combustion.
3.3 Existing Control Strategies Used for Coal-fired Electric Utility Boilers

       Electric utilities must comply with applicable Federal standards and programs that
specifically regulate criteria air emissions from coal-fired electric utility boilers.  These
regulations and programs include New Source Performance Standards (NSPS), the CAA Title IV
Acid Rain Program, and the CAA Title V Operating Permits Program. The EPA has delegated
authority to individual state and local agencies for implementing many of these regulatory
requirements.  In addition, individual states have established their own standards and
requirements for those power plants that operate within their jurisdictions. Electric utility
companies use one or a combination of the following three control strategies to comply with the
specific set of requirements applicable to a given coal-fired boiler.

       Pre-combustion Controls. Control measures in which fuel substitutions are made or fuel
       pre-processing is performed to reduce pollutant formation in the combustion unit.
                                         5-4

-------
       Combustion Controls.  Control measures in which operating and equipment
       modifications are made to reduce the amount of pollutants formed during the combustion
       process; or in which a material is introduced into the combustion unit along with the fuel
       to capture the pollutants formed before the combustion gases exit the unit.

       Post-combustion Controls: Control measures in which one or more air pollution control
       devices are used at a point downstream of the furnace combustion zone to remove the
       pollutants from the post-combustion gases.

       Table 3-1 shows the distribution of emissions control strategies for PM, SO2, and NOX
used for coal-fired electric utility boilers in 1999 as reported in the Part II EPA ICR data.6  All
coal-fired electric utility boilers in the United States are controlled for PM emissions by using
some type of post-combustion controls. These particulate emission  control types are discussed in
Section 3.4. Approximately two-thirds of the total coal-fired electric utility boilers use add-on
controls for SC>2 emissions.  Most of these controlled units use either a pre-combustion or a post-
combustion control strategy for SO2 emissions. The  methods used for controlling SO2 emissions
from coal-fired electric utility boilers  are discussed in Section 3.5. Although approximately two-
thirds of the coal-fired electric utility boilers are controlled for NOX  emissions, these units are not
necessarily the same units controlled for SO2 emissions. The predominant strategy for
controlling NOX emissions is to use combustion controls. Section 3.6 discusses the application of
NOX emission controls to coal-fired electric utility boilers.
3.4 Particulate Matter Emission Controls

       Four types of control devices are used to collect PM emissions from coal-fired electric
utility boilers: electrostatic precipitators, fabric filters, mechanical collectors, and particle
scrubbers.  Table 3-2 presents the 1999 nationwide distribution of PM controls on coal-fired
electric utility boilers by total number of units and by percentage of nationwide electricity
generating capacity.  Electrostatic precipitators are the predominant control type used on coal-
fired electric utility boilers both in terms of number of units (84 percent) and total generating
capacity (87 percent). The second most common control device type used is a fabric filter.
Fabric filters are used on about 14 percent of the coal-fired electric utility boilers. Particle
scrubbers are used on approximately three percent of the boilers.  The least used control device
type is a mechanical  collector.  Less than one percent of the coal-fired electric utility boilers use
this type of control device as the sole PM control. Other boilers equipped with a mechanical
collector use this control device in combination with one of the other PM control device types.
 3.4.1 Electrostatic Precipitators
                                4,7
       Electrostatic precipitator (ESP) control devices have been used to control PM emissions
for over 80 years.  These devices can be designed to achieve high PM collection efficiencies
(greater than 99 percent), but at the cost of increased unit size. An ESP operates by imparting an
electrical charge to incoming particles, and then attracting the particles to oppositely charged
                                           5-5

-------
Table 3-1.  Criteria air pollutant emission control strategies as applied to
coal-fired electric utility boilers in the United States for the year 1999 as reported
in the Part II EPA ICR data (source:  Reference 6).
Criteria
Air Pollutant
Particulate
matter
Sulfur
dioxide
Nitrogen
oxides
Percentage of Coal-fired Electric Utility Boilers Using Control Strategy
as Reported in Phase II EPA ICR Data ""
Meet Applicable
Standards
Without
Additional
Controls
0%
37%
40%
Pre-combustion
Controls
0%
40%
0%
Combustion
Controls
0%
3%
57%
Post-combustion
Controls
100%
20%
3%
    (a) Approximately 1.5 % of the boilers use a combination of pre-combustion and post-combustion SO2 controls.
    (b) Approximately 1% of the boilers using post-combustion NOx controls also use some type of combustion
       controls.
                                         5-6

-------
Table 3-2.  Nationwide distribution of existing PM emission controls used for
coal-fired electric utility boilers for the year 1999 as reported in the Part II EPA
ICR data (source: Reference 6).
PM
Control Type
Electrostatic precipitator
(Cold-side)
Electrostatic precipitator
(Hot-side)
Fabric filter
Particle scrubber
Mechanical collector (d)
Multiple control device
combinations (e)
Abbreviation
Code
CS- ESP
HS-ESP
FF
PS
MC

Nationwide Total
Phase II EPA ICR Data
Number
of Boilers
822 (a)
122
155 (b)
23 (c)
5
13
1,140(f)
Percent of
Nationwide
Total Number
of Units
72.1 %
10.8%
13.6%
2.0%
0.4 %
1.1 %
100%
Percent of
Nationwide
Electricity
Generating
Capacity
74.7 %
11.3%
9.4 %
3.0 %
0.2 %
1 .4 %
100%
    (a) Includes 10 boilers with cold-side ESP in combination with upstream mechanical collector.
    (b) Includes eight boilers with baghouse in combination with upstream mechanical collector.
    (c) Includes two boilers with particle scrubber in combination with upstream mechanical collector.
    (d) Boilers using  mechanical collector as only PM control device.
    (e) Boilers using  a combination of two or more different control device types other than mechanical
       collectors. Includes two  boilers that use a hot-side ESP in series with a cold-side ESP.
    (f) Does not include the three IGCC units.
                                            5-7

-------
metal plates for collection.  Periodically, the particles collected on the plates are dislodged in
sheets or agglomerates (by rapping the plates) and fall into a collection hopper. The dust
collected in the ESP hopper is a solid waste that must be disposed of.

       The effectiveness of particle capture in an ESP depends largely on the electrical resistivity
of the particles being collected.  An optimum value exists for a given ash. Above and below this
value, particles become less effectively charged and collected. Table 3-3 presents the PM
collection efficiency of an ESP compared with the other control device types. Coal that contains
a moderate to high amount of sulfur (more than approximately three percent) produces an easily
collected fly ash.  Low-sulfur coal produces a high-resistivity fly ash that is more difficult to
collect.  Resistivity of the fly ash can be changed by operating the boiler at a different
temperature or by conditioning the particles upstream of the ESP with sulfur trioxide, sulfuric
acid, water, sodium, or ammonia. In addition, collection efficiency is  not uniform for all particle
sizes. For coal fly ash, particles larger than about 1 to 8 |im and smaller than about 0.3  |im (as
opposed to total PM) are typically collected with efficiencies from 95 to 99.9 percent. Particles
near the 0.3 |im size are in a poor charging region that reduces collection efficiency to 80 to 95
percent.

       An ESP can be used at one of two locations in a coal-fired electric utility boiler  system.
For many years, every ESP  was installed downstream of the air heater where the temperature of
the flue gas is between 130  and 180 C (270 and 350 F).  An ESP installed at this location is
referred is as a "cold-side" ESP.  However, to meet SO2 emission requirements, many electric
utilities switched to burning low-sulfur coal (discussed in the Section 3.5.1). These coals have
higher electrical ash resistivities, making the fly ash more difficult to capture downstream of the
air heater.  Therefore, to take advantage of the lower fly-ash resistivities at higher temperatures,
some ESPs are installed upstream of the air heater, where the temperature of the flue gas is in the
range of 315 to 400 C (600 to 750 F). An ESP installed upstream of the air heater is referred to
as a "hot-side" ESP.

3.4.2 Fabric Filters4'8

       Fabric  filters (FF) have been used for fly ash control from coal-fired electric utility boilers
for about 30 years. This type  of control device collects fly ash in the combustion gas stream by
passing the gases through a porous fabric material.  The buildup of solid particles on the fabric
surface forms a thin, porous layer of solids or a filter, which further acts as a filtration medium.
Gases pass through this cake/fabric filter, but the fly ash is trapped on  the cake surface.  The
fabric material used is typically fabricated in the shape of long, cylindrical bags.  Hence, fabric
filters also are frequently referred to as  "baghouses."

       Gas flow through a FF becomes excessively restricted if the filter cake on the bags
becomes too thick. Therefore, the dust collected on the bags must be removed periodically. The
type of mechanism used to remove the filter cake classifies FF design types. Depending on the
FF design type, the dust particles will be collected either on the inside or outside of the bag. For
designs in which the dust is collected on the inside of the bags, the dust is removed by either

-------
Table 3-3. Comparison of PM collection efficiencies for different PM control
device types (source: Reference 4)
PM
Control Type
Electrostatic precipitator
(Cold-side)
Electrostatic precipitator
(Hot-side)
Fabric filter
Particle scrubber
Mechanical collector
Representative PM
Mass Collection Efficiency Range
Total
PM
99 to 99. 7%
99 to 99. 7%
99 to 99. 9%
95 to 99 %
70 to 90 %
PM
less than 0.3 \im
80 to 95 %
80 to 95 %
99 to 99. 8%
30 to 85 %
Oto15%
                                   5-9

-------
mechanically shaking the bag (called a "shaker type" FF) or by blowing air through the bag from
the opposite side (called a "reverse-air" FF). An alternate design mounts the bags over internal
frame structures, called "cages" to allow collection of the dust on the outside of the bags. A
pulsed jet of compressed air is used to cause a sudden stretching then contraction of the bag
fabric dislodging the filter cake from the bag.  This design is referred to as a "pulse-jet" FF. The
dislodged dust particles fall into a hopper at the bottom of the baghouse. The dust collected in
the hopper is a solid waste that must be disposed of.

       An FF must be designed and operated carefully to ensure that the bags inside the collector
are not damaged or destroyed by adverse operating conditions. The fabric material must be
compatible with the gas  stream temperatures and chemical composition. Because of the
temperature limitations of the available bag fabrics, location of an FF for use in a coal-fired
electric utility boiler is restricted to downstream of the air heater.  In general, fabric filtration is
the best commercially available PM control technology for high-efficiency collection of small
particles (see Table 3-3).

       Electrostatic stimulation of fabric filtration (ESFF) involves a modified fabric filter that
uses electrostatic charging of incoming dust particles to increase collection efficiency and reduce
pressure drop compared to fabric filters without charging. Filter bags are specially made to
include wires or conductive threads, which produce an electrical field parallel to the fabric
surface. Conductors can also be placed as a single wire in the center of the bag. When the bags
are mounted in the baghouse, the conductors are attached to a wiring harness that supplies
electricity.  As particles enter the field and are charged, they form a porous mass or cake of
agglomerates at the fabric surface. Greater porosity of the cake reduces pressure drop, while the
agglomeration increases efficiency of small particle collection. Cleaning is required less
frequently, resulting in longer bag life. For felted or nonwoven bags, the field promotes
collection on the outer surface of the fabric, which also promotes longer bag life.  Filtration
velocity can be increased so that less fabric area is required in the baghouse. The amount of
reduction is based on an economic balance among desired performance, capital cost, and
operating costs. A number of variations exist on the ESFF idea of combining particle charging
with fabric filtration.

       The University of North Dakota, Energy and Environmental Research Center
(UND/EERC) has developed another type of combined control device called the Advanced
Hybrid Collector (AHC).9 A charging (and collection) section can also be placed ahead of the
bags in a fabric filter. This approach is used in the AHC  along with the use of membrane fabrics
(woven or felted fabrics having a membrane laminated to the filtration surface of the fabric).
The membrane is typically polytetrafluoroethylene (PTFE). With about 90 percent of the mass of
particles collected in the electrostatic charging and collection section of the AHC, the load on the
fabric filter part of the system is much reduced. With a membrane fabric for the bags, it is likely
that filtration velocity can be increased significantly.
                                         5-10

-------
3.4.3 Particle Scrubbers4

       Particle scrubbers operate by shattering streams of water into small droplets that collide
with and trap solid particles contained in the flue gas or by forcing the gases into intimate contact
with water films. The particle-laden droplets or water films coalesce and are collected in a sump
at the bottom of the scrubber. The three basic types of particle scrubbers are venturi scrubbers,
preformed spray scrubbers, and moving-bed scrubbers. Venturi scrubbers are the type most
commonly used for coal-fired electric utility boilers.  This scrubber design transports the particle-
laden flue gas through a constriction where violent mixing takes place. Water is introduced
either at or upstream of the constriction. Preformed spray scrubbers are usually vertical cylinders
with flue gas passing upward through droplets sprayed from nozzles near the top of the unit.
Moving-bed scrubbers have an upper chamber in which a bed of low-density spheres (often
plastic) is irrigated by streams of water from above. Gas passing upward through the bed agitates
the wetted spheres, which continually expose fresh liquid surfaces for particle transfer.
Regardless of the scrubber design, all particle scrubber systems generate wastewaters from the
scrubber blowdown that must be treated and discharged.

       Particle scrubbers are more sensitive to particle size distribution in the flue gas than either
an ESP or an FF. In general, particle scrubbers are not as effective as these other control devices
at collecting small particles (see Table 3-3). Also, while a venturi particle scrubber will have a
lower initial cost for a given boiler unit application than either an ESP or an FF, the high pressure
drop required for the scrubber to achieve a high collection efficiency results in high operating
costs.  These factors, in large part, account for the low use of particle scrubbers at coal-fired
utilities.

3.4.4 Mechanical Collectors4

       Mechanical collectors are the oldest, simplest, and least efficient of the four types of PM
control devices. The collectors used for utility boilers are generally in the form  of groups of
cylinders with conical bottoms (multicyclones). Flue gas entering the cylinder tangentially to the
wall is imparted with a circular motion around the cylinder's axis. Particles in the gas  stream are
forced toward the wall by centrifugal force, then downward through a discharge at the bottom of
the cone.  Collection efficiency for a typical multicyclone can be about 70 to 75 percent for
10-|im particles, but can drop to less than 20 percent for smaller l-|im particles. Mechanical
collectors can be efficient for relatively large particles because their settling velocity is high
compared to fine particles.  In a cyclone, larger particles are forced through the gas stream
towards the outer wall because of their mass and inertia, while small particles have insufficient
mass to be much affected. Electrically charging particles tends to agglomerate them, especially
small particles, with the resulting larger agglomerates having increased mass over the individual
small particles.  In charged mechanical collectors, a charging section is placed ahead of a
mechanical collector, and collection efficiency for smaller particles is significantly increased.
                                          5-11

-------
3.5 SOi Emission Controls

       Sulfur dioxide emissions from most coal-fired electric utility boilers are controlled using
either of two basic approaches. The first approach is to use pre-combustion measures, namely,
the firing coal that contains lower amounts of sulfur.  The low-sulfur coal may be naturally
occurring or the result of coal cleaning.  The other approach is to remove the sulfur compounds
from the flue gas before the gas is discharged to the atmosphere. These post-combustion
processes are collectively called "flue gas desulfurization" or "FGD" systems.  All FGD systems
can be further classified as wet or dry flue gas scrubbing systems.  A third control approach
available for those coal-fired electric utility boilers using a fluidized-bed combustor is to burn the
coal together with limestone.  An FBC can be characterized as a boiler type with inherently lower
SC>2 emissions.  In this report, however, combustion of coal in fluidized-bed with limestone is
also considered to be an SC>2 combustion control method.  The SC>2 control  approaches include a
number of different technology subcategories that are now commercially used in the United
States, Europe, or Pacific Rim countries.

       Table 3-4 presents the 1999 nationwide distribution of SC>2 controls used for coal-fired
electric utility boilers by total number of units and by percentage of nationwide electricity
generating capacity. For approximately one-third of the boilers, no SC>2 controls were reported in
the Part n EPA ICR data. The other two-thirds of the units reported using some type of control
to meet the SC>2 emission standards applicable to the  unit.  Pre-combustion  control by burning a
low-sulfur content coal was reported for approximately 40 percent of the boilers. Post-
combustion control devices for SC>2 removal are used for approximately 20 percent of the boilers.
Wet FGD systems  are the most commonly used post-combustion control technique.  The newer
technologies of spray dryer systems or dry injection are limited in their application to existing
units. The remaining 3 percent of the boilers use fluidized-bed combustion with limestone.

3.5.1  Low-sulfur Coal

       A coal with sufficiently low sulfur content that when burned in the boiler meets the
applicable SO2 emission standards without the use of additional controls is  sometimes referred to
as "compliance coal." Coals naturally low in sulfur content may be mined directly from the
ground.  Alternatively, the sulfur  content of coal fired in the boiler may be lowered first by
cleaning the coal or blending coals obtained from several sources.  However, burning low-sulfur
coal may not be a technically feasible or economically practical SO2 control alternative for all
boilers.  In some cases, a coal with the required sulfur content to meet the applicable standard
may not be available or cannot be fired satisfactorily  in a given boiler unit design. Even if such a
coal is available, use of the low-sulfur coal that must be transported long distances from the mine
may not be cost-competitive with burning higher sulfur coal supplied by closer mines and using a
post-combustion control device.

       Various coal cleaning processes  may be used  to reduce the sulfur content of the coal.  A
significant portion of the pyritic sulfur minerals mixed with the mined coal  can usually be
                                         5-12

-------
Table 3-4.  Nationwide distribution of existing SO2 emissions controls used for
coal-fired electric utility boilers for the year 1999 as reported in the Part II EPA
ICR data (source:  Reference 6).
SO2 Control Type
Burn low-sulfur coal
("compliance coal")
Wet FGD system
Spray dryer system
Fluidized-bed coal combustion
with limestone (a)
Dry injection
No controls reported (d)
Abbreviation
Code
LSC
FGD
SDA
FBC
Dl

Nationwide Total
Phase II EPA ICR Data
Number
of Boilers
455
173 (a)
52 (b)
37 (c)
2
421
1,140(e)
Percent of
Nationwide
Total Number
of Units
39.9 %
15.2%
4.6%
3.2%
0.2 %
36.9 %
100%
Percent of
Nationwide
Electricity
Generating
Capacity
38.2 %
23.8 %
3.4 %
1.1 %
<0.1 %
33.5 %
100%
   (a) Includes one FBC boiler unit using a wet FGD system.
   (b) Includes three FBC boilers using spray dryer systems.
   (c) FBC boilers using no downstream post-combustion SO2 controls.
   (d) Entry in ICR response indicated none or was left blank.
   (e) Does not include the three IGCC units.

-------
removed by physical gravity separation or surface property (flotation) methods. However,
physical coal cleaning methods are not effective for removing the organic sulfur bound in coal.
Another method of reducing the overall sulfur content of the coal burned in a given boiler unit is
to blend coals with different sulfur contents to meet a desired or target sulfur level.

3.5.2  Fluidized-bed Combustion with Limestone

       One of the features of FBC boilers is the capability to control SO2 emissions during the
combustion process.  This is accomplished by adding finely crushed limestone to the fluidized
bed. During combustion, calcination of the limestone (reduction to lime by subjecting to heat)
occurs simultaneously with the oxidation of sulfur in the coal to form SO2. The SO2, in the
presence of excess oxygen, reacts with the lime particles to form calcium  sulfate.  The sulfated
lime particles are removed with the bottom ash or collected with the fly ash by a downstream PM
control device.  Fresh limestone is continuously fed to the bed to replace the reacted limestone.

3.5.3  Wet FGD Systems

       The SO2 in flue gas can be removed by reacting the sulfur compounds with a solution of
water and an alkaline chemical to form insoluble salts that are removed in the scrubber effluent.
These processes are called "wet FGD systems" in this report. Most wet FGD systems for control
of SO2 emissions from coal-fired electric utility boilers are based on using either limestone or
lime as the  alkaline source. At some of these facilities, fly ash is mixed with the limestone or
lime.  Several other scrubber system designs (e.g., sodium carbonate, magnesium oxide, dual
alkali) are also used by a small percentage of the total number of boilers.

       The basic wet limestone scrubbing process is simple and is the type most widely used for
control of SO2 emissions from coal-fired electric utility boilers. Limestone sorbent is
inexpensive and generally locally available throughout the United States.  In a wet limestone
scrubber, the flue gas containing SO2 is brought into contact with a limestone/water slurry. The
SO2 is absorbed into the slurry and reacts with limestone to form an insoluble sludge.  The
sludge, mostly calcium sulfite hemihydrate and gypsum, is disposed of in  a pond specifically
constructed for the purpose or is recovered as a salable byproduct.

       The wet lime scrubber operates in a  similar manner to the wet limestone scrubber.  In a
wet lime scrubber, flue gas containing SO2 is contacted with a hydrated lime/water slurry; the
SO2 is absorbed into the slurry and reacts with hydrated lime to form an insoluble sludge. The
hydrated lime provides greater alkalinity (higher pH) and reactivity than limestone.  However,
lime-scrubbing processes require appropriate disposal of large quantities of waste sludge.

       The SO2 removal efficiencies of existing wet limestone  scrubbers  range from 31 to
97 percent,  with an average of 78  percent. The SO2 removal efficiencies of existing wet lime
scrubbers range from 30 to 95 percent.  For both types of wet scrubbers, operating parameters
affecting SO2 removal efficiency include liquid-to-gas ratio, pH of the scrubbing medium, and
the ratio of calcium sorbent to SO2.  Periodic maintenance is needed because of scaling, erosion,
                                         5-14

-------
and plugging problems.  Recent advancements include the use of additives or design changes to
promote SO2 absorption or to reduce scaling and precipitation problems.

3.5.4 Spray Dryer Adsorber

       A spray dryer adsorber (sometimes referred to as wet-dry or semi-dry scrubbers) operates
by the same principle as wet lime scrubbing, except that the flue gas is contacted with a fine mist
of lime slurry instead of a bulk liquid (as in wet scrubbing).  For the spray dryer absorber process,
the combustion gas containing SO2 is contacted with fine spray droplets of hydrated lime slurry
in a spray dryer vessel.  This vessel is located downstream of the air heater outlet where the gas
temperatures are in the range of 120 to 180 C (250 to 350 F). The SO2  is absorbed in the slurry
and reacts with the hydrated lime reagent to form solid calcium sulfite and calcium sulfate as in a
wet lime scrubber. The water is evaporated by the hot flue gas and forms dry, solid particles
containing the reacted sulfur. These particles are entrained in the flue gas, along with fly ash,
and are collected in a PM collection device. Most of the SO2 removal occurs in the spray dryer
vessel itself, although some additional  SO2 capture has also been observed in downstream
particulate collection devices, especially fabric filters.  This process produces dry reaction waste
products for easy disposal.

       The primary operating parameters affecting SO2 removal are the calcium-reagent-to-
sulfur stoichiometric ratio and the approach to saturation in  the spray dryer. To increase overall
sorbent use, the solids collected in the spray dryer and the PM collection  device may be recycled.
The SO2 removal efficiencies of existing lime spray dryer systems  range from 60 to 95 percent.

3.5.5 Dry Injection

       For the dry injection process, dry powdered lime (or another suitable sorbent) is directly
injected into the ductwork upstream of a PM control device. Some systems use spray
humidification followed by dry injection. This dry process eliminates the slurry production and
handling equipment required for wet scrubbers and spray dryers, and produces dry reaction waste
products for easier disposal.  The SO2 is adsorbed and reacts with the powdered sorbent. The dry
solids are entrained in the combustion gas stream, along with fly ash, and then collected by the
PM control device. The SO2 removal efficiencies of existing dry injection systems range from
40 to 60 percent.

3.5.6 Circulating Fluidized-bed Adsorber

       In the circulating fluidized-bed adsorber (CFBA), the flue gas flows upward through a
bed of sorbent particles to produce a fluid-like condition in the bed. This condition is obtained
by adjusting gas flow rate sufficiently to support the particles, but not carry them out of the
system.  Characteristics of the bed are high heat and mass transfer, because of high mixing rates,
and particle-to-gas contact.  These conditions allow the CFBA's bed of sorbent particles to
remove a sorbate from the gas stream with high effectiveness. In a CFBA, material is withdrawn
from the bed for treatment (such as desorption) then re-injected into the bed.  Currently, CFBAs
                                         5-15

-------
are used with limestone and ash as sorbents for SC>2 control, but they also have the capability to
remove Hg from the flue gas.  The SC>2 removal ranges for CFBAs from 80 to 98 percent.
3.6 NOX Emission Controls

       Control techniques used to reduce NOX formation include combustion and post-
combustion control measures. Combustion measures consist of operating and equipment
modifications that reduce the peak temperature and excess air in the furnace. Post-combustion
control involves converting the NOX in the flue gas to molecular nitrogen and water using either a
process that requires a catalyst (selective catalytic reduction) or a process that does not use a
catalyst (selective noncatalytic reduction).

       Table 3-5 presents the 1999  nationwide distribution of NOX controls used for coal-fired
electric utility boilers by total number of units and by percentage of nationwide electricity
generating capacity. Approximately one-third of the boilers do not use additional NOX controls.
The other two-thirds of the units use additional controls to meet the applicable NOX standards.
The predominant control NOX strategy is to use one or more combustion control techniques.
Post-combustion NOX reduction technologies (both catalytic and noncatalytic) accounted for only
a small percentage of the NOX emission controls used in 1999 (approximately three percent of the
total units).  However, a number of electric utilities are considering the addition of these types of
controls to their coal-fired boilers to comply with new NOX emission control requirements.

3.6.1  Combustion Controls

       A variety of combustion control practices can be used including low NOX burners,
overfire air, off-stoichiometric firing, selective or biased burner firing, reburning, and
burners-out-of-service.  Control of NOx also can be achieved through staged combustion (also
called air staging).  With staged combustion, the primary combustion zone is fired with most of
the air needed for complete combustion of the coal. The remaining air needed is introduced  into
the products of the partial combustion in a second combustion zone. Air staging lowers the peak
flame temperature, thereby reducing thermal NOX, and reduces the production of fuel NOX by
reducing the oxygen available for combination with the fuel nitrogen.  Staged combustion may be
achieved through methods that require modifying equipment or operating conditions so that a
fuel-rich condition exists near the burners (e.g., using specially designed low-NOx burners,
selectively removing burners from service, or diverting a portion of the combustion air). In
cyclone boilers and some other wet  bottom designs, combustion  occurs with a molten ash layer
and the combustion gases flow to the main furnace; this design precludes the use of low NOX
burners and air staging.  Low-NOx burners may be used to lower NOX emissions by about 25 to
55 percent.  Use of overfire air (OFA) as a single NOX control technique reduces NOX by 15 to
50 percent.  When OFA  is combined with low-NOx burners, reductions of up to 60 percent may
result.  The actual NOX reduction achieved with a given combustion control technique may vary
from boiler to boiler.
                                         5-16

-------
Table 3-5.  Nationwide distribution of existing NOX emissions controls used for
coal-fired electric utility boilers for the year 1999 as reported in the Part II EPA
ICR data (source:  Reference 6).
NOx Control Type
Combustion controls -
low-NOx burners
Combustion controls -
low-NOx burners + overfire air
Combustion controls -
overfire air
Other combustion controls (a)
Selective noncatalytic reduction
Selective catalytic reduction
No controls reported (b)
Abbreviation
Code
CC-LNB
CC-LNB/OFA
CC-OFA
CC
SNCR
SCR

Nationwide Total
Phase II EPA ICR Data
Nationwide
Number
of
Boilers
404
84
79
83
32
6
452
1,140(c)
Nationwide
Percentage
of
Boilers
35.4 %
7.4 %
6.9 %
7.3 %
2.8 %
0.5 %
39.7%
100%
Percent of
Nationwide
Electricity
Generating
Capacity
43.0 %
10.4%
10.6%
5.6 %
0.6 %
1.3%
28.5 %
100%
    (a) Combustion controls other than low-NOx burners or overfire air. The controls include burners-out-of service,
      flue gas recirculation, off-stoichiometric firing, and fluidized-bed combustion.
   (b) Entry in ICR response indicated "none," "not applicable," or was left blank.
   (c) Does not include the three IGCC units.
                                          5-17

-------
       Just as the combustion air to the primary combustion zone can be reduced, part of the
fuel may be diverted to create a secondary flame with fuel-rich conditions downstream of the
primary combustion zone. This combustion technique is termed reburning and involves injecting
10 to 20 percent of the fuel after the primary combustion zone and completing the combustion
with overfire air. The fuel injected downstream may not necessarily be the same as that used in
the primary combustion zone.  In most applications of reburning, the primary fuel is coal and the
reburn fuel is natural gas (methane).

       Other ways to reduce NOX formation by reducing peak flame temperature include using
flue gas recirculation (FOR), reducing boiler load, injecting steam or water into the primary
combustion zone, and increasing spacing between burners. By using FGR to return part of the
flue gas to the primary combustion zone, the flame temperature and the concentration of oxygen
in the primary combustion zone are reduced.

       Temperatures can also be reduced in the primary combustion zone by increasing the space
between burners for greater heat transfer to heat-absorbing surfaces. Another combustion control
technique involves reducing the boiler load. In this case, the formation of thermal NOX generally
decreases directly with decreases in heat release rate; however, reducing the load may cause poor
air and fuel mixing and increase CO and soot emissions.

3.6.2  Selective Catalytic Reduction

       The selective catalytic reduction (SCR) process uses a catalyst with ammonia gas (NH3)
to reduce the NO and NO2 in the flue gas to molecular nitrogen and water. The ammonia gas is
diluted with air or steam, and this mixture is injected into the flue gas upstream of a metal
catalyst bed (composed of vanadium, titanium, platinum, or zeolite).  In the reactor, the reduction
reactions occur at the catalyst surface. The SCR catalyst bed  reactor is usually located between
the economizer outlet and air heater inlet, where temperatures range from 230 to 400 C (450 to
750 F).

3.6.3  Selective Noncatalytic Reduction

       The selective noncatalytic reduction (SNCR) process is based on the same basic
chemistry of reducing the NO and NO2 in the flue gas to molecular nitrogen and water but does
not require the use of a catalyst to prompt these reactions. Instead, the reducing agent is injected
into the flue gas stream at a point where the flue gas temperature is within a very specific
temperature range.  Currently, two SNCR processes are commercially available: the THERMAL
DeNOx7 and the NOXOUT7.  The THERMAL DeNOx7 uses ammonia gas as the reagent and
requires the gas be injected where the flue gas temperature is  in the range  of 870 to 1090 C
(1,600 to 2,000 F).  Consequently, the ammonia gas is injected at a location upstream of the
economizer. However, if the ammonia is injected above 1,090 C (2,000 F), the ammonia will
oxidize and form more NOX. Once the flue gas temperature drops below the optimum
temperature range, the effectiveness of the process drops significantly.  By adding hydrogen gas
or other chemical enhancers, the reduction reactions can be sustained to temperatures down to
                                        5-18

-------
approximately 700 C (1,300 F). The NOXOUT7 is a similar process but uses an aqueous urea
solution as the reagent in place of ammonia.

       Using nitrogen-based reagents requires operators of SNCR systems to closely monitor
and control the rate of reagent injection. If injection rates are too high, NOX emissions may
increase, and stack emissions of ammonia in the range of 10 to 50 ppm may also result.  A
portion (usually around 5 percent) of the NO reduction by SNCR systems results from
transformation of NO to N2O, which is a global warming gas.
3.7 Emission Control Configurations for Coal-fired Electric Utility Boilers

       Mercury can exist in several forms in the flue gas from a coal-fired electric utility boiler
(discussed in Chapter 5).  The distribution of these Hg forms in the flue gas stream can be altered
when reagents for post-combustion pollutant control processes are introduced into the flue gas.
Also, as will be discussed in Chapter 6, some of the existing post-combustion control devices
already in use at coal-fired electric utility power plants to meet PM and SO2 emission standards
also control Hg emissions with varying levels of effectiveness.  Control measures can be
implemented that may enhance the capture of Hg by these control devices. Other Hg control
measures can be implemented in conjunction with control devices already in place at a given
facility. Therefore, understanding which types of post-combustion control devices how electric
utilities currently are implementing at their coal-fired power plants is useful when investigating
potential Hg control measures for these facilities.

       Table 3-6 presents the 1999 nationwide distribution of post-combustion control device
configurations used for coal-fired electric utility boilers. For approximately 70 percent of the
boilers, the only control device used downstream of the furnace is an ESP. If the unit is subject
to SO2 and/or NOX emission limit standards, these units do burn low-sulfur coals to meet the SO2
emission limit and use some type of NOX combustion controls to meet the NOX emission limit.
Approximately 25 percent of the boilers use some combination of post-combustion control
devices. The most common configuration used is an ESP with a downstream wet scrubber for
SO2 control.  Less than 2 percent of the units use a combination of PM, SO2, and NOX post-
combustion control devices.
                                         5-19

-------
Table 3-6. Nationwide distribution of post-combustion emission control
configurations used for coal-fired electric utility boilers for the year 1999 as
reported in the Part II EPA ICR data (source: Reference 6).
Post-combustion
Control Strategy
Post-combustion
PM controls
only
Post-combustion
PM controls
and
SO2 controls
Post-combustion
PM controls
and
NOX controls
Post-combustion
PM controls,
SO2 controls,
and
NOX controls
Post-Combustion Emission Control Device Configuration
PM control
E
S
P
/

/

/

/
/



/

/
/

/
/



/
/
F
F

/
/





/

/



/
/


/
/
/


P
S



/
/




/













M
C





/






/










SO2 control
W
S







/

/
/

/

/





/
/
/
S
D
A








/


/






/
/



D
I













/









NOX control
S
C
R

















/

/

/

S
N
C
R















/
/

/

/

/
Total
Phase II EPA ICR Data
Number
of boilers
791
80
6
5
4
2
2 (a)
133
38
18
13
4
3
2
1
12
11
1
6
4
2
1
1
1,140(b)
Percent of
nationwide
total number
69.4%
7.0%
0.5%
0.4%
0.4%
0.2%
0.2%
1 1 .7 %
3.3%
1.6%
1.1 %
0.4%
0.2%
0.2%
0.1 %
1.0%
0.9%
0.1 %
0.5%
0.4%
0.2%
0.1 %
0.1 %
100%
 (a) Units using hot-side ESP in series with a cold-side ESP. Counted asAmultiple control device combination in Table 3-2.
 (b) Does not include the three IGCC units.
                                        5-20

-------
3.8 References
    1.   Nizich, S.V., A. A. Pope, and L.M. Driver. National Air Pollutant Emissions Trends,
        1900-1998, U.S. EPA and the States: Working Together for Cleaner Air, EPA-454/R-
        00-002 (NTIS PB2000-108054). Office of Air Quality Planning and Standards,
        Research Triangle Park, NC. March 2000.

    2.   French, C.L., W.H. Maxwell, W.D. Peters, G.E. Rice, O.R. Bullock, A.B Vasu, R.
        Hetes, A. Colli, C. Nelson, and B.F. Lyons. Study of Hazardous Air Pollutant Emissions
        from Electric Utility Steam Generating Units  Final Report to Congress, Volume 1.
        EPA-453/R-98-004a.  Office of Air Quality Planning and Standards, Research Triangle
        Park, NC. February 1998. Available at:
        < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

    3.   U. S. Environmental Protection Agency. Air Quality Criteria for Paniculate Matter and
        Sulfur Oxides,  Volunes 1-3, EPA/600/8-82/029a-c. (NTIS PB84-156777).  Office of
        Health and Environmental Assessment, Environmental Criteria and Assessment Office,
        Research Triangle Park, NC. 1982.

    4.   Buonicore, A.J., and W.T. Davis (eds.). Air Pollution Engineering Manual. Air&
        Waste Management Association. Van  Nostrand Reinhold, New York, NY. 1992.

    5.   U.S. Environmental Protection Agency. Air Quality Criteria for Oxides of Nitrogen,
        Volumes 1-3, EPA/600/8-9l/049a-c (NTIS PB92-176361; 95-124525; 95-124517),
        Office of Health and Environment Assessment, Environmental Criteria and Assessment
        Office, Research Triangle Park, NC. 1991.

    6.   U.S. Environmental Protection Agency. Database of information collected in the
        Electric Utility Steam Generating Unit  Mercury Emissions Information Collection
        Effort.  OMB Control No. 2060-0396.  Office of Air Quality Planning and Standards.
        Research Triangle Park, NC.  April 2001.  Available at:
        < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

    7.   Woodward, K. Stationary Source Control Techniques Document for Fine Particulate
        Matter, EPA/425/R-97-001 (NTIS PB99-116493). Office of Air Quality Planning and
        Standards, Research Triangle Park, NC. October 1998.

    8.   Turner, J.H., and J.D. McKenna.  Fabric Filter Baghouses I - Theory, Design, and
        Selection.  ETS, Inc., Roanoke, VA. 1989.

    9.   Center for Air Toxic Metals (C ATM).  Technical Focus - Advanced Hybrid Particulate
        Collector, Fourth Annual Meeting. Grand Forks, ND.  September 16-17, 1997.
                                        5-21

-------
                                      Chapter 4
                              Measurement of Mercury
4.1 Introduction
       Accurate measurements of the various forms of Hg present in flue gas from a coal-fired
electric utility boiler are important: to characterize and determine facility and/or fuel-type
absolute emissions, for understanding the behavior of Hg in combustion processes and
combustion configurations, and to evaluate the removal efficiency of control technologies for Hg.
A variety of measurement techniques, both manual and continuous monitoring, are available for
measuring total Hg and select, speciated forms.  It is the latter need and ability that is most
critical to supporting the understanding of Hg behavior and its control.

       Because of the importance of these measurements, particularly speciated Hg
measurements, research on Hg measurement techniques and performance is an integral
component of the overall Hg control research strategy. The science of speciated Hg
measurements from coal-fired electric utility boilers has only recently been investigated, with the
majority of research on the subject occurring within the last 5 years.  This research has examined
the development and performance of both manual and continuous emission monitor
measurements.  Much of this work began with examining and understanding measurement
performance under very controlled and  simplistic conditions, primarily through the use of
blended gases in a laboratory setting. This afforded the ability to investigate specific
measurement variables and issues individually.  Based on this knowledge, experimentation
expanded to pilot-scale combustion systems where gases/Hg species of interest could be doped
into the combustion system, and measurement performance characterized.  Though still
simplistic, this approach results in a measurement environment that more closely represents real-
world measurement scenarios. Ultimately, investigations moved to pilot-scale coal combustion
test units, and finally to full-scale, field applications. At each step, the measurement complexity
increases.  The complexities associated with the combustion of different coal types, relative
amounts of coal combustion emissions  (e.g.,  SOx, NOx, HC1, C^, PM), and pollution control
device availability and configuration  all have an impact on the ability to perform quality Hg
measurements.1

       The purpose of this chapter is to provide an understanding of the principles, applications,
and limitations of Hg measurement methodologies, particularly with respect to understanding
and interpreting the Part in EPA ICR data. This chapter also serves to introduce principles and
                                          4-1

-------
issues related to Hg CEMs and their use as a valuable research tool.  The following sections
provide a summary of the approaches and state-of-the art of manual and continuous emission
measurement methods and issues associated with performing Hg measurements from coal-fired
electric utility boilers.
4.2 Manual Methods for Hg Measurements

       Manual methods are well established for measuring total Hg emissions from a variety of
combustion sources.  The EPA Method 101 A2 and Method 293 were developed to measure total
Hg emissions (particulate phase and gas phase) from combustion sources such as sewage sludge
incinerators and municipal waste combustors. These reference methods were developed and
used to support total Hg regulatory needs.  A reference method for speciated Hg measurement
does not exist, essentially because there are no regulations requiring speciated Hg emissions
measurements. However, a valid, accepted methodology was needed to characterize the
emissions from coal-fired electric utility power plants to better assess the contribution from this
category as well as potential risk.  The Ontario-Hydro Method 4 (called the OH Method in this
report) presently is the method of choice for measuring Hg species in the flue gas from coal-fired
electric utility plants. This method has been submitted to the American Society for Testing and
Materials (ASTM) for acceptance as a standard reference method.1  The Hg emission data
collected for the Part HI EPA ICR were measured using the OH Method.

       Generally, all sampling trains consist of the same sampling components: a nozzle and
probe operated isokinetically for extracting a representative sample from the stack or duct, a filter
to collect particulate matter, and a liquid solution and/or reagent to capture gas-phase Hg. After
sampling, the filter and sorption media are prepared and analyzed for Hg in a laboratory.
Figure 4-1 shows a diagram of the sampling train used for the OH Method.

       Several of the manual methods, including the OH Method, being developed for speciated
Hg measurements from combustion sources have been adapted/modified from accepted test
methods for measuring total Hg. Measurement of total Hg is based on the concept that all forms
of gaseous Hg can be captured with a strong oxidizing solution such as potassium permanganate.
The speciation is accomplished relying on the solubility and insolubility of the gaseous Hg
species. To speciate gaseous Hg into the oxidized Hg (Hg2+) and elemental Hg (Hg) forms,
multiple solutions/reagents are used.  The Hg2+ form is considered to be readily soluble in
aqueous solutions, while Hg is essentially insoluble.1 When the aqueous solutions are
positioned immediately after the filter, the Hg2+ is captured and the Hg passes through to the
oxidizing solution where it is then captured.  These solutions are analyzed separately to
determine the distribution of oxidized and Hg within the sampling train. Table 4-1  presents a
comparison of the different manual test methods, their configuration, and the solutions used that
have been investigated for measuring speciated Hg.

       The OH Method, along with the other test methods listed in Table 4-1, were  thoroughly
evaluated to determine their appropriateness for performing speciated Hg measurements from
                                          4-2

-------
        Thermocouple
               ^_\_
        Probe-^^E

Wall     Liner-
         Pilot-
                         Manometer
                   Heated
                    Area
                             Filter
                            Holder
                                                                       Thermometer
                            Check
fey fcy jf & J^ M & |?y  , Valve
-U  P pi,  *-  ?!  Sii ?<'  B,  lce
                         U-Bath
             HP  ? it  B-L
                F /*  ^Ksitica
                           Gel
                                               KCi
=1?  rr  ft' r  n  Pii
: ;i  ^i,  u^ : ;   s..:.  ':.$ .
 H*.  h^  ;
w
            Orifice
                                  Bypass    Vacuum
                                   Valve     Gauge
                                                    HNO, /HA,
                                                                 Vaeoym Line

Figure 4-1.  Diagram of sampling train for Ontario-Hydro Method  (source:
Reference 4).
                                          4-3

-------
Table 4-1.  Summary of selected manual test methods evaluated for measurement of Hg in combustion gases.
Manual
Test
Method
EPA Method 29
EPA Method 101 A"
EPA Method 101B
(draft)
Ontario-Hydro Method
Tris-Buffer Method
MESA Method
Sampling Train Configuration
Front-half collection
(PM and Hgp)
glass fiber filter
glass fiber filter
glass fiber filter
glass fiber filter
glass fiber filter
glass wool d
Back-half collection
(gaseous Hg)
impinger solutions
impinger solutions
impinger solutions
impinger solutions
impinger solutions
sorbent beds
Impinger Configuration
(number of impingers - impinger set solution)
First Set
2 HNO3-H2O2
3 H2SO4-KMnO4
2 deionized water
3 KCI
2 tris solution c
2 KCI-soda lime
Second Set
1 dry
none used
1 HN03-H202
1 HN03-H202
2 H2SO4-KMnO4
2 iodated carbon
Third Set
2 H2SO4-KMnO4
none used
2 H2SO4-KMnO4
3 H2SO4-KMnO4
none used
none used
Analytical
Method
CVAA"
CVAA
CVAA
CVAA
CVAA
CVAFS e
a. Test method developed and validated by EPA for measuring Hg emissions from chlor-alkali plants.
b. CVAA = Cold-vapor atomic absorption.
c. Tris solution is tris(hydroxymethyl) aminomethane in a solution of ethylenediaminetetracetic acid in water.
d. Glass wool is only used to trap particulate matter and prevent its carry-over to the sorbent beds. The glass fiber filter in the other test methods is used to
  collect and quantitate particulate matter.
e.  CVAFS = Cold-vapor atomic fluorescence spectrometry.

-------
coal-fired combustion sources.1  The University of North Dakota, Energy and Environmental
Research Center (UND/EERC) performed a thorough, parametric evaluation of these methods
under a variety of laboratory and pilot-scale test conditions, including the combustion of
multiple, representative coal varieties.  A detailed presentation of these tests and their results are
contained in two comprehensive reports.1'5

       Initial experimental work focused on EPA Method 29. These results indicated that
Method 29 exhibited speciation measurement biases under some conditions.1 The testing
expanded to include the Mercury Speciation Adsorption (MESA) Method, Tris-Buffer Method,
draft EPA Method 10IB, and OH Method.1 Pilot-scale coal combustion experiments were then
performed in conjunction with the dynamic spiking of Hg or mercuric chloride into the duct at
various locations within the post-combustion facility. Samples by the respective methods were
collected at sampling locations both upstream and downstream of particulate control systems.
These tests were used to isolate the most appropriate methods for further, more definitive testing.

       It was during the initial dynamic Hg spiking tests that effects from fly ash on the quality
of speciated measurements were observed.  Speciated Hg measurements using the OH Method
and Tris-Buffer Method where the gas sampling and dynamic spiking of Hgtook place at the
inlet and outlet of the PM control device indicated that significant oxidation of the Hg occurred
as a result of reactivity with the coal fly ash (see Figures 4-2 and 4-3).

       The effects of PM on Hg speciation can be significant, particularly at sampling locations
upstream of PM control devices. The flue gas upstream of a PM control device contains a high
concentration of PM (relative to flue gas downstream of a PM control device).  When sampling
takes place upstream of a PM control device, the sampling train filter has the potential to collect
a high loading of fly ash (due to the high concentration of PM in the flue gas).  The speciated Hg
measurement can be biased in two ways. The fly ash on the filter can adsorb gaseous Hg from
the flue gas as it passes through the filter. Reactive fly ashes  can also oxidize gaseous Hg
entering the  filter. When adsorption  and/or oxidation occur across the filter, they alter the
distribution of total Hg and/or gaseous Hg measured.  For example,  if particles on the filter
adsorb gaseous Hg, the filter will contain a greater amount of Hgp than if no adsorption had taken
place; in this case, the sampling-train method will overestimate the amount of Hgp in the flue gas
and underestimate the gaseous Hg, thus, the total distribution of Hg  will be altered.
Alternatively, fly ash on the filter can oxidize gaseous Hg to Hg2+ (without adsorption)
overestimating the amount of Hg2+ in the flue gas. Thus, the  distribution of gaseous Hg will be
altered. The rates of these transformations are dependent on the properties of the coal and
resulting fly ash, the amount of fly ash, the temperature, the flue gas composition, and the
sampling duration.  As a result, the magnitude of these biases varies significantly and cannot be
uniformly assessed.  It is for this reason, that ICR measurements performed at the inlet of PM
control systems possess a large degree of uncertainty. A more detailed  discussion of the
implications of fly ash speciation biases on the ICR data is presented in Chapter 6.

       A final series of pilot-scale tests were conducted to more definitively evaluate the two
most promising methods identified as a result of the initial dynamic  spiking experiments
                                           4-5

-------
   O)
       20  1
       18 
       16 
       14-|
       10 I
   UJ    8 
   O      ^
        6H
        4 
        2 
                                             ; Total vapor-phase Hg as measured by method


                                              Oxidized Hg as measured by method

                                             : Elemental Hg as measured by method


                                              Particle-bound Hg as measured by method


                                          Baseline = 3-run average for Test 26 (PTC run no. 550)

                                          Spiked = 3-run average for Test 28 (PTC run no. 551)
                   8.2 ug/Nm3 Hg Spiked
                                      8.2 ug/Nm3 Hg Spiked
                 Baseline
                                                        8.2 ug/Nm3 Hg Spiked
    EPA
Method 101A
                                       Method
                              MANUAL TEST METHOD
tta#5-Hydro
 Method
Figure 4-2. Comparison of Hg speciation measured by manual test methods from
UND/EERC pilot-scale evaluation tests firing Blacksville bituminous coal and
sampling and spiking Hg at FF inlet (source: graph prepared using test data
presented in Appendix B to Reference 1).
                                        4-6

-------
       201
               Total vapor-phase Hg as measured by method

               Oxidized Hg as measured by method

               Elemental Hg as measured by method

           Baseline = 3-run average for Test 27 (PTC run no. 550)

           Spiked = 3-run average for Test 31 (PTC run no. 552)
                     EPA
                  Method 101A
        ^Method

MANUAL TEST METHOD
Figure 4-3. Comparison of gaseous Hg speciation measured by manual test
methods from UND/EERC pilot-scale evaluation tests firing Blacksville
bituminous coal and sampling and spiking Hg at FF outlet (source: graph
prepared using test data presented in Appendix B to Reference 1).
                                       4-7

-------
discussed above.1 Both Draft EPA Method 101B and the OH Method were selected for formal
EPA Method 301 validation testing. Method 301 is EPA's accepted guidance for validation of
source testing methodologies.6 For these validation tests, all sampling and dynamic spiking of
Hg and HgCl2 into a flue gas stream were performed at the outlet of the high efficiency fabric
filter (FF), while burning a blend of Ohio No. 5 and Ohio No. 6 coals.1 Validation testing was
not performed at the PM control device inlet location.

       A summary of the Method 301 validation results is shown in Table 4-2.  The tests
verified that both the OH Method and the  draft EPA Method 101B achieved acceptable
performance as defined by Method 301.l The precision of the OH Method for total gaseous Hg
was determined to be less than 11 percent relative standard deviation (RSD) for Hg
concentrations greater than 3 |ig/Nm3 and less than 34 percent RSD for Hg concentrations less
than 3 |ig/Nm3.  These values were within the acceptable range, based on the criteria established
in EPA Method 301 (less than 50 percent RSD). In all cases, the laboratory bias for these tests
based on a calculated correction factor was not statistically significant, though some oxidation
(less than 15 percent) of the  Hg spike was observed even when spiking and sampling was done
at the outlet of the fabric filter. The draft EPA Method 101B also met Method 301 validation
requirements, though it did not perform as well as the OH Method.1  As a result, the OH Method
was selected as the most appropriate method for Hg speciation measurements in coal
combustion gases.1

       Final approval by the ASTM of the OH Method as an international test procedure is still
pending as of the date of this report. The OH Method, in  its current draft form, is available from
the EPA Office of Air Quality Planning and Standards (OAQPS) Emission Measurement Center
(EMC).4 The draft version  of the OH Method submitted to ASTM states that the method is
applicable for sampling elemental,  oxidized, and particle-bound Hg at the inlet and outlet of
emission control devices and is suitable for measuring Hg concentrations ranging from
approximately 0.5 to 100 |ig/Nm3.4  Measurement sensitivity/detection levels can be extremely
important where control technology performance is  being determined in relatively low Hg coal
content applications.

       In summary, while several manual methods for Hg speciating measurements exist, the
OH Method is the most thoroughly examined and accepted of these methods, and has met EPA
Method 301 validation requirements. Application to air pollution control device inlet locations
should be considered with caution due to the known catalytic and sorptive effects of certain coal
fly ash PM.  These measurement artifacts  do not affect the use of the OH Method for total Hg
measurements.
4.3 Continuous Emission Monitors for Hg Measurements

       Continuous emission monitors (CEMs) are preferable for multiple reasons to using
manual methods for measuring Hg.  A CEM is capable of providing a real-time or near-real-time
response for Hg measurements. A CEM can be used to obtain continuous Hg measurements
                                          4-8

-------
Table 4-2.  Results from EPA Method 301 evaluation tests for the Ontario-Hydro
Method (sources: References 1 and 4).
Ontario-
Hydro
Method a
Baseline
Hg Spike
(15.0|ig/Nm3)
HgCI2 Spike
(19.9|ig/Nm3)
Total Vapor-Phase Hg
Mean",
|j,g/Nm3
23.35
38.89
42.88
Standard
Deviation
2.05
2.00
2.67
RSDC,
%
8.79
5.13
6.23
Oxidized Hg
Meanb,
|j,g/Nm3
21.24
23.32
40.22
Standard
Deviation
2.13
2.08
2.87
RSD,
%
10.02
8.94
7.14
Elemental Hg
Mean",
|j,g/Nm3
2.11
15.57
2.66
Standard
Deviation
0.65
1.09
0.89
RSD,
%
30.69
6.97
33.31
a.  The correction factor in all cases was not statistically significant and is not shown.
b.  For each mean result, there were 12 replicate samples (four quad trains).
c.  RSD = Relative standard deviation.
                                           4-9

-------
over long periods in time. Conversely, manual methods are capable of only infrequent
"snapshot" Hg measurements over time. As a result, CEMs are able to distinguish the magnitude
and duration of short-term emission characteristics as well as perform long-term emission
measurements to truly characterize a process's temporal emissions. Again, manual methods are
not capable of performing these functions. It is for these reasons that Hg CEMs are extremely
valuable tools supporting the understanding and control of Hg emissions from coal-fired electric
utility power plants. This section discusses the state-of-the-art of using CEMs for Hg
measurements and the associated measurement issues.

       In general, Hg CEMs are a relatively new and yet unproven technology. Although CEMs
that measure total Hg only are used to support regulatory applications in several European
countries, the use of these CEMs is limited. Several total Hg CEMs are available commercially
and are primarily of European origin.7'8 In the United States., Hg CEMs have been limited to
research applications with respect to coal-fired combustion emissions monitoring.  As with the
manual methods, CEMs capable of Hg speciation measurement are of the most value to
supporting research on the characterization and control of Hg emissions from coal-fired electric
utility boilers. The speciating Hg CEMs currently available should be considered prototypes.

       The CEMs being developed for measuring Hg are similar to most other types  of CEMs
used for combustion processes in that the combustion gas sample typically must be extracted
from the stack and then transferred to  the analyzer for detection. However, continuous Hg
monitoring is complicated by the fact that Hg exists in different forms (i.e., Hg, Hg2+, and Hgp)
and that quantitative transport of all these forms is difficult.
       Typically, Hg CEMs measure (i.e., detect) only Hg.  These CEMs measure total Hg
through the use of a conversion system that converts (reduces) the gaseous non-elemental or Hg
.2+
           0
forms to Hg for detection. Mercuric chloride is considered to be the primary oxidized form of
Hg, though recent research suggests that other oxidized forms of Hg do indeed exist.9'10
Although particulate-bound Hg can also be reduced to the gaseous elemental form, particulate
sample delivery issues make this impractical.  As a result, for most commercially available
CEMs, the total Hg measured is in fact total gaseous Hg (TGM).

       The conversion of gaseous, non-Hg is commonly accomplished using a liquid reducing
agent (e.g., stannous chloride).  This technique is least preferable, though more established. The
use of wet chemical reagents is considered to be a limitation to Hg CEM use. The wet chemicals
typically possess corrosive properties and require frequent replenishment. The spent reagents
may possess hazardous properties that result in waste disposal concerns. In addition, the
reducing ability of reagents such as stannous chloride can be affected by high levels of SO2.n

       In addition to the more established wet chemistry conversion methods, dry conversion
methods are also available. These techniques use high temperature catalysts or thermal reduction
units to not only convert non-Hg, but also condition the sample for analysis by removing
selective interferants.  This approach does much to minimize the size of the conversion system as
well as maintenance requirements. However, these systems have not been well characterized for
                                          4-10

-------
coal combustion gas Hg measurement applications.

       Because the particulate form is difficult to transfer and is also often a measurement
interferant, the particulate is typically filtered out and Hgp remains unmeasured.  This could
potentially impart a negative bias to the total Hg measurements.  This bias could be further
amplified as certain types of particulate may actually capture gas-phase Hg. This may not be a
significant issue for sources where Hgp is not present in appreciable quantities, but may be a
significant issue for high particulate-emitting sources (e.g., sources with minimal PM control) or
in cases where the Hg measurements are conducted upstream of PM control devices. Therefore,
the  capability of a CEM to measure Hgp is important and should not be ignored.

       Similarly, there  are known  complications with the quantitative transfer of mercuric
chloride.9 Mercuric chloride (HgQ2) is water soluble and reactive with many surfaces. Losses
due to adsorption are the major concern. As a result, recent emphasis has been placed on
locating the non-Hg conversion systems as close as possible to the source so that the elemental
form is transferred from the source to the detection unit instead of transporting the oxidized
forms long distances.

       In general, Hg CEMs can be distinguished by their Hg measurement detection principle.
Detection systems include: cold-vapor atomic absorption spectrometry (CVAAS); cold-vapor
atomic fluorescence spectrometry (CVAFS); in-situ ultraviolet differential optical absorption
spectroscopy (UVDOAS); and atomic emission spectrometry (AES).1'7'8'9

       The majority of Hg CEM systems employ CVAAS or CVAFS as the detection technique.
These detection techniques are susceptible to measurement interferences resulting from the
presence of common combustion process emissions. Gases such as NOX, SO2, HC1, and C12 can
act as measurement interferants as  well as degrade the performance of concentrating devices
(e.g., gold amalgams).  As such, conditioning systems and/or techniques that remove or negate
the  effects of these interfering gases prior to sample delivery to the detector are required.  The
SO2 is a major spectral  interferant with most CVAA detection systems. The effects of SO2 are
commonly negated through the use of a gold trap.  The sample gas is directed through a gold
trap, where the Hg amalgams with the gold surface. Once the trap is loaded, it is heated and
flushed with a SO2-free carrier gas to the detector. The trapping also serves to improve
measurement sensitivity by concentrating the sample. A trapping device is required of CVAFS
systems to achieve optimum sensitivity; not because of the concentrating aspect, but because the
carrier gas will enable maximum sensitivity. Oxygen and nitrogen have spectral quenching
effects that suppress measurement  sensitivity. Conditioning of the sample gas prior to reaching
the  gold trap is often required. HC1 and NOX in combination can poison the gold surface,
preventing amalgamation with the  Hg. Removal of both or either of these constituents is
required.

       An alternative to the Hg measurement approach is AES. With this technique, the Hg is
ionized by a high-energy source (e.g., plasma) and the emission energy detected.  The advantage
to this technique is that all forms of Hg, including particulate-bound Hg, are capable of being
                                          4-11

-------
ionized and detected. Although this technology is not quite as developed, another major
advantage of AES is that the ionization source and detector can be located directly at the source,
avoiding sample delivery issues. In addition, AES is not as susceptible to spectral interferences
from common flue gas constituents.

       Speciated Hg measurements are important to characterize combustion process emissions
and evaluate Hg control strategies. While there are no commercially available CEMs that
directly measure the various speciated forms of Hg, several total gaseous Hg CEMs, both
commercial and prototype, have been enhanced to indirectly measure speciated Hg (the elemental
and oxidized forms) by determining the difference between Hg and total gaseous Hg. This
difference is recognized as the oxidized form. Separate Hg measurements are made before and
after the conversion step in order to calculate the oxidized form.  This indirect speciation method
is referred to as "speciation by difference."  Based on the current understanding that the oxidized
species of primary interest is mercuric chloride and that mercuric chloride is the dominant form
of oxidized Hg present, the "speciation by difference" technique is considered an acceptable
approach to obtaining speciated Hg measurements.

       A key to performing the speciated Hg measurement is being able to perform reliable Hg
measurements.  The Hg2+ must be removed without adding to the true amount of Hg in the
sampled gas stream. This is often accomplished using a liquid reagent to remove the water-
soluble Hg2+. These reagents also may serve to  neutralize the effects of measurement
interferants.  The greatest concern is the reliability of the speciated Hg measurement.
Measurement artifacts exist that bias the speciation, primarily by over-reporting the level of the
oxidized species. The largest cause of this bias  comes from the reactivity of certain types of PM
(as discussed in Section 4.2). The PM may possess catalytic properties whereby,  at the
conditions of Hg CEM PM filtering environments, Hg can be oxidized across the PM surface.
This is not an issue from a TGM measurement standpoint (unless transport of oxidized Hg is an
issue). However, it may have major implications when  measuring Hg in gas streams possessing
high PM loadings.  This bias is minimized in low PM loading gas streams, consistent with Hg
measurements downstream of PM control devices. Another potentially significant source of
speciated Hg measurement bias takes place in the liquid phase. In combustion gases where C^ is
present, under certain conditions the C12 may react in the liquid phase to oxidize Hg0.12 There is
evidence that this problem can be mitigated.

       As stated previously, the current, primary application of Hg CEMs is as a research
tool/process monitor. Speciating Hg CEMs are  integral to the DOE/EPA/EPRI Hg control
technology development and evaluation research program. These Hg CEMs are used to
characterize existing Hg emissions and distributions, including control technology performance.
More importantly, these speciating Hg CEMs are used to better understand and optimize
potential Hg control technologies so that absolute emissions can be established through OH
sampling. Ultimately, it is desired to accept the quality  and performance of Hg CEMs and
measurements data so as to replace the reliance on OH measurements.  Several pilot-scale and
field tests have been performed specifically to evaluate and determine the measurement
performance of both total and speciating Hg CEMs.
                                          4-12

-------
       Several tests have been conducted specifically to evaluate total Hg CEMs as a compliance
assurance tool.  The first such test, sponsored by the EPA Office of Solid Waste (OSW),
evaluated the performance of three Hg CEMs to measure total Hg emissions from a cement kiln
that burned hazardous waste as a fuel.13 Measurement performance was evaluated following the
proposed "Performance Specification 12  Specifications and Test Procedures for Total
Mercury Continuous Monitoring Systems in Stationary Sources" (PS-12).14 At the time, this was
a relatively new test procedure and had yet to be implemented. In fact, the guidance called for
Hg and HgCl2 gas standards that had yet to be developed and proven. The tests were only
marginally successful. None  of the Hg CEMs tested met the performance test requirements.
Based on the test results, the EPA/OSW concluded that Hg CEMs should not be considered as a
compliance tool for hazardous waste combustors.13  In retrospect, the harshness of the cement
kiln's exhaust gas stream was concluded as a major cause of the test program's lack of
success.8'13 The cement kiln chosen for the EPA/OSW Hg CEM testing was not equipped with
acid gas controls and had relatively high PM loading, resulting in severe interferences and
operational difficulties for the CEMs.

       The DOE Mixed Waste Focus Area (MWFA) has sponsored several tests determining the
measurement performance of a single total Hg CEM under hazardous waste incineration
conditions.15'16  Measurement performance was also evaluated following PS 12. These tests
demonstrated not only Hg CEM performance, but also that additional elements of the PS 12 test
procedures could be implemented. A prototype Hg compressed gas standard was used for the
first time. While these tests have been relatively successful, they are still limited in scope and
application.

       The EPA's Environmental Technology Verification (ETV) Program, in collaboration
with the NRMRL, has completed testing of four commercially available Hg CEMs from three
vendors using the unique capabilities of NRMRL's pilot-scale combustion test facility. These
tests examined the measurement performance of both total and speciated Hg CEMs under two
distinct and diverse combustion conditions. Coal and chlorinated waste combustion conditions
were simulated. These verification tests used PS 12 as guidance, but also considered specific
measurement issues of interest and innovative approaches that better examined these issues.  The
pilot-scale tests were unique in that specific measurement issues were investigated as variables.
The pilot-scale combustion facility enabled independent control  of Hg concentration and species.
As a result, the total Hg measurement could be challenged by the distribution of oxidized and
Hg. Interference flue gas constituents were also independently examined. The ETV testing
made use of several new quality assurance and quality control (QA/QC) tools.  Newly developed
Hg compressed gas standards were used to determine Hg CEM calibration drift and system bias.
As a result, not only were Hg CEMs evaluated, but also improved techniques for evaluating Hg
CEMs were demonstrated.  Performance data for the participating Hg CEMs are not yet
available.

       The UND/EERC has evaluated the performance of Hg CEMs during field tests at eight
different coal-fired electric utility power plants representing facilities that burn lignite,
subbituminous coal, or bituminous coal.11'17  A variety of air pollution control devices and
                                         4-13

-------
configurations were encountered, including ESPs, FFs, wet FGD scrubbers, spray dryer
absorbers, and venturi scrubbers. For these tests, the Hg CEMs evaluated demonstrated the
ability to measure total gaseous Hg within +20 percent of the OH Method measurements. The
field-testing also examined the measurement performance of several Hg CEMs at low stack Hg
emissions levels. These tests demonstrated a distinct advantage of the AF-based systems over
the AA-based system (see Figure 4-4). Below concentrations of 5 |ig/m3, the AA-based systems
exhibited higher signal to noise ratios. At these concentrations, the AF-based systems are a
better choice.

       The EPA/OAQPS/EMC has recently initiated a study to determine the measurement
performance of two commercially available total Hg CEMs at a coal-fired electric utility power
plant. Measurements of performance will be recorded to determine potential monitoring
applications based on measurement performance achieved. Data from this study, and future
studies of Hg CEM measurement performance at additional source categories, should aid in the
future crafting of a performance specification for application of total Hg CEMs to a variety of
different Hg emission source categories.

       Performance testing of Hg CEMs has focused primarily on total Hg CEMs; total Hg
CEMs are the most widely available commercially. However, with respect to the development
and evaluation of Hg control technologies for coal-fired electric utility power plants,  the most
urgent need is for a speciating Hg monitor. As stated previously, the primary use of speciating
Hg CEMs is as a research tool though application as a  process monitor is also appealing. Of
those speciating Hg CEMs in use, most are commercially available total or Hg CEMs modified
for use as a speciating Hg CEM.  Very few speciating Hg CEMs are available commercially.
The major distinction among speciating Hg CEMs is not the analyzer or detection principle, but
the approach for managing potential interferants and method for converting oxidized forms of Hg
to the detectable, elemental form.

       Performance testing of speciating Hg CEMs to support Hg control technology research
has also been performed under pilot- and field-scale operations and research continues in this
area. Work performed by the UND/EERC has also focused on the research and development of
speciating Hg CEMs, particularly the development and evaluation of pretreatment/conversion
systems that can be used with multiple, commercially available Hg CEMs. The EERC has used
speciating Hg CEMs to support field measurement activities in conjunction with OH Method
measurements.  Figure 4-5  compares the measurement performance of several speciating Hg
CEMs to OH Method measurements made during testing at a coal-fired electric utility power
plant.

       A key to assessing measurement performance and validating measurement data quality is
the development Quality  Assurance/Quality  Control (QA/QC) tools such as elemental and
oxidized Hg gas standards. The tools are needed for instrument calibration, continuing
calibration or drift checks, and system bias checks. The EPA/ORD has been active in the
development of both elemental and HgCl2 gas standards. A commercial compressed  gas standard
for Hg has been evaluated for stability and accuracy. While the stability of the Hg compressed
                                         4-14

-------
                    11
       '>
       -T  0.1 -I
                                  4         B
                                   Time, fit
Figure 4-4.  Comparison of total Hg results for CEMs at low Hg levels.

(Reprinted from "State-of-the-Art of Mercury Continuous Emission Monitors for Coal-Fired Systems."
Conference on Air Quality II Mercury, Trace Elements, and Particulate Matter, McLean, VA,  September
2000, by D. L. Laudal and N. B. French, with permission of the University of North Dakota Energy &
Environmental Research Center as copyright owner.)
                                          4-15

-------
         
         D>
         3.
                                                              u.
                             Unit 5
-o-
         H
""                  Hg)
Figure 4-5. Comparison of Hg speciation results for CEMs at low Hg levels.

(Reprinted from "State-of-the-Art of Mercury Continuous Emission Monitors for Coal-Fired Systems."
Conference on Air Quality II Mercury, Trace Elements, and Particulate Matter, McLean, VA, September
2000, by D. L. Laudal and N. B. French, with permission of the University of North Dakota Energy &
Environmental Research Center as copyright owner.)
                                         4-16

-------
gas standard has been confirmed, techniques for establishing the standard's true concentration
have not. As a result, quantitative use of the standard is limited. Similarly, acceptance of a
HgCl2 standard is valuable: this standard is used to assess Hg conversion system effectiveness as
well as overall sampling system delivery efficiency and reactivity, parameters not challenged by
an Hg gas standard.  This is particularly relevant in measurement applications where oxidized
Hg may be the predominant Hg form present. Moreover,  several Hg CEMs vendors have
developed QA/QC capabilities to perform their own instrument calibration drift and system bias
checks from internal Hg gas sources.  These capabilities are needed for routine  daily operational
performance verification.

      In summary, Hg CEMs are currently the tool of choice for evaluating the performance of
candidate Hg control  technologies.  As different control technologies are evaluated, the
associated measurement issues are encountered and addressed. Measurement issues are primarily
associated with the oxidized Hg conversion systems as well as particulate bias effects,
particularly at pollution control device inlet measurement locations. Both wet chemistry and dry
conversion/conditioning systems are used to support these control technology research programs.
It is the  conversion/conditioning system that requires the most attention during operation of Hg
CEM systems.  It is also this frequent need for attention that limits their application to short
measurement intervals. As a result, consideration as a compliance assurance tool is hindered.
Clearly, in order to function as a dedicated process monitor and/or compliance tool, additional
research is needed to  develop and/or evaluate more reliable and less labor intensive Hg
conversion/sample conditioning systems.  These objectives are likely to be furthered as a result
of control technology demonstration and evaluation activities.
4.4 Summary, Conclusions, and Recommendations

       Valid and reliable Hg measurements, by either manual methods or using CEMs, are
critical to the characterization and future reduction of Hg emissions from coal-fired electric
utility power plants. Although these measurement techniques are tools that support a larger
research objective, the quality, applicability, and specificity of these measurements directly
impact the ability to conduct Hg emission control research. Measurement techniques that
determine both the Hg2+ and Hg gaseous forms of Hg are preferred over those techniques that
can measure only total gaseous Hg. Conversely, speciated Hg measurement techniques are more
complex  and more susceptible to measurement biases. Although viable measurement techniques
exist and measurement performance has been demonstrated for certain measurement situations,
acceptable measurement techniques are not available to meet all measurement needs.  Additional
research and development is still needed to enable quality measurements from all necessary
measurement environments.

       The OH Method is the only manual method that is currently recognized in the United
States for speciated Hg measurements in coal combustion gases. The OH Method appears to
provide valid speciation results at sampling locations downstream of PM control devices in
                                          4-17

-------
which most of the fly ash has been removed from the gas stream. However, measurements made
upstream of PM control devices are susceptible to measurement artifacts that bias the
measurements of the different Hg species causing potential uncertainty in results.  However,
these artifacts do not affect the measurement of total Hg.

       A limited number of both private prototype and commercial Hg CEMs are available for
the measurement of total gas-phase Hg and to a lesser extent, speciated gas-phase Hg.  Because
of the diversity and severity of associated measurement environments, numerous measurement
obstacles exist (e.g., PM artifacts, interferences, conversion systems, sample
conditioning/delivery) that have not been adequately addressed, particularly with respect to
speciated measurements. While Hg CEMs are used being used as a tool by researchers, these
devices are not yet suitable for routine Hg monitoring applications at coal-fired electric utility
power plants. As a research tool, Hg CEMs are suitable for short-term measurement needs.
However, the technology has not advanced to the extent that acceptable, long-term measurement
performance has been demonstrated. This must be accomplished for Hg CEMs to be considered
suitable for any purpose beyond use as a research tool. The primary obstacle is the lack of
sample conditioning/conversion systems suitable  for long-term, minimal attention operation.

       Improved methods for the sampling and analysis are critical to support the development
of Hg emission control technologies, for use for Hg monitoring and control (process control), and
for potential use as compliance tools. Specifically, research is  needed to:

       1.  Develop improved sample conditioning/conversion systems (particularly dry, non-wet
          chemical)  capable of long-term, minimal maintenance, operation,

       2.  Develop and demonstrate improved Hg CEM measurement techniques that address
          known and potential measurement obstacles (e.g., PM artifacts, interferences/biases,
          conversion systems, sample conditioning/delivery),

       3.  Develop accepted QA/QC tools (e.g., elemental and oxidized Hg gas standards) for
          validating instrument performance and data quality,

       4.  Develop and verify a manual test method suitable for measuring total and speciated
          Hg at sampling locations upstream of PM control devices,

       5.  Develop and verify a manual test method (e.g., modified OH Method) that can
          simultaneously measure speciated Hg  and other trace metals,

       6.  Develop and demonstrate measurement techniques that are capable of directly
          identifying and quantifying trace levels of individual ionic species of Hg [e.g., HgCl2,
          HgCl, HgS, HgS04, Hg (N03)2],
                                          4-18

-------
       7.  Verify the ability of Hg CEMs to accurately measure total gas-phase Hg and speciated
          gas-phase Hg at diverse stack conditions representative of fuel type and pollution
          control device configurations (e.g., downstream of PM control devices and wet FGD
          scrubbers),

       8.  Verify the ability of Hg CEMs to accurately measure total gas-phase Hg and speciated
          gas-phase Hg at measurement locations upstream of PM control devices,

       9.  Demonstrate Hg CEM long-term monitoring performance, including operational
          requirements,

       10. Identify and evaluate alternative, cost-effective semi-continuous methods for
          measuring the stack emission of total Hg, and

       11. Demonstrate the use of Hg CEMs and semi-continuous monitoring methods as
          potential Hg emission compliance tools.
4.5 References
1.   Electric Power Research Institute. Evaluation of Flue Gas Mercury Speciation Methods,
    Final Report TR-108988, Palo Alto, CA, December 1997.

2.   U.S. Environmental Protection Agency. "Method 101ADetermination of Parti culate and
    Gaseous Mercury Emissions from Stationary Sources." Code of Federal Regulations, Title
    40, Part 61, Appendix B.

3.   U.S. Environmental Protection Agency .  "Method 29Determination of Metals Emissions
    from Stationary Sources." Code of Federal Regulations, Title 40, Part 60, Appendix A.

4.   "Standard Test Method for Elemental, Oxidized, Particle-Bound, and Total Mercury in Flue
    Gas Generated from Coal-Fired Stationary Sources (Ontario-Hydro Method), October 27,
    1999. Available at: < http://www.epa.gov/ttn/emc/prelim/pre-003.pdf >.

5.   Electric Power Research Institute. A State-of-the-Art Review of Flue Gas Mercury
    Speciation Methods,  Final Report TR-107080, Palo Alto, CA, December 1996.

6.   U.S. Environmental Protection Agency. "Method 301 - Field Validation of Pollutant
    Measurement Methods from Various Waste Media." Code of Federal Regulations, Title 40,
    Parts 63,  Appendix A.

7.   Ryan, J.V. "Development and Evaluation of Mercury CEMS for Combustion Emissions
    Monitoring." In Proceedings of 17th Annual Waste Testing and Quality Assurance
    Symposium, Arlington, VA. August 15, 2001.
                                         4-19

-------
8.   French, N., S. Priebe, and W. Haas, Jr. "State-of-the-art mercury CEMS." Analytical
    Chemistry News& Features, 470-475A (July 1, 1999).

9.   Hedges, S., J. Ryan, and R. Stevens. Workshop on Source Emission and Ambient Air
    Monitoring of Mercury, Bloomington, MN, September 13-14, 1999.  EPA/625/R-00/002
    (NTIS PB2001-100963).  National Risk Management and National Exposure Research
    Laboratory, Cincinnati, OH. June 2000.

10.  Brown, T. D., D.N. Smith, R.A. Hargis, Jr., and W.J. O'Dowd. "1999 Critical Review:
    Mercury Measurement and Its Control:  What We Know, Have Learned, and Need to
    Further Investigate," Journal of the Air & Waste Management Association, June 1999.  pp.
    1-97.  Available at:
    < http://www.lanl. gov/proj ects/cctc/resources/pdfsmisc/haps/CRIT991 .pdf >.

11.  Laudal, D. L., T. D. Brown, and P. Chu, "Testing of a Mercury Continuous Emission
    Monitor at Three Coal-Fired Electric Utilities." Paper presented at the 92nd Annual
    Meeting and Exposition of the Air & Waste Management Association, St. Louis, MO, June
    1999.

12.  Linak, W. P., J. V. Ryan, B.S. Ghorishi, and J. O. L. Wendt. Issues Related to Solution
    Chemistry in Mercury Sampling Impingers. Journal of the Air & Waste Management
    Association,  51: 688-698 (2001).

13.  U. S. Environmental  Protection Agency, Draft Mercury Continuous Emissions Monitor
    System Demonstration, Volume I: Holnam, Inc., Hazardous Waste Burning Kiln, Holly Hill,
    SC.  Office of Solid Waste and Emergency Response, Washington, DC.  March 1998.

14.  U. S. Environmental  Protection Agency. Draft Performance Specification 12-
    Specifications and Test Procedures for Total Mercury Continuous Monitoring Systems in
    Stationary Sources,  Office of Air Quality Planning and Standards, Emission Measurement
    Center, Research Triangle Park, NC. Proposed April  19, 1996.  Available at:
    < http://www.epa.gov/ttn/emc/propperf.html >.

15.  Gibson, L. V., J. E. Dunn, R. L.  Baker, W. Sigl, and I. Skegg, "Field Evaluation of a Total
    Mercury Continuous Emission Monitor at a U. S. Department of Energy Mixed Waste
    Incinerator." Paper presented at the 92nd Annual Meeting and Exposition of the Air and
    Waste Management Association, St. Louis, MO, June 1999.

16.  Baker, R. L.  "Are We Ready for Meeting Continuous Emission Monitoring Requirements
    for Total Mercury Combustion Sources?" Paper presented at the 93rd Annual Meeting and
    Exposition of the Air and Waste Management Association, Salt Lake City, UT,  June 2000.
                                         4-20

-------
17.  Laudal, D.L., and N.B. French, "State-of-the-Art of Mercury Continuous Emission Monitors
    for Coal-Fired Systems." Conference on Air Quality II Mercury, Trace Elements, and
    Particulate Matter, McLean, VA, September 2000.
                                         4-21

-------
                                      Chapter 5
                         Mercury Speciation and Capture
5.1 Introduction
       The source of Hg emissions from coal-fired electric utility boilers is the Hg that naturally
exists in coal and is released during the combustion process.  As discussed in Chapter 2, the Hg
content of a coal varies by coal type and where it is mined. When the coal is burned in an
electric utility boiler, most of the Hg bound in the coal is released into the combustion product
gases.  This chapter provides an introduction to Hg chemistry and behavior of Hg as it leaves the
combustion zone of the furnace and  passes in the flue gas through the downstream boiler
sections, air heater, and air pollution control devices. Recent research on Hg chemistry in coal-
fired electric utility boiler flue gas is summarized.
5.2 General Behavior of Mercury in Coal-fired Electric Utility Boilers

       The majority of Hg in coal exists as sulfur-bound compounds and compounds associated
with the organic fraction in coal. Small amounts of elemental Hg may also be present in the
coal. Figure 5-1 presents a simplified schematic of the coal combustion process.  The primary
products of coal combustion are carbon dioxide (CC^) and water (H2O). In addition, as
discussed in Chapter 3, significant quantities of the pollutants sulfur dioxide (862) and nitrogen
oxides (NOx) are also formed. When the coal is burned in an electric utility boiler, the resulting
high combustion temperatures in the vicinity of 1,500 C (2,700 F) vaporize the Hg in the coal
to form gaseous elemental Hg. Subsequent cooling of the combustion gases and interaction of
the gaseous elemental Hg with other combustion products result in a portion of the Hg being
converted to other forms.

       There are three basic forms of Hg in the flue gas from a coal-fired electric utility boiler:
(1) elemental Hg (represented by the symbol Hg in this report); (2) compounds of oxidized Hg
(collectively represented by the symbol Hg2+ in this report); and (3) particle-bound mercury
(represented by the symbol Hgp in this report). Oxidized mercury compounds in the flue gas
from a coal-fired electric utility boiler may include mercury chloride (HgCb), mercury oxide
(HgO), and mercury sulfate (HgSO/j).  Some researchers refer to oxidized mercury compounds
collectively as ionic mercury. This is because, while oxidized mercury compounds may not exist
as mercuric ions in the boiler flue gas, these compounds are measured as ionic mercury by the
                                          5-1

-------
        COAL
        HgS
     Organic Hg
       APCD
       INLET

I       Major
      Mercury
      Species


140 C  HHgC,2
        HgO
       HgS04
Figure 5-1.  Mercury species distribution in coal-fired electric utility boiler flue
gas.
                                        5-2

-------
speciation test method used to measure oxidized Hg (discussed in Chapter 4).  Similarly,
particle-bound Hg is referred to ^paniculate mercury by some researchers.  The term particle-
bound mercury is the preferred and is used in this report to emphasize that the mercury is bound
to a solid particle.

       The term speciation is used to denote the relative amounts of these three forms of Hg in
the flue gas of the boiler.  At present, speciation of Hg in the flue gas from a coal-fired electric
utility is not well understood. A number of laboratory and field studies have been conducted, or
are ongoing, to improve the understanding of the transformation of Hg to the other Hg forms in
the flue gas downstream of the boiler furnace. Data obtained to date indicate that combinations
of site-specific factors affect the speciation of Hg in the flue gas. These factors include:

         Type and properties of the coal burned.
         Combustion conditions in the boiler furnace.
         Boiler flue gas temperature profile.
         Boiler flue gas composition.
         Boiler fly ash properties.
         Post-combustion flue gas cleaning technologies used.

       The current understanding of the mechanisms by which Hg transforms to Hg2+ and Hgp
in the flue gas from coal-fired electric utility boilers is discussed in subsequent sections of this
chapter. It is important to understand how Hg speciates in the boiler flue gas because the overall
effectiveness of different control strategies for capturing Hg often depends on the concentrations
of the different forms of Hg present in the boiler flue gas.  This topic will be discussed in detail
in Chapters 6 and 7.

5.3  Speciation of Mercury

       As mentioned above,  high temperatures generated by combustion in the boiler furnace
vaporize Hg in the coal. The resulting gaseous Hg exiting the furnace combustion zone can
undergo subsequent oxidation in the flue gas by several mechanisms. The predominant oxidized
Hg species in boiler flue gases is believed to be HgCb. Other possible oxidized species may
include HgO, HgSO/i, and mercuric nitrate monohydrate Hg(NO3)2*H2O. The potential
mechanisms for oxidation of Hg in the boiler flue gas include:

         Gas-phase oxidation.
         Fly ash mediated oxidation.
         Oxidation by post-combustion NOX controls.

       Each of these oxidation mechanisms is discussed in the following sections.
                                          5-3

-------
5.3.1  Gas-phase Oxidation

       As mentioned above, Hg in coal is believed to completely vaporize and convert into
gaseous Hg in the combustion zone of a boiler system. As gaseous Hg travels with the flue gas
in the boiler, it can undergo gas-phase oxidation to form gaseous Hg2+, most of which is believed
to be HgCb. Recent researchl has speculated that the major gas-phase reaction pathway to form
gaseous HgCb is the reaction of gaseous Hg with gaseous atomic chlorine (Cl).  The latter is
formed when chlorine in coal vaporizes during combustion.

       At the furnace exit, the temperature of the flue gas is typically in the vicinity of 1400 C
(2552 F). The flue gas cools as it passes through the heat exchanging equipment in the post-
combustion region. At the outlet of the air heater (the last section of heat exchanging
equipment), the temperature of the flue gas ranges from 127 to 327 C (261 to 621F).  Chemical
equilibrium calculations predict that gas-phase oxidation of Hg to Hg2+ starts at about 677 C
(1251 F) and is essentially complete by 427 C (801 F). Based on these results, Hg should  exist
entirely as Hg2+ downstream of the air heater. However, flue-gas measurements of Hg at air
                                  n                                      94-
heater outlets indicate that gaseous Hg  is still present at this location, and that Hg  ranges from
5 to 95 percent of the gas-phase Hg.  These data suggest that, due to kinetic limitations, the
oxidation of Hg does not reach completion.

       As mentioned previously, gas-phase oxidation of Hg is believed to take place via
reaction with gaseous Cl.  At furnace flame temperatures, a major portion of the chlorine in the
coal exists as gaseous chlorine atoms, but as gas cools in post-combustion, the  chlorine atoms
combine to form primarily hydrogen chloride (HC1) and minor amounts of molecular chlorine
(Cb). The rapid decrease in Cl concentration results in "quenched" Hg2+  concentrations
corresponding to equilibrium values around 527 C (981 F).

       Figures 5-2 and  5-3 show predicted distributions of Hg species in  coal-fired electric
utility flue gas as a function of flue gas temperature. The predicted distributions are based on
equilibrium calculations of gas-phase oxidation of Hg in flue gas from the combustion of a
bituminous coall and a  subbituminous  coal2, respectively. Figure 5-2 shows that 80 percent of
gaseous Hg is oxidized to HgCb by 527 C (981F). Figure 5-3 indicates no oxidation of Hg at
or above 527 C (981F).  As mentioned above, the gas-phase oxidation of Hg is believed to be
kinetically limited, proceeding only to equilibrium levels around 527 C (981 F).

       The difference in the equilibrium oxidation levels at 527 C (800 K) in  Figures 5-2 and
5-3 is attributed to the different chlorine levels in the model coals used in the calculations. The
calculated data in  Figure 5-2 are based  on a bituminous coal with a relatively high chlorine
concentration of several hundred parts per million by weight (ppmw). In  contrast, the calculated
data in Figure 5-3 are based on a typical western  subbituminous coal with a relatively low
chlorine content of 26 ppmw.  Research indicates that coals with relatively high chlorine
contents tend to produce more Hg2+ than coals with relatively low chlorine contents.3
                                           5-4

-------
                      600
700
BOO
BOO
1000
1100
                                                K
Figure 5-2.  Predicted distribution of Hg species at equilibrium, as a function of
temperature for a starting composition corresponding to combustion of a
bituminous coal (Pittsburgh) in air at a stoichiometric ratio of 1.2 (source:
Reference 2).
                                     5-5

-------
        100
           500
eoo

                                   Temperature, K
Figure 5-3.   Predicted distribution of Hg species at equilibrium, as a function of
temperature for a starting composition corresponding to combustion of a
subbituminous coal (Powder River Basin) in air at a stoichiometric ratio of 1.2
(source:  Reference 2).
                                    5-6

-------
       In addition to being kinetically limited by Cl concentration, recent research conducted at
EPA has found that gas-phase oxidation of Hg is also inhibited by the presence of 862 and
water vapor. 4 As shown in Figure 5-1, 862 and water vapor are constituents in the flue gas from
coal-fired electric utility boilers. Figure 5-4 shows results from bench-scale experiments
examining the effects of SC>2 and water vapor on the oxidation of gaseous Hg These
experiments were carried out using a simulated flue gas containing a base composition of 40
parts per million by volume (ppmv) Hg, 5 mole % carbon dioxide (62), 2 mole % oxygen (O2\
and a balance of nitrogen (TS^); the temperature of the flue gas was 754 C (1,389 F). The
effects of SC>2, water vapor, and HC1 were studied by adding these constituents to the base flue
gas.  HC1 was added to the simulated flue gas at three concentrations typical of coal combustion
flue gas (50, 100, and 200 ppmv); 862 and water vapor were added with the HC1 at 500 ppmv
and 1.7 mole %, respectively.

       As shown in Figure 5-4, the oxidation of Hg was inhibited by the presence of 862 and
water vapor. HC1 is not believed to react directly with Hg to cause its oxidation (a chlorinating
agent such as atomic chlorine or C\2 is needed).  HC1 may produce trace quantities of the
chlorinating agent in the flue gas.  It is speculated that SC>2 and water vapor may inhibit gas-
phase oxidation of Hg by scavenging the chlorinating agent.

       In addition to experimental studies, research has also been reported on the development
of a kinetic model that is used to better understand the reaction mechanism involved  in gas-phase
Hg oxidation. A detailed chemical kinetics model using a chemical mechanism consisting of 60
reactions and 21 chemical species was developed recently to predict Hg speciation in combustion
flue gas. 5  The speciation model accounts for the chlorination and oxidation of key flue gas
components, including Hg. The performance of the model is very sensitive to temperature. For
low reaction temperatures (< 630 C), the model produced only trace amounts of Cl and C\2 from
HC1, leading to  a drastic under-prediction of Hg chlorination compared with experimental data.
For higher reaction temperatures, model predictions were in good accord with experimental data.
For conditions that produce an excess of Cl and C\2 relative to Hg, chlorination of Hg is
determined by the competing influences of the initiation step, Hg + Cl   HgCl, and  the
recombination reaction, 2C1   C^.  If the Cl recombination is faster, Hg chlorination will
eventually be determined by the slower pathway Hg + C\2  >
       Another attempt has been made to formulate an elementary reaction mechanism for gas-
phase Hg oxidation.6 The proposed eight-step Hg oxidation mechanism quantitatively describes
the reported extents of Hg oxidation for broad ranges of HC1 and temperature.  In the proposed
mechanism, Hg is oxidized by a Cl atom recycle process, and, therefore, the concentrations of
both Cl and C\2 are important. Once a pool of Cl atoms is established, Hg is first oxidized by Cl
into HgCl, which, in turn, is oxidized by C\2 into HgC^. The second step regenerates Cl atoms.
Since the concentrations of Hg species are small in coal combustion flue gases, independent
reactions establish and sustain the pool of Cl atoms. The pool is governed by the chemistries of
moist CO oxidation, Cl  species transformations, and nitrogen oxide (NO) production.  The model
predictions show that O2 weakly promotes homogeneous Hg oxidation, whereas moisture is a
strong inhibitor as it inhibits the decomposition of HC1 to C^. NO was identified as an effective
inhibitor for Hg oxidation through its effect on reducing the concentration of hydroxyl (OH)
                                          5-7

-------
                30
500 ppmv SO2
1.7% H2O

no SO2/H2O
                                             500 ppmv SO2
                         50          100          200
                           HCI Concentration (ppmv)
Figure 5-4. Effects of SO2 and water vapor on the gas-phase oxidation of Hg at
754 C and at three different HCI concentrations.
                                  5-8

-------
in the flue gas. The formation of HOC1 from OH and Cl is essential for the oxidation of Hg,
which oxidizes HgCl into HgCl2 and OH. The elimination of OH via OH+NO+M = HONO+M
is believed to inhibit Hg oxidation.

5. 3.2  Fly Ash Mediated Oxidation

       In fabric filtration, flue gas penetrates a layer of fly ash as it passes through the filtering
unit.  The intimate contact between the flue gas and the fly ash on the filter provides an
opportunity for the latter to oxidize some of the incoming gaseous Hg. However, this
phenomenon does not occur across ESPs because the flue gas does not pass through a collected
layer of fly ash (see Chapter 3 for a description of the operation of FFs and ESPs).

       Certain fly ashes have been shown to promote oxidation of Hg across a FF more actively
than others.  For example, fly ashes from bituminous coals tend to oxidize Hg at higher rates
than fly ashes from subbituminous coals and lignite. Differences in oxidation appear to be
attributable to the composition of the fly ash, the presence of certain flue gas constituents, and
the operating conditions of FFs.

       Bench-scale tests were conducted at EPA to investigate the effects of fly ash composition
and flue gas  parameters on the oxidation of gaseous Hg0.4'7  In these experiments, a simulated
flue gas containing Hg (and  other species) was passed through a fixed bed of simulated or actual
coal fly ash,  and oxidation of Hg was measured across the reactor. Experimental results
indicated two possible reaction pathways for fly-ash-mediated oxidation of Hg. One possible
pathway is the oxidation of gaseous Hg by fly ash in the presence of HC1, and the other is the
oxidation of gaseous Hg by fly ash in the presence of NOx.  The research also reflected that the
iron content  of the ash appeared to play a key role in oxidation of Hg. This EPA research is
described in  the ensuing paragraphs.

       Coal  fly ash is a mixture of metal oxides found in both crystalline and amorphous forms.
Glasses are common ash constituents, composed primarily of the oxides of silicon and aluminum
(known as aluminosilicate glasses) that can contain a significant amount of cations such as iron,
sodium, potassium, calcium,  and magnesium.  Iron oxide (in the form of magnetite or hematite)
is also as commonly found in ash as calcium oxide and calcium sulfate. In the presence of
sufficiently high flue-gas concentrations of HC1 or Cb, metallic oxides in fly ash may be
converted to metal chlorides such as cuprous chloride (CuCl).  Three-component model fly ashes
were prepared by adding Fe2O3 or CuO at various weights to a base mixture of A^Os and SiO2.
An additional three-component fly ash was prepared by adding CuCl to a base mixture of A^Os
and SiO2. Municipal waste combustion fly ashes contain significant amounts of copper
compared to coal combustion fly ashes that contain only trace levels of copper.  Model fly ashes
were prepared and tested in order to understand the effect of differences in copper content on the
oxidation of Hg. Four-component fly ashes were prepared by adding various weights of CaO,
and Fe2O3 or CuO to a base mixture of A12O3 and SiO2. Actual coal fly ashes were obtained
from the combustion of three different coals (two subbituminous and one bituminous) from a
pilot-size, pulverized-coal-fired furnace.
                                          5-9

-------
       Model flue gas compositions were simulated to represent the temperature and
composition of coal-fired electric utility flue gas as it enters a FF. The temperature of coal
combustion flue gas as it enters a FF typically ranges from 150 C (302 F) to 250 C (482 F).
Potentially important flue gas species (in terms of Hg oxidation) include chlorine (primarily in
the form of HC1 at FF temperatures), NOx (primarily in the form of NO at FF temperatures), 862,
and water vapor. The base flue gas consisted of 40 ppbv Hg, 2 mole % 62, 5 mole % CC>2, and
the balance N2 at a temperature of 250 C (482 F). HC1 (50 ppmv), NO (200 ppmv), SO2
(500 ppmv), and/or water vapor (1.7 mole %) were added to the base gas to determine their
effect on oxidation. About 10 percent of NC>2 (10 ppmv) was measured when 200 ppmv of NO
was added to the base flue gas which contains 2 mole % of 2. The mixture of NO and NO2 in
flue gas is referred to collectively as NOx.  Table 5-1 shows the simulated and actual fly ashes
and simulated flue gas tested.

       Oxidation Behavior of Model Fly Ashes. HC1 and NOx were identified as the active
components  in flue gases for the oxidation of Hg. NOx were more active than HC1. Cupric oxide
(CuO) and ferric oxide (Fe2O3) were identified as the active components in model fly ashes for
Hg oxidation. In the presence of NOx, inert components of model fly ashes such as alumina
(A12O3) and silica (SiO2) appeared to become active in oxidation of Hg.  Steady-state oxidation
of Hg promoted by the four-component model fly ashes (containing calcium oxide, CaO) was
reached at much slower rates compared to those obtained using the three-component model fly
ashes that contained no CaO (Figures 5-5 and 5-6). The partial removal of gas-phase HC1 by
CaO in the CaO-containing model fly ashes may have reduced the available chlorinating agent
and resulted in slower oxidation of Hg.

       Oxidation Behavior of Actual Coal Fly Ashes. As shown in Table 5-1, the Blacksville fly
ash (derived from a bituminous coal) completely oxidized Hg in the presence of NO (base +
NO), but showed little oxidation in the presence of HC1 (base + HC1). 7  The Comanche fly ash
(derived from a subbituminous coal) did not oxidize Hg in the presence of NO or HC1.  The
Absaloka coal (derived from a subbituminous coal) showed 30 to 35 percent oxidation of Hg in
the presence of NO, but no oxidation in the presence of HC1.  It is believed that the high
reactivity of the Blacksville coal in NO is related to its relatively high Fe2O3 concentration (22
percent); this observation is in agreement to that seen for the high iron (approximately 14
percent) three- and four-component model fly ashes.

       More tests were conducted recently at EPA on actual fly ash samples with different coal
ranks and iron contents in order to get a better understanding of the effects of iron in coal fly
ashes on speciation of Hg. 8 It was observed that one subbituminous (3.7 percent iron) and three
lignite coal fly ash (1.5 to 5.0 percent iron) samples tested with low iron content did not oxidize
Hg in the presence of NO and HC1.  However, a bituminous coal fly ash sample (Valmont
Station) with a low iron content (2.3 percent iron) completely oxidized Hg in the presence of
NO and HC1. It was also found that, upon adding Fe2O3 to the low iron content subbituminous
and lignite fly ash sa
significant Hg oxid
iron-doped samples.
and lignite fly ash samples to reach an iron content similar to that of the Blacksville sample,
significant Hg oxidation reactivity was measured (33 to 40 percent oxidation of Hg) for these
                                         5-10

-------
 
 I
 2
 'x
 e
m
 I
               80
               80
              40
               20
                                  *  *  **   ^p^^
                                             *
                                  10     20        50
                                          tint e |m inj
                                                100
200
                    3-Component: silica/alum in a (3.5/1) and 14 wt% Fe203
                    4-Component: silica/alum in a (3.5/1), 13 vt% Fe203, and 8 wt% CaO
Figure 5-5. Hg oxidation in the presence of the three- and four-component model
fly ashes containing iron at a bed temperature of 250 C (source: Reference 4).
                                    5-11

-------
              100  -
 .2
 S3
 "SS
  INK
 X
 o
s>
               60
               3D
                               4-component
                                  10     20        SO
                                               100
200
                                           ) and 1    CyQ
                                        (3.511), 1    CyO,    B
Figure 5-6. Hg oxidation in the presence of the three- and four-component model
fly ashes containing copper at a bed temperature of 250 C (source: Reference 4).
                                    5-12

-------
Table 5-1.  Percent oxidation of Hg by simulated and actual coal-fired electric
utility boiler fly ash (source: Reference 4).
Fly Ash Composition
(by weight percentages)
2-Component Model Fly Ash
22% AI2O3 + 78% SiO2
% Oxidation of Hg by fly ash
Base"
Base
+ HCI
Base
+
HCI,
S02
Base
+
HCI,
S02,
HP
Base
+ NO
Base
+ NO,
S02

b
0


39
4
5-Component Model Fly Ashes
19% AI2O3, + 67% SiO2 + 14% Fe2O3
22% AI2O3 + 77% SiO2 + 1% Fe2O3
22% AI2O3 + 78% SiO2 + 0.1% Fe2O3
22% AI2O3 + 77% SiO2 + 1% CuO
22% AI2O3 + 78% SiO2 + 0.1% CuO
22% AI2O3 + 72% SiO2 + 7% CaO
22% AI2O3 + 78% SiO2 + 0.1% CuCI
4-Component Model Fly Ashes
21% AI2O3 + 71% SiO2, + 1% CuO + 7% CaO
18% AI2O3, + 63% SiO2 + 13% Fe2O3 + 6% CaO
Actual Fly Ash Samples
Blacksville coal fly ash (bituminous)
22% Fe2O3, 6% CaO
Comanche coal fly ash (subbituminous)
5% Fe2O3, 32% CaO
Absaloka coal fly ash (subbituminous)
4% Fe2O3, 24% CaO
0




0
87
92
67
15
93
92
0
77
88
43

89
86
13

54
37

84
63
14
23
93
48
11
70
35
0.86

80
26
3
16
3
13




91
87
82
93
43
49








6
0







100
0
30-35



   (a)  Base gas consisted of 40 ppbv Hg, 2 mole% O2, 5 mole% CO2, and balance N2 at a temperature of 523 K.
       HCI, NO, SO2, and water vapor were added to the base gas in the following concentrations 50 ppmv, 200
       ppmv, 200 ppmv, and 1.7 mole%, respectively.
   (b) Blank cells mean test not conducted.
                                           5-13

-------
       The physical, chemical, and carbon properties of the Blacksville and Valmont samples
were also characterized. It was found that the two fly ash samples have different unburned
carbon contents (3.4 percent for Valmont and 16.8 percent for Blacksville). Based on this
finding, it appears that iron content may not be the only ash-related factor that affects the Hg
oxidation reactivity of bituminous coal fly ashes. The effect of physical properties, such as
surface area, and the effects of chemical properties, such as  sodium content and alkalinity, in the
fly ash may also determine the propensity of different fly ashes to oxidize Hg in flue gas.

       Research for obtaining a better understanding of the  roles of NOx and Fe2C>3 in the
heterogeneous oxidation of Hg was reported recently by UND/EERC.9 In UND/EERC's
reported research, the effects of NOx and hematite (oc-Fe2O3) on Hg transformations were studied
by injecting them into actual coal combustion flue gases produced from burning bituminous
(Blacksville), subbituminous (Absaloka), and lignite (Falkirk) coals in a 7-kW combustion
system. It was found that the Blacksville fly ash has high Fe2C>3 content (12.1 percent),  and the
Absaloka and Falkirk fly ashes have significantly lower Fe2C>3 contents (4.5 and 7.9 percent,
respectively). Portions of the Fe2C>3 in Blacksville and Falkirk fly ashes are present as
maghemite (7^6263), and a portion of the Fe2C>3 in Absaloka fly ash is present as hematite (oc-
Fe2O3).  The flue gas generated from the combustion of Blacksville coal contained Hg2+ as the
predominant Hg species (77 percent), whereas Absoloka and Falkirk flue gases contained
predominantly Hg (84 and 78 percent, respectively). Injections of NC>2 (80 to 190 ppm) at 440
to 880 C and oc-Fe2O3 (6 and  15 percent) at 450 C into Absoloka and Falkirk coal  combustion
flue gases did not change Hg speciation.  The UND/EERC researchers suggested that the lack of
                      n      9+
transformation from Hg to Hg  in the 7-kW combustion system was possibly due to
components of either Absoloka and Falkirk coal combustion flue gases, or their fly ashes,
inhibiting the oc-Fe2O3 catalyzed heterogeneous oxidation of Hg by NOx.  The researchers also
                              94-
believed that an abundance of Hg  in Blacksville coal combustion flue gas and y-Fe2O3 in the
corresponding fly ash, and the inertness of injected oc-Fe2O3 with respect to heterogeneous Hg
oxidation in Absoloka and Falkirk flue gases, are indications that y-Fe2O3 rather than oc-Fe2O3
catalyzes Hg2+ formation.

       A study of the role of fly ash in the speciation of Hg in coal combustion flue gases was
reported by Iowa State University.10 In this study, bench-scale laboratory tests were performed
in a simulated flue gas stream  using two fly ash samples obtained from the ESPs of two full-
scale coal-fired electric utility boilers.  One fly ash was derived from burning a western
subbituminous coal (Powder River Basin) while the other was derived from an eastern
bituminous coal (Blacksville). Each of the two  samples was separated into three subsamples
with particle sizes greater than 10, 3, and 1 |im using three cyclones.  The amount of sample
collected in these three size ranges was 85 to 90 percent, 10 to!5 percent, and 1 percent of the
total ash, respectively.  Only the two largest sized subsamples were tested for Hg oxidation
reactivity. The Blacksville sample was also separated into strongly magnetic (20 percent),
weakly magnetic (34 percent), and nonmagnetic (46 percent) fractions using a hand magnet for
testing Hg oxidation reactivity of the individual fractions. Since  magnetism of the fly ash
samples is contributed mainly by iron oxides in the samples, the iron oxide content of the
magnetically separated samples is in the following order: strongly magnetic > weakly magnetic >
nonmagnetic.  The low iron content PRB fly ash is nonmagnetic and was not magnetically
                                          5-14

-------
separated for testing. Scanning electron microscopy with energy-dispersive x-ray analysis
(SEM-EDX) was used to examine the surface morphology and chemical composition of the fly
ash samples. X-ray diffraction (XRD) was also used to examine the mineralogical composition
of the whole and fractionated fly ash samples.  XRD identifies only crystalline components of
the  samples. This is important since coal combustion fly ashes typically contain a considerably
amount of glassy, amorphous material.

       It was observed that, although the fly ashes tested were chemically and mineralogically
different, there were no large differences in the catalytic potential for oxidizing Hg.10  The
Blacksville fly ash tended to show somewhat more catalytic reactivity (16 to 19 percent Hg
oxidation) than the PRB fly ash (4 to 10 percent Hg oxidation).  The researchers of this project
suggested that the difference in reactivity could be due largely to the larger surface area (3.4
m2/g) of the Blacksville fly ash compared to  that (1.5 m2/g) of the PRB fly ash. It was found
from the SEM-EDX analyses that the iron-rich (highly magnetic) phases in the greater than 10
|im size fraction  of the Blacksville sample contained about 25 percent (atomic) Fe, 10 percent
each of Al and Si, 2 percent Ca, and lesser amounts of Na, S, K, and Ti. The nonmagnetic
Blacksville fly ash fraction in the greater than 10 |im size range contained only 4 percent Fe,  10
percent Al, 20 percent Si, and lesser amounts of Na, S, K, and Ti. For the PRB fly ash (all
nonmagnetic), both the greater than 10 |im and greater than 3 |im fractions contained about 3
percent Fe, 10- 20 percent Al and Si, about 10  percent Ca, and 2 percent or less of Mg, S, K, and
Ti.  The XRD results showed that the whole  Blacksville ash contained primarily quartz (SiO2),
mullite (3Al2O32SiO2),  magnetite (FesO/i), hematite (Fe2C>3), and a trace of lime (CaO). The
PRB fly ash contained mostly quartz and lesser amounts of lime, periclase (MgO), and calcium
aluminum oxide  (CasAbOe). No magnetite or hematite was found in this ash.  It is interesting to
note that the nonmagnetic fractions actually  showed substantially higher amounts of oxidized Hg
than the magnetic fractions. The reported test results of this study indicated that the nonmagnetic
fraction resulted  in 24 percent of the Hg being  oxidized, while 3 percent of the Hg oxidized when
using the magnetic ash.  It has been suspected that the magnetic (iron-rich) fraction in fly ash
would be more catalytic  than the nonmagnetic  (aluminosilicate-rich) fraction because of its
mineralogy (predominantly iron oxides), and possibly because the magnetic phase tends to be
enriched in transition metals that could also serve as Hg oxidation catalysts. However, under
the  experimental conditions employed in this study, the test results do not support this.  It was
found that the surface area of the nonmagnetic fraction is about four times that of the magnetic
fraction. From this study it appears that surface area is a dominant factor in determining the
ash's Hg oxidation reactivity.

       Because major differences were not observed with the two fly ashes, a set of tests
involving a full factorial design was conducted using only the Blacksville fly ash in order to
apply statistical techniques for identifying the important factors in determining Hg oxidation.10
The statistical analysis results indicated that  the composition of the simulated flue gas used in the
tests and whether or not ash was present in the gas stream were the two most important factors.
The presence of HC1, NO,  NC>2, and SC>2 and all two-way gas interactions of the four gases listed
above were found statistically significant for Hg oxidation. The HC1, NC>2, and  SC>2 appeared to
contribute to Hg oxidation, while the presence of NO appeared to suppress Hg oxidation. NO2
was found to be the most important of the four reactive gases tested; next were HC1 and NO.
                                          5-15

-------
However, the effect of NO depended on whether NC>2 was present. Although the presence of
NC>2 was statistically significant as a main factor, it was found more important in its interactions
with other gas components. Based on the statistical analysis results, the researchers of this
project concluded that the interactions of flue gases with fly ash to cause Hg oxidation are
extremely complex, and the difficulty in understanding the Hg chemistry in coal combustion flue
gases is not surprising. It is noted that the EPA study showed significant Hg oxidation reactivity
of the Blacksville ash, while studies at UND/EERC and Iowa State University show little Hg
oxidation reactivity of Blacksville ash. Since the ash samples used in the above studies were
generated at three different plant operating conditions, these conditions may play an important
role in contributing to the reactivity of the ashes.

5.3.3 Oxidation by Post-combustion NOx Controls

       There are indications that post-combustion NOx controls SCR and SNCR may oxidize
some of the Hg in the flue gas of a coal-fired electric utility boiler. The research on this issue is
ongoing. For current understanding of this subject, the reader is referred to Chapter 6.

5.3.4 Potential Role of Deposits, Fly Ash, andSorbents on Mercury Speciation

       Gaseous Hg (both Hg and Hg2+) can be adsorbed by the solid particles in the coal-fired
electric  utility boiler flue gas.  Adsorption is the phenomenon where a vapor molecule in a gas
stream contacts the surface of a solid particle and is held there by attractive forces between the
vapor molecule and the solid.  Solid particles are present in all coal-fired electric utility boiler
flue gas as a result of the ash that is generated during combustion of the coal.  Ash that exits the
furnace  with the flue gas is called fly ash. Other types of solid particles may be introduced into
the flue  gas stream (e.g., lime, powdered activated carbon) for pollutant emission control. Both
types of particles may adsorb gaseous Hg in the boiler flue gas. This section addresses the
adsorption of gaseous Hg by fly ash. Adsorption of Hg by sorbent particles introduced into the
flue gas stream and subsequently captured in a downstream PM control device is discussed in
Chapter 6 as related to specific control technologies that  may be implemented to increase overall
Hg removal from the boiler flue gas.

       Gaseous Hg can be adsorbed by fly ash in the flue gas (sometimes called "in-flight"
adsorption).  In-flight adsorption of gaseous  Hg by fly ash occurs in the post-combustion region
where the flue gas contains its highest concentration of fly ash (i.e., prior to the first PM control
device).  The  type of coal from which a fly ash originates appears to strongly influence its ability
                       11             19
to adsorb Hg. Pilot-scale   and field data  have indicated that fly ashes from subbituminous
coals (specifically, those from the Powder River Basin in Wyoming) adsorb more gaseous Hg
than fly ash from lignite and bituminous coals. Test data show 30 percent in-flight adsorption of
gaseous Hg by fly ashes from boilers burning these subbituminous coals compared to 10 to 20
percent  adsorption by the fly ashes from boilers burning  lignite or bituminous coals. It has been
suggested that the measured removals of Hg by fly ash can be inflated based on the sampling
method, but in most cases are below 15 percent. General trends indicate that in-flight field
capture  of Hg from combustion of subbituminous coals is higher than from combustion of
bituminous coals.13
                                          5-16

-------
       The carbon content of fly ash is another parameter that may influence adsorption of
gaseous Hg (the carbon in fly ash is unburned coal). Conditions that result in increased amounts
of carbon in fly ash tend to increase the amount and subsequent capture of particle-bound Hg.
Hg has been found to concentrate in the carbon-rich fraction of fly ash.14'15  For similar coals,
both laboratory16 and pilot- and large-scale data u have shown a positive correlation between
adsorption of gas-phase Hg and carbon content in fly ash. A research project conducted at full-
scale coal-fired electric utility boilers in Colorado indicates that certain fly ashes adsorb
significant levels of Hg from flue gas. Chapter 7 describes the methodology and results of this
study in detail.  Many of these fly ashes have carbon content greater than 7 percent, but one low-
carbon content fly ash has also been identified.  This research project and the possibility of using
fly ash re-injection for Hg control is discussed in Chapter 6.

       Gaseous Hg also can be adsorbed by fly ash collected on the surface of a FF. In a FF,
there is contact of gaseous Hg in the flue gas with the collected layer of fly ash on the FF bags as
the gases flow through the FF.  Pilot-scale tests of a low-carbon fly ash (less than 0.5 percent
carbon) showed that the fly ash adsorbed 65 percent of the gaseous Hg entering  a FF; the data
indicate that fly ash properties other than just carbon content may affect adsorption.  The tested
fly ash was produced from the combustion of a subbituminous coal from the Powder River Basin
in Wyoming. Western subbituminous coals generally contain high concentrations of CaO and
tend to adsorb high levels of Hg. At this time, the mechanisms by which these Western coals
adsorb Hg are not known; however, the CaO content may be a factor.  It has been shown in a
pilot-scale study that combustion of western coals tends to produce relatively high particle-bound
Hg emissions.17
5.4 Capture of Mercury by Sorbent Injection

       Mercury can be captured and removed from a flue gas stream by injection of a sorbent
into the exhaust stream with subsequent collection in a PM control device such as an electrostatic
precipitator or a fabric filter.  The implementation of this type of Hg control strategy requires the
development, characterization, and evaluation of low-cost and efficient Hg  sorbents.
Experimental methods for characterization and evaluation are presented below.  Further, efforts
to develop better sorbents, with greater capacity and lower cost, are also discussed.

5.4.1  Sorbent Characterization

       Sorbents are characterized by their physical and chemical properties. The most common
physical characterization is surface area.  The interior of a sorbent particle is highly porous. The
surface area of sorbents is determined using the Brunauer, Emmett, and Teller (BET) method of
             1 &
N2 adsorption.   Nitrogen is adsorbed at the normal boiling point of-195.8 C  and the surface
area is determined based on mono-molecular coverage.  Surface areas of sorbents range from 5
m2/g for Ca-based sorbents to over 2000 m2/g for highly porous activated carbons. Mercury
capture often increases with increasing surface area of the sorbent.  However, recent research19
has suggested that pore surface area in the micropores is more important than the total surface
area for the removal of part per billion concentrations of Hg from coal combustion flue gases.
                                          5-17

-------
       Particle size distribution is another physical characteristic that is used to describe
sorbents. Activated carbons that are used for Hg control are powdered with a size on the order of
44 |im or less. Particle size is measured using sieves or a scanning electron microscope (SEM).
Generally, the smaller the particle size of an activated carbon, the better the access to the surface
area and the faster the rate of adsorption kinetics.  Careful consideration of particle size
distribution can provide significant operating benefits, both in fabric filter applications, where
pressure drop must be considered, and in ESP (or duct injection) applications, where mass
transfer limitations in the short residence time mean that adsorption is a function of sorbent
particle size.

       Determination of the pore size distribution of an activated carbon is an extremely useful
way of understanding the performance characteristics of the material. Pore sizes are based on the
diameter of the pore and are categorized using the following IUPAC conventions: micropores
<2 nm, mesopores 2-50 nm, and macropores >50 nm.  Micropore volume can be estimated from
CC>2 adsorption at 273 K using the Dubinin-Radushkevich (DR) equation. Total pore volume
can be determined using N2 adsorption.

       Some of the chemical properties of activated carbons that influence Hg capture include
sulfur content, iodine content, chlorine content, and water content. Functional groups of a
sorbent have been shown to play an important role in adsorption behavior. Many carbon-oxygen
functional groups have been identified in activated carbon including carbonyl, carboxyl, quinone,
lactones, and phenol  groups. Many methods have been used to study the functional groups
present in carbonaceous materials including neutralization of bases, direct analysis of the oxide
layer by chemical reaction, infrared spectroscopy, and x-ray photoelectron spectroscopy. For
example, specific surface oxygen functional groups can be estimated by using the data measured
from the base titration based on the following assumptions: NaHCOs titrates carboxyl groups;
NaOH titrates carboxyl, lactone, and phenol groups; CC>2 is a decomposition product of carboxyl
                                                                               90 	
and lactone groups; and CO is a decomposition product of phenol and carbonyl  groups.  The
NaOH and HC1 titration values can estimate the acidity and basicity of a carbon, respectively.

5.4.2  Experimental Methods Used in Sorbent Evaluation

       In order to evaluate the performance of a specific Hg sorbent, several types of
experimental reactors are used. The first step is testing in a bench-scale reactor system, which
may be a fixed-bed, entrained-flow, or a fluidized-bed system.  Sorbents that perform well in
bench-scale tests are  then tested in a pilot-scale system and may eventually be tested in a full-
scale system.  These  systems are discussed below.

5.4.2.1  Bench-scale Reactors

       Bench-scale reactors are the smallest category of reactors, hence the term "bench-scale."
There are several types of bench-scale reactors that are used to evaluate Hg sorbents.  The first
type that will be discussed is a fixed-bed or packed-bed system.  This type of reactor simulates
Hg capture that would occur in a FF. Another type of bench-scale reactor is an entrained-flow
                                          5-18

-------
reactor, which simulates in-flight capture of Hg upstream of an ESP. It is important to highlight
the major differences between these two reactors as shown in Table 5-2.

       Fixed-bed Reactor.  A schematic of the experimental apparatus used by EPA to study the
capture of Hg and HgCb is shown in Figure 5-7. A detailed description of the apparatus can be
found elsewhere.21 In this system the Hg vapor generated is carried into a manifold by a nitrogen
stream where it is mixed with 862, HC1, CC>2, 62, and water vapor (as required by each
particular experiment). The sorbent to be studied (approximately 0.02 g diluted with 2 g inert
glass beads; bed length of approximately 2 cm) is placed in the reactor and maintained at the
desired bed temperature by a temperature controller. A furnace kept at 850 C is placed
downstream of the reactor to convert any Hg2+ (as in HgCb) to Hg.  According to
thermodynamic predictions, the only Hg species that exists at this temperature is Hg0.22 Quality
control experiments, in the absence of HC1 in the simulated flue gas, also showed that all the
HgCb could be recovered as Hg across this furnace. The presence of the furnace enables
detection of non-adsorbed HgCb as Hg by the on-line ultraviolet (UV) Hg analyzer, thus
providing actual, continuous Hg or HgCb capture data by the fixed bed of sorbent.  The UV Hg
analyzer used in this system responds to SC>2 as well as Hg. Signal effects due to SC>2 are
corrected by placing an on-line 862 analyzer (UV) downstream of the Hg analyzer and
subtracting the measured 862 signal from the total response of the Hg analyzer; the  862 analyzer
is incapable of responding to Hg in the  concentration range generally used.

       In each test, the fixed bed is exposed to the Hg-laden gas for 7 hours or until 100 percent
breakthrough (saturation) is achieved (whichever comes first). During this period the exit
concentration of Hg is continuously monitored.  The instantaneous removal of Hg or HgCb at
any time (t) is obtained as follows:

       Instantaneous removal at time t (%) = 100*[(mercury)in-(mercury)out]/(mercury)in.

       The specific amount of Hg uptake (q, cumulative removal up to time t; weight Hg
species/weight sorbent) is determined by integrating and evaluating the  area under the removal
curves. Selected experiments conducted using this experimental setup have been run in duplicate
and indicated a range of+10% about the mean in the experimental results. It was found that
differences in equilibrium Hg/HgCl2 capacities, at 200-300 mg/Nm3 inlet concentration, are
statistically significant if the Hg/HgCl2 capacities are at least +10 percent different from one
another.

                                                                            9^
       Entrained-flow Reactor.  An example of a bench-scale entrained-flow reactor   is shown
in Figure 5-8. This EPA reactor is constructed of quartz and is 310.5  cm long with an inside
diameter of 4 cm. Three gas-sampling ports are located along the length of the reactor and are
labeled SP1, SP2, and SP3.  The reactor is heated with three Lindberg, 3-zone  electric furnaces
in series.  The baseline Hg concentration is measured in the absence of activated carbon using
an ultraviolet (UV) analyzer (Buck Scientific, model 400A). Once the baseline is established,
activated carbon is fed into the top of the reactor using a fluidized-bed feeder (0.2-0.5 std.
L/min). The gas-phase Hg concentration is then measured at one of the sample ports by pulling
a gas sample (0.5 std. L/min) through a 1 |im filter to remove any particles, then through a
                                          5-19

-------
Table 5-2.  Comparison of bench-scale fixed-bed with entrained-flow reactors
           Test Condition
     Fixed-Bed Reactor
Entrained-Flow Reactor
        Simulation of capture in
        Fabric filter
  Upstream of an ESP
           Sorbent exposure
     Minutes/Hours/Days
  Less than 4 seconds
      Sorbent evaluation based on
Breakthrough or uptake capacity
      Reactivity
                                            5-20

-------
                    Mercury Generation Carbon Trap   Manifold
> -
'iijlll&r N2 '
'Vv^^x*-'
~^s
Water Bath

Carbon Trap < 
I4
Rotameter



T1 v. >
On-off Vs
S\ *" t
Ives K

N2^*Q-
Humidifie
S02
Analyzer



i i
< c>
^
k j
<\ >
L J
<1 >
4CJO I
' "' F FICI/NO/NC
so2
Air/CO2
Elemental
Mercury
Analyzer

Nation Dryer


By-Pass " 1
Oxidation Reactor N
k N.
l>


 ," *
	 1
Lindberg
Furnace
Figure 5-7.  Schematic of bench-scale fixed-bed reactor (source: Reference 21).
                                       5-21

-------
Fluidized Bed Feeder , n
J
CH4 1 1
1 1

T
Lindberg
3 -Zone 	 ^-
Furnaces


\



\ i
1^
I





J
0


^ 	 1X1 	 Dilution N2
_ -*-  [XI 	 NO
-^ 	 [XI 	 Air/H2O


SI> ^ Buck Hg SU2/U2
Analyzer Analyzer
T " n ^^H i 	 i n  i n  i
U ^^| ' ' L_i_J L_s_J
Furnace Drier J-,
Carbon
Trap
T
^ r Pump W
| 	 | ^ Exhausl ^
Carbon
Trap
Figure 5-8.  Schematic of bench-scale flow reactor with methane burner (source:
Reference 23).
                                   5-22

-------
reducing furnace to convert any oxidized Hg to Hg.  The reduction method is described
elsewhere.21 After the reducing furnace, the gas is dried using a Nafion gas sample dryer
(Perma Pure, Inc.) and is finally sent to a Buck analyzer.

       Initial tests  are conducted using nitrogen (TS^) as the carrier gas with later tests performed
in a flue gas from a methane flame. In the N2 carrier gas tests, industrial grade N2 (1 std. L/min)
flows over a Hg permeation tube that is housed in a permeation oven (VICI Medtronic's, model
190) to generate a Hg-laden gas stream. The N2/Hg stream is diluted with a second N2 stream
(12 std. L/min) to the desired concentration before entering the top of the reactor.  Other gases
(SC>2, NOx, O2, water vapor) can be blended into the N2 carrier gas in the mixing manifold.

       A fluidized-bed feeder is used to inject sorbent into the reactor. An inlet line of N2 is
used to fluidize and carry the activated carbon to the reactor.  The carbon feed rate is adjusted by
varying the amount of N2 (0.2 to 0.5  std. L/min) entering the feeder.

       Because the UV analyzer used to detect Hg is sensitive to particles, a filter is used to
remove any carbon that may have been carried with the gas.  Tests have been conducted to
determine if carbon particles accumulate on the filter, as this would act like a packed bed and the
reactor's removal of Hg would be a combination of in-flight and filter (packed-bed) capture. In
these tests, activated carbon was injected in the absence of Hg, and a gas sample was pulled
through the filter. After 1 minute, Hg was added to the gas stream to see if there was a lag in
the time it takes for the baseline to return.  The results were the same as for a blank filter,
suggesting that the  filter does not have an effect on the results.

       The total flow through the reactor is typically 13 std. L/min, which gives residence times
of 5.2, 11.5, and 17.7 s at ports SP1,  SP2, and SP3, respectively. The velocity of the particles
through the reactor is assumed to be the same as that  of the gas flow since the terminal velocity
of the particles is smaller than the velocity of the gas  through the reactor by a factor of 3.

       Fluidized-bed Reactor.  Another type of bench-scale reactor that is used to evaluate
sorbents is a fluidized-bed reactor,24 shown in Figure 5-9.  The advantage of this type of reactor
is the extended contact time between the sorbent and  the Hg-laden gas. Bench-scale Hg removal
tests can be performed on a fluidized-bed reactor apparatus. In a typical experiment, an
Hg/NO/SO2 mixture, nitrogen, and dry air are metered through rotameters to produce 12 scfh of
a dry simulated flue gas of 300 ppmv NOx, 600 ppmv SO2, 8 percent 2, and varying Hg
concentrations. This gas is preheated to reaction temperature (80 C) and humidified with
vaporized water to  an average 10.5 mol  % water.  The resulting wet simulated flue gas is then
passed through a vertical reactor loaded with fluidized sorbent and sand, and then passed through
a filter to remove any entrained particulate to protect  the downstream equipment.  The reactor
and filter assembly are housed in an oven maintained at 80 C.  The test stand is equipped with a
bypass of the reactor and filter assembly to allow for  bias  checks. Sorbent is exposed to
simulated flue  gas for 30 minutes.  Water is removed from the spent flue gas with  a NAFION
Dryer. Dry gas is then  serially analyzed with Hg, SO2, and NOX continuous emission monitors
(CEMs).
                                          5-23

-------

  Nirogwi
  Air
Figure 5-9. Schematic of bench-scale fluidized-bed reactor system (source:
Reference 24).
                                     5-24

-------
5.4.2.2 Pilot-scale Systems

       Initial design and testing is done in bench-scale reactors.  Once the fundamentals of Hg
capture have been tested in a bench-scale system, the next step is to move up to a larger or pilot-
scale system. The main difference between bench- and pilot-scale systems involves testing
sorbents in a more realistic situation involving coal combustion flue gas. This gas is generated in
a pilot-scale combustor that contains a FF or ESP for paniculate control. An example of this is
the pilot-scale combustor operated by DOE (see Figure 7-3). This system burns coal at a rate of
500 Ib/hr and is equipped with a FF. Sorbents, such as activated carbon, are injected upstream of
the PM control device. Mercury removal is determined by gas-phase sampling upstream of the
sorbent injection point and downstream of the PM control device.

       Pilot-scale Hg removal can also be examined using a flue gas  slipstream from a full-scale
unit. An ESP or FF is attached to the slipstream and tested.  A portable FF was developed by
EPRI and  called a COHPAC (COmpact Hybrid PArticulate Collector) unit. 26 This unit was
tested for Hg removal using activated carbon.  The URS Corporation  (formerly Radian
International) also developed a reactor system that uses a slipstream of actual flue gas withdrawn
from a power plant to evaluate sorbents or catalysts in a fixed bed. 2?  It should be noted that the
slipstream reactor, which uses actual coal combustion flue gas, does not always produce the
same Hg captive behavior of a sorbent that a similar laboratory system does using simulated flue
    9R
gas.    It is important to perform pilot-scale tests prior to conducting full-scale tests to eliminate
uncertainties and costly redesign of a process.  With the data collected in the pilot-scale studies,
full-scale tests can be initiated.

5.4.2.3 Full-scale Tests

       Most of work to date in Hg control has been done in bench- or pilot-scale systems. These
reduced-scale systems provide insight into many issues, but cannot fully account for the impacts
that additional control technologies have on plant-wide equipment. Therefore, it is necessary to
scale up and perform full-scale tests to document  actual performance  in a full-scale boiler. These
tests are based on the results obtained in bench- and pilot-scale tests.  Screening tests in bench-
and pilot-scale systems identify sorbents that are effective in capturing Hg. These sorbents are
then tested in a full-scale coal-fired electric utility power plant  to determine full-scale
performance.

       Each full-scale unit is unique in terms of the pollution control  equipment that is present
as well as  the operating conditions.   Some of the factors that are evaluated include:

          Type of paniculate control equipment  that is used (ESP or FF),

          Impact of cake thickness and cleaning frequency in  a FF, and

          Removal of Hg by the fly ash in the system.  Subbituminous coal ashes have been
           shown to be effective in capturing Hg.
                                          5-25

-------
5.4.3  Research on Sorbent Evaluation

5.4.3.1 Sorbent Evaluation Using Enhanced-flow Reactors

       A flow reactor was designed to simulate Hg capture through a duct or ESP and to obtain
kinetic rate constants for the adsorption of Hg onto sorbents.  Several researchers have predicted
that, under certain conditions, dispersed-phase capture would be limited by mass transfer.29'30
Calculations were performed to determine the required operating conditions to minimize external
mass transfer effects in the flow reactor, and experimental tests were performed to verify these
           01 1110
calculations. '  '   The first test involved changing the diffusion coefficient by changing the gas
in the system from N2 to helium (He) and to argon (Ar) while holding all other parameters
constant (particle size, residence time, temperature, and Hg concentration). The diffusion
coefficient increased by an order of magnitude by changing the gas from N2 to He.  Using a
lignite-based commercially available carbon (Norit FGD) at 100 C and a Hg concentration of
86 ppb, Hg removal was 6 percent at a carbon to Hg ratio (C:Hg) of 1,500:1 and increased to 30
percent at a C:Hg of 8,000:1. Experimental results were similar when He was used as compared
to N2.  If external mass transfer were controlling, then a higher Hg removal would have been
obtained using He, since the mass transfer coefficient increased.

       A second test involved using two commercially available activated carbons, Norit FGD
and Calgon WPL at 100 C and 124 ppb Hg in dry N2. Removal for the FGD carbon ranged
from 9 percent (C:Hg=2200:1) to 23 percent (C:Hg=6400:1). Removal for the WPL carbon
ranged from 11 percent (C:Hg=340) to 94 percent (C:Hg=5000:1). If dispersed-phase capture in
the flow reactor were film-mass-transfer limited, the two activated carbons would have removed
similar amounts of Hg at a given C:Hg, assuming  each carbon had sufficient Hg capacity.

       The flow reactor has been used to examine the effect of temperature, particle size,
residence time, carbon type, and gas composition on Hg removal.31"33 The effect of particle
size on Hg removal for Darco FGD at 100 C and a Hg concentration of 86 ppb is shown in
Figure 5-10. Several particle sizes (4-8, >8-16, >16-24, and >24-44 |im) were injected into the
flow reactor at C:Hg ratios ranging from 2000 to 11,000:1.  The gas was sampled at SP2,
resulting in a gas contact time of 8.4 s. Figure 5-11 shows that greater Hg removal is achieved
by increasing the feed rate and by decreasing the particle size. At a C:Hg of 5000:1, a 5 percent
reduction was obtained with the >24-44 |im size fraction as compared to a 20 percent reduction
with the 4-8 |im fraction. Thus by using a smaller particle a higher removal can be obtained at a
given C:Hg. Both external and internal mass transfers are dependent on particle size: the effect
of mass transfer increases with an increase in particle size.

5.4.3.2 Sorbent Evaluation Using Packed-bed Reactors

       Recent bench-scale studies at the University of North Dakota=s Energy and
Environmental Research Center (UND/EERC) have focused on the interactions of gaseous flue
gas constituents on the adsorption capacity of activated carbon for Hg.34 Bench-scale studies
were performed using a fixed bed of carbon.  The tested carbon was a commercially available
lignite-based activated carbon (LAC) commercially known as Darco FGD from Norit
                                         5-26

-------
             el
             o
              40


              35


              30-


              *25-
              10
                                   Carbon to Mercury Ratio
                                                             10000
12000
Figure 5-10.  Effect of particle size on adsorption for Darco FGD at 100 C,
86 ppb Hg concentration, and 8.4 s contact time (source: Reference 31).
                                      5-27

-------
Americas, Inc. A simulated flue gas containing a nominal concentration of 15 |ig/Nm3 of
gaseous Hg was passed through the fixed bed of carbon. In addition to Hg, the baseline test gas
contained 6 percent C>2, 12 percent CC>2, 8 percent H^O, and the balance N2. Various flue gas
constituents (SC>2, HC1, NO, and NO2) were added individually and in combination to the
baseline test gas to determine the effects of flue gas constituents on Hg adsorption.  Temperature
effects were also examined. Table 5-3 shows the various compositions of gas tested.

       For each adsorption test, a Hg CEM was used to monitor total or elemental Hg.
Measurements were alternated between the inlet and outlet locations of the test bed. For a given
test, measurements took place primarily at the outlet location; however, occasionally the inlet
location was tested to confirm that a constant concentration of gaseous Hg was entering the test
bed. For each test, the analyzer was set to measure total gaseous Hg at the outlet; however,
occasionally the analyzer was set to measure only gaseous Hg at the outlet. The purpose of
measuring only gaseous Hg at the outlet was to determine if any incoming gaseous Hg was
being oxidized by carbon in the bed (evident if the concentration of gaseous Hg in the outlet gas
was less than the concentration of total gaseous Hg in the outlet gas).

       For adsorption to take place (assuming attractive forces exist between a particular
gaseous specie and sorbent), the adsorbing specie must have sufficient time to reach the surface
of a sorbent and diffuse into its pores (where most adsorption takes place).  If any of the
adsorbing specie in a gas stream passing through a fixed bed of sorbent cannot reach the surface
of the sorbent (mainly its pore surfaces), the specie will pass through the bed unadsorbed.
Researchers conducted preliminary tests to show that the gaseous Hg in the test gas had
sufficient time (under the conditions tested) to contact the sorbent and to diffuse into its pores.
Proving this point was important since some of the adsorption tests showed immediate
breakthrough of Hg in the outlet gas.  In these cases, immediate breakthrough was not due to
insufficient contact time but rather the carbon's inability to adsorb all of the gaseous mercury.

       Figure 5-11 shows an example of the sampling and measurements taken during testing of
the  baseline test gas with HC1, NC>2,  and 862  (as noted in the graph, 862 was added to the
baseline test gas 2.5 hours after the start of the test).  Except where noted, the Hg concentrations
in Figure 5-11 are those in the outlet test gas and represent concentrations of total gaseous Hg.
Mercury concentrations in the graph are quantified as a percentage of the inlet concentration of
gaseous Hg.  The percentage of Hg in the outlet test  gas is called percent breakthrough. Figure
5-11 indicates that the analyzer sampled and measured total gaseous Hg in the outlet gas at all
times during testing except at approximately 5.2 hours, at which time the analyzer sampled and
measured Hg in the inlet gas. At approximately 5.15  hours the analyzer measured gaseous Hg
instead of total gaseous Hg in the outlet test gas; the drop in the concentration curve at this time
from approximately 150 percent to zero percent indicates that Hg in the outlet test gas consisted
                    94-  	                         f\
entirely of gaseous Hg  .  Thus, while only gaseous Hg was in the test gas entering the carbon
bed, the Hg was oxidized to Hg2+  as it passed through the bed. (Why some of the outlet
concentrations of total gaseous  Hg exceeded 100 percent of the inlet Hg concentration for this
run is explained further on in this section.)
                                          5-28

-------
Table 5-3.  Composition of test gases to simulate coal combustion flue gas used
for UND/EERC bench-scale study (source: Reference 34).
SO2 ppmv
HCI ppmv
NO ppmv
NO2 ppmv
Baseline test gas"
0
0
0
0
Baseline test gas plus 1 additional gas
1600
0
0
0
0
50
0
0
0
0
300
0
0
0
0
20
Baseline test gas plus 2 additional gases
1,600
1,600
1,600
0
0
0
50
0
0
50
50
0
0
300
0
0
300
300
0
0
20
20
0
20
Baseline test gas plus 3 additional gases
1,600
1,600
1,600
0
50
50
0
50
300
0
300
300
0
20
20
20
Baseline test gas plus 4 additional gases
1600
50
300
20
     (a) Prior to adding SO2, HCI, NO, and/or NO2, the baseline test gas contained 15 |ig/nm3 of gaseous Hg;
        6 percent O2; 12 percent CO2; 8 percent H2O; and the balance N2.
                                       5-29

-------
      o
      'a
      ^
      
      <
      JB
      c
      s
      *fMMi
      Cl$
      e
      
-------
       Graphs of the adsorption tests with the 15 remaining gases in Table 5-3 can be found
elsewhere;13 the cited graphs are similar to Figure 5-11 in that Hg concentrations (primarily
outlet concentrations of total gaseous Hg) are plotted versus the time of the adsorption test.

       The following summarizes the detailed test results:

         When the sorbent was exposed to the baseline gas only, the sorbent initially captured
          10 to 20 percent of the incor
          bed (i.e., was not adsorbed).
10 to 20 percent of the incoming gaseous Hg; the rest of the Hg passed through the
         When the sorbent was exposed to SC>2 in addition to the baseline gas, Hg capture
          improved slightly.

         Under exposure of the sorbent to HC1, NO, or NC>2 added one at a time to the baseline
          gas, the Hg capture of the sorbent improved to 90 to 100 percent.

         An apparently significant interaction between SO2 and NC>2 gases and the sorbent
          caused a rapid breakthrough of Hg as well as conversion of the Hg to its volatile
          oxidized form. This effect occurred at both 107 and 163 C (225 and 325 F) and with
          or without the presence of HC1 and NO.

         In the presence of all four acid gases (SO2, HC1, NO, and NO2), rapid breakthrough
          and oxidation of the Hg occurred at both 107 and 163 C (225 and 325 F).  This
          suggests that the interactions between the sorbent and NO2 and SO2 gases produced
          poor sorbent performance, which may be a major effect.  This may be likely to occur
          over a variety of conditions typical  of coal-fired electric utility boilers, and represents
          a hurdle that must be overcome to achieve effective Hg control by carbon adsorption.

       The UND/EERC is continuing to investigate the interactions of gaseous flue gas
constituents on the adsorption capacity of activated carbon for Hg. In addition, other types of
sorbents are being developed and investigated under similar simulated flue gas conditions. Other
gaseous flue gas constituents are also being examined to assess their impact on the adsorption of
Hg.

5.4.3.3 Sorbent Evaluation Using Fluidized-bed Reactors

       Under DOE=s Small Business Innovative Research (SBIR) Program, Environmental
Elements Corporation (EEC) has been developing a circulating fluidized bed (CFB)24 to promote
agglomeration of fine PM, allowing for its capture in an ESP. In addition, a single injection  of
iodide-impregnated activated carbon was added to the fluidized bed to adsorb gaseous Hg. High
residence time, as a result of particle recirculation, allows for effective utilization of the carbon
and high collection of the fine particles. Laboratory tests with heated air indicate that, with a high
density of fly ash at a 4-second residence time within the bed, fine particle emissions are reduced
by an order of magnitude.
                                          5-31

-------
       Results from the laboratory-scale testing indicate that spiked gaseous Hg was
significantly reduced when passed through the fluidized bed of fly ash (50 percent Hg removed)
with a further reduction to essentially zero, when activated carbon was injected into the bed
(25 |ig/m3 to zero) at 110 C (230 F). The iodide-impregnated activated carbon was fully
utilized after greater than 2 hours within the bed. An adsorption capacity was calculated to be
770 |ig/g for the carbon and 480 |ig/g for the bed of ash.  Other field tests were conducted at
Public Service Electric and Gas= Mercer Station with similar results.24
5.5 Sorbent Development

       The implementation of an effective and efficient Hg control strategy using sorbent
injection requires the development of low-cost and efficient Hg sorbents. Of the known Hg
sorbents, activated carbon and calcium-based sorbents have been the most actively studied.
However, improved versions of these sorbents and new classes of Hg sorbents can be expected,
as this is still a very active field.

5.5.1  Powdered Activated Carbons

       Activated carbons have been  extensively studied for their Hg capture capability.
Activated carbon is the reference sorbent for Hg control in municipal waste combustors.  Many
factors may affect the adsorptive capability of the activated carbon sorbent.  These include the
temperature and composition of the flue gas, the concentration of Hg in the exhaust stream, and
the physical and chemical characteristics of the activated carbon (or functionalized/impregnated
carbon). Some specific efforts at understanding these effects are given below.

5.5.1.1 Effects of Temperature, Mercury Concentration, and Acid Gases

       The effects of bed temperature, Hg concentration, presence of acid gases (HC1 and 802),
and presence of water vapor on the capture  of Hg and HgQ2 by thermally activated carbons
(FGD and PC-100) and Ca-based sorbents [Ca (OH) 2 and a mixture of Ca(OH)2 and fly ash]
were examined in a fixed-bed, bench-scale system.21 Sorption studies indicated an abundance of
HgQ2 adsorption sites in calcium-based sorbents. Increasing the HgCl2 concentration increased
its uptake, and increasing the bed temperature decreased its uptake. Gas-phase HgQ2
concentration had a very strong effect on its adsorption, while bed temperature had a small
influence on adsorption. The observed temperature and concentration trends suggest that the
process is adsorption-controlled and that the rate of HgQ2 capture is determined by how fast
molecules in the vicinity of the active sites are being adsorbed.  Mixtures of Ca(OH)2 and fly  ash
with 7 times higher surface area than Ca(OH)2 and a totally different pore size distribution
exhibited identical HgCl2 capture to that of Ca(OH)2. The presence of acid gases (1000 ppm  SO2
and 50 ppm HC1) drastically decreased the uptake of HgCl2 by Ca(OH)2. The inhibition effect of
SO2 was more drastic that HC1, and essentially controlled the HgCl2 uptake. It was hypothesized
that the inhibition effect is due to competition between these acid gases and HgCl2 for the
available alkaline sites.
                                          5-32

-------
       Sorption studies further indicated that the available active sites for capturing Hg in the
activated carbons are limited, suggesting that it is more difficult to control Hg emissions than
HgCb emissions.  Increasing the Hg inlet concentration and decreasing the bed temperature
increased the saturation capacities of the activated carbons, the time needed to reach this
capacity, and the initial rate of Hg uptake.  Unlike HgCb capture by Ca(OH)2, bed temperature
had a very strong effect on the Hg adsorption by the activated carbons, and gas-phase Hg
concentration had a small influence on such adsorption. PC-100, with twice the surface area of
FGD, consistently exhibited higher saturation capacities (3-4 times higher) than FGD. The
presence of acid gases had a positive effect on the capture of Hg by a lignite-coal-based
activated carbon (FGD) and had no influence on Hg capture by a bituminous-coal-based
activated carbon (PC-100). This difference was related to a higher concentration of Ca (acid gas
sorbent) in FGD. It appears that adsorption of these acid gases by FGD creates active S and Cl
sites, which are instrumental in capturing Hg, through formation of S-Hg and Cl-Hg bonds in
the solid phase (chemisorption). These results indicate that the optimum region for the control of
Hg by injection of activated carbon is upstream of the acid gas removal system.

5.5.1.2 Role of Surface Functional Groups

       The content of oxygenated acidic and alkaline surface functional groups (SFGs) on the
surface of two activated carbons was manipulated to investigate their role in Hg and HgCb
capture.35 Acidic SFGs on the surface of activated carbons were neutralized by a variety  of
alkaline washes. The alkaline-treated activated carbon showed no enhancement in Hg and
HgCb capture, thus indicating that acidic SFGs play no role in capturing Hg species. The
alkaline SFGs content was increased by  a thermal treatment process. The thermally treated
activated carbons did not exhibit any improvement with regard to their Hg and HgCb capture
capabilities as compared to the untreated ones.  The activated carbons were then treated with a
very dilute HC1 solution to decrease their alkaline SFGs content. The HCl-treated activated
carbon showed a very significant  improvement in its Hg and HgCb capture capabilities.  This
observation was contrary to the initial hypothesis that alkaline sites are needed to capture acidic
HgCb from the flue gas. It was then hypothesized that HC1 treatment increases the number of
active surface chlorine sites, which subsequently enhance Hg and HgCb capture.  An analytical
technique, Energy-Dispersive X-ray Spectroscopy (EDXS), was used to quantify surface  Cl sites.
A strong correlation between the increased amount of surface Cl and Hg/HgCl2 uptake
enhancement was observed. The  role  of SFGs containing Cl atoms in providing Hg/HgCl2
active sites was established. Future investigation using SEM/EDXS and Fourier Transform
Infrared (FTIR) will focus on understanding the nature of Cl bonds on the surface of carbon, so
that more effective Hg species sorbents can be manufactured.

5.5.1.3 In-flight Capture of Mercury by a Chlorine-impregnated Activated  Carbon

       Activated carbon duct injection seems to be the most promising Hg control technology
for coal-fired electric utility boilers equipped with ESPs. In this technology, the injected
activated carbon removes Hg only while contacting the flue gas during very limited sorbent/gas
contact time (<3 seconds). Prior investigations have shown that very high, and rather costly,
carbon-to-Hg weight ratios (>50,000)  are needed to achieve adequate Hg removal. In order to
                                          5-33

-------
reduce the operating cost of the carbon injection process, either a more efficient sorbent that can
operate at a lower carbon-to-Hg weight ratio or a lower-cost activated carbon (or possibly both)
are required.  In this study33, a cost-effective Cl-impregnation process was successfully
implemented on an inexpensive virgin activated carbon.  The Cl-impregnated carbon was
produced in a 5 pound large batch, and its in-flight Hg removal efficiency was evaluated in a
flow reactor (as previously discussed in Section 5.4.2.1) with gas/solid contact times of 3 to
4 seconds. The Hg removal efficiency of more than 80 percent was obtained in a flue gas
containing the effluent of natural gas combustion doped with coal combustion levels of NOx and
SC>2 at carbon-to-Hg weight ratios of about 3000.  Hg removal was rather insensitive to the
adsorption temperature in the range of 100-200 C. Cost analysis showed that this Cl-
impregnation process can produce a very active and cost-effective activated carbon that can be
used as a practical sorbent in a duct injection control technology in ESP-equipped coal-fired
electric utility boilers.  Preliminary  cost estimates indicated that approximately  53 percent
reduction of the total annual cost of Hg control could be possible when using Cl-impregnated
FGD in lieu of virgin activated carbon. Future investigations would be focused on evaluating the
Cl-impregnated activated carbon in a pilot-scale, 21-kW (90,000-Btu/hr) refractory-lined,
furnace fired  with pulverized coal.33

5.5.2  Calcium-based Sorbents

       Work conducted by EPA and ARCADIS Geraghty & Miller, Inc. [funded by the Illinois
Clean Coal Institute (ICCI)] indicates that the injection of calcium-based sorbents into flue  gas
can result in significant removal of Hg.36'37 Researchers examined the high-temperature/short-
gas-phase residence time removal of Hg using injection of lime while burning an Illinois #6 coal
in a pilot-scale combustor.  The lime was injected as a slurry at a calcium-to-sulfur (Ca:S) ratio
of 2.0 mol/mol at 968 C (1775 F). Under these conditions, 77 percent of the total Hg was
removed from the flue gas (Table 5-4). Based on these results, they concluded, "injection of
lime in the high temperature regions of coal-fired processes upstream of air pollution control
systems can efficiently transfer Hg from the gas to the solid phase." Summaries of work follow.

5.5.2.1 Capture of Low Concentrations of Mercury Using Calcium-based Sorbents

       The capture of Hg and mercuric chloride (HgCb), the  Hg species identified in coal  flue
gas, by three types of calcium-based sorbents differing in their internal structure, was examined
in a packed-bed, bench-scale study under simulated flue gas conditions for coal-fired electric
utility boilers.38 The results obtained were compared with Hg and HgCb capture by an
activated carbon (FGD) under identical conditions. Tests were conducted with  and without SC>2
to evaluate the effect of 862 on Hg and HgCb control by each of the sorbents.

       The Ca-based sorbents showed insignificant removal of Hg in the absence of SC>2.
However, in the presence of SC>2, Hg capture was enhanced for the three Ca-based sorbents.  It
was postulated that the reaction of hydrated lime with 862 would result in pore mouth closure as
evidenced by the sharp drop in the 862 removal rate after the initial 10 minutes of exposure.
Despite the loss of internal  surface area, the relatively high uptake of Hg, observed for these
sorbents in the presence of SO2, suggests that Hg and SO2 do  not compete for the same active
                                          5-34

-------
Table 5-4.  Mercury removal by lime sorbent injection as measured by EPA bench-
scale tests (source:  Reference 36).
Test
Baseline
Lime sorbent injection
Total Hg Concentration,
ug/dscm
5.7
8.0
Total Gaseous Hg,
percent
100
23
Total Particle-bound Hg,
percent
0
77
                                   5-35

-------
sites, and that the sites for Hg capture are influenced positively by the presence of SC>2.
Moreover, the capture of Hg in the presence of 862 increased with sorbent surface area and
internal pore structure.

       Conversely, the three Ca-based sorbents showed decreased removal of HgCb in the
presence of 862. In the absence of 862, roughly 25 percent of the incoming HgCb was captured.
The alkaline sites in the Ca-based sorbents were postulated to be instrumental in the capture of
acidic HgC^. SO2 not only competed for these alkaline sites but also, as mentioned, likely
closed pores with subsequent reduction in accessibility of the interior of the Ca-based sorbent
particles to the HgCb molecules.

       It was hypothesized that the capture of Hg in the presence of SO2 may occur through a
chemisorption mechanism, while the nature of the adsorption of HgCb molecules may be
explained through a physisorption mechanism. The effect of temperature studies further
supported this hypothesis.  Increasing the system temperature caused an increase in Hg uptake
by the sorbents in the presence of SC>2. However, the increase in temperature resulted in a
significant decrease in the HgCb uptake in the absence or presence of SC>2. Increased sorbent
surface area and internal pore structure had no observable effect on HgCb capture in the
presence of 862.

       With the relatively large quantities of Ca needed for SC>2 control at coal-fired electric
utility boilers, the above results suggest that  Ca-based sorbents, modified by reaction with fly
ash, can be used to control total Hg emissions and 862 cost effectively.  The most effective
calcium-based sorbents are those with significant surface area (for SO2 and HgCb capture) and
pore volume (for Hg capture).

5.5.2.2 Simultaneous Control of Hg, 862, andNOxby Oxidized-calcium-based Sorbents
                                                                       n       -i-9
       Multipollutant sorbents have been developed that can remove both Hg and Hg   as
effectively as FGD activated carbon in fixed-bed simulations of coal-fired electric utility boiler
flue gas at 80 C.39 Oxidant-enriched, calcium-based sorbents proved far superior to activated
carbon with respect to SC>2 uptake on the same fixed-bed simulations. These oxidant-enriched,
calcium-based sorbents also performed better with respect to NOx and SC>2 uptake than baseline
lime hydrates for fixed- and fluid-bed simulations at 80 C.

       Preliminary economic analyses suggest that silicate sorbents with oxidants are 20 percent
of the cost of activated  carbon for Hg removal, while oxidant-enriched lime hydrates offer
reduced, but significant savings. Credits for 862 and NOx increase the savings for multipollutant
sorbents over activated carbon.

       The apparent superiority of multipollutant lime and silicate hydrates enhanced with
oxidants has been confirmed at conditions typical of gas-cooled, semi-dry  adsorption processes
on boilers; performance of sorbents at higher-temperature conditions of duct sorbent injection
technologies remains to be evaluated. Planned field evaluations of both semi-dry adsorption and
                                          5-36

-------
duct sorbent injection will allow better economic and performance comparisons of activated
carbon sorbents to that of oxidant-enriched lime and silicate hydrates.

       A technology for the efficient capture of Hg through in furnace injection of a calcium-
based sorbent has been developed by McDermott Technologies recently.  A discussion of the
full-scale tests of the technology is presented in Chapter 7.

5.5.3 Development of Low-cost Sorbents

       Since 1995, EPRI has supported a sorbent development program for removal of Hg
emissions from coal-fired electric utility power plants at several research organizations including
Illinois State Geological Survey (ISGS), University of Illinois (UI), and URS Corporation.  The
development of effective Hg sorbents that can be produced at lower costs than existing
commercial  activated carbons is the main objective of the program. The development efforts
were documented in three EPRI Reports.40"42 A significant number of sorbents were derived
from a variety  of precursor materials, including coal, biomass, waste tire,  activated carbon fibers,
fly ash, and zeolite,  through this work.  Different preparation methods were used to determine
the effects of sorbent properties, such as pore size distribution, pore volume, surface area,
particle size, and sulfur content, on the ability to remove Hg. The effects of different processing
methods, including steam activation, grinding,  size-fractionation, and sulfur-impregnation, on
sorbent performance were also  investigated in laboratory tests. Promising low-cost sorbents
were further evaluated in actual flue gas at several full-scale coal-fired electric utility power
plants.

       Results of the EPRI sorbent development work showed that effective sorbents can be
prepared from  inexpensive precursor materials using simple activation steps. One notable
example is that a char produced from corn fiber, a by-product from a corn-to-ethanol production
process, showed a Hg adsorption capacity over twice that of the commercial FGD carbon
sorbent, after the char was activated in CC>2 at 865 C for 3.5 hours.40 Inactivated corn char had
no capacity for HgCb, and only a low capacity for Hg. It appears that the composition of the
flue gas has  a significant effect on the Hg adsorption capacities of the coal-derived activated
carbons.41 The EPRI-funded study found that the presence of acid gases (SC>2 and HC1) inhibits
Hg and HgCb adsorption for both lignite- and bituminous-coal-derived activated  carbons.
However,  research conducted by EPA showed that the presence of acid gases enhanced the
capture of Hg by a  lignite activated  carbon and had no influence on the adsorption by a
bituminous-coal-derived activated carbon.21  In a later more extensive follow-up study funded by
EPRI and  ICCI, the  effects of acid gases on the HgCb and Hg adsorption capacities of activated
carbons were found to vary, depending on the precursor materials and characteristics of the
carbons.43 For example, carbons derived from tire and corn fiber had much higher HgCb and
Hg adsorption capacities when they were tested in a high-SO2 concentration flue gas simulating
the combustion of Eastern bituminous coals compared to those when they were tested in the low-
SC>2  concentration flue gas simulating Western subbituminous coal combustion. Complex
interactions occurring between  the characteristics of the carbons and the acid gases may lead to
the observed varying effects of the acid  gases on Hg adsorption behaviors of the carbon sorbents.
                                          5-37

-------
More fundamental research is needed to understand and predict the effects of acid gases on the
performance of sorbents derived from different precursor materials.

       The most effective sorbents were obtained by the sulfur-impregnation of activated
carbons derived from waste material and carbon fibers.40 Researchers at the University of
Pittsburgh demonstrated that impregnation of heteroatoms such as sulfur44 and chloride45 is an
effective method to improve the vapor-phase Hg adsorption capacities of activated carbons.  It
has been suggested that sorbent-impregnation studies should focus on highly microporous
sorbents since the presence of active surface functional groups, sulfur as an example, in the
micropores through impregnation is likely to provide the most reactive sites for Hg adsorption
from coal combustion flue gas.19  They stressed that the micropore surface area of sorbent is an
important physical property for vapor-phase Hg adsorption. Most of the commercial activated
carbons are used for liquid-phase applications and contain a large mesopore surface area, in
addition to micropores, that are less effective for adsorption of ppb levels of Hg from coal
combustion flue gases. EPA researchers46 have observed the importance of active functional
groups in the micropores for vapor-phase Hg adsorption. After treating an activated carbon with
an aqueous sulfuric solution, they found that most of the mesopores of the carbon are filled with
water due to the presence of the hydroscopic sulfuric acid, and the carbon becomes a highly
microporous sorbent. The Hg adsorption capacity of the sulfuric-acid-treated carbon is much
higher than that of the as-received carbon due to the presence of the active sulfuric acid
functional groups in the micropores of the treated carbon.

       The most recent research conducted by ISGS, UI, and URS Corporation showed that
relatively low surface area microporous biomass-based carbon sorbents, such as those derived
from pistachio nut shells and from corn fiber, are as effective as the commercial FGD carbon
sorbent for Hg adsorption.43  They found that the Hg adsorption capacities of the biomass-based
carbon sorbents, which contained negligible (0.09 percent) sulfur, are comparable to those of the
coal- and tire-derived carbons that have substantial sulfur contents (0.98 to 2.1 percent). The
biomass-based carbon sorbents also have very little chlorine functional groups. It appears that
more oxygen, another heteroatom, remained in the biomass-based carbon sorbents after the
pyrolysis of the oxygen-rich biomass from the carbon-making process contributing to the
significant Hg adsorption capacities of such sorbents.  It has been suggested recently by EPA
researchers47 that the Hg adsorption capacity of an activated carbon is correlated to the
concentrations of the oxygen functional groups of the carbon.  They changed the oxygen
functional group concentrations of a carbon by heating the carbon sample to 900 C in an inert
atmosphere to remove the functional groups. Also, more oxygen functional groups were
introduced to the carbon sample by oxidizing the carbon sample in an aqueous nitric acid
solution. They suggested that lactone and carbonyl groups introduced during the oxidization of
the carbon by nitric acid treatment might be the active sites for Hg adsorption.

5.5.4  Modeling of Sorbent Performance

       The Hg adsorption data produced from bench-scale tests provide a relative indication of
performance for different sorbents; however, the actual Hg removal performance of the sorbents
in full-scale systems cannot be predicted based on bench-scale results alone. To predict Hg
                                          5-38

-------
removal in full-scale systems, mass transfer considerations have to be combined with laboratory
data.  Such an approach was applied by by EPRI recently to develop a model for predicting
sorbent performance in full-scale systems.48 The model is also capable of determining when
mass transfer limits Hg removal and when it is limited by sorbent capacity.  By incorporating the
appropriate mass transfer expressions, the model relates the adsorption characteristic data for a
given sorbent tested under a given set of flue gas conditions in the laboratory to the expected Hg
removal performance across a FF or an ESP.

       Results of the sorbent performance predicted by the model agree reasonably well with
data of the same sorbent measured by pilot-scale tests for both ESP and FF applications. The
pilot-scale facilities used for the tests consisted of an ESP with a 160-acfm wire-tube ESP, and a
FF with a 4000-acfm transportable pulse-jet FF operating in the COHPAC configuration.
Results of the pilot-scale tests and modeling both showed that a carbon sorbent with 15 |im
diameter and 1000 |ig/g Hg adsorption capacity achieved about 80 percent Hg removal in a FF
operated at about 140 C (280 F) with 3 Ib/Macf sorbent injection rate and cleaning cycle of 45
min. However, test and modeling results both showed that Hg removal decreases to less than 20
percent when the same sorbent was injected upstream of an ESP under conditions similar to the
above.

       Laboratory tests which have been conducted to  evaluate the adsorption characteristics of
potential sorbents for Hg removal seem to  suggest that  reactivity of the sorbent might be more
important than its equilibrium adsorption capacity for sorbent injection.  Currently, an ESP is
more widely used than a FF as a PM control device for coal-fired electric utility boilers in the
United States.  Sorbent reactivity is the important parameter determining Hg removal  when
injecting a powdered sorbent upstream of an ESP, where adsorption of Hg occurs mainly in-
flight with short residence times (about 2 seconds).  When injecting sorbent upstream of a FF,
additional Hg removal can occur due to the presence of accumulated sorbent in the filter cake,
resulting in improved mass transfer and sorbent utilization.  Sorbent capacity becomes a more
important parameter than reactivity in such cases.
5.6 Capture of Mercury in Wet FGD Scrubbers

5.6.1  Wet Scrubbing

       Mercuric chloride is readily soluble in water. Thus, the oxidized fraction of Hg vapors in
flue gas is efficiently removed when a power plant is operated with a wet scrubber for removing
SC>2. The elemental fraction, on the other hand, is insoluble and is not removed to any
significant degree.  A DOE-funded study49 conducted by CONSOL, Inc. showed that the nominal
Hg removal for wet FGD systems on units firing bituminous coals is approximately 55 percent,
with the removal of Hg2+ between 80 and 95 percent.  Studies conducted by McDermott
Technologies, Inc. at its  10-MWe research facility suggested a possible conversion of the Hg2+
captured in the scrubbing media and reemissions as Hg0.50 McDermott Technologies performed
follow up tests to investigate the use of additives to prevent the conversion of adsorbed Hg2+ to
gaseous Hg0.51 These tests are described in more detail in Chapter 7.
                                          5-39

-------
5.6.2  Oxidation

       The challenge to Hg removal in wet scrubbers for SC>2 is to find some way to oxidize the
elemental Hg vapor before it reaches the scrubber or to modify the liquid-phase of the scrubber
to cause oxidation to occur there.

       URS Radian International has conducted various laboratory and field-test studies to
investigate adsorption and catalytic oxidation of gaseous Hg in coal-fired electric utility flue
gas.  The results of the bench-scale testing are discussed below. The additional pilot- and full-
scale testing conducted by URS Radian International are discussed in Chapter 7.

       Different compositions of catalysts and fly ashes were tested in a bench-scale, fixed-bed
                                                                  n ^9
configuration to identify materials that adsorb and/or oxidize gaseous Hg .   Mixing sand with a
particular catalyst or fly ash created fixed beds of sorbents.  A simulated coal-fired electric utility
boiler flue gas containing gaseous Hg was then passed through the bed.  The flue gas was tested
at the inlet and outlet of each sorbent bed to determine Hg adsorption and/or oxidation across the
bed.  Table 5-5 lists the simulated flue gas conditions and the most active catalysts and fly ashes
identified during testing for oxidation of gaseous Hg.

       Figure 5-12 is an example of the adsorption/oxidation of gaseous Hg with time by one of
the iron catalysts in Table 5-5.  In this figure, the oxidation of gaseous Hg increases as the
breakthrough of Hg from the catalyst bed increases (breakthrough is quantified as a percentage
of the incoming Hg). At 100 percent breakthrough when the catalyst is no longer adsorbing any
of the incoming Hg (i.e., the catalyst has reached its equilibrium adsorption capacity for the
incoming Hg), all of the Hg passing through the bed is being oxidized to Hg2+.

       Figure 5-13 shows adsorption/oxidation results for all of the catalysts in Table 5-5.
Adsorption and oxidation of gaseous Hg was greater at  149 C (300  F) than at the higher
temperature of 371 C (700 F).  The adsorption and oxidation activity of the activated carbon
was considered the highest among the materials tested because a lower mass was utilized during
the tests compared to the other materials.

       Figure 5-14 shows the adsorption/oxidation results for the fly ashes from Table 5-5.  Like
the catalysts, the fly ashes showed higher adsorption and oxidation of gaseous Hg at 149  C
(300 F) than at 371 C  (700 F); for this reason, only the lower temperature results are shown in
Figure 5-14.  The subbituminous and bituminous coal fly ashes generally showed higher
oxidation rates than the lignite coal fly ashes. As seen, the #2 bituminous coal  fly ash had
varying adsorption and  oxidation rates depending upon where the fly ash samples were collected.
Samples collected from the hoppers of the first field of the ESP indicated lower oxidation of
gaseous Hg but a higher adsorption of Hg compared to the finer fly ash collected in the fifth and
final field of the ESP. Although not shown, fly ash captured by a cyclone in the Hg speciation
sampling train indicated a higher adsorption but no oxidation of the gaseous  Hg.  Fly ash from
the fifth field of the ESP indicated the highest rate of oxidation and the lowest size-fractionated
particles. This may be associated with the size differences of the fly ash and/or the surface
                                          5-40

-------
Table 5-5.  Simulated flue gas conditions with the most active catalysts and fly
ashes indicated for oxidation of gaseous Hg to gaseous Hg2+(source:
Reference 52)
Parameter
Fixed-bed Temperature
Hg Injection
Oxygen
Carbon Dioxide
Moisture
Sulfur Dioxide
HCI
Gas Flow Rate
Baseline
Conditions
300 and 700 F
45 to 60 jag/Mm3
7 percent
12 percent
7 percent
1600 ppmv
50 ppmv
1 L/min
Most Active
Catalysts
Fe#1 (1000mg)
Pd#1 (1000 mg)
Fe #2 (200 mg)
Fe #3 (200 mg)
NOX Catalysts
(1000 mg)
Fe#4(1000mg)
Pd#2(1000mg)
Carbon (20 mg)
Most Active
Fly Ashes
Subbituminous #1
Subbituminous #2
Bituminous #1
Bituminous #2-Field 1a
Bituminous #2-Field 5
Bituminous #3
Lignite #1
Oil-Fired #1
      (a) Fly ash collected at the first and fifth field of the ESP at the EPRI ECTC.
                                      5-41

-------
  TJ
  0)
  N

  !5

O    (D
    CO

  O T3
  I 
  S 2

  < 5
  O  <
  O>
  X
                                                 % Hg
400
                                     600      800


                                      Time, hr
1000
                                  1200
Figure 5-12. Adsorption and subsequent oxidation of gaseous Hg in a simulated

flue gas at 149C (300 F) (source: Reference 52).
                                    5-42

-------
I UU.U 
90.0 '
.8 80.0-
'  70.0 -
0 B 60.0 
u bO.il 
S 1 40.0-
3 b
 < 30.0 -
g 20.0-
10.0-
-



;X
-










-:sryv*r TI
****
! M













-

























P^P



















-w













MH









l~~1 A- AdSC'MC
cm % Oxiaizfrd

















apag
















.,,







,.,

















,,.
T-





_














i









 








*







,,,
j i
-
 H
"1
IA *"*r ^I-BTI r*iir TAI^BT* -'s.n^i^r  A "r -.riA a'r ^^^ "*r ^ri^ '*r -r.Vr ";A^ klr -^i".-> 'v itii^ '*r ->f"i.^'T isiv^.,l?r
               F**1     Frj*i     Fe*2     Fe#?   N
              lOOOmg    lOOCimg    230 mg     MC'iig    tuOOrnn
K ng
 PC V2   Caitxin
IUGO mg  33 mg
Figure 5-13. Adsorption and oxidation of gaseous Hg by various catalysts at
149 C (300 F) and 371 C (700 F) (source: Reference 52).
                                      5-43

-------
 T3
 0)
 tfl "O
 O 0)
 < CD
 T3 ><
 0) U-
 T3
 "X

 
o
 CD
 X
   CO

100-

80-
60-
40-


on _

0-











1  1


pn 2

p'^
p*










"^7
//


'"/;.,
;
c c .S
D5 O3 LL
6




Figure 5-14. Adsorption and oxidation of gaseous Hg by various coal fly ashes at

            149 C (300 F) and 371 C (700 F) (source: Reference 52).
                                     5-44

-------
chemistry of the finer fly ash being enriched in trace metals or other condensed or adsorbed
compounds from the flue gas during the combustion of the bituminous coal.

5.6.3  Gas and Liquid Oxidation Reagents

       Argonne National Laboratory has been investigating the use of oxidizing agents that
could potentially convert gaseous Hg into more soluble species that would be absorbed in wet
FGD systems.53 Current research is focused on a process concept that involves introduction of
an oxidizing agent into the flue gas upstream of the scrubber.  The oxidizing agent employed is
NOXSORB J , which is a commercial product containing chloric acid and sodium chlorate.
When a dilute solution of this agent was introduced into a gas stream containing gaseous Hg and
other typical flue-gas species at 300 F (149 C), it was found that nearly 100 percent of the
gaseous Hg was removed from the gaseous phase and recovered in process liquids. A
significant added benefit was that approximately 80 percent of the NO was removed at the same
time.  Thus, the potential exists for a process that combines removal of 862, NO, Hg, and,
perhaps, PM.

       Continuing laboratory research efforts are acquiring the data needed to establish a mass
balance for the process. In addition, the effects of such process parameters as reagent
concentration, 862 concentration, NO concentration, and reaction time (residence time) are being
studied. For example, SO2 has been found to decrease slightly the amount of gaseous Hg
oxidized while appearing to increase the removal of NO from the gaseous phase. Preliminary
economic projections, based on the results to date, indicate that the chemical cost for NO
oxidation could be less than $5,000/ton NO removed; while for gaseous Hg oxidation, it would
be about $20,000/lb  Hg removed.  These results will be refined as additional experimental
results are obtained.
5.7 Observations and Conclusions

       When coal is burned in an electric utility boiler, the resulting high combustion
temperatures in the vicinity of 1500 C (2700 F) vaporize the Hg in the coal to form gaseous Hg.
Subsequent cooling of the combustion gases and interaction of the gaseous Hg with other
combustion products result in a portion of the Hg being converted to other forms, viz., Hg + and
Hgp. The term speciation is used to denote the relative amounts of these three forms of Hg in the
flue gas of the boiler. It is important to understand how Hg speciates in the boiler flue gas
because, as discussed in Chapters 6 and 7, the overall effectiveness of different control strategies
for capturing Hg often depends on the concentrations of the different forms of Hg species present
in the boiler flue gas.

       The speciation of Hg results from oxidation of Hg in the boiler flue gas, with the
predominant oxidized Hg species believed to be HgC^.  The mechanisms for this oxidation
include gas-phase oxidation, fly-ash-mediated oxidation,  and oxidation by post-combustion NOx
controls. Data reveal that gas-phase oxidation is kinetically limited and occurs due to reactions
                                          5-45

-------
of Hg with oxidizers such as Cl and C^. Research also suggests that gas-phase oxidation may be
inhibited by the presence of NO, 862, and water vapor.

       Certain fly ashes have been shown to promote oxidation of Hg more than others.  The
differences in oxidation appear to be attributable to the composition of the fly ash and the
presence of certain flue gas constituents.  The results of bench-scale research conducted at EPA
reflect that the presence of HC1 and NOx in flue gas and iron in fly ash assists in oxidation.
                                                   9-1-
Other research indicates that y-Fe2O3 may be causing Hg   formation, and that surface area may
be a dominant factor in this regard. Also, there are indications that HC1, NO2, and 862 in the
flue gas may contribute to Hg oxidation, while the presence of NO may suppress Hg oxidation.

       The understanding of Hg speciation in the flue gas of coal-fired electric utility boilers is
far from being mature, and research and development efforts are currently underway to develop
more information.

       Mercury can be captured and removed from a flue gas stream by injection of a sorbent
into the exhaust stream with subsequent collection in a PM control device such as an  electrostatic
precipitator or a fabric filter. However, adsorptive capture of Hg from flue gas is a complex
process that involves many variables. These include the temperature and composition of the flue
gas, the concentration of Hg in the exhaust stream, and the physical and chemical characteristics
of the sorbent (and associated functional group). The implementation of an effective  and
efficient Hg control strategy using sorbent injection requires the development of low-cost and
efficient Hg sorbents.  Of the known Hg sorbents, activated carbon and calcium-based sorbents
have been the most actively studied.  However, improved versions of these sorbents and new
classes of Hg sorbents can be expected, as this is still a very active field of study.

       Adsorption of elemental Hg is enhanced by the presence of functional groups  and/or
catalytically active sites (that oxidize the Hg to Hg2+).  Oxidation of elemental Hg to ionic
species by the catalytic components that may be present in fly ash (especially iron compounds) is
a critical step before adsorption of the species by the fly ash or some injected sorbents. Both the
oxidant and binding sites for the adsorption of elemental Hg may also be provided by the
injected sorbents.  Also, alkaline components of the fly ash exhibit sorptive properties for
oxidized Hg.  Fly ashes, which contain higher unburned carbon contents, such as those produced
from low-NOx burner systems, may have significant catalytic and sorptive properties. The
unburned carbon appears to have some oxidant/adsorption sites similar to those that existed in
the activated carbon sorbents.

       Activated carbon binding sites may be enhanced by impregnation with an active additive
(e.g., 8, Cl, I) or pretreatment (e.g., with H2SO4 or HC1). It appears that the presence of
hetroatoms, such as sulfur and chlorine, on the activated carbon surfaces greatly enhance the
adsorption of Hg.  Other non-carbon-based sorbents may be enhanced by oxidant/catalyst
additions. The enhancement is caused by the oxidation of the elemental Hg either by the added
oxidant or by the added catalyst to the sorbents. A promising alternative appears to be the
replacement of the coal-based activated carbons with a low cost, high-capacity, reactive sorbent.
Such sorbents are currently under development.
                                          5-46

-------
       Oxidized Hg is readily absorbed by alkaline solutes/slurries or adsorbed by alkaline PM
(or by sorbents). Flue gas desulfurization systems, which use alkaline materials to neutralize the
acidic 862 gas, remove oxidized Hg effectively in the flue gas.  Current research is focusing on
optimization of the existing desulfurization systems as a retrofit technology for controlling
oxidized Hg emissions and on development of new multipollutant control technologies for
simultaneously controlling both 862 and oxidized Hg emissions.
5.8 References

1.   Senior, C.L., A.F. Sarofim, T. Zong, JJ. Helble, and R. Mamani-Paco.  Gas phase
    transformation of mercury in coal-fired power plants.  Fuel Processing Technology, 63,
    (2-3): 197-214 (2000).

2.   Senior, C.L., L.E. Bool, J. Morency, F. Huggins, G.P. Huffman, N. Shah,  J.O.L.Wendt, F.
    Shadman, T. Peterson, W. Seames, B. Wer, A.F. Sarofim, I. Olmeze, T. Zeng, S. Growley,
    A. Kolker, C. A. Palmer, R. Finkelman, JJ. Helble, and MJ. Wornat.  Toxic substances from
    coal combustion - a comprehensive assessment. Physical Science, Inc., Final Report
    (Contract No. DE-AC-22-95, PC 951011, U.S. Department of Energy, Federal Energy
    Technology Center).  September 1997.

3.   Senior, C.L., JJ. Helble, and A.F. Sarofim.  "Predicting the speciation of mercury emissions
    from coal-fired power plants." Paper presented at the Conference on Air Quality II: Mercury,
    Trace Elements, andParticulate Matter, McLean, VA. September 19-21, 2000.

4.   Ghorishi, S.B., C.W. Lee, and J.D. Kilgroe. "Mercury speciation in combustion systems:
    studies with simulated flue gases and model fly ashes." Paper presented at the 92nd Annual
   Meeting of Air & Waste Management Association, St. Louis, MO. June 20-24, 1999.

5.   Edwards, J.R., R.K. Srivastava, and J.D. Kilgroe. A study of gas-phase mercury speciation
    using detailed chemical kinetics.  Published in Journal of Air & Waste Management
   Association, 5:  869-877(2001).

6.   Niksa, S., JJ. Helble, and N. Fujiwara. "Interpreting laboratory test data on homogeneous
    mercury  oxidation in coal-derived exhausts." Paper presented 94th Annual Meeting of the
    Air & Waste Management Association, Paper # 86, Orlando, FL. June 24 -28, 2001.
7.  Lee, C. W., J.D. Kilgroe, and S.B. Ghorishi.  "Speciation of mercury in the presence of coal
   and waste combustion fly ashes." Presented at the 93rd Annual Meeting of the Air & Waste
   Management Association, Salt Lake City, UT. June 18-22, 2000.
   Lee, C.W., R.K. Srivastava, J.D. Kilgroe, and S.B. Ghorishi. "Effects of iron content in coal
   combustion fly ashes on speciation of mercury." Paper presented at the 94th Annual Meeting
   of the Air & Waste Management Association, Paper # 156, Orlando, FL. June 24 -28, 2001.
                                         5-47

-------
9.  Galbreath, K.C., CJ. Zygarlicke, D.L. Toman, and R.C. Schulz. "Effects of NOX and oc-
   Fe2O3 on mercury transformations in a 7-kW coal combustion system." Paper presented at
   the 94th Annual Meeting of the Air & Waste Management Association, Paper # 767,
   Orlando, FL. June 24 -28, 2001.

10. Norton, G.A., H. Yang, R.C. Brown, D.L. Laudal, G.E. Dunham, J. Erjave, and J.M. Okoh.
   "Role of fly ash on mercury chemistry in simulated flue gas streams." Paper presented at the
   94th Annual Meeting of the Air & Waste Management Association, Paper # 164, Orlando,
   FL. June 24-28, 2001.

11. Haythornthwaite, S., S. Sjostrom, T. Ebner, J. Ruhl, R. Slye, J. Smith, T. Hunt, R. Chang,
   and T.D. Brown. "Demonstration of dry carbon-based sorbent injection for mercury control
   in utility ESP's and baghouses." In Proceedings of the EPRI/DOE/EPA Combined Utility
   Air Pollutant Control Symposium, EPRI TR-108683-V3; Washington, DC. August 25-29,
   1997.

12. Laudal, D.L., M.K. Heidt, T.D. Brown, and B.R. Nott.  "Mercury speciation: a  comparison
   between method 29 and other sampling methods." Presented at the 89th Annual Meeting of
   the Air & Waste Management Association, Nashville, TN, Paper 96-W64A.04. June 1996.

13. Brown, T. D., D.N.  Smith, R.A. Hargis,  Jr., and W.J. O'Dowd. "1999 Critical Review:
   Mercury Measurement and Its Control:  What We Know, Have Learned, and Need to Further
   Investigate," Journal of the Air & Waste Management Association, June  1999. pp. 1-97.
   Available at: < http://www.lanl.gov/projects/cctc/resources/pdfsmisc/haps/CRIT991.pdf>.

14. Li, Z., and J.Y. Hwang. "Mercury distribution in fly ash compounds." Presented at the Air
   & Waste Management Association Annual Meeting,  Toronto, Ontario, Canada. June 8-13,
   1997.

15. Huggins, F.E., N. Yap, G.P. Huffman, and J.K. Neathery.  "Investigation of mercury
   adsorption on Cherokee fly-ash using XAFS spectroscopy." Presented at the 93rd Annual
   Meeting of the Air & Waste Management Association, Salt Lake City, UT. June 18-22, 2000.

16. Carey, T.R., O.W. Hargrove, Jr., C.F. Richardson, R. Chang, F.B. Meserole. Performance of
   Activated Carbon for Mercury Control in Utility Flue Gas Using Sorbent Injection, In
   Proceedings of the EPRI/DOE/EPA Combined Utility Air Pollutant Control Symposium,
   Washington, DC; EPRI TR-108683-V3. August 25B29, 1997.

17. Galbreath, K.C., and C.J. Zygarlicke.  Mercury transformation in coal combustion flue gas.
   Fuel Processing Technology,  65-66: 289-310 (2000).
                                         5-48

-------
18. Brunauer, S., P.H. Emmett, and E. Teller, J. Am. Chem. Soc., 60, 309. 1938.

19. Hsi C.,  J. Rood, M. Rostam-Abadi, S. Chen, and R. Chang.  "Effects of sulfur impregnation
   temperature on the properties and mercury adsorption capacities of activated carbon fibers
   (ACFs)."  Environmental Science and Technology, 35, 2785-2791. 2001.

20. Bansal R.C., J.B. Donnet, and F. Stoecki. Active Carbon. New York, NY, and Basel,
   Switzerland: Marcel Dekker. 1988.

21. Ghorishi,  S.B., and B.K. Gullett. Sorption of mercury species by activated carbons and
   calcium-based sorbents: effect of temperature, mercury concentration and acid gases.  Waste
   Management & Research, 16: 6: 582-593.  1998.

22. Krishman, S.V., B.K. Gullett, and W. Jozewicz. Sorption of Elemental Mercury by
   Activated Carbons. Environmental Science and Technology, 28(8):  1506-1512(1994).

23. Serre, S.D., B.K. Gullett, and S.B. Ghorishi. Entrained-flow adsorption of mercury using
   activated carbon. Journal of the Air & Waste Management Association, 51: 733-741 (May
   2001).

24. Helfritch, D.G., P.L. Feldman, and S.J. Pass. "A circulating fluid bed fine particulate and
   mercury control concept." Presented at the EPRI/DOE/EPA Combined Utility Air Pollutant
   Control Symposium, Washington DC. August 1997.

25. Hargis, R.A., WJ. O'Dowd, and H.W. Pennline. "Sorbent injection for mercury removal in
   a pilot-scale coal combustion unit."  Presented at the 93rd Annual Meeting of the Air &
   Waste Management Association, Salt Lake City, UT. June 18-22, 2000.

26. Waugh, E.G., B.K. Jenson, L.N. Lapatrick, F.X. Gibbons, S. Sjostrom, J. Ruhl, R. Slye, and
   R.A. Chang. "Mercury control in utility ESP's and FFs through dry  carbon based sorbent
   injection pilot-scale demonstration."  In Proceedings of the EPRI/DOE/EPA Combined
   Utility Air Pollutant Control Symposium. Washington, DC, EPRI TR-108683-V3). August
   23-29, 1997.

27. Carey, T.R., C. Richardson, R. Chang, and F.B. Meserole.  "Assessing sorbent injection
   mercury control effectiveness." Paper presented at the 1999 Spring National Meeting of the
   American Institute of Chemical Engineers, Houston, TX. March 14-18, 1999.

28. Sjostrom, S., T. Ebner, T. Ley, R. Slye, C.  Richardson, T. Machalek, M. Richardson, R.
   Chang, and  F. Meserole. "Assessing the performance of mercury sorbents in coal
   combustion  flue gas." Paper presented at the 94* Annual Meeting of the Air & Waste
   Management Association, Orlando, FL. June 24-28, 2001.
                                         5-49

-------
29. Rostam-Adadi, M., S.G. Chen, H-C. His, M. Rood, R. Chang, T.Carey, B. Hargrove, C.
   Richardson, W. Rosenhoover, and F. Meserole.  "Novel vapor phase mercury sorbents." In
   Proceedings of the First EPRI/DOE/EPA Combined Utility Air Pollution Control Symposium
   (The Mega Symposium), Washington, DC, August 25-29, 1997.

30. White, D.M., W.E. Kelly, MJ. Stucky, J.L. Swift, and M.A. Palazzolo. Emission test report:
   field test of carbon injection for mercury control, Camden County Municipal Waste
   Combustor, EPA/600/R-93/181 (NTIS PB94-101540), U.S. EPA, Air and Energy
   Engineering Research Laboratory, Research Triangle Park, NC. September 1993.

31. Serre, S.D., B.K. Gullett, and S.B. Ghorishi. "Elemental mercury capture by activated carbon
   in a flow reactor." Paper presented at 93rd Annual Meeting of the Air & Waste Management
   Association, Salt Lake City, UT. June 18-22, 2000.

32. Serre, S.D., B.K. Gullett, and Y. H. Li. "The effect of water (vapor-phase and carbon) on
   elemental mercury removal in a flow reactor." Paper presented at 94th Annual Meeting of
   the Air & Waste Management Association, Paper # 164, Orlando, FL. June 24 -28, 2001.

33. Ghorishi, S.B., R. Keeney, W. Jozewicz, S. Serre, and B. Gullett. "In-flight capture of
   elemental mercury by a chlorine-impregnated activated carbon."  Paper #731 presented at
   the 94th Annual Meeting of Air & Waste Management Association, Orlando, FL. June 24-
   28,2001.

34. Brown, T. D., D.N. Smith, R.A. Hargis, Jr., and W.J. O'Dowd. "1999 Critical Review:
   Mercury Measurement and Its Control:  What We Know, Have Learned, and Need to Further
   Investigate," Journal of the Air &  Waste Management Association., June 1999. pp. 1-97.

35. Ghorishi, S.B., R.M. Keeney, and B.K. Gullett.  "Role of surface functional groups in the
   capture of elemental mercury and mercuric chloride by activated carbons." In Proceedings
   of the Air Quality II Conference, McLean, VA. September 19-21, 2000.

36. Gullett, B.K., S.B. Ghorishi, K. Raghunathan, and K. Ho.  Removal of Coal-Based Volatile
   Trace Elements: Mercury and Selenium, Final Technical Report. September 1, 1995, through
   August 31, 1996.

37. Ghorishi, S.B., and C.B. Sedman.  "Combined Mercury and Sulfur Oxides Control Using
   Calcium-Based Sorbents." Paper presented at the EPRI/DOE/EPA Combined Utility Air
   Pollutant Control Symposium, Washington, DC. August 25-29, 1997.

38. Ghorishi, S.B., and C.B. Sedman.  Low concentration mercury sorption mechanisms and
   control by calcium-based sorbents: application in coal-fired processes, Journal of the Air &
   Waste Management Association, 48:1191-1198,1998.
                                         5-50

-------
39. Ghorishi, S.B., C. Singer, W. Jozewicz, C. Sedman, andR. Srivastava. "Simultaneous control
   of Hg, SC>2, and NOxby novel oxidized calcium-based sorbents." Paper # 243, Presented at
   the 94th Annual meeting of the Air & Waste Management Association. June 24-28, Orlando,
   FL, 2001.

40. EPRI Report 1000454. Development and evaluation of low cost mercury sorbents.
   November 2000.

41. EPRI Report TE-114043.  Development and evaluation of mercury sorbents. November
   1999.

42. EPRI Report TR-110532.  Development and evaluation of low-cost sorbents for removal of
   mercury emissions from coal combustion flue gas. September 1998.

43. Rostam-Abadi, M., S. Chen, A.A. Lizzio, H-C. His, C.M.B. Lehmann, M. Rood, R. Chang,
   C. Richardson, T. Machalek, and M. Richardson. "Development of low-cost sorbents for
   mercury removal from utility flue gas." Paper presented at U.S. EPA/DOE/EPRI Combined
   Power Plant Air Pollutant Control Symposium, and the Air & Waste Management
   Association Specialty Conference on Mercury Emissions: Fate, Effects, and Control,
   Chicago, IL. August 20-23, 2001.

44. Korpiel, J.A., and R.D. Vidic. Effect of sulfur impregnation method on activated carbon
   uptake of gas-phase mercury. Environmental Science and Technology, 31: 2319-2326
   (1997).

45. Vidlic, R. D. and D.P. Siler.  Vapor-phase elemental mercury adsorption by activated carbon
   impregnated with chloride and chelating agents. Carbon.  3-14 (2001).

46. Li, Y. H., S.D. Serre, C.W Lee, and B.K. Gullett. "Elemental mercury adsorption by
   activated carbon treated with sulfuric acid."  Presented at the U.S. EPA/DOE/EPRI
   Combined Power Plant Air Pollutant Control Symposium, and the Air & Waste Management
   Association Specialty Conference on Mercury Emissions: Fate, Effects, and Control,
   Chicago, IL. August 20 -23,  2001.

47. Li, Y. H., C.W. Lee, and B.K. Gullett. "Characterization of activated carbons' physical and
   chemical properties in relation to their mercury adsorption." Presented at the American
   Carbon Society CARBON '01, An International Conference on Carbon, University of
   Kentucky Center for Applied Energy Research, Lexington, KY. July 14 - 19, 2001.

48. Meserole, F., C. F. Richardson, T. Machalek, M. Richardson, and R.  Chang. "Predicted
   Costs of Mercury Control at Electric Utilities Using Sorbent Injection." Presented at U. S.
   EPA/DOE/EPRI Combined  Power Plant Air Pollutant Control Symposium, and the Air &
   Waste Management Association Specialty Conference on Mercury Emissions: Fate, Effects,
   and Control, Chicago, IL, August 20-23, 2001.
                                         5-51

-------
49. DeVito, M.S., and W.A. Rosenhoover.  "Hg flue gas measurements from coal-fired utilities
   equipped with wet scrubbers." Presented at 92nd Annual Meeting of the Air & Waste
   Management Association, St. Louis, MO. June 20-24, 1999.

50. Redinger, K.E., A. Evans, R. Bailey, and P. Nolan. "Mercury emissions control in FGD
   systems." Presented at the EPRI/DOE/EPA Combined Air Pollutant Control System,
   Washington, DC. August 25-29, 1997.

51. McDermott Phase III Study Section, McDermott Technologies, Inc. Advanced Emissions
   Control Development Program Phase III - Approved Final Report, prepared for the U. S.
   Department of Energy (US DOE-FETC contract DE-FC22-94PC94251-22) and  Ohio Coal
   Development Office (grant agreement CDO/D-922-13). July 1999.  Available at:
   < http://www.osti.gov/dublincore/servlets/purl/756595-LACvcL/webviewable/756595.pdf>.

52. Hargrove, O.W., Jr., T.R. Carey, C.F. Richardson, R.C.  Sherupa, F.B. Meserole, R.G. Rhudy,
   and T.D. Brown. "Factors affecting control of mercury by wet FGD." Paper presented at the
   EPRI/DOE/EPA Combined Utility Air Pollutant Control Symposium. Washington DC.
   August 25-29, 1997.

53. Livengood, C.D., and M.H. Mendelsohn. "Process for combined control of mercury and
   nitric oxide." Presented at the EPRI/DOE/EPA Combined Utility Air Pollutant Control
   Symposium, Atlanta, GA. August 16-20, 1999.
                                        5-52

-------
                                     Chapter 6

    MERCURY CAPTURE BY EXISTING CONTROL SYSTEMS USED BY
                 COAL-FIRED ELECTRIC UTILITY BOILERS


6.1    INTRODUCTION

       Existing coal-fired electric utility boilers in the United States use a variety of emission
control technologies to meet air standards for sulfur dioxide (SO2), nitrogen oxides (NOX),
and particulate matter (PM).  The EPA's ICR data presented in Chapter 3 of this report
indicate that most electric utilities are controlling NOX emissions from their coal-fired boilers
by combustion modification techniques and controlling SO2 emissions by burning low-sulfur
coal. All of the coal-fired electric utility boilers use some type of post-combustion control
device to meet PM emission standards. Of these PM controls, electrostatic precipitators
(ESPs) are the predominant control type used on coal-fired boiler units (83 percent) with the
second most common control device  being a fabric filter (14 percent).  Use of post-
combustion SO2 controls is less common: approximately 20 percent of the boiler units use
either wet flue gas desulfurization (FGD) systems (15 percent) or spray dryer absorber (SDA)
systems (5 percent). While the use of either selective non-catalytic reduction (SNCR) or
selective catalytic reduction (SCR) on coal-fired electric utility boilers for NOx emission
control presently is very limited (less than 4 percent), the application of these post-
combustion NOx controls is becoming more prevalent.

       The implementation of post-combustion controls is not specifically intended to
control mercury emissions from coal-fired utility boilers.  However,  these controls capture
mercury in varying degrees depending on the control technologies used and  the mercury
speciation at the inlet to the control device(s). This chapter discusses mercury capture by
existing post-combustion control  systems used by coal-fired utility boilers.   An estimate of
nationwide mercury emissions from existing coal-fired utility boilers is presented.  The
mechanisms by which existing post-combustion control systems capture mercury are
reviewed. The ICR mercury emission test data for mercury capture by the existing post-
combustion control systems used  for  coal-fired utility boilers are presented and discussed.
6.2 EPA ICR PART HI DATA

       As introduced in Chapter 1 of this report, the EPA conducted a three-part data
collection effort to gather information about the coal-fired utility boilers operating in the
United States in 19991. The Part I ICR data consist of information on the coal types burned,
the boiler furnace types, and the air pollutant control devices used for the 1,143 coal-fired
utility boilers in the United States having a capacity equal to or greater than 25 MWe. These
data are summarized and discussed in Chapters 2 and 3 of this report. The Part II ICR data

                                         6-1

-------
consist of information on the quantity, mercury content, and other selected properties of coal
burned by each of the identified 1,143 boiler units during calendar year 1999. A summary and
evaluation of these data are presented in Section 2.7 and Appendix A of this report.  For Part
in of the information collection effort, the EPA selected a subset of the coal-fired electric
utility boilers for which field source testing was performed to obtain mercury emission data
for the air pollutant control devices now being used for these units.  This chapter presents a
summary and analysis of the emissions data collected by Part in of EPA's information
collection effort.

       The EPA ICR Part III data are composed of mercury emission source test results for 80
coal-fired electric utility boilers.  These boiler units were selected by the EPA to be generally
representative of the nationwide population of coal-fired utility boilers according to the type of
boiler used, the type of coal burned, and the air emission controls used. For each of the tested
boiler units, the flue gas mercury measurements were generally made at the inlet and outlet of
control device(s). The mercury measurements were made using the OH Method for speciated
mercury (this test method is discussed in Section 4.1 of this report).  Also, samples of the coal
being burned in the boiler unit during the source test were collected and analyzed for mercury
content.

       For boiler units that use a control configuration consisting of a  single PM control
device, the flue gas samples were collected at the inlet to the PM control  device and in the
stack. For units using SDA systems, the flue gas measurements were made at the inlet to the
SDA and in the stack. For units using an ESP or FF followed by a wet FGD scrubber, the flue
gas measurements were taken at the inlet to the wet scrubber inlet (i.e., downstream of the PM
control device) and in the stack. For units equipped with a PS and a wet FGD scrubber,
measurements were made at the inlet to the PS device and in the stack.

       Of the three IGCC plants located in the United States, two of the plants (Polk Power
Station and Wabash River Repowering Project) were included as part of the Part HI ICR test
program. At both facilities, combustion gas measurements using the OH Method were made
at the exhaust stack of the gas turbines.  During testing, coal feed rates to the coal-
gasification units were recorded. Coal samples were collected during testing and analyzed
for total mercury.

       A summary of 81 boiler and coal type configurations for which mercury emission data
were collected is given in Table 6-1.  Of these boiler units, 65 were pulverized-coal-fired
(PC-fired) boilers.  Such boilers account for the vast majority of the 1,143 coal-fired electric
utility boilers operating in the United States in terms of both total units and nationwide
generating capacity as shown Table 2-4.
                                         6-2

-------
Errata Page 6-3, dated 3-21-02
                                  Table 6-1
              Distribution of ICR Mercury Emission Test Data
                     By Boiler-coal Type Configurations
Boiler Unit
Type
Pulverized-coal-
fired
Cyclone-fired
Fluidized-bed
Combustor
Stoker-fired
IGCC (b)
Total Number
of Units Tested
Number of Boiler Units Tested
Fuel Burned In Boiler Unit
Bituminous
Coal
26
3
1
2
2
34
Subbituminous
Coal
29
2
0
0
0
31
Lignite
9
2
2
0
0
13
Other(a)
1
0
2
0
0
3
Total
Number of
Units
Tested
65
7
5
2
2
81
      (a) Some units used coal wastes or a blend of fuels.
      (b) Integrated coal gasification combined cycle unit.
      A summary of the flue gas cleaning devices installed on the PC-fired test units is given
in Table 6-2 as a function of type of fuel burned in each unit in 1999. These data show that:
        A total of 28 test units were equipped with a CS-ESP (14), HS-ESP (8), or FF (6).
        The 11 dry FGD units were equipped with either a SDA/ESP (3) or SDA/FF (8).
        The 20 wet FGD units were equipped with a PS + Wet FGD (6), CS-ESP + Wet
         FGD (6), HS-ESP + Wet FGD (6), or FF + Wet FGD (2).
        Two units were equipped with a CS-ESP + FF.
        One was equipped with a PS.
                                      6-3

-------
Errata Page 6-4, dated 3-21-02

                                    Table 6-2
 Distribution of ICR Mercury Emission Test Data for Pulverized-coal-fired
    Boilers By Post-combustion Emission Control Device Configuration
Post-
combustion
Control
Strategy
PM Control Only
PM Control and
Dry SO2 Scrubber
System
PM Control and
Wet SO2 Scrubber
System
Post-combustion
Emission Control
Device
Configuration
CS-ESP
HS-ESP
FF
CS-ESP + FF
PS
SDA + CS-EP
SDA + FF
DI + CS-ESP
PS + wet FGD
CS-ESP + wet FGD
HS-ESP + wet FGD
FF + wet FGD
Other Control Device Configuration
Number of Units Tested
Number of Boiler Units Tested
Fuel Burned In Boiler Unit
Bituminous
Coal
7
4
4
0
0
0
3
1
1
1
1
2
2
27
Subbituminous
Coal
5
4
2
0
1
3
3
0
4
3
5
0
0
29
Lignite
1
0
0
2
0
0
2
0
1
2
0
0
0
8
Other
1
0
0
0
0
0
0
0
0
0
0
0
0
1
Total
14
8
6
2
1
3
8
1
6
6
6
2
2
65
   PM Controls
   CS-ESP = cold-side electrostatic precipitator
   HS-ESP = hot-side electrostatic precipitator
   FF = fabric filter
   PS = particle scrubber
SO? Controls
DI = dry injection
FGD = flue gas desulfurization system
SDA = spray dryer adsorber system
6.3 MERCURY CONTENT OF UTILITY COALS BURNED IN 1999

       The analysis results of more than 39,000 coal samples were reported in the Part II ICR
data. These results include the mercury content of as-fired coals and supplemental fuels
burned in electric utility boilers in 1999.  A comparison of the mercury contents of the
different major coal types and supplemental fuels burned by electric utilities in 1999 and
normalized by fuel heating value is shown Figure 6-1.  Waste bituminous coal and waste
anthracite had the highest mercury contents expressed in Ib Hg/1012 Btu. The mercury content
of the bituminous coal, subbituminous coal, and lignite (the three most commonly used fuels)
was generally less than 15 lb/1012 Btu. Statistical information on each type of fuel burned in
coal-fired utility boilers is presented in Table 6-3.
                                       6-4

-------
ON
200

150
CO
*1s

1 00
,a A ww
"wo
M


50


-












X
x



T X x x
-I J L
_^^__^^__^^_^^^___J__ F^q r^n







x

T

n

J_

x


















>




:







J
	 \







:







L
^













T
a j u
T  r
x

















Legend
~ I001h Percentik
X 99th Percentile
T 951h Percentile

n Mean

Median
5th Percentile
X 1st Percentile
0 Percentile






              0
                Bituminous          Subbituminous          Lignite
                            South             Indonesian            Petroleum
                           American                                 Coke
                                                  Tires                Waste
                                                           Waste    Bituminous   Waste
                                                          Anthracite           Subbituminous
   No. of Analyses  27,8
270
78       1,047     1,149

       Fuel Type
149
377
575
53
                                       Figure 6-1. 1999 ICR data analyses - mercury in fuels.

-------
                                      Table 6-3
            Comparison of Mercury Content Normalized By Heating Value
      In As-fired Coals and Supplemental Fuels for Electric Utility Boilers in 1999

Fuel Type
Anthracite coal
Bituminous coal
South American bituminous coal (a)
Subbituminous coal
Indonesian subbituminous coal (b)
Lignite
Waste anthracite coal
Waste bituminous coal
Waste subbituminous coal
Petroleum coke
Tire-derived fuel
Number
of
Analyses
114
27,884
270
8,193
78
1,047
377
575
53
1,149
149
Ratio of Mercury to Fuel Heat Content
(Ib Hg per 1012 Btu)
Range
5.02-35.19
0.04-103.81
0.70-66.81
0.39-71.08
0.79-4.61
0.93-75.06
2.49-73.02
2.47 -172.92
5.81-30.35
0.06-32.16
0.38-19.89
Mean
15.28
8.59
5.94
5.74
2.51
10.54
29.31
60.50
11.42
23.18
3.58
Median
13.37
7.05
4.91
5.00
2.39
7.94
27.77
53.32
10.79
2.16
2.79
Standard
Deviation
6.23
6.69
5.28
3.59
0.86
9.05
11.94
44.35
4.66
3.18
2.78
(a) Bituminous coal imported from South America and burned at one power plant in Florida and one power
   plant in Texas.
(b) Subbituminous coal imported from Indonesia and burned at a coal-fired power plant in Hawaii.
6.4  POTENTIAL MERCURY CAPTURE IN EXISTING UNITS

       Mercury capture in existing units depends on Hg speciation at the inlet to the control
device(s) and the type(s) of control technologies used. Units that burn bituminous coals have
relatively high concentrations of Hg2+ at the inlet to the control device(s).  Units that burn
subbituminous coal or lignite typically have relatively low concentrations of Hg2+ and high
concentrations of Hg at the inlet to the control device(s).

       The effects of coal and combustion conditions are attributed primarily to the flue gas
composition and properties of fly ash that  affect the speciation and capture of Hg. While OH
measurements made upstream of PM control devices do not always provide quantitatively
accurate information on Hg speciation, they do provide semi-quantitative information relative
to the amounts of Hgp, Hg2+,  and Hg in flue gas from the combustion of different types of

-------
coals. They also provide useful information on the potential for the oxidization of the Hg and
the capture of the resulting reaction products in downstream control devices.

       The relatively high concentrations of chlorine in bituminous coals are believed to
result in the oxidization of Hg to form Hg2+, primarily HgC^. By contrast, both subbituminous
coal and lignite have lower amounts of chlorine and higher amounts of alkaline material
(calcium and sodium) than bituminous coals. Chlorine from the combustion of subbituminous
coal and lignite tends to react with the alkaline materials in flue gas, and little if any chlorine
is available for the oxidization of Hg.  Therefore, flue gas from combustion of subbituminous
coal and lignite tends to have relatively low concentrations of Hg2+.

6.4.1  Units with an ESP or FF

       Approximately 77 percent of the coal-fired utility boilers currently operating in the
United States are equipped with only an ESP or an FF. Gaseous mercury (both Hg and Hg2+)
can potentially be adsorbed on fly ash and be collected in a downstream ESP or FF. The
modern ESPs or FFs that are now used on most coal-fired units achieve very high capture
efficiencies for total paniculate matter (see Table 3-3). As a consequence, these PM control
devices are also effective in capturing Hgp in the boiler flue gases.

       The degree to which mercury can be adsorbed onto fly ash for subsequent capture in
PM control is dependent on the speciation of mercury, the flue gas concentration of fly ash,
and the properties of fly ash.  It is currently believed that mercury is primarily adsorbed onto
the unburned carbon in fly ash (see Section 5.3). Approximately 80 percent of the coal ash in
PC-fired boilers is entrained with the flue gas as fly ash.  PC-fired boilers with low-NOx
burners have higher levels of carbon in the fly ash with a correspondingly higher potential for
mercury adsorption. Cyclone and stoker boilers tend to have high levels of carbon in the fly
ash, but have lower flue gas concentrations of fly ash than PC-fired boilers.  Fly ash
concentrations in fluidized-bed combustors tend to be higher than those in PC-fired boilers.
Also,  the carbon content of fluidized-bed combustor fly ash is generally higher than that of
PC-boiler fly ash.

       The syngas from a coal gasifier is composed mainly of hydrogen, carbon monoxide,
carbon dioxide, and nitrogen. This gas also contains vaporous trace elements, such as
mercury, as well as dust and aerosols containing trace elements.  The source of mercury in
syngas is the mercury that is naturally present in coal  and is released during the gasification
processes, which typically takes place at 950 C (1750 F). Mercury that is not retained in the
solid residue from the gasification  process is released almost exclusively as Hg.

       Gas-phase mercury in units equipped with an ESP can be adsorbed on the entrained
fly ash upstream of the ESP.  The gas-phase mercury in units equipped with a FF can be
adsorbed by entrained fly ash or it  can be adsorbed  as the flue gas passes through the filter
cake on the surface of the FF. The degree to which gaseous mercury adsorbs on the filter
cake typically depends on the speciation of gaseous mercury in the flue gas; in general,
gaseous Hg2+ is easier to adsorb than gaseous Hg (see discussion in Section 5.3.1).

                                         6-7

-------
6.4.2  Units with SPA Systems

       An SDA system operates by the same principle as a wet FGD system using a lime
scrubbing agent, except that the flue gas is mixed with a fine mist of lime slurry instead of a
bulk liquid (as in wet scrubbing). The SO2 is absorbed in the slurry and reacts with the
hydrated lime reagent to form solid calcium sulfite and calcium sulfate. The heat of the flue
gas, leaving dry solid particles of calcium sulfite and calcium sulfate, evaporates the water in
the mist. Entrained particles (unreacted sorbent particles, reaction products, and fly ash) are
captured in the downstream PM control device (either an ESP or FF).

       The performance of SDA systems in controlling SC>2 emissions is dependent on the
difference between the SDA outlet temperature and the corresponding flue gas water vapor
saturation temperature.  SDA systems on coal-fired boilers typically operate about 20 F
(11 C) above the saturation temperature (i.e.,  a 20 F [11 C] approach to saturation
temperature). The relatively low flue gas temperatures afforded by SDA systems increase the
potential for mercury capture.  The caking or buildup of moist fly ash deposits, which can
plug the SDA reactor and coat downstream surfaces, dictates the minimum flue gas
temperatures, which can be employed at the outlet of SDAs.

       Hgp is readily captured in SDA systems. Both Hg and Hg2+ can potentially be
adsorbed on  fly ash,  calcium sulfite, or calcium sulfate particles in the SDA. They can also
be adsorbed and captured as the flue gas passes through the ESP or FF, whichever is used for
PM control.  In addition, gaseous Hg2+ may be absorbed in the slurry droplets and react with
the calcium-based sorbents within the droplets. Nearly all of the Hgp can be captured in the
downstream  PM control device.  If the PM control device is a FF, there is the potential for
additional capture of gaseous mercury as the flue gas passes through the bag filter cake
composed of fly ash and dried slurry particles.

6.4.3  Units with Wet FGD Systems

       Approximately 15 percent of coal-fired utility boilers in the United States use wet
FGD systems to control SO2 emissions.  In each of these systems, a PM control device is
installed upstream of the wet FGD scrubber. PM control devices used with wet FGD
scrubbers include particulate scrubbers  (PS), CS-ESPs, HS-ESPs, and FF baghouses. As
described in  Chapter 3, wet FGD systems remove gaseous SC>2 from flue gas by absorption.
In wet scrubbers, gaseous species are mixed with a liquid in which they are soluble.  For SO2
absorption, gaseous  SO2 is  mixed with a caustic slurry, typically water  and limestone or water
and lime.

       Gaseous compounds of Hg2+ are generally water-soluble and can absorb in the
aqueous slurry of a wet FGD system. However, gaseous Hg is insoluble in water and
therefore does not absorb in such slurries.  When gaseous compounds of Hg2+ are absorbed in
the liquid slurry of a wet FGD system, the dissolved species are believed to react with

-------
dissolved sulfides from the flue gas, such as H2S, to form mercuric sulfide (HgS); the HgS
precipitates from the liquid solution as sludge. In the absence of sufficient sulfides in the
liquid solution, a competing reaction that reduces/converts dissolved Hg2+ to Hg is believed
to take place. When this conversion takes place, the newly formed (insoluble) Hg is
transferred to the flue gas passing through the wet FGD system. The transferred Hg
increases the concentration of Hg in the flue gas passing through the wet FGD (since the
incoming Hg is not absorbed), thereby resulting in a higher concentration of gaseous Hg in
the flue gas exiting the wet FGD compared to that entering.  Transition metals in the slurry
(originating from the flue gas) are believed to play an active role in the conversion reaction
since they can act as catalysts and/or reactants for reducing oxidized species.

       Recent research on the capture of mercury in wet scrubber systems is discussed in
Section 5.6.

6.4.4 Units with Other Control Devices

       Some units use PS systems, primarily venturi scrubbers, to control PM emissions.
Capture of Hg in these systems is limited to soluble Hg compounds such as HgC^. PS
systems are typically poor fine PM collectors and, if Hgp in the flue gas is associated with
fine PM, capture of Hgp by such scrubbers may be poor. Hg is insoluble and will not
typically be captured by the scrubber. It is possible to capture Hg2+ in the wet scrubbers, but
the scrubber chemistry, and the manner in which the scrubber is operated, will determine
whether it is effectively removed, or whether it is stripped, from the scrubbing liquor.
Stripping can occur if the Hg2+ is not adsorbed on the particles, or reacted chemically with
liquid-phase reactants within the scrubber.

       Mechanical collectors such as cyclones do a poor job of capturing fine PM, and
mercury capture in these control devices should be limited to the capture of Hgp associated
with particles larger than  10 |im.

6.5  EPA'S  PART IE ICR DATA EVALUATION APPROACH

       The methods used to evaluate the Part HI ICR data were based on two interrelated
objectives. The first objective was to estimate the amount, speciation, and geographical
distribution of national mercury emissions from coal-fired power plants in 1999.  The second
was to characterize the effects of coal properties, combustion conditions, and flue gas cleaning
methods on the speciation and capture of mercury.  The satisfaction of the first objective
involved the development of mercury emission factors as a function of the type of coal burned,
the type of boiler, and the air pollution control device(s) used.

6.5.1  Evaluation Method

       The development of emission factors for different classes of coal-fired units was based
on hypotheses derived from current understanding of mercury speciation and capture, as
discussed in Chapter 5. The hypotheses are:

                                        6-9

-------
         Mercury speciation and capture are dependent on the coal properties, combustion
          conditions, and flue gas cleaning methods that are used for any specific test unit,

         Hg2+ is more readily absorbed in aqueous media than Hg, and therefore can be
          captured in wet scrubbers, while Hg cannot,

         Gas-phase mercury can be adsorbed onto the unburned carbon in fly ash, which can
          catalyze oxidation of Hg,

         Hgp can be readily captured in an ESP or an FF,

         The potential for mercury capture increases with decreasing flue gas temperatures, and

         Flue gas from combustion of bituminous coals typically has a higher fraction of
          Hg2+ than the gas from subbituminous and lignite coals.

       Combinations of coal, boiler, and control technologies that are expected to behave in
a similar manner with respect to speciation and capture of mercury can be grouped into data
sets called coal-boiler-control technology classes or bins. Many of these data sets in the ICR
database consist of tests at one or two units, and this small number of samples results in
relatively large uncertainties concerning the central values and variability of the underlying
populations. However, the mean values and statistical behavior of the classes with a large
number of test units can be investigated, and the results can be compared with the results of
classes with a small number of test sites. If the relative behavior of the large and small data
sets is consistent with our theoretical expectations, then we can have some confidence that
the speciation and capture estimates for the smaller sets are reasonable.

       The ICR Part in emission data were sorted into coal-boiler-control classes. Next, the
data in each class were evaluated for consistency, and the data between classes were evaluated
according to the postulated behavior criteria given above. With few exceptions, the differences
in speciation and capture of mercury between the different classes were consistent with the
above-hypothesized behavior.  Based on this observation, emission factors were developed for
use in estimating the amount and speciation of mercury emissions from coal-fired electric
utility boilers in 1999.  The data in the coal-boiler-control classes were also used to conduct
further evaluations of the effects of coal properties, combustion conditions, and flue gas
cleaning conditions on the control of mercury emissions at existing coal-fired power plants.

6.5.2 Measures of Performance

       Measures used to evaluate the effect of the coal, boiler, and control device variables
on the capture of mercury included the inlet and outlet concentrations of Hgp, Hg2+, Hg,  and
HgT, and the reduction of HgT. Emission factors, defined in this report to be  the fraction  of
mercury emitted to the atmosphere relative to the amount that enters the  first air pollution
control device,  were also calculated and used to evaluate the emission of speciated Hg  and

                                         6-10

-------
HgT.

       The fraction of HgT captured in air pollution control device(s) can be used
interchangeably with the emission factor for HgT [EMFT]:

       EMFT = 1 - Capture HgT

Where the fractional capture is:

       Capture HgT = [  HgT (inlet) - HgT(outlet)]/HgT(inlet) = 1 - HgT(outlet)/HgT(inlet)

And the percentage reduction (%Red) across the control device(s) is:

       %Red = 100 x [1 - HgT(outlet)/HgT(inlet) ]

       The %Red can be determined from either (1) the inlet and outlet concentrations of
HgT as measured by the  OH Method, or (2) inlet concentration estimates made from Part in
coal samples and outlet  concentrations obtained with the OH Method. When the OH
measurements are used to evaluate the reduction in emissions or emission factors, the inlet
and outlet concentrations must be expressed on a common basis |im/dscm at 3% O2) or Ib of
Hg/1012 Btu of coal burned to account for air in-leakage through fans or across the air
pollution control device(s).

       The results of the OH Method emission tests for HgT are shown in Figures 6-2  and
6-3. Figure 6-2 is a scatter plot of the inlet versus the outlet concentrations of HgT.  In
general, the outlet HgT concentration increases with increasing inlet HgT concentrations.  The
increasing outlet HgT concentrations that appear linear with respect to HgT inlet
concentrations  are indicative of a constant percentage reduction across the control device(s).
ESPs exhibit this type of performance for the control of PM.  These types of devices are
called constant reduction devices. Note that there are also a number of data points distributed
just above the x-axis; i.e.,  zero outlet concentration. These data points are indicative of
constant outlet  devices with low emission concentrations. FF baghouses tend to operate like
constant outlet  devices.

       Figure 6-3 is a scatter plot showing inlet HgT concentration versus percent reduction
in HgT across the control device(s).  There are no discernable trends in the capture of HgT as a
function of inlet concentration.  The negative emission reductions represent cases for which
the outlet HgT concentration is higher than the inlet concentration. This can result from one
or a combination of factors.  For example, negative emission reductions can occur when (1)
temperature changes within the test unit increase the desorption of Hg, (2) ESP rapping
cycles result in the reentrainment of Hgp, and (3) small differences between Hg inlet and
outlet concentrations cannot be accurately quantified because of imprecision in the OH
Method.
                                         6-11

-------
60

s5) 50 "
d
0
'-& 40 -
c<3
u
a
o
0
O
H
"an 20 -
ffi

0

00
f>O

0 
o
ijLodfo
>A O
&
,ga v
iiiaCR>.jfc
o S*S5?*oo- M^ A A A
















0 20 40 60 80 100 120 140 160
Inlet Hg(T) Concentration, (ig/m
      Fig. 6-2.  Inlet versus outlet mercury concentration for all tests.
              150
          .0   50
           O
          O
              -50
              -100
              -150
                              e	o-
                                 o<>
                             40     60     80     100    120    140    li
                           Inlet Hg(T) Concentration, (ig/m
Fig. 6-3. Inlet mercury concentration versus percent reduction for all tests.
                                6-12

-------
       Emission factors for speciated Hg can be developed by dividing or normalizing the
stack Hg species by the concentration of HgT at the inlet to the first control device. In the
development of these emission factors, it is assumed that all of the Hg in the as-burned coal is
equal to the value of HgT measured at the inlet sampling location by the OH method. The
emission factors for Hgp (EMFP), Hg2+ (EMF2+), and Hg (EMF) are calculated by:

                 EMFP = Hgp  (outlet) / HgT (inlet),

                 EMF2+ = Hg2+ (outlet) / HgT (inlet), and

                 EMF = Hg (outlet) / HgT (inlet).

For situations where HgT (outlet) is higher than HgT (inlet), the stack emission factors are
calculated by replacing the HgT (inlet) value with the corresponding HgT (outlet) value:

                 EMFp = Hgp (outlet)/HgT (outlet),   [for HgT (outlet) > HgT (inlet)],

                 EMF2+ = Hg2+ (outlet)/HgT (outlet),  [for HgT (outlet) > HgT (inlet)], and

                 EMF = Hg (outlet)/HgT (outlet),   [for HgT (outlet) > HgT (inlet)].

In the latter case, it should be noted that EMFP + EMF2+ + EMF = 1.

       In addition to the above  emission factors, speciation factors (SPFs) are calculated and
used to characterize Hg speciation at both the inlet and outlet sampling locations.  The SPFs
represent the fractions of HgT in the inlet or outlet samples that are present as Hgp, Hg2+, or
Hg. For the inlet sampling train:

                 SPFp  =  Hgp (inlet) / HgT (inlet),

                 SPF2+ =  Hg2+ (inlet) / HgT (inlet), and

                 SPF =  Hg (inlet) / Hgx (inlet).

For the outlet sampling train:

                 SPFp =  Hgp (outlet) / Hgx (outlet),

                 SPF2+ =  Hg2+ (outlet) / HgT (outlet),  and

                 SPF =  Hg (outlet) / HgT (outlet).

In all cases:
                 SPFp + SPF2+ + SPF=l.
                                         6-13

-------
       Emission factors and speciation factors for units equipped with an ESP, FF, PM
scrubber, mechanical collector, SDA/ESP, or SDA/FF were calculated using inlet OH
measurements for HgT and outlet OH measurements for speciated and HgT. For units with
wet FGD scrubbing systems, emission factors were determined by multiplying the average
emission factor for the PM control device that precedes the scrubber by the emission factors
for the scrubber as determined by  OH measurements. For example, the estimated EMFs for a
PC-fired boiler burning subbituminous coal and equipped with cold-side ESP and wet FGD
system are calculated as follows:

       The class average CS-ESP EMFT for a PC-boiler firing subbituminous coal is 0.91,
       and the class average wet FGD EMFT for a PC-boiler firing subbituminous coal is
       0.71. The EMFT across both control devices is therefore:

                    EMFT (CS-ESP + FGD) = EMFT (CS-ESP) x EMFT (FGD)

                    = 0.91x0.71 = 0.65.

       The corresponding level of control across both devices is:

                    %Reduction (CS-ESP + FGD) = 100  * [1- EMFT (CS-ESP + FGD)]

                    = 100 (1-0.65) = 35%.

       Emission factors for coal gasification units were calculated using the Hg content of
the feed coal and the OH measurements made in the stack.

6.5.3  Comparisons of HgT (Inlet) Using OH Measurement and Coal Hg Data

       Emission factors for speciated and total Hg relative to inlet Hg concentrations can be
determined using two methods. The first  method uses the HgT inlet concentrations from OH
sampling train measurements.  The second method involves the calculation of total Hg inlet
values using coal Hg data and sampling train data (flue gas flow rate, moisture concentration,
O2 concentration, and temperature).

       Emission factor estimates  determined using the OH Method train data and the ICR Part
II coal data often give significantly different results. The best estimate can sometimes be
obtained by discarding outliers, by reviewing the test reports for tests conditions that can lead
to questionable results, and by comparison of the results relative to tests at other test sites.  In
some cases, it is not possible to arrive at a best estimate, and there is a significant amount of
uncertainty leading to a range of estimates.

       Mercury capture (percent reduction in emissions) and emission factors for Hgp, Hg2+,
Hg, and HgT were then calculated using the average stack values for each data set as
determined by both coal and OH Method  sample train data.  Emission factors based on the
OH Method sampling train data provided  the most consistent results. The inlet

                                        6-14

-------
 concentrations and percentage reduction reflected in the body of this report correspond
 primarily to test results obtained using the OH Method.

 6.5.4  Development of Data Sets for Coal-boiler-control Classes

        As described earlier, unit classes are defined as those combinations of coal, boiler,
and control technologies that are expected to provide similar results in the speciation and
capture of Hg. Data sets for different classes of units were developed by sorting the unit tests
by coal type, boiler type, NOx control method, PM control method, and SC>2 control method.
Data sets were consolidated whenever the joint sets appeared to provide the same results as
the initial groupings.  Thus, wall- and tangentially fired PC boilers were consolidated into a
single conventional PC boiler set. Units that reported no NOX controls were consolidated with
low-NOx burners, overfire-air staging, and concentric firing systems.

 6.5.5  Questionable Nature of OH Speciation Measurements Upstream of PM Controls

        Initial evaluations of the Part HI ICR data dealt with comparisons of the coal-boiler-
 control classes using the results of OH speciation measurements at both the inlet and outlet
 sampling locations. Comparisons were also made of the results obtained using either the Part
 in ICR coal data or the inlet OH data to evaluate emission reduction trends. The comparison
 of speciation at the inlet and outlet locations produced, in some cases, results contrary to the
 expected behavior of Hg between the inlet and outlet of the control devices.

        Previous research has shown that the OH sampling method provides valid
 measurements for HgT at both the inlet to flue gas cleaning devices and in the stack.  Also, the
 OH Method has been shown to provide valid Hg speciation measurements when samples are
 taken downstream of an efficient PM control device. However, the  OH Method can give
 erroneous speciation measurements for locations upstream of PM control devices.

        The OH sampling train consists of a probe, a particulate filter,  a series of impingers, a
 gas flow meter, and a sample pump.  The filter captures particulate matter and Hgp, while the
 downstream impingers separate Hg2+from Hg. Fly ash captured by the sampling train filter
 can absorb gas-phase Hg (Hg2+ and Hg) and oxidize Hg resulting in physical and chemical
 transformations within the sampling train. The rates of these transformations are dependent
 on the properties of fly ash, the amount of fly  ash, the temperature, the flue gas composition,
 and the sampling duration. Samples collected downstream of efficient PM control devices do
 not contain enough fly ash to significantly alter Hg speciation within the sampling train, but
 samples collected upstream of PM control devices can give erroneous results because of fly-
 ash-induced transformations.
                                         6-15

-------
                                                         Table 6-4
                  ICR Mercury Emission Test Allocations by Coal-boiler-control Class
   No.

 Ea Group
Coal-boiler Control Class
No. of
 Test
Runs
           Test Unit IName
(Bold numbers in parentheses indicate no.
             of test runs)
         POST-COMBUSTION CONTROLS: COLD-SIDE ESPS

         Bituminous Coal, PC Boiler with CS-ESP

         Bituminous Coal and Pet Coke, PC Boiler with CS-ESP
         Bituminous Coal, PC Boiler with SNCR and CS-ESP
         Subbituminous Coal, PC Boiler with CS-ESP
         Subbituminous/ Bituminous Coal, PC Boiler with CS-ESP
         Lignite, PC Boiler with CS-ESP
         POST-COMBUSTION CONTROLS: HOT-SIDE ESPS
         Bituminous Coal, PC Boiler with HS-ESP
         Subbituminous Coal, PC Boiler (Dry Bottom) with HS-ESP
         Subbituminous Coal, PC Boiler (Wet Bottom) with HS-ESP
         Subbituminous/ Bituminous Coal, PC Boiler with HS-ESP
         POST-COMBUSTION CONTROLS: FF BAGHOUSES
         Bituminous Coal, PC Boiler with FF Baghouse
         Bituminous Coal/Pet. Coke, PC Boiler with FF Baghouse (Measurements not valid, disregard^
         Bituminous/Subbituminous Coal, PC Boiler with FF Baghouse
         Subbituminous Coal, PC Boiler with FF Baghouse
         POST-COMBUSTION CONTROLS: MISCELLANEOUS PM CONTROLS
         TX Lignite, PC  Boiler with CS-ESP and FF (COHPAC)
         Subbituminous Coal, PC Boiler with PM Scrubbers
         POST-COMBUSTION CONTROLS: DRY FGD SCRUBBERS
         Bituminous Coal, PC Boiler with DSI and CS-ESP
         Subbituminous Coal, PC Boiler with CS-ESP/SDA
         Bituminous Coal, PC Boiler with SDA/FF
         Bituminous Coal, PC Boiler with SCR and SDA/FF
         Subbituminous Coal, PC Boiler with SDA/FF
         ND Lignite, PC  Boiler with SDA/FF
         Bituminous Coal, Stoker with SDA/FF
         POST-COMBUSTION CONTROLS: WET FGD SCRUBBERS
         Bituminous Coal, PC Boiler with PS and Wet FGD Scrubbers
         Subbituminous Coal, PC Boiler with PS and Wet FGD Scrubbers
         ND Lignite, PC  Boiler with PS and Wet FGD Scrubbers
         Bituminous Coal, PC Boiler with CS-ESP and Wet FGD Scrubbers
         Subbituminous Coal, PC Boiler with CS-ESP and Wet FGD Scrubbers
         TX Lignite, PC  Boiler with CS-ESP and Wet FGD Scrubbers
         Bituminous Coal, PC Boiler with HS-ESP and Wet FGD Scrubbers
         Subbituminous Coal, PC Boiler with HS-ESP and Wet FGD Scrubbers
         Bituminous Coal, PC Boiler with FF and Wet FGD Scrubber
         CYCLONE-FIRED BOILERS
         Lignite, Cyclone Boiler with CS-ESP
         Subbituminous Coal/Pet Coke, Cyclone Boiler with HS-ESP
         Lignite, Cyclone Boiler with Mechanical Collector
         Lignite, Cyclone Boiler with SDA/FF
         Bituminous Coal, Cyclone Boiler with PS and Wet FGD Scrubbers
         Bituminous Coal, Cyclone Boiler with CS-ESP and Wet FGD Scrubbers
         FLUIDIZED-BED COMBUSTORS
         Lignite, FBC with CS-ESP
         Anthracite Coal Waste, FBC with FF
         Bituminous Coal Waste, FBC with FF
         Bituminous Coal/Pet. Coke, FBC with SNCR and FF
         Subbituminous Coal, FBC with SCR and FF
         Lignite, FBC with CS-FF
                                                         Brayton Point 1 (3), Brayton Point 3 (3), Gibson 0300
                                                         (3), Gibson 1099 (3), Meramec (3), Jack Watson (3),
                                                         Widow Creek (3)
                                                         Presque Isle 5 (3), Presque Isle 6 (3)
                                                         Salem Harber (3)
                                                         Montrose (3), George Neal South (3), Newton (3)
                                                         St. Clair (3)
                                                         Stanton 1 (3)

                                                         Cliffside (3), Gaston (3), Dunkirk (3)
                                                         Cholla 3 (3), Columbia (3)
                                                         Platte (3), Presque Isle 9 (3)
                                                         Clifty(3)

                                                         Sammis (3), Valmont (3)
                                                         Valley (3)
                                                         Shawnee (3)
                                                           .swell 2 (3), Comanche (3)

                                                         Bigbrown (3), Monticello 1-2 (3)
                                                           .swell 3 (3)

                                                         Washington (3)
                                                         GRDA (3), Laramie 3 (3), Wyodak (3)
                                                         Mecklenburg (3)
                                                         Logan (3), SEI (3)
                                                         Craig 3 (3), Rawhide (3), NSP Sherburne (3)
                                                         Antelope Valley (3), Stanton 10 (3)
                                                         Dwayne Collier (3)

                                                         Bruce Mansfield (3)
                                                           .swell 4 (3), Cholla 2 (3), Colstrip (3), Lawrence (3)
                                                         Lewis and Clark (3)
                                                         AES Cayuga (3), Big Bend (3)
                                                         Jim Bridger (3), Laramie River 1 (3), Sam Seymore (3)
                                                         Monticello 3 (3), Limestone (3)
                                                         Charles Lowman (3), Morrow (3)
                                                         Coronado (3), Craig 1 (3), Navajo (3), San Juan (3)
                                                         Clover (3), Intermountain (3)

                                                         Leland Olds (2)
                                                         Nelson Dewey (3)
                                                         Bay Front (3)
                                                         Coyote (2)
                                                         Lacygne (3)
                                                         Baffly (3)

                                                         R.M.Heskett(3)
                                                         Kline Township (3)
                                                         Scrubgrass (3)
                                                         Stockton Cogen (3)
                                                         AES Hawaii (3)
                                                         TNP(3)
          The effects of filtered solids on a filter in the OH sampling train are shown in Figure 6-4.
These test results were obtained from pilot-scale coal combustion experiments conducted by the
DOE Federal Energy Technology Center (FETC) [now the National Energy Technology
Laboratory  (NETL)].  The OH sampling train speciation data shown in Figure 6-4 were
collected simultaneously in two different manners.   In the first, tests designated by the symbols
OH-n (n=l, 2, 3...),  samples were collected by  running the sampling train in the prescribed
method by collecting an isokinetic sample with the probe nozzle facing upstream.  In the second
manner, tests designated by MOH-n (n=l, 2,  3,  ...) were run with the probe nozzle facing
downstream  so that the PM entering the train would be minimal2.
                                                            6-16

-------
         12
         10
       O
       s?
       
o

I
d

O
O)
             i
                      I
                                                   if.
P
                                            O"

                                               ^U
              Figure 6-4. Effect of OH sample filter solids on Hg speciation.

    The results of these experiments show that, for each of the simultaneous runs, the values of
HgT can be considered to be equal when taking into account sample variations resulting from the
imprecision of the OH Method.  However, the samples taken with the probe facing upstream
indicated higher concentrations of Hgp and Hg2+ than the samples with the nozzles facing
downstream.  This provides evidence that PM collected on the filter of the train facing upstream
resulted in the oxidization and adsorption of Hg as  flue gas passed through the sampling train.
This and other evidence indicate that in some cases the use of the OH Method to collect
speciation samples upstream of PM control devices provides questionable results3.

6.6 FUEL, BOILER, AND CONTROL TECHNOLOGY EFFECTS

       Based on current understanding of speciation and capture of mercury, it is believed that
the ICR data represent a number of subpopulations corresponding to fuel-boiler-control
combinations. Sections 6.6 and 6.7 provide an interpretation of physical and chemical
phenomena that can be used to characterize the roles that coal, combustion, and flue gas
cleaning variables play in the speciation and capture of Hg. Section 6.8 provides a summary
of national  emission estimates that were based on data described in Sections 6.6 and 6.7.
Conclusions are provided in Section 6.9.

       The interpretations in Sections 6.6  and 6.7 are based on previous bench-, pilot-, and
full-scale tests, plus a number of different modeling efforts related to speciation and capture of
Hg in coal-fired boilers. While we have attempted to provide an internally consistent
interpretation of the data, some of the observed results are inconsistent with the current
theories on the behavior of Hg. In these instances,  either our interpretations  may be incorrect
and other factors may account for the apparent discrepancies in results, or the data may be
incorrect.  It is believed that some discrepancies result from questionable OH Method or from
                                        6-17

-------
errors in reporting test results.

       The evaluation of ICR Phase HI data indicates that air pollution control technologies
now used on coal-fired utility boilers exhibit levels of control that range from 0 to 99 percent
reduction of HgT. The level of control varies with the coal, combustion conditions, and flue
gas cleaning methods used at individual sites.  In some instances, there is substantial variation
in the three tests conducted at individual sites.  The run-to-run variations at any given site can
result from actual variations in emissions or with problems associated with the measurement
method.

       The OH Method is relatively complex,  and measurement method problems can result
from errors that occur:
       during the collection of samples,
       in extracting samples from the sampling train,
       from the chemical extraction of Hg from the nozzle and probe wash, from the sample
       train filter, and from the different impingers,
       from Hg analysis, and
       from data reduction and transcription.

Some errors are inevitable in spite of the best efforts of everyone  involved in the measurement
process.

       In statistical terms, the OH data represent a very small number of samples of the
underlying population.  Each individual test represents the average of flue gas concentration of
speciated Hg during a short "snapshot" in time. Run-to-run variations at any given site result
from temporal variations in coal properties, combustion conditions, and emission control
technology process conditions. There are also  site-to-site variations within a given coal-
boiler-control class and variations between classes.  Even considering these sample population
variations, the ICR data provide a great deal of information, when evaluated in the context of
current knowledge  on the behavior of Hg in coal-fired electrical generating units.

       Table 6-5 shows differences in the average reduction  in HgT emissions for coal-boiler-
control classes that burn pulverized coal. Plants that employ only post-combustion PM
controls display class average HgT emission reductions ranging from 1 to 90 percent. Units
with FFs obtained the highest average levels of control. Decreasing average levels of control
were generally observed for units equipped with a CS-ESP, HS-ESP, and PS. For units
equipped with dry scrubbers, the class average HgT emission  reductions ranged from 2 to 98
percent. The estimated class average reductions for wet FGD scrubbers were similar and
ranged from 10 to 98 percent.

       For PC-fired boilers, the amount of Hg captured by a  given control technology is
greater for bituminous coal than for either subbituminous coal or  lignite. For example, the
average capture of Hg based on OH inlet measurements in PC-fired plants equipped with a
CS-ESP is 36 percent for bituminous coal, 9 percent for subbituminous coal, and 1 percent for
lignite.

                                         6-18

-------
Errata Page 6-19, dated 3-21-02

                                     Table 6-5
      Average Mercury Capture by Existing Post-combustion Control
                   Configurations Used for PC-fired Boilers
Post-
combustion
Control
Strategy
PM Control Only
PM Control and
Spray Dryer
Adsorber
PM Control and
Wet FGD
System(a)
Post-combustion
Emission
Control Device
Configuration
CS-ESP
HS-ESP
FF
PS
SDA+ESP
SDA+FF
SDA+FF+SCR
PS+FGD
CS-ESP+FGD
HS-ESP+FGD
FF+FGD
Average Mercury Capture by Control Configuration
Coal Burned in Pulverized-coal-fired Boiler Unit
Bituminous Coal
36%
9%
90%
not tested
not tested
98%
98%
12%
74%
50%
98%
Subbituminous
Coal
3%
6%
72%
9%
35%
24%
not tested
-8%
29%
29%
not tested
Lignite
-4%
not tested
not tested
not tested
not tested
0%
not tested
33%
44%
not tested
not tested
       (a) Estimated capture across both control devices

       CS-ESP = cold-side electrostatic precipitator
       FF = fabric filter
       SDA = spray dryer adsorber system
HS-ESP = hot-side electrostatic precipitator
PS = particle scrubber
6.6.1 Coal Effects

       While OH speciation measurements may not provide an accurate characterization of
the speciation at the inlet sampling location, transformations within the sampling train provide
an indication of the fly ash reactivity, and potential for Hg adsorption.  SPFs for selected coal-
boiler-control classes are summarized in Table 6-6.  The data in Table 6-6 are class average
SPFs for PC-fired boilers at the inlet and outlet sampling locations. Data are shown for
bituminous, subbituminous, ND lignite, and TX lignite.  Relatively high levels of SPFP at the
inlet indicate that the Hg was either present as Hgp in the flue gas, or it was readily absorbed
by fly ash on the sampling train filter. Relatively high levels of Hg2+ at the inlet indicate that
Hg at the inlet sampling location was either already  oxidized or oxidized as the flue gas passed
through the sampling train. Relatively high levels of measured Hg indicate that there were
relatively high levels of Hg in the inlet flue gas.
       	                                                                          9-1-
       The units burning bituminous coal exhibited relatively high levels of SPFP and SPF
in the inlet samples. It is hypothesized that high levels of SPFP+ SPF2+, or alternatively low
SPF, in the inlet sampling train indicates a high probability that Hg can be readily captured in
downstream APCD(s). For the bituminous-coal-fired units, values of SPFP and SPF 2+ ranged
from 0.03 to 0.92, while values of SPF ranged from 0.01 to 0.37. The HS-ESP unit exhibited
the highest level of Hg followed by units equipped  with SDA/FF systems. HS-ESP units
                                        6-19

-------
operate at temperatures where Hg is not easily oxidized or captured.  The SDA/FF units
exhibited a 98 percent capture of HgT; and the relative concentrations of the SPF2+ and SPF
measurements at the stack sampling location were 0.22 and 0.77, respectively. This could
result from the efficient capture of Hg2+ in these units.

       The PC-fired units burning subbituminous coal exhibit inlet SPF values ranging from
0.44 to 0.84.  The summed SPFP + SPF2+ values for the CS-ESP and HS-ESP units were
similar. Both of these classes of units exhibited HgT captures of 9 percent. The moderately
low HgT captures for the SDA/ESP (38 percent) and SDA/FF (25 percent) are reflected by the
summed inlet SPFP + SPF 2+ values for these units.  The units with FF systems (72 percent
average capture) had measured average inlet SPF values of 44 percent.

       There were a limited number of tests for units firing lignite. The units burning ND
lignites tend to have a higher SPF values than units burning TX lignites.  The CS-ESP units
burning ND lignite exhibited an average inlet  SPF value of 0.98.  While there was no
comparable test unit that fired TX lignite, a unit equipped with a CS-ESP + FF exhibited an
average inlet SPF of 0.60. While the inlet measurements for the CS-ESP + FF unit were
taken downstream of the CS-ESP, a higher SPF would have been expected if the TX lignite
were to provide similar speciation results as the ND lignite. Moderate to average SPF values
(0.47) were also noted for the CS-ESP + wet FGD units using TX lignite. Inlet measurements
for these units were also made downstream  of a CS-ESP.

       The similarities between inlet and outlet SPF values can also be used to identify
instances where the measured inlet speciation  values provide a good estimate of the true Hg
speciation in the flue gas at the inlet sampling location. Units with similar inlet and outlet
SPFs  are identified by an (a) in Table 6-6.  These cases correspond to tests in which the
capture of HgT is < 25 percent for many of the units firing subbituminous coals and ND
lignite (e.g., comparison of the respective inlet and outlet values for SPFP).
                                        6-20

-------
 Errata Page 6-21, dated 3-21-02
                                    Table 6-6
        Effects of Coal and Control Technology Inlet and Outlet SPF
                       and Capture for PC-fired Boilers
Coal-Control Class
Bituminous
CS-ESP
SNCR and CS-ESP
HS-ESP (a)
FF
SDA/FF
SCR and SDA/FF

Subbituminous
CS-ESP (a)
HS-ESP (a)
FF
SDA/ESP
SDA/FF (a)

ND Lignite
CS-ESP (a)
SDA/FF (a)

TX Lignite
CS-ESP + FF
CS-ESP + Wet FGD
Inlet
SPFD

0.35
0.92
0.09
0.92
0.59
0.82


0.05
0.02
0.33
0.13
0.01


0.01
0.03


0.09
0.00
SPF2+

0.58
0.03
0.53
0.04
0.28
0.17


0.25
0.15
0.23
0.26
0.06


0.01
0.04


0.31
0.52
SPFU

0.07
0.05
0.37
0.04
0.15
0.01


0.70
0.83
0.44
0.61
0.84


0.98
0.93


0.60
0.47
Outlet
SPFD

0.02
0.20
0.04
0.01
0.01
0.05


0.00
0.00
0.01
0.00
0.01


0.00
0.00


0.00
0.01
SPF2+

0.78
0.35
0.59
0.52
0.22
0.46


0.31
0.17
0.87
0.05
0.05


0.04
0.03


0.70
0.14
SPFU

0.20
0.45
0.37
0.47
0.77
0.48


0.69
0.83
0.12
0.94
0.94


0.96
0.97


0.30
0.85
% Red
HgT

36
91
9
90
98
98


3
6
72
35
24


-4
0


NA
44
 (a) Units with similar inlet and outlet SPF values.

6.6.2  Control Technology Effects

       Control technology effects are inseparable from coal and boiler effects.  In the
following sections, post-combustion control technology effects will be evaluated in terms of
the three major types of controls currently used for coal-fired utility boilers: PM controls, dry
FGD scrubbing controls, and wet FGD controls. These evaluations will be discussed initially
in terms of control technology and coal effects on PC-fired boilers. The speciation and capture
of Hg from cyclone-fired combustors, FBCs, and IGCC units will then be discussed.

       A summary of test results for each of the coal-boiler-control classes for which ICR Hg
emission data were collected is given in Table 6-7. The data include information on the
                                                              94-    n
number of tests for each class, the average emission factors for Hgp, Hg  , Hg , and Hgx, and
the average and range of Hgx emission reductions.
                                       6-21

-------
                                                          Table 6-7

                    Average Mercury Emission Factors and Percent Reduction for Coal-boiler-control Classes
No.
Ea Group

1
2
3
4
5
6


1
2
3
4


1
2
3
4


1
2


1
2
3
4
5
6
7


POST-COMBUSTION CONTROLS: COLD-SIDE ESPs
Bituminous Coal, PC Boiler with CS-ESP
Bituminous Coal and Pet Coke, PC Boiler and CS-ESP
Bituminous Coal, PC Boiler with SNCR and CS-ESP
Subbitummous Coal, PC Boiler with CS-ESP
Subbituminous/ Bituminous Coal, PC Boiler with CS-ESP
Lignite, PC Boiler with CS-ESP

POST-COMBUSTION CONTROLS: HOT-SIDE ESPs
Bituminous Coal, PC Boiler with HS-ESP
Subbituminous Coal, PC Boiler (Dry Bottom) with HS-ESP
Subbituminous Coal, PC Boiler (Wet Bottom) with HS-ESP
Subbituminous/ Bituminous Coal, PC Boiler with HS-ESP

POST-COMBUSTION CONTROLS: FF BAGHOUSES
Bituminous Coal, PC Boiler with FF Baghouse
Bituminous Coal/Pet. Coke, PC Boiler withFF Baghouse (Measurements not valid, disregard)
Bituminous/Subbituminous Coal, PC Boiler with FF Baghouse
Subbituminous Coal, PC Boiler with FF Baghouse

POST-COMBUSTION CONTROLS: MISCELLANEOUS CONTROLS
TX Lignite, PC Boiler with CS-ESP and FF (COHPAC)
Subbituminous Coal, PC Boiler with PM Scrubbers

POST-COMBUSTION CONTROLS: DRY FGD SCRUBBERS
Bituminous Coal, PC Boiler with DSI and CS-ESP
Subbituminous Coal, PC Boiler with CS-ESP/SDA
Bituminous Coal, PC Boiler with SDA/FF
Bituminous Coal, PC Boiler with SCR and SDA/FF
Subbituminous Coal, PC Boiler with SDA/FF
ND Lignite, PC Boiler with SDA/FF
Bituminous Coal, Stoker with SDA/FF
No. of
Tests

21
6
3
9
3
3


9
6
6
3


6
3
3
6


6
3


3
9
3
6
9
6
3
HgT, (Ig/dscm
Inlet

13.82
4.47
4.41
10.05
6.79
11.67


9.07
9.12
10.63
14.51


8.13
2.20
4.61
7.80


50.05
6.18


17.03
12.64
13.59
15.22
9.56
9.65
2.39
Outlet

10.31
1.73
0.41
9.57
5.36
12.06


7.95
8.41
10.92
9.57


0.64
2.31
1.38
2.42


59.65
5.63


9.32
7.78
0.24
0.28
7.39
9.69
0.14
Average Bin EMF (a)
HgD

0.04
0.01
0.02
0.00
0.00
0.00


0.05
0.00
0.00
0.02


0.00
0.02
0.00
0.00


0.00
0.00


0.00
0.01
0.00
0.00
0.01
0.00
0.01
Hg2+

0.48
0.19
0.03
0.31
0.20
0.04


0.53
0.16
0.09
0.32


0.07
0.77
0.13
0.24


0.75
0.01


0.37
0.05
0.00
0.01
0.03
0.04
0.01
Hg

0.15
0.20
0.04
0.66
0.59
1.00


0.33
0.77
0.95
0.32


0.03
0.19
0.16
0.04


0.40
0.90


0.18
0.99
0.02
0.01
0.72
0.96
0.03
HgT

0.64
0.40
0.09
0.97
0.79
1.04


0.91
0.94
1.03
0.66


0.10
0.98
0.30
0.28


1.15
0.91


0.55
1.04
0.02
0.02
0.76
1.00
0.06
Red in HgT, %
Range

81.01 -0.00
70.84-50.29
93.06-87.07
17.46-(-)0.10
35.63-8.71
4.42 - 0.00


42.51 -0.00
27.34-0.00
26.93-0.00
36.99-29.51


93.04-84.15
5.67-(-)25.15
72.62-66.73
87.45-52.67


28.69-0.00
13.81 -5.25


52.61 -40.68
62.53-0.00
99.23-96.91
98.72-96.56
47.31 -0.00
8.49 - 0.00
95.43-92.84
Ave.

36.03
60.14
90.90
8.75
21.33
1.47


15.09
8.80
4.50
34.03


89.67
-6.73
69.95
72.43


4.93
8.74


44.89
37.94
97.91
98.05
25.40
1.95
94.25
to
to
                                                                                                       (continued)

-------
                                                       Table 6-7 (cont'd)
                    Average Mercury Emission Factors and Percent Reduction for Coal-boiler-control Classes
No.
Ea Group

1
2
3
4
5
6
7
8
9


1
2
3
4
5
6


1
2
3
4
5
6
COAL-BOILER-CONTROL CLASS
POST -COMBUSTION CONTROLS: WET FGD SCRUBBERS
Bituminous Coal. PC Boiler with PS and Wet FGD Scrubbers
Subbitummous Coal. PC Boiler with PS and Wet FGD Scrubbers
ND Lmmte. PC Boiler with PS and Wet FGD Scrubbers
Bituminous Coal. PC Boiler with CS-ESP and Wet FGD Scrubbers
Subbitummous Coal. PC Boiler with CS-ESP and Wet FGD Scrubbers
TX Lmmte. PC Boiler with CS-ESP and Wet FGD Scrubbers
Bituminous Coal PC Boiler with HS-ESP and Wet FGD Scrubbers
Subbitummous Coal. PC Boiler with HS-ESP and Wet FGD Scrubbers
Bituminous Coal. PC Boiler with FF and Wet FGD Scrubbers

CYCLONE-FIRED BOILERS
Lmmte. Cvclone Boiler with CS-ESP
Subbituminous Coal/Pet Coke Cvclone Boiler with HS-ESP
Lianite. Cvclone Boiler with Mechanical Collector
Lmmte. Cvclone Boiler with SDA/FF
Bituminous Coal Cvclone Boiler with PS and Wet FGD Scrubbers
Bituminous Coal. Cvclone Boiler with CS-ESP and Wet FGD Scrubbers

FLUIDIZED-BED COMBUSTORS fFBCsl
Lmmte. FBC with CS-ESP
Anthracite Coal Waste. FBC with FF
Bituminous Coal Waste FBC with FF
Bituminous Coal/Pet. Coke. FBC with SNCR and FF
Subbitummous Coal. FBC with SCR and FF
Lienite FBC with CS-FF
No. of
Tests

3
12
3
6
9
6
6
12
6
















HST* Ug/dscm
Inlet

11 15
630
23 64
777
11.22
4403
1046
5.12
1 88


6 80
277
3 33
1782
9 88
561


1046
4469
128 83
2.24
1 71
35 98
Outlet

975
6.42
15 09
264
824
25 00
5 83
374
0 50


5 60
3 00
498
1634
757
311


5 95
0 12
0 13
0.12
073
15 11
Average Bin EMF fa)
Hgr

000
002
000
000
000
000
001
0.01
001


000
002
0.25
000
001
000


003
000
000
0.01
000
000
Hg2+

0 17
0.06
002
003
002
008
0 18
003
008


0 16
008
076
001
004
0.06


0.06
000
000
002
001
031
Hg

0 70
1 00
065
031
0.72
048
0 39
0.74
0 19


077
1 00
0 56
0 90
072
049


0 53
0 00
0 00
002
041
0 12
HgT

0 88
1 08
067
034
0.75
0 56
058
078
028


093
1 11
1 57
091
077
0 55


0.62
000
000
006
043
043
Red in Hg
Range

14 94.7 42
74.27-0.00
5075-8 81
7606-6401
5753-1 51
5607-21 31
59 20-27 96
41.48-M16.05
98 95-96 78


33 26-0 00
0 13-000
0 34-0 00
11 81-548
2453-22 17
5495-5411


54 49 - n 09
99 74 _ 99 73
99 92 - 99 85
96 09 - 92 1 6
6491 - 51 35
61 88 -5407
T,/o
Ave.

1239
1015
3277
7068
2678
43 73
4463
18.17
9780


1663
0 04
0.11
8 64
23 29
54.43


38 29
9974
99 89
9425
5737
5705
to
   (a)  See Section 6.5.2 for discussion of emission factors.

-------
6.6.3  Post-combustion PM Controls

       In 1999, 72 percent of the coal-fired electric utility boilers in the U.S. used post-
combustion controls that consisted only of PM controls. The Phase IICR revealed that there
were 890 units that used only post-combustion PM controls.  This included 791 units using
either CS- or HS-ESPs and 80 units that used FF baghouses.  The number of boiler units in the
U.S. equipped only with PM controls is shown in Table 6-8 along with the number of test
units in each PM control category.

                                      Table 6-8
  Number of Coal-fired Utility Boilers Equipped with Particulate Matter Controls Only
Particulate Matter
Control

CS-orHS-ESP(a)
Two ESPs in series
Fabric Filter
ESP w/ Fabric Filter
Particulate Scrubber
ESP w/ Particulate
Scrubber
Mechanical Collector
Number of Units
Utility Industry
791
2
80
6
5
4
2
Test Units
25
2
12
2
1
0
1
                 (a) 14 CS-ESPs and 9 HS-ESPs were tested
6.6.3.1 Cold-side ESPs
       A total of 14 PC-fired units equipped with CS-ESPs were tested. The types of fuels
that were used in these tests are given in Table 6-9.

                                      Table 6-9
             Type of Fuel Used in PC-fired Units Equipped with CS-ESP
Type of Fuel
Bituminous
Bituminous & Pet. Coke
Subbituminous
Subbituminous/Bituminous
Total
No. of Test Units
8
2
3
1
14
       One of the units burning bituminous coal was also equipped with an SNCR system for
NOx control.  One cyclone-fired unit that burned lignite was also tested. The results of Hg
emission tests on PC-fired units equipped with a CS-ESP are given in Table 6-10.
                                        6-24

-------
               Table 6-10
Post-combustion Controls: Cold-side ESPs
Hg Speciation at Inlet and Outlet (ug/dscm(o), 3%O2) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In Hg In HgT In HgT In Hgp Out Hg2+ Out Hg Out HGT Out %RHT %RHgT
No. OH OH OH OH Coal OH OH OH OH OH Coal
Bituminous Coal. PC Boiler with CS-ES
Brayton Point 1
Brayton Point 1
Brayton Point 1
Average
Brayton Point 3
Brayton Point 3
Brayton Point 3
Average
Gibson 0300
Gibson 0300
Gibson 0300
Average
Gibson 1099
Gibson 1099
Gibson 1099
Average
Meramec
Meramec
Meramec
Average
Jack Watson
Jack Watson
Jack Watson
Average
Widows Creek
Widows Creek
Widows Creek
Average
1
2
3

1
2
3

1
2
3

1
2
3

1
2
3

1
2
3

1
2
3

2.01
2.61
2.17
2.27
3.14
1.83
1.40
2.12
1.94
1.25
1.75
1.65
5.53
27.57
4.60
12.57
7.61
9.34
5.65
7.53
3.60
4.91
4.64
4.38
3.36
2.98
2.87
3.07
3.34
3.69
3.50
3.51
3.67
3.14
3.26
3.36
31.74
38.06
44.44
38.08
10.33
3.78
11.02
8.38
0.49
1.36
1.93
1.26
1.22
1.16
0.60
0.99
0.44
0.45
0.47
0.45
j
0.32
0.25
0.26
0.28
0.36
0.34
1.60
0.77
4.39
2.92
1.65
2 99
2.34
1.25
1.58
1.72
0.14
0.44
0.62
0.40
0.92
0.25
0.23
0.47
0.54
0.51
0.50
0.51

5.68
6.55
5.93
6.05
7.17
5.31
6.26
6.25
38.08
42.23
47.85
42.72
18.20
32.60
17.20
22.67
8.23
11.15
8.19
9.19
5.74
6.32
5.46
5.84
4.34
3.94
3.83
4.04

6.80
4.21
5.01
5.34
8.55
5.30
5.58
6.48
13.69
13.33
13.53
13.52
14.00
15.09
14.69
14.59
8.46
10.72
5.89
8.36
4.70
5.67
6.20
5.52
3.11
2.67
2.15
2.64

0.77
0.75
0.77
0.76
0.78
0.96
0.01
0.59
0.00
0.01
0.01
0.01
0.03
0.05
0.03
0.04
0.00
0.01
0.00
0.00
0.05
0.05
0.06
0.05
0.14
0.01
0.01
0.06

3.83
3.19
3.02
3.35
3.18
2.47
3.43
3.03
32.03
32.21
42.87
35.70
6.06
8.41
11.03
8.50
0.76
2.20
1.51
1.49
2.57
2.99
2.92
2.83
1.48
1.28
0.65
1.14

0.23
0.25
0.24
0.24
0.46
0.37
1.70
0.85
7.51
5.80
4.17
5.83
5.03
5.00
4.65
4.90
0.80
1.13
0.79
0.91
1.87
0.89
0.89
1.22
0.78
0.68
0.67
0.71

4.84
4.18
4.02
4.35
4.43
3.80
5.15
4.46
39.54
38.01
47.05
41.54
11.12
13.46
15.71
13.43
1.56
3.35
2.30
2.40
4.49
3.94
3.88
4.10
2.40
1.97
1.34
1.90

14.73
36.11
32.19
27.68
38.21
28.47
17.70
28.13
-3.85
9.98
1.66
2.60
38.92
58.69
8.68
35.43
81.01
69.97
71.96
74.32
21.71
37.70
29.04
29.48
44.75
50.00
65.11
53.29

28.86
0.68
19.64
16.39
48.20
28.27
7.71
28.06
-188.83
-185.13
-247.76
-207.24
20.62
10.76
-6.93
8.15
81.54
68.77
60.99
70.43
4.39
30.53
37.45
24.13
22.95
26.25
37.81
29.00
Average 4.80 8.00 1.02 13.82 8.06 0.22 8.00 2.09 10.31 35.85 -4.44
Minimum
Maximum
STDEV



1.25
27.57
5.62
0.44
44.44
13.05
0.14
4.39
1.09
3.83
47.85
13.86
Bituminous Coal and Pet Coke. PC Boiler with CS-ESP
Presque Isle 5
Presque Isle 5
Presque Isle 5
Average
Presque Isle 6
Presque Isle 6
Presque Isle 6
Average
1
2
3

1
2
3

4.56
3.60
5.06
4.40
2.73
2.97
2.96
2.89
0.48
0.66
0.45
0.53
0.63
0.72
0.62
0.65
0.14
0.57
0.12
0.27
0.17
0.25
0.17
0.20
5.17
4.82
5.63
5.21
3.52
3.94
3.75
3.74
2.15
15.09
4.35

4.27
3.48
3.93
3.89
2.29
4.34
3.85
3.49
0.00
0.96
0.34

0.01
0.00
0.02
0.01
0.06
0.03
0.03
0.04
0.65
42.87
12.01

0.72
0.82
0.71
0.75
0.84
1.00
0.73
0.86
0.23
7.51
2.24

1.06
1.02
0.91
1.00
0.70
0.93
0.81
0.81
1.34
47.05
13.67

1.80
1.84
1.64
1.76
1.60
1.96
1.57
1.71
-3.85
81.01
23.90

65.29
61.87
70.84
66.00
54.54
50.29
58.00
54.28
-247.76
81.54
88.31

57.92
47.14
58.19
54.42
30.10
54.87
59.17
48.05
Average 3.65 0.59 0.24 4.47 3.69 0.02 0.81 0.90 1.73 60.14 51.23
Minimum
Maximum
STDEV



2.73
5.06
0.96
0.45
0.72
0.11
0.12
0.57
0.17
3.52
5.63
0.86
2.29
4.34
0.75
0.00
0.06
0.02
0.71
1.00
0.11
0.70
1.06
0.13
1.57
1.96
0.15
50.29
70.84
7.44
30.10
59.17
11.25
                                     (continued)
                 6-25

-------
                                  Table 6-10 (cont'd)
                      Post-combustion Controls: Cold-side ESPs
Hs Soeciation at Inlet and Outlet Cus/dscmfS), 3% 02) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In Hg In HgT In
No. OH OH OH OH
Bituminous Coal. PC Boiler M
Salem Harbor
Salem Harbor
Salem Harbor
Average
1
2
3

4.12
4.09
3.96
4.06
ith SNCR and CS-ESP
0.32 0.32
0.04 0.16
0.06 | 0.15
0.14 0.21
Subbituminous Coal. PC Boiler with CS-ESP
Montrose
Montrose
Montrose
Average
George Neal So.
George Neal So.
George Neal So.
Average
Newton
Newton
Newton
Average
1
2
3

1
2
3

1
2
3

1.94
0.91
1.63
1.49
0.17
0.07
0.02
0.09
0.04
0.04
0.08
0.05
1.85
2.52
2.85
2.41
4.78
4.35
3.53
4.22
0.58
0.63
1.65
0.95
6.00
4.93
4.68
5.20
6.34
8.24
3.77
6.12
9.70
9.85
9.26
9.61
4.76
4.29
4.17
4.41
9.79
8.36
9.16
9.10
11.29
12.66
7.32
10.42
10.32
10.52
11.00
10.61
HgT In
Coal

3.44
2.35
3.27
3.02
44.90
51.99
47.76
48.21
8.96
7.82
10.19
8.99
9.07
8.05
9.34
8.82
Hgp Out
OH

0.07
0.10
0.08
0.08
0.03
0.02
0.02
0.02
0.03
0.06
0.02
0.04
0.00
0.00
0.00
0.00
Hg2+ Out
OH

0.28
0.07
0.08
0.14
.57
.60
.30
.49
4.07
4.60
4.74
4.47
2.26
1.66
2.04
1.99
HgOut
OH

0.27
0.15
0.14
0.19
5.48
5.94
5.69
5.70
5.47
6.87
6.39
6.24
8.07
7.13
8.03
7.74
HGT Out
OH

0.62
0.32
0.29
0.41
8.08
8.56
8.01
8.22
9.58
11.53
11.15
10.75
10.33
8.80
10.07
9.73
%R H T
OH

87.07
92.57
93.06
90.90
17.46
-2.31
12.54
9.23
15.18
8.89
-52.29
12.04
-0.10
16.33
8.46
8.23
%R Hg T
Coal

82.11
86.40
91.16
86.55
82.01
83.54
83.22
82.92
-6.90
-47.37
-9.36
-21.21
-14.00
-9.28
-7.82
-10.36
Average 0.54 2.53 6.98 10.05 22.01 0.02 2.98 6.56 9.57 2.69 17.12
Minimum
Maximum
STDEV
SPF




0.02
1.94
0.76
0.05
0.58
4.78
1.50
0.25
Subbituminous/ Bituminous Coal. PC Be
St Clair
StClair
StClair
Average
Lignite, PC Bo
Stanton 1
Stanton 1
Stanton 1
Averaae
1
2
3

2.53
2.87
0.98
2.13
2.29
2.13
1.94
2.12
ler with CS-ESP
i
2
3
0.04
0.13
0.08
0.15
0.13
0.05
3.77
9.85
2.34
0.69
7.32
12.66
1.61
1.00
iler with CS-ESP
1.97
1.40
4.28
2.55
11.96
10.81
11.66
6.79
6.39
7.20
6.79
12.15
11.06
11.79
7.82
51.99
19.75
1.00

16.26
14.36
17.71
16.11
31.51
41.24
19.94
0.00
0.06
0.02
0.00

0.01
0.01
0.01
0.01
0.04
0.02
0.01
1.66
4.74
1.16
0.31

1.35
1.39
1.33
1.35
0.42
0.43
0.45
5.47
8.07
1.02
0.69

3.01
3.74
5.24
4.00
11.16
11.68
11.97
8.01
11.53
1.30
1.00

4.37
5.14
6.57
5.36
11.62
12.14
12.43
-52.29
17.46
21.75


35.63
19.65
8.71
21.33
4.42
-9.70
-5.41
-47.37
83.54
50.89


73.13
64.24
62.89
66.75
63.13
70.56
37.67
0.08 0.11 11.48 11.67 30.89 0.02 0.44 11.60 12.06 -3.57 57.12
       The test units with a CS-ESP display significant run-to-run differences (variations) in
the HgT (inlet), HgT (outlet), and % HgT reduction. These differences may result from the
changing HgT inlet concentrations, changing boiler and control device operating conditions, or
sampling and analysis problems.  Two important variables that affect Hg capture are changes
in Hg inlet concentration and unit operating temperatures.

       Run-to-run variations for test units burning bituminous coal in PC-fired boilers
equipped with CS-ESPs are shown in Figure 6-5. While the class average HgT reduction for
these units was 36 percent, the run-to-run emission reductions in HgT range from 0 to 81
percent. All inlet and outlet HgT concentrations for the Widow Creek, Jack Watson, Brayton
3, and Brayton 1 were similar. The Meramec plant exhibited relatively high HgT reductions as
did run 2 on Gibson 1099.  Gibson 0300 exhibited high stack gas concentrations of HgT, and
run 1 on Gibson 0300 had a higher outlet HgT concentration than at the inlet. The unit-to-unit
variations in HgT emission reductions for these same units are shown in Figure 6-6.  The
average emission reduction for the seven 3-run tests  shown in Figure 6-6 is  still 37 percent,
but unit-to-unit emission reductions range  from 3 percent for Gibson 0300 to 74 percent for
Meramec. The speciation of Hg for the bituminous coals is predominantly Hg2+.
                                        6-26

-------
       In Figure 6-6, there are two unit test averages given for Gibson. Both averages are for
the same unit, Gibson 0300.  The unit average for Gibson 1099 is for tests conducted in
October 1999, while the average for Gibson 0300 is for tests conducted in March 2000. The
tests in October and March used coal from the same source. Average unit reductions in HgT
for the October and March tests were 35 and 3 percent, respectively.  The apparent
discrepancy in the test results led plant engineers to investigate. The investigation indicated
that steam-cleaning of the air preheater during the collection of OH samples was the probable
cause of these inconsistencies.

       The Hg speciation and HgT reductions for PC-fired units equipped with CS-ESPs and
burning subbituminous coal and lignite are shown in Figure 6-7.  Hg emission reductions for
the units range from -4 to 12 percent, exhibiting little if any Hg capture. The relative
concentrations of Hg in the stack gas are higher than those observed for units firing
bituminous coal.
      Widow Creek
      Jack Watson
        Meramec
      Gibson 1099
      Gibson 0300
        Brayton 3
        Brayton 1
                          10          20           30           40           50
                            Total Mercury Concentration,|o,g/dscm @ 3% Qz
                                                                                    60
Figure 6-5. Inlet and outlet mercury concentrations for bituminous PC-fired boilers with
                                       CS-ESP.
                                         6-27

-------
o
'B
   Widow Creek
      [53%]

   Jack Watson
      [30%]

      Meramec
       [74%]
a   Gibson 1099
       [35%]
  Gibson 0300
              7///////////A
      Brayton 3
        [28%]

      Brayton 1
        [28%]
                                              Hg(p)Out 0Hg(2+)Out DHg(0)Out
0
                                                                             40
                                                                                     45
                    5       10       15       20       25       30       35
                             Mercury Concentration, jig/dscm @ 3% O2

Figure 6-6. Mercury emissions from bituminous-coal-fired PC boilers with CS-ESP.
         Stanton 1 [-4%]




    |      Newton [8%]
   '^n
   o

   jjf George Neal South
   r       [12%]
    3


         Montrose [9%]
                      0       2       4       6       8      10      12

                               Mercury Concentration, |^g/dscm @ 3% Oi
                                                                             14
  Figure 6-7. Mercury emissions for subbituminous- and lignite-fired PC boilers with
                                        CS-ESP.
                                          6-28

-------
       Run-to-run variations on a given unit can be attributed to operating variables such as
inlet Hg concentrations, operating temperature, soot blowing, reentrainment losses within an
ESP, or the imprecision of the OH Method.

       Mercury outlet concentrations can be expressed by:

       HgT (outlet) = HgT (inlet) - Hgp (captured in the control device)

              + Hgp (reentrained and escapes the control device)

              - Hg or Hg2+ (adsorbed and captured within the control device)

              + Hgp, Hg2+, or Hg (desorbed or is reentrained and escapes capture)

       Deposits or captured fly ash between the inlet and outlet sampling location (the stack)
can adsorb or desorb gas-phase Hg, depending on time-dependent changes in the inlet Hg
concentration and operating temperatures downstream of the inlet sampling location.
Temperature effects can be understood by considering the  deposits and collected fly ash
between the inlet and stack locations to be a complex system that adsorbs and desorbs Hg.  If
the system has reached equilibrium in terms of operating conditions, there will be a constant
relationship between the inlet and outlet concentrations of Hg.  Increases in operating
temperatures within the system can increase the rate at which Hg is desorbed,  resulting in
increased  outlet concentrations relative to the inlet concentrations. Temperature decreases can
increase Hg  adsorption within the system. This can cause a decrease in the Hg outlet
concentrations relative  to the inlet concentrations.

       Temporal changes in inlet and outlet Hg concentrations  are the result of hysteresis or
history effects.  Hypothetical changes in Hg reduction for three  tests on a single unit that could
occur because of the time lag between changing inlet and outlet Hg concentrations are
illustrated in Figure 6-8. In this illustration, Hg emission reductions during runs 1, 2, and 3
averaged 30, -15, and 40 percent, respectively. The -15 percent indicates that the measured
outlet Hg  concentrations were higher than the  inlet concentrations.
                                         6-29

-------
                                            Time
 Figure 6-8. Hypothetical effect of inlet and outlet HgT concentration changes on run-to-
                                  run HgT capture.

       Changes in the fly ash carbon content, changes in unit operating conditions such as
load, and diurnal changes in temperature may also result in hysteresis effects.  The ICR tests
for each unit represent a snapshot in time. Additional OH Method tests or tests with Hg
CEMs are needed over an extended period of time to more fully characterize the effects of
coal, combustion, and control technology variables on stack emissions of Hg.
6.6.3.2 Hot-side ESPs

       Eight ICR units that burn pulverized coal and that were equipped with an HS-ESP
were tested.  Three of these units burned bituminous coal; four burned subbituminous coal;
and one burned subbituminous and bituminous coal.  A ninth, a cyclone-fired unit equipped
with an HS-ESP, burned subbituminous coal and petroleum coke. Hg test data for the eight
PC-fired units are given in Table 6-11.
                                        6-30

-------
                                     Table 6-11
                       Post-combustion Controls: Hot-side ESPs
Hg Speciation at Inlet and Outlet (ng/dscm(3> 3% O2 ) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In Hg In HgT In HgT In Hgp Out Hg2+ Out Hg Out HGT Out %RHT %RHgT
No. OH OH OH OH Coal OH OH OH OH OH Coal
Bituminous Coal. PC Boiler with HS-ES
Cliffside
Cliffside
Cliffside
Average
Gaston
Gaston
Gaston
Average
Dunkirk
Dunkirk
Dunkirk
Average
1
2
3

1
2
3

1
2
3

0.17
0.09
0.08
0.11
4.28
2.57
0.43
2.42
0.09
0.01
0.01
0.04
3.72
3.54
4.15
3.80
0.86
0.71
3.94
1.84
8.56
8.91
9.15
8.87
j
3.31
3.33
7.27
4.63
2.64
3.56
2.83
3.01
2.82
1.43
3.20
2.48

7.20
6.95
11.49
8.55
7.77
6.84
7.20
7.27
11.47
10.36
12.36
11.40

5.43
3.84
8.80
6.02
5.20
6.27
4.70
5.39
10.06
10.30
9.65
10.00

0.41
0.10
0.10
0.20
0.74
0.40
1.15
0.76
0.21
0.08
0.03
0.11

2.79
2.27
3.97
3.01
4.70
5.80
4.73
5.08
6.89
4.57
6.40
5.95

3.95
1.95
2.54
2.81
2.34
3.47
2.04
2.62
3.67
2.46
3.82
3.32

7.14
4.31
6.61
6.02
7.78
9.66
7.92
8.46
10.77
7.12
10.25
9.38

0.86
38.00
42.51
27.12
-0.19
-41.37
-10.00
-17.19
6.08
31.27
17.08
18.14

-31.58
-12.17
24.94
-6.27
-49.56
-54.19
-68.41
-57.39
-7.09
30.90
-6.26
5.85
Average 0.86 4.84 3.38 9.07 7.14 0.36 4.68 2.92 7.95 9.36 -19.27
Minimum
Maximum
STDEV



0.01
4.28
1.52
0.71
9.15
328
Subbituminous Coal. PC Boiler (Drv Bo
Cholla 3
Cholla 3
Cholla 3
Average
Columbia
Columbia
Columbia
Average
1
2
3

1
2
3

0.07
0.51
0.45
0.34
0.01
0.01
0.01
0.01
0.37
0.32
0.43
0.37
0.93
5.82
0.46
2.41
1.43
7.27
1.59
6.84
12.36
230
torn) with HS-ESP
1.93
0.46
0.61
1.00
14.27
13.40
14.65
14.11
2.37
1.28
1.49
1.71
15.22
19.24
15.12
16.52
3.84
10.30
2.55

51.98
54.43
40.48
48.96
9.85
10.30
10.35
10.17
0.03
1.15
037

0.01
0.01
0.01
0.01
0.00
0.00
0.00
0.00
2.27
6.89
54

0.51
0.01
0.39
0.30
.74
.16
.65
.51
1.95
3.95
080

1.87
1.00
1.27
1.38
11.71
11.82
12.68
12.07
4.31
10.77
2.02

2.40
1.02
1.67
1.70
14.45
13.98
15.34
14.59
-41.37
42.51
26.41

-1.30
20.42
-12.28
2.28
5.02
27.34
-1.47
10.30
-68.41
30.90
3454

95.39
98.12
95.87
96.46
-46.78
-35.71
-48.18
-43.56
Average 0.18 1.39 7.55 9.12 29.57 0.01 .41 6.73 8.14 6.29 26.45
Minimum
Maximum
STDEV



0.01
0.51
0.23
0.32
5.82
2.18
0.46
14.65
7.21
Subbituminous Coal. PC Boiler ("Wet Bottom) with
Platte
Platte
Platte
Average
Presque Isle 9
Presque Isle 9
Presque Isle 9
Average
1
2
3

1
2
3

0.03
0.02
0.03
0.03
0.04
0.01
0.01
0.02
4.15
1.92
4.39
3.48
0.14
0.14
0.10
0.13
9.82
11.31
11.63
10.92
6.70
6.89
6.43
6.68
1.28
19.24
8.26
HS-ESF
14.00
13.25
16.04
14.43
6.89
7.05
6.55
6.83
9.85
54.43
21.77

11.10
9.65
6.05
8.93
9.86
8.92
9.91
9.56
0.00
0.01
0.00

0.01
0.01
0.01
0.01
0.00
0.00
0.00
0.00
0.01
2.74
1.24

1.45
0.78
1.51
1.25
0.57
0.67
0.54
0.59
1.00
12.68
5.87

8.76
16.86
14.90
13.51
6.30
6.74
6.38
6.47
1.02
15.34
7.09

10.23
17.65
16.43
14.77
6.88
7.41
6.92
7.07
-12.28
27.34
14.88

26.93
-33.20
-2.40
-2.89
0.10
-5.23
-5.76
-3.63
-48.18
98.12
76.82

7.88
-82.85
-171.57
-82.18
30.22
16.87
30.11
25.73
Average 0.02 1.80 8.80 10.63 9.25 0.01 0.92 9.99 10.92 -3.26 -28.22
Minimum
Maximum
STDEV
Subbituminous
Clifty
Cliftv
Clifly
Average



/Biti
3
1
2

0.01
0.04
0.01
0.10
4.39
2.03
millions Coal. PC Bo
0.01
0.40
0.02
3.41
2.35
3.58
6.43
11.63
2.41
ler with Fl
11.46
11.17
11.13
6.55
16.04
4.27
S-ESP
14.87
13.92
14.73
6.05
11.10
1.72

7.84
8.02
7.66
0.00
0.01
0.01

0.07
0.70
0.01
0.54
1.51
0.44

5.50
3.60
5.04
6.30
16.86
4.69

3.80
4.67
5.34
6.88
17.65
4.91

9.37
8.96
10.39
-33.20
26.93
19.13

36.99
35.58
29.51
-171.57
30.22
82.08

-19.53
-11.78
-35.57
       As shown in Figure 6-9, the units that fired bituminous coal exhibited average
emission reductions of 18 percent (Dunkirk), -17 percent (Gaston), and 27 percent (Cliffside).
In Figure 6-10, the HS-ESP units that burned Subbituminous coal and lignite exhibit Hg
emission reductions of 2 percent (Cholla), -1 percent (Columbia), -3 percent (Platte), and -6
percent (Presque Isle). Stack concentrations of Hg were substantially higher for the units
burning Subbituminous coal than for those burning bituminous coal.

       Hot-side ESPs tend to exhibit poor capture because they operate over a temperature
range where the oxidization and adsorption of Hg is limited.
                                        6-31

-------
                                              Hg(p)Out HHg(2+)Out DHg(0)Out
                          0         2          4         6          8          10

                                   Mercury Concentration, jig/dscm @ 3% O2



     Figure 6-9. Mercury emissions from bituminous-fired PC boilers with HS-ESP.
      Presque Isle 9 [-6%]
    g      Platte [-3%]

         Columbia [-1%]
          Cholla 3 [2%]
                               IHg(p)Out DHg(2+)Out DHg(0)Out
                    0     2     4     6      8     10    12     14     16
                            Mercury Concentration, (ig/dscm @ 3% O2
Figure 6-10. Mercury emissions for subbituminous- and lignite-fired PC boilers with HS-
                                           ESP.
                                           6-32

-------
6.6.3.3 FF Baghouses

       Six PC-fired units with FF baghouses were tested. The results of one test unit (Valley)
were omitted from the results because of data quality problems. The unit name, type of coal
burned, and reduction in HgT are given in Table 6-12 for the five units with valid test data.

                                     Table 6-12
            Mercury (HgT) Reduction at PC-fired Units with FF Baghouses
Unit
Sammis
Valmont
Shawnee
Boswell 2
Comanche
Coal
Bituminous
Bituminous
Bituminous/subbituminous
Subbituminous
Subbituminous
Reduction in HgT,
%
92
87
70
83
62
       Detailed test results for the five units listed in Table 6-12 are given in Table 6-13. The
average run-to-run HgT reductions for the FF units ranged from 53 to 92 percent.  The class
average emission reductions for the two bituminous-coal-fired units was 90 percent, the
average for the single unit that fired bituminous and Subbituminous coals was 70 percent, and
the class average for the two units that fired Subbituminous coal was 72 percent. There were
generally high stack concentrations of Hg2+ for all FF units. Hg can be oxidized as it passes
through the FF, either from reactions with fly ash on the filter cake or from reactions  with bag
filter material. This can lead to relatively low concentrations of Hg in the stack gas.   These
observations may not apply to all bag filter materials, or units that burn either lignite  or
Subbituminous coal.

6.6.3.4  Comparison of ESPs and FFs

       The average unit-to-unit reductions in HgT in the inlet and outlet of PC-fired units
equipped with a CS-ESP, HS-ESP, or FF baghouse are shown in Figure 6-11. Stack
concentrations and speciation results are shown in Figure 6-12. SPF results are shown in
Figure 6-13.

       The best Hg capture is exhibited for units equipped with a FF (72 to 90 percent
average reductions). This is followed by units that are equipped with a CS-ESP and that burn
bituminous coal or bituminous coal and petroleum coke (35 to 54 percent average reductions).
Poor capture (-4 to 9 percent average reductions) is shown for all units that are equipped with
a HS-ESP and for units that are equipped with a CS-ESP and burn either Subbituminous coal
or lignite.  Units, which exhibit poor HgT capture, display higher SPF values than units that
have good HgT capture. In units that burn bituminous coal or bituminous  coal and petroleum
                                        6-33

-------
        2+
coke, Hg  constitutes more than half of the total Hg in the stack gas. This is also true for the
unit that is equipped with a FF and burns subbituminous coal.
                                     Table 6-13
                       Post-combustion Controls: FF Baghouses
Hg Speciation at Inlet and Outlet (ug/dscm(S> 3% O2) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In Hg In HgT In HgT In Hgp Out Hg2+ Out Hg Out HGT Out %RHT %RHgT
No. OH OH OH OH Coal OH OH OH OH OH Coal
Bituminous Coal. PC Boiler \\
Sammis
Saminis
Saminis
Average
Valmont
Valmont
Valmont
Average
1
2
3

1
2
3

11.78
15.35
14.62
13.92
0.92
0.92
1.23
1.02
ith FF Baghouse
0.48
0.50
0.51
0.50
0.12
0.07
0.10
0.10
0.61
0.54
0.52
0.55
0.18
0.14
0.17
0.17

12.86
16.38
15.65
14.97
1.22
1.12
1.51
1.29

6.64
9.54
9.55
8.58
0.80
0.44
0.60
0.61

0.01
0.01
0.02
0.01
0.00
0.00
0.00
0.00

0.49
0.58
0.51
0.53
0.12
0.10
0.21
0.14

0.61
0.55
0.57
0.57
0.04
0.02
0.03
0.03

1.11
1.14
1.10
1.12
0.16
0.12
0.24
0.17

91.37
93.04
92.97
92.46
86.98
89.53
84.15
86.89

83.28
88.05
88.48
86.60
80.04
73.26
60.16
71.16
Average 7.47 0.30 0.36 8.13 4.59 0.01 0.34 0.30 0.64 89.67 78.88
Minimum
Maximum
STDEV



0.92
15.35
7.16
0.07
0.51
0.22
Bituminous Coal/Pet. Coke. PC Boiler w
Valley
Valley
Valley
Average
Bituminous/Su
Shawnee
Shawnee
Shawnee
Average
1
2
3

liliiti
i
2
3

0.04
0.05
0.04
0.04
jminous C
3.18
3.01
3.44
3.21
1.44
1.49
1.22
1.38
oal. PC Bo
0.58
0.98
0.57
0.71
0.14
0.61
0.22
1.12
16.38
7.59
0.44
9.55
4.49
0.00
0.02
0.01
0.10
0.58
0.22
0.02
0.61
0.30
ith FF Baghouse < Measurements not valid. disreeareD
1.21
0.45
0.67
0.78
iler with F
0.72
0.66
0.67
0.68
Subbituminous Coal. PC Boiler with FF Baehouse
Boswell 2
Boswell 2
Boswell 2
Average
Comanche
Comanche
Comanche
Average
2
3
1

1
3
2

1.99
0.83
2.75
1.85
1.81
5.27
2.59
3.23
1.26
1.15
1.81
1.41
3.93
1.28
1.45
2.22
1.46
2.49
1.60
1.85
5.71
3.67
5.77
5.05
2.69
1.99
1.92
2.20
F Bagho
4.48
4.65
4.68
4.61
4.71
4.46
6.16
5.11
11.46
10.22
9.82
10.50
0.95
1.33
1.52
1.27
use
2.39
4.29
2.66
3.11
4.35
5.20
8.35
5.97
15.91
14.24
17.08
15.74
0.11
0.04
0.00
0.05
0.01
0.02
0.01
0.01
0.00
0.00
0.07
0.03
0.00
0.00
0.00
0.00
2.02
1.55
1.89
1.82
0.63
0.61
0.60
0.61
0.35
0.58
1.26
0.73
3.33
3.20
3.99
3.51
0.41
0.42
0.52
0.45
0.84
0.75
0.68
0.76
0.23
0.12
0.14
0.16
0.27
0.33
0.65
0.42
0.12
1.14
0.52

2.54
2.00
2.41
2.31
1.49
1.37
1.28
1.38
0.59
0.70
1.47
0.92
3.60
3.52
4.65
3.92
84.15
93.04
3.54

5.67
-0.70
-25.15
-6.73
66.73
70.51
72.62
69.95
87.45
84.32
76.06
82.61
68.58
65.52
52.67
62.26
60.16
88.48
10.76

-165.84
-50.84
-58.75
-91.81
37.66
68.03
51.82
52.50
86.43
86.54
82.34
85.10
77.37
75.26
72.80
75.14
Average 2.54 1.81 3.45 7.80 10.86 0.01 2.12 0.29 2.42 72.43 80.12
Minimum
Maximum
STDEV



0.83
5.27
1.50
1.15
3.93
1.06
1.46
5.77
1.94
4.46
11.46
3.06
4.35
17.08
5.59
0.00
0.07
0.03
0.35
3.99
1.57
0.12
0.65
0.20
0.59
4.65
1.72
52.67
87.45
12.91
72.80
86.54
5.85
                                        6-34

-------
Coal/APCD [Hg Reduction,%]
Sub(wet)/HS-ESP
[-3%]
Sub/HS-ESP
Bit/HS-ESP
Lig/CS-ESP
[-4%]
Sub/CS-ESP
Bit/CS-ESP
[36%]
Sub/FF
[72%]
Bit/FF
[90%]
l
-------
  u
 a
 a
 u
 1
  cS
  o
 U
Sub(wet)/HS-ESP
[-3%]
Sub/HS-ESP
[6%]
Bit/HS-ESP
[9%]
Lig/CS-ESP
[-4%]
Sub/CS-ESP
P%]
Bit/CS-ESP
[36%]
Sub/FF
[72%]
Bit/FF
[90%]
DHg(p) Out HHg(2+) Out DHg(O) Out


! 	 1

c::::::::i

IIIIIIZIIIIIIIZ'IIIIIIIZI]

in

iiiiiiiiiiiiiiiiiiB^^



c:::::::::::::::^^^^^

\ 	 i

                             20           40          60           80

                                Relative Mercury Concentration in Stack, %
100
       Figure 6-13. Relative mercury speciation for PC-fired boilers with ESPs and FFs.
6.6.3.5 Other PM Controls

       Other PM control methods that were tested included two units firing TX lignite and
equipped with a CS-ESP followed by a pulse-jet FF baghouse, and one PC-fired unit burning
subbituminous coal and equipped with a PM scrubber (see Table 6-14).  The three-run average
HgT reduction across the PM scrubber on this latter unit was 9 percent.

       At the Bigbrown and Monticello units, the inlet and outlet Hg measurements were
made across the baghouse. There is little consistency between three runs for the Monticello
unit, and the data may not be valid.  Bigbrown exhibited negligible HgT capture across the FF.
While some Hgp and Hg2+ may have been captured in the upstream ESP, the low amounts of
fly ash captured in the downstream FF probably account for the lack of HgT capture in the
baghouse.
                                        6-36

-------
                                     Table 6-14
                Post-combustion Controls: Miscellaneous PM Controls
Hg Speciation at Inlet and Outlet (us/dscm(o), 3% O2) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In Hg In HgT In HgT In Hgp Out Hg2+ Out Hg Out HGT Out %RHT %RHgT
No. OH OH OH OH Coal OH OH OH OH OH Coal
TX Lignite. PC Boi
Bigbrown
Bigbrown
Bigbrown
Average
Monticello 1-2
Monticello 1-2
MonticeMo 1-2
Average
1
2
3

1
2
3

ler with CS-ESP and FF fCOHl
2.59
0.54
0.14
1.09
15.97
0.37
7.97
8.10
8.35
10.37
14.14
10.95
22.54
14.82
22.74
20.03
31.24
27.31
21.93
26.83
8.82
46.29
44.19
33.10
PACT)
42.18
38.21
36.21
38.87
47.34
61.48
74.90
61.24

50.86
49.95
46.92
49.24
53.79
54.09
84.65
64.18

0.01
0.01
0.01
0.01
0.17
0.11
0.08
0.12

16.58
17.66
18.49
17.58
32.01
78.08
86.89
65.66

25.20
25.47
22.12
24.26
1.58
14.93
18.51

41.80
43.13
40.62
41.85
33.76
93.11
105.48
11.67 77.45

0.92
-12.88
-12.20
-7.68
28.69
-51.46
-40.84
-21.20

17.82
13.65
13.42
14.96
37.23
-72.13
-24.61
-19.83
Average 4.60 15.49 29.96 50.05 56.71 0.07 41.62 17.97 59.65 -13.63 -2.44
Minimum
Maximum
STDEV



0.14
15.97
6.31
8.35
22.74
6.03
Subbituminous Coal. PC Boiler with PM
Boswell 3
Boswell 3
Boswell 3
Average
1
2
3

0.01
0.01
0.06
0.03
0.25
0.31
0.62
0.39
8.82
46.29
14.07
Scrubbei
6.06
6.00
5.21
5.76
36.21
74.90
15.16

6.32
6.32
5.89
6.18
46.92
84.65
13.94

5.00
6.38
5.79
5.72
0.01
0.17
0.07

0.00
0.00
0.00
0.00
16.58
86.89
32.27

0.05
0.06
0.06
0.06
1.58
25.47
8.99

5.82
5.39
5.51
5.57
33.76
105.48
31.13

5.87
5.45
5.58
5.63
-51.46
28.69
26.49

7.16
13.81
5.25
8.74
-72.13
37.23
39.61

-17.51
14.54
3.60
0.21
 6.6.4 Hg Capture in Units with Dry FGD Scrubbers

       Thirteen units with dry scrubbing systems were tested.  One unit uses dry sorbent
injection in combination with a CS-ESP, three units use SDA/ESP systems, and the
remaining nine units are equipped with SDA/FF systems. Two of the units equipped with
SDA/FFs were also equipped with a SCR system.  Hg emission test results for the dry
scrubber units are summarized in Table 6-15.

       At the Port Washington unit, sorbent is injected downstream of the air preheater.  OH
inlet measurements were made upstream of the preheater, and outlet measurements were
made in the duct downstream of the CS-ESP. The average capture of HgT for the Port
Washington dry sorbent injection unit was 45 percent.  The SPF2+ and SPF values for this
unit fell within the range of values exhibited by PC-fired boilers that are equipped with a CS-
ESP and burn bituminous coal. The three pulverized subbituminous-coal-fired units
equipped with a SD A/ESP system exhibited average HgT captures of 25 percent (GRDA), 40
percent (Laramie 3), and 41 percent (Wyodak).

       As mentioned above, nine units equipped with a SD/FF system were tested. One unit
firing bituminous coal had a HgT capture of 98 percent.  The two units firing bituminous coal
and also equipped with an SCR system had a class average HgT capture of 99 percent. Three
SDA/FF units fired with subbituminous coal had HgT captures of 36, 32, and 5 percent.
                                       6-37

-------
                 Table 6-15
Post-combustion Controls: Dry FGD Scrubbers
Hg Speciation at Inlet and Outlet (|ig/dscm(3! 3% O2) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In HgIn HgT In HgT In Hgp Out Hg2+ Out Hg Out HGT Out %RHT %RHgT
No. OH OH OH OH Coal OH OH OH OH OH Coal
Bituminous Coal. PC Boiler with DSI and CS-ESP
Washington
Washington
Washington
Average
1
2
3

0.00
0.00
0.00
0.00
Subbituminoiis Coal. PC Boil
GRDA
GRDA
GRDA
Average
Laramie 3
Laramie 3
Laramie 3
Average
Wyodak
Wyodak
Wyodak
Average
1
2
3

1
2
3

1
2
3

4.33
7.75
6.40
6.16
erwithCS
0.13 1 4.42
0.53 2.97
0.51
0.39
0.03
1.69
4.55
2.09
2.49
3.05
2.25
260
8.78
5.39
0.22
0.52
0.44
0.39
3.88
4.71
3.57
4.05
11.63
11.02
9.95
10.86
ESP/SDA
7.77
6.50
3.71
5.99
0.63
8.53
9.28
6.15
11.63
9.42
11.51
1085

15.97
18.77
16.36
17.03
12.31
9.99
13.01
11.77
0.88
10.75
14.27
8.63
18.00
17.17
17.34
1750

13.01
13.36
13.33
13.24
11.22
10.73
12.24
11.40
15.03
17.67
14.94
15.88
4.46
6.41
8.17
634

0.00
0.00
0.04
0.02
0.01
0.01
0.01
0.01
0.03
0.03
0.03
0.03
0.05
0.05
0.05
005

6.41
5.84
6.52
6.26
1.55
1.28
0.34
1.06
0.10
0.04
0.04
0.06
0.07
0.17
0.25
0 16

3.06
3.05
3.04
3.05
5.58
11.12
5.41
7.37
3.87
4.52
5.27
4.56
9.97
10.11
10.11
1006

9.47
8.90
9.59
9.32
7.13
12.42
5.76
8.44
4.00
4.58
5.35
4.64
10.09
10.32
10.41
10.27

40.68
52.61
41.37
44.89
42.06
-24.23
55.72
24.51
0.00
57.35
62.53
39.96
43.95
39.87
39.99
41.27

27.22
33.42
28.06
29.56
36.42
-15.70
52.94
24.55
73.40
74.05
64.22
70.56
-126.38
-61.16
-27.41
-71 65
Average 1 69 3 28 7 67 12 64 11 21 0 03 0 43 7 33 7 78 35 25 7 82
Minimum
Maximum
STDEV



0.03
4.55
1.54
0.22
8.78
2.72
Bituminous Coal. PC Boiler with SDA/F
Mecklenburg
Mecklenburg
Mecklenburg
Average
1
2
3

11.34
5.66
6.90
7.97
3.40
4.21
3.04
3.55
Bituminous Coal. PC Boiler with SCR ai
Logan
Logan
Logan
Average
SEI
SEI
SEI
Average
1
2
3

1
2
3

12.87
12.74
12.45
12.69
13.48
9.47
12.01
11.66
7.22
4.36
4.59
5.39
0.30
0.25
0.25
0.26
0.63
11.63
3.60

6.16
0.02
0.02
2.07
id SDA/FI
0.21
0.35
0.25
0.27
0.14
0.18
0.16
0.16
0.88
18.00
5.28

20.91
9.89
9.96
13.59

20.31
17.46
17.29
18.35
13.92
9.90
12.42
12.08
4.46
17.67
4.32

11.52
13.28
11.50
12.10

18.28
18.14
17.51
17.98
11.79
11.74
11.97
11.83
0.01
0.05
0.02

0.00
0.00
0.00
0.00

0.02
0.02
0.01
0.02
0.01
0.01
0.02
0.01
0.04
1.55
0.57

0.07
0.07
0.01
0.05

0.08
0.13
0.04
0.09
0.34
0.09
0.08
0.17
3.87
11.12
2.91

0.09
0.23
0.23
0.18

0.16
0.17
0.17
0.16
0.13
0.12
0.11
0.12
4.00
12.42
3.06

0.16
0.31
0.24
0.24

0.26
0.32
0.22
0.27
0.48
0.22
0.21
0.30
-24.23
62.53
28.72

99.23
96.91
97.60
97.91

98.71
98.16
98.72
98.53
96.56
97.79
98.34
97.56
-126.38
74.05
70.07

98.60
97.70
97.92
98.07

98.57
98.23
98.74
98.51
95.94
98.13
98.28
97.45
Average 12.17 2.83 0.22 15.22 14.90 0.02 0.13 0.14 0.28 98.05 97.98
Minimum
Maximum
STDEV



9.47
13.48
1.41
0.25
7.22
2.98
0.14
0.35
0.08
9.90
20.31
3.82
11.74
18.28
3.38
0.01
0.02
0.00
0.04
0.34
0.11
0.11
0.17
0.03
0.21
0.48
0.10
96.56
98.72
0.81
95.94
98.74
1.03
                                              (continued)
                    6-38

-------
                                  Table 6-15 (cont'd)
                   Post-combustion Controls: Dry FGD Scrubbers
Hg Speciation at Inlet and Outlet (ug/dscm(S> 3% O2) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In Hg In HgT In HgT In Hgp Out Hg2+ Out Hg Out HGT Out %RHT %RHgT
No. OH OH OH OH Coal OH OH OH OH OH Coal
Subbituminous Coal. PC Boil
Craig 3
Craig 3
Craig 3
Average
Rawhide
Rawhide
Rawhide
Average
NSP Sherburne
NSP Sherburne
NSP Sherburne
Average
1
2
3

1
2
3

1
2
3

0.57
0.92
0.90
0.80
0.25
1.92
3.76
1.98
0.03
0.03
0.03
0.03
er with SDA/FF
0.65 0.20
0.50 0.17
0.23
0.46
1.38
0.83
0.46
0.89
0.53
0.23
0.19
0.32
0.12
0.16
12.46
12.85
14.79
13.37
10.92
10.92
10.24
10.69

1.42
1.60
1.25
1.42
14.09
15.59
19.01
16.23
11.48
11.18
10.46
11.04

1.20
1.06
0.93
1.06
8.09
7. 3
9. 4
8. 2
8. 9
8. 7
7.73
8.10

0.00
0.00
0.00
0.00
0.12
0.01
0.03
0.05
0.12
0.14
0.27
0.18

0.04
0.04
0.03
0.04
0.76
0.69
0.98
0.81
0.20
0.18
0.24
0.20

0.90
0.89
0.82
0.87
10.80
9.91
9.00
9.90
8.42
12.09
9.84
10.12

0.94
0.93
0.86
0.91
11.68
10.60
10.01
10.76
8.74
12.40
10.35
10.50

33.79
41.83
31.67
35.76
17.16
32.03
47.31
32.16
23.81
-10.94
1.05
4.64

21.49
11.88
7.38
13.58
-44.36
-44.61
-8.41
-32.46
-5.43
-49.92
-34.01
-29.78
Average 0.93 0.56 8.07 9.56 5.79 0.08 0.35 6.96 7.39 24.19 -16.22
Minimum
Maximum
STDEV



0.03
3.76
1.23
ND Lignite, PC Boiler with SI
Antelope Valley
Antelope Valley
Antelope Valley
Average
StantonlO
StantonlO
StantonlO
Average
1
2
3

1
2
3

0.16
0.21
0.16
0.18
0.22
0.27
0.50
0.33
0.19
1.38
0.37
1A/FF
0.38
0.42
0.16
0.29
0.24
0.36
0.69
0.43
0.12
14.79
6.08

7.80
7.82
7.67
7.75
10.23
9.86
9.45
9.85
1.25
19.01
6.63

8.34
8.45
8.00
8.22
10.70
10.49
10.64
10.61
0.93
9.24
3.59

13.85
16.03
12.50
14.27
12.82
15.63
9.45
12.63
0.00
0.27
0.09

0.01
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.03
0.98
0.36

0.25
0.79
0.33
0.56
0.40
0.17
0.01
0.19
0.82
12.09
4.68

omit
8.16
6.97
7.56
10.14
10.58
10.81
10.51
0.86
12.40
4.97

NA
8.98
7.32
8.15
10.56
10.76
10.83
10.72
-10.94
47.31
18.96

NA
-6.27
8.49
1.11
1.24
-2.54
-1.77
-1.02
-49.92
21.49
27.38

NA
44.01
41.45
42.73
17.63
31.15
-14.61
11.39
Average 0.27 0.38 9.01 9.65 13.29 0.02 0.34 9.33 9.69 -0.17 23.93
Minimum
Maximum
STDEV



0.16
0.50
0.13
0.16
0.69
0.20
Bituminous, Stoker with SDA/FF
Dwayne Collier
Dwayne Collier
Dwayne Collier
Average
1
2
3

2.19
2.14
1.99
2.11
0.03
0.18
0.03
0.08
7.67
10.23
1.18

0.06
0.42
0.11
0.20
8.00
10.70
1.32

2.28
2.75
2.13
2.39
9.45
16.03
2.67

3.37
3.48
3.29
3.38
0.01
0.02
0.01

0.06
0.03
0.01
0.03
0.01
0.79
0.29

0.02
0.03
0.03
0.03
6.97
10.81
1.69

0.08
0.09
0.06
0.08
7.32
10.83
1.53

0.16
0.15
0.10
0.14
-6.27
8.49
5.53

92.84
94.48
95.43
94.25
-14.61
44.01
23.91

95.16
95.64
97.04
95.95
       The average HgT captures in two units firing lignite were 1 and -1 percent.  A single
stoker-fired boiler burning bituminous coal had a total average Hg capture of 94 percent.

       The reduction in emissions for each SDA test class is shown in Figure 6-14. The
stack concentrations of Hgp, Hg2+, Hg, and HgT are shown in Figure 6-15 along with the
average total HgT capture for each SDA class. The relative Hg speciation for the same coal-
fired boiler classes is shown in Figure 6-16. The predominance of Hg in the stack emissions
from units fired with subbituminous coal and lignite is attributed to low levels of Hg
oxidization and the relative ease of Hg2+ capture.
                                        6-39

-------
    Sub, SDA/ESP
       [35%]
     Lig, SDA/FF

  Sub, SDA/FF
P      [24%]
g    Bit, SDA/FF
U       [98%]
4        6        8        10       12

Mercury Concentration, (ig/dscm @ 3% O2
14
                                                                                      16
                    Figure 6-14. Mercury control for dry FGD scrubbers.
Sub, SDA/ESP
[35%]

o
S Lig, SDA/FF
% [0%]
&
&
p Sub, SDA/FF
U [24%]
1
e
U Bit, SDA/FF
[98%]



g

^

|-| BHg^Out K|Hg(2+)OutDHg(0)Out
U
                          0        2       4       6        8       10      12

                                Mercury Concentration, jig/dscm @ 3% O2


              Figure 6-15. Mercury speciation for PC boilers with SDA.
                                         6-40

-------
        Sub, SDA/ESP
           [35%]   liiiL
  o
  'B
  a
  u
  o
  U
Lig, SDA/FF
                                       \ Hg(p) Out H Hg(2+) Out D Hg(0) Out
Sub, SDA/FF
   [24%]
         Bit, SDA/FF
           [98%]
                  0     10    20     30     40     50     60    70     80    90     100

                               Relative Mercury Concentration in Stack, %


          Figure 6-16. Relative mercury speciation for PC boilers with SDA.


6.6.5 Hg Capture in Units with Wet FGD Scrubbers

       The wet FGD scrubber systems that were tested consisted of units equipped with four
PM control device configurations.  These different configurations are expected to have
different effects on the speciation and capture of Hg. These different configurations included
units equipped with a PS,  a CS-ESP,  an HS-ESP,  or a FF baghouse.  Inlet and outlet
measurements on the PS + wet FGD units were made across both control devices. Inlet
measurements on the systems with an ESP or FF were made between the PM control device
and the FGD scrubber. Outlet measurements were made in the stack.

       A total of 23 units with wet FGD systems  were tested. Seven units used PM scrubber
systems to control particulate emissions, eight used CS-ESPs, six used HS-ESPs, and two
used FF baghouses.  Twenty-one of the test units burned pulverized coal. The other two test
units burned bituminous coal in cyclone-fired boilers. One unit was equipped with a PM
scrubber, and the other had a CS-ESP. The number of PC-fired test units in each coal-control
class is shown in Table 6-16. (Also see Tables 6-4 and 6-6.)
                                        6-41

-------
                                     Table 6-16
                     PC-fired Boiler PM Controls for Wet FGD Systems
PM
Control
PS
CS-ESP
HS-ESP
FF

Number of Test Units
Bit.
1
2
2
2

Subbit.
4
3
4
0

Lignite
1
2
0
0

Totals
6
7
6
2
21
       The results of emission tests on wet FGD systems are summarized in Table 6-17.  The
next to last column in Table 6-17 shows the percent reduction in HgT across the wet FGD
scrubber as determined by the OH sampling train measurements. The last column is an
estimate of the reduction in HgT across the PM control device and wet FGD scrubber.  These
estimates were made using the class PM average coal-boiler-control EMF that is applicable to
each test unit (see Section 6.5.2).

       Class average emission test results for the PC-fired boilers with wet FGD units are
shown in Figures 6-17, 6-18, and 6-19. Each of these figures is based on capture estimates
across the PM control device and wet FGD scrubber combinations.  Figure 6-17 shows the
class average stack concentrations of Hgp, Hg2+, and Hg. Figure 6-18 shows the average
inlet and outlet concentrations of HgT and percent reduction for each class. Figure 6-19
shows the relative mercury speciation for PC-boilers with wet FGD  scrubbers.
                                        6-42

-------
Errata Page 6-43, dated 3-21-02
Table 6-1 7
Post-combustion Controls: Wet FGD Scrubbers
Hg Speciation at Inlet and Outlet ( u g/dscm@3%O2): %
Plant ID
Hgpm
Run No. OH
Bituminous Coal, PC
Bruce Mansfield
Bruce Mansfield
Bruce Mansfield
Average
Subbituminous
Boswell 4
Boswell 4
Boswell 4
Average
Cholla 2
Cholla 2
Cholla 2
Average
Colstrip
Colstrip
Colstrip
Average
Lawrence
Lawrence
Lawrence
Average
Average
Minimum
Maximum
STDEV
1
2
3

Coal,
1
2
3

1
2
3

1
2
3

1
2
3





ND Lignite, PC Boiler
Lewis and Clark
Lewis and Clark
Lewis and Clark
Average
1
2
3

Hg" In
OH
Hg" In
OH
MgT in
OH
Reduction for
M9T Mgpuut
In Coal OH
Hg" Out
OH
OH Train and Coal Data
Hg" Out
OH
MgT uut
OH
VoK HCJf VoK HCJf
OH Coal
Boiler with PS and Wet FGD Scrubber
0.27
0.73
0.27
0.42
PC Boiler
0.11
2.98
2.75
1.95
0.42
1.11
0.41
0.65
1.78
1.94
1.63
1.78
0.23
0.53
0.24
0.33
1.18
0.11
2.98
1.02
8.65
9.84
8.34
8.94
1.58
2.08
1.70
1.79
with PS and Wet
0.33
1.07
0.55
0.65
0.97
0.93
2.06
1.32
2.29
2.37
2.86
2.51
1.65
0.63
0.65
0.98
1.36
0.33
2.86
0.85
with PS and Wet
1.15
1.68
1.41
1.41
16.47
13.64
6.28
12.13
5.05
1.47
1.16
2.56
4.68
2.62
2.99
3.43
1.08
6.37
5.39
4.28
4.99
4.41
4.96
4.79
3.76
1.08
6.37
1.82
10.50
12.65
10.31
11.15
FGD
5.48
5.53
4.45
5.15
6.07
4.66
5.46
5.40
5.15
10.68
9.88
8.57
6.86
5.58
5.86
6.10
6.30
4.45
10.68
1.96
10.93
8.93
11.82
10.56
Scrubber
6.98
6.63
7.93
7.18
6.99
6.37
5.09
6.15
7.63
7.98
7.93
7.85
6.24
5.47
6.03
5.91
6.77
5.09
7.98
0.98
0.04
0.06
0.04
0.05

0.02
0.20
0.28
0.17
0.15
0.19
0.11
0.15
0.05
0.02
0.02
0.03
0.01
0.08
0.09
0.06
0.10
0.01
0.28
0.09
1.89
2.73
1.22
1.95

0.10
0.44
0.59
0.38
0.21
0.14
0.14
0.16
0.42
0.45
0.39
0.42
0.49
0.53
0.51
0.51
0.37
0.10
0.59
0.17
7.01
7.96
8.29
7.76

5.53
5.89
5.57
5.66
3.93
4.67
4.22
4.27
9.13
11.03
2.13
7.43
6.37
6.71
6.20
6.42
5.95
2.13
11.03
2.34
8.95
10.76
9.55
9.75

5.65
6.53
6.43
6.21
4.29
5.01
4.46
4.59
9.60
11.51
2.54
7.88
6.87
7.32
6.81
7.00
6.42
2.54
11.51
2.40
14.81 18.11
14.94 -20.57
7.42 19.25
12.39 5.60

-3.08 19.00
-18.25 1.41
-44.40 18.91
-21.91 13.11
29.30 38.59
-7.51 21.38
18.29 12.27
13.36 24.08
-86.54 -25.89
-7.74 -44.19
74.27 67.94
-6.67 -0.71
-0.07 -10.01
-31.14 -33.75
-16.21 -12.96
-15.81 -18.91
-7.76 4.39
-86.54 -44.19
74.27 67.94
39.47 31.96
FGD Scrubber
11.65
8.43
10.20
10.09
29.27
23.75
17.89
23.64
15.33
15.54
18.96
16.61
0.06
0.00
0.00
0.02
0.50
0.35
0.50
0.45
13.86
14.19
15.81
14.62
14.42
14.55
16.31
15.09
50.75 5.98
38.74 6.41
8.81 13.94
32.77 8.78
CONTINUED
                               6-43a

-------
Errata Page 6-43b, dated 3-21-02
Table6-17(cont'd)
Post-combustion Controls: Wet FGD Scrubbers
Hg Speciation at Inlet and Outlet ( M


Plant ID

Run
No.
Bituminous Coal, PC
AES Cayuga
AES Cayuga
AES Cayuga
Average
Big Bend
Big Bend
Big Bend
Average
Average
Minimum
Maximum
STDEV
Subbituminous
Jim Bridger
Jim Bridger
Jim Bridger
Average
Laramie River 1
Laramie River 1
Laramie River 1
Average
Sam Seymour
Sam Seymour
Sam Seymour
Average
Average
Minimum
Maximum
STDEV
2
1
3

1
2
3





Coal,
1
2
3

1
2
3

1
2
3






Hgpln
OH
Boiler
0.00
0.00
0.00
0.00
0.09
0.05
0.02
0.05
0.03
0.00
0.09
0.03

Hg2+ In
OH

g/dscm@3%O2): % Reduction for OH Train and Coal Data

Hg In HgT In
OH
with CS-ESP
6.40
5.87
5.55
5.94
4.86
4.92
4.26
4.68
5.31
4.26
6.40
0.78
PC Boiler with
0.05
0.44
0.07
0.19
0.25
0.04
0.02
0.10
0.03
0.01
0.01
0.01
0.10
0.01
0.44
0.15
*Note the column title changes from
2.49
2.04
1.78
2.10
3.14
2.16
3.08
2.79
3.00
4.08
5.39
4.16
3.02
1.78
5.39
1.13
coal to Wet
2.58
2.24
2.95
2.59
2.40
2.31
2.13
2.28
2.43
2.13
2.95
0.30
OH

HgT HgpOut
In Coal OH

Hg2+ Out
OH

Hg Out
OH

HgT
Out OH
7oK HgT
Wet
FGD*
7oK HgT
PM+FGD
*
and Wet FGD Scrubber
8.98
8.11
8.50
8.53
7.34
7.29
6.41
7.01
7.77
6.41
8.98
0.94
CS-ESP and
5.21
5.64
4.50
5.12
7.52
8.35
7.53
7.80
9.10
13.10
11.96
11.38
8.10
4.50
13.10
2.94
7.74
8.12
6.35
7.41
10.91
10.55
10.63
10.70
12.13
17.19
17.35
15.56
11.22
6.35
17.35
3.88
11.87 0.00
10.70 0.00
10.80 0.00
11.12 0.00
17.52 0.05
11.25 0.00
12.01 0.03
13.59 0.03
12.36 0.01
10.70 0.00
17.52 0.05
2.59 0.02
0.18
0.36
0.18
0.24
0.21
0.12
0.23
0.19
0.22
0.12
0.36
0.08
2.70
2.73
3.08
2.83
2.18
1.75
2.05
1.99
2.41
1.75
3.08
0.50
2.88
3.09
3.26
3.08
2.44
1.87
2.31
2.21
2.64
1.87
3.26
0.53
67.91
61.88
61.63
63.81
66.70
74.37
64.01
68.36
66.08
61.63
74.37
4.78
76.06
71.56
71.38
73.00
75.16
80.88
73.15
76.39
74.70
71.38
80.88
3.56
Wet FGD Scrubber
no coal flow 0.06
no coal flow 0.05
no coal flow 0.03
not included 0.05
13.52 0.02
15.45 0.00
15.71 0.01
14.90 0.01
60.48 0.06
43.20 0.11
51.04 0.06
51.58 0.07
33.24 0.04
13.52 0.00
60.48 0.11
20.84 0.03
0.25
0.29
0.20
0.25
0.29
0.12
0.03
0.15
0.24
0.29
0.35
0.29
0.23
0.03
0.35
0.10
6.63
6.51
5.92
6.36
4.86
5.73
4.48
5.02
12.25
13.33
11.99
12.53
7.97
4.48
13.33
3.50
6.95
6.85
6.15
6.65
5.18
5.85
4.52
5.18
12.54
13.74
12.39
12.89
8.24
4.52
13.74
3.59
FGDandPM+FGD
10.32
15.64
3.06
9.68
52.57
44.54
57.53
51.55
1.51
23.90
31.99
19.13
26.78
1.51
57.53
21.09
14.60
19.67
7.69
13.99
54.83
47.18
59.56
53.86
1.51
23.90
31.99
19.13
28.99
1.51
59.56
20.83
CONTINUED
                               6-43b

-------
Errata Page 6-43c, dated 3-21-02
                            (Intentionally Blank)
                                  6-43c

-------
             Table 6-17 (cont'd)
Post-combustion Controls: Wet FGD Scrubbers
Hg Speciation at Inlet and Outlet (us/dscm(o), 3%O2 ) : % Reduction for OH Train and Coal Data
Plant ID Run Hgp In Hg2+ In Hg In HgT In HgT In Hgp Out Hg2+ Out Hg Out HGT Out %RHT %RHgT
No. OH OH OH OH Coal OH OH OH OH Wet FGD PM+FGD
TX Lignite, PC Boi
Monticello 3
Monticello 3
Monticello 3
Average
Limestone
Limestone
Limestone
Average
1
2
3

1
2
3

ler with CS-ESP and wet FGD Scrubbei
0.19
0.11
0.13
0.14
0.01
0.01
0.02
0.02
16.49
19.77
25.83
20.70
23.55
24.55
28.15
25.42
29.39
28.15
27.21
28.25
13.38
13.11
14.11
13.54
46.07
48.03
53.16
49.09
36.94
37.68
42.29
38.97

61.96
63.13
76.52
67.20
14.49
20.84
15.29
16.87

0.31
0.18
0.24
0.24
0.04
0.33
0.12
0.17

6.50
0.44
7.26
4.73
2.69
3.18
1.27
2.38

29.45
25.52
23.10
26.02
15.96
16.23
17.18
16.46

36.25
26.14
30.60
31.00
18.69
19.74
18.58
19.01

21.31
45.57
42.44
36.44
49.40
47.59
56.07
51.02

21.31
45.57
42.44
36.44
49.40
47.59
56.07
51 02
Average 0.08 23.06 20.89 44.03 42.04 0.20 3.56 21.24 25.00 43.73 43.73
Minimum
Maximum
STDEV



0.01
0.19
0.07
Bituminous Coal. PC Boiler \\
Charles Lowman
Charles Lowman
Charles Lowman
Average
Morrow
Morrow
Morrow
Average
1
2
3

1
2
3

2.64
1.55
3.45
2.55
0.05
0.01
0.03
0.03
16.49
28.15
4.24
ith HS-ES
3.33
3.98
3.55
3.62
10.80
8.31
6.98
8.70
13.11
29.39
8.09
36.94
53.16
6.28
14.49
76.52
28.12
-* and wet FGD Scrubber
2.09
2.17
2.02
2.09
4.41
4.10
3.32
3.94
8.06
7.69
9.01
8.26
15.27
12.42
10.33
12.67
23.49
21.50
23.94
22.98
5.48
5.42
5.38
5.43
0.04
0.33
0.11

0.06
0.07
0.05
0.06
0.05
0.03
0.04
0.04
0.44
7.26
2.76

1.68
1.86
2.06
1.87
2.06
1.79
1.12
1.65
15.96
29.45
5.63

3.39
3.50
3.19
3.36
5.00
4.50
4.55
4.68
18.58
36.25
7.32

5.13
5.44
5.30
5.29
7.11
6.31
5.71
6.38
21.31
56.07
11.89

36.44
29.31
41.18
35.64
53.46
49.18
44.70
49.11
21.31
56.07
11.89

44.29
38.03
48.44
43.58
59.20
55.45
51.52
55.39
Average 1.29 6.16 3.02 10.46 14.20 0.05 1.76 4.02 5.83 42.38 49.49
Minimum
Maximum
STDEV



0.01
3.45
1.50
3.33
10.80
3.05
2.02
4.41
1.08
7.69
15.27
2.91
5.38
23.94
9.65
Subbituminous Coal. PC Boiler with HS-ESP and wet FGD Scrubber
Coronado
Coronado
Coronado
Average
Craig 1
Craig 1
Craig 1
Average
Navaio
Navaio
Navaio
Average
San Juan
San Juan
San Juan
Average
1
2
3

1
2
3

1
2
3

1
2
3

0.03
0.03
0.03
0.03
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.03
0.02
0.08
0.02
0.04
0.99
0.82
1.09
0.96
0.33
0.29
0.16
0.26
2.91
0.45
0.62
1.33
6.25
3.31
5.07
4.87
2.19
1.86
1.87
1.97
3.61
2.52
1.99
2.71
3.55
3.93
3.50
3.66
5.81
4.26
3.62
4.56
3.20
2.71
2.99
2.96
3.97
2.85
2.19
3.01
6.49
4.41
4.16
5.02
12.08
7.65
8.70
9.47
4.45
4.76
3.86
4.36
2.45
2.79
2.30
2.51
4.37
2.63
2.63
3.21
7.94
8.69
11.00
9.21
0.03
0.07
0.02

0.02
0.08
0.11
0.07
0.00
0.00
0.01
0.01
0.05
0.02
0.01
0.03
0.05
0.08
0.05
0.06
1.12
2.06
0.35

0.04
0.07
0.13
0.08
0.13
0.11
0.09
0.11
0.04
0.04
0.04
0.04
0.45
0.38
0.31
0.38
3.19
5.00
0.75

3.56
1.83
3.08
2.82
2.13
2.09
2.03
2.08
3.67
3.79
3.77
3.75
7.14
4.79
4.66
5.53
5.13
7.11
0.75

3.61
1.98
3.32
2.97
2.26
2.20
2.14
2.20
3.76
3.85
3.82
3.81
7.64
5.25
5.02
5.97
29.31
53.46
8.74

-12.95
26.82
-11.30
0.86
43.05
22.93
2.44
22.81
42.00
12.65
8.25
20.97
36.74
31.35
42.31
36.80
38.03
59.20
7.66

-0.87
34.64
0.60
11.46
49.14
31.17
12.87
31.06
48.20
21.99
18.06
29 42
43.50
38.69
48.48
43.56
Average 0.03 1.86 3.23 5.12 4.82 0.04 0.15 3.54 3.74 20.36 28.87
Minimum
Maximum
STDEV



0.02
0.08
0.02
0.16
6.25
2.05
1.86
5.81
1.19
2.19
12.08
3.02
Bituminous Coal. PC Boiler with FF and wet FGD scrubbei
Clover
Clover
Clover
Average
Intermountain
Intermountain
Intermountain
Average
1
2
3

1
2
3

0.06
0.03
0.08
0.06
0.01
0.01
0.01
0.01
1.00
1.11
1.16
1.09
1.01
1.08
1.36
1.15
1.11
1.99
0.62
1.24
0.20
0.24
0.22
0.22
2.17
3.13
1.86
2.39
1.22
1.33
1.58
1.38
2.30
11.00
2.86

29.21
41.19
49.02
39.81
2.00
1.97
3.09
2.35
0.00
0.11
0.04

0.05
0.02
0.06
0.04
0.01
0.01
0.01
0.01
0.04
0.45
0.14

0.42
0.34
0.05
0.27
0.03
0.07
0.08
0.06
1.83
7.14
1.52

0.42
0.17
0.14
0.24
0.25
0.46
0.41
0.37
1.98
7.64
1.64

0.88
0.53
0.25
0.55
0.29
0.54
0.50
0.44
-12.95
43.05
20.32

59.42
83.13
86.76
76.43
76.15
59.67
68.68
68.16
-0.87
49.14
18.15

96.78
98.66
98.95
98.13
98.11
96.80
97.52
97.48
Average 0.03 1.12 0.73 1.88 21.08 0.03 0.16 0.31 0.50 72.30 97.80
Minimum
Maximum
STDEV



0.01
0.08
0.03
1.00
1.36
0.13
0.20
1.99
0.71
1.22
3.13
0.70
1.97
49.02
21.47
0.01
0.06
0.02
0.03
0.42
0.17
0.14
0.46
0.14
0.25
0.88
0.23
59.42
86.76
11.66
96.78
98.95
0.92
                    6-44

-------
       Sub, HS-ESP+wet FGD [20%]




        Bit, HS-ESP+wet FGD [42%]




        Lig, CS-ESP+wet FGD [44%] b^^




        Sub, CS-ESP+wet FGD [27%]
^
o
'B


I
M    Bit, CS-ESP+wet FGD [66%]
  P
  u

  I

  "3
  o
  U
          Bit,FF+wetFGD[72%]




         Lig, PM+wet FGD [33%]  I




         Sub, PM+wet FGD [-8%]  I




         Bit, PM+wet FGD [12%]
                                               Hg(p)0ut SHg(2+)Out DHg(0)Out
                              0        5       10       15       20       25      30



                                    Mercury Concentration, jig/dscm @ 3% O2






            Figure 6-17. Mercury speciation for PC boilers with wet FGD.

M



Q

U
a.

<

IS
o
U
 Sub, HS-ESP+wet FGD [20%]




 Bit, HS-ESP+wet FGD [42%]




 Lig, CS-ESP+wet FGD [44%]




 Sub, CS-ESP+wet FGD [27%]




 Bit, CS-ESP+wet FGD [66%]




     Bit, FF+wet FGD [72%]




    Lig, PM+wet FGD [33%]




    Sub, PM+wet FGD [-8%]




    Bit, PM+wet FGD [12%]
                       0
                                 10    15    20    25    30     35     40    45
                              Mercury Concentration, ng/dscm @ 3% O2


            Figure 6-18. Mercury emissions for PC boilers with wet FGD.
50
                                          6-45

-------
                                               DHg(p)Out SHg(2+)Out DHg(0)Out
 I
 Q
 U
 a.
 a
Sub, HS-ESP+wet FGD [20%]

Bit, HS-ESP+wet FGD [42%]

Lig, CS-ESP+wet FGD [44%]

Sub, CS-ESP+wet FGD [27%]

Bit, CS-ESP+wet FGD [66%]

    Bit, FF+wet FGD [72%]

   Lig, PM+wet FGD [33%]

   Sub, PM+wet FGD [-8%]

   Bit, PM+wet FGD [12%]
0     10    20    30    40    50    60     70   80

           Relative Mercury Concentration in Stack, %
                                                                        90
                                                                                   100
              Figure 6-19. Relative mercury speciation for PC boilers with wet FGD.
       The best levels of HgT capture are exhibited by units burning bituminous coal and
equipped with a FF (98 percent), CS-ESP (75 percent), or HS-ESP (50 percent). The higher
capture levels for bituminous-fired boilers equipped with the CS-ESP, HS-ESP, or FF control
devices are consistent with the high levels of Hg oxidization associated with these coal-boiler
control classes (see Figures 6-12 and 6-13).  The very high levels of Hg capture exhibited by
the bituminous-coal-fired boiler units with a FF and wet FGD system can be attributed to high
levels of Hg oxidization and to the capture or conversion of Hgp and Hg2+ as flue gas passes
through the FF cake.  Estimates of HgT capture across the wet FGD and PM + wet FGD
combinations are shown in Table 6-18 for units burning bituminous coal.  Detailed data for
these units are given in Table 6-17. The best control is exhibited by wet FGD systems
equipped with a FF followed by units equipped with a CS-ESP and a HS-ESP.

       The HgT capture in one test unit burning bituminous coal and equipped with a PM
scrubber + wet FGD system averaged 12 percent. Hg at the outlet of the  scrubber was
predominantly Hg.
                                         6-46

-------
                                      Table 6-18
                    Wet FGD Scrubbers Burning Bituminous Coal
Controls and Test Unit
FF + Wet FGD
Clover
Intermountain
Average
CS-ESP + Wet FGD
Big Bend
AES Cayuga
Average
HS-ESP + Wet FGD
Charles R. Lowman
Morrow
Average
Reduction in HgT, %
FGD
76
68
72

68
64
66

36
49
43
PM + FGD
98
97
98

76
73
75

44
55
50
       The estimated capture of HgT in wet FGD units burning subbituminous coals is given
in Table 6-19. The four PS units were Boswell 4, Cholla 2, Colstrip, and Lawrence. The inlet
and outlet HgT data appeared reasonable except for runs  1 and 2 on Colstrip. All tests on
Lawrence and Boswell 4 had HgT outlet concentrations higher than the corresponding HgT
inlet concentrations. Cholla 2, which had HgT emission reductions ranging from -8 to 29
percent, appeared to exhibit hysterisis effects.  One unit,  Lewis and Clark, burned a ND
lignite. This unit also appeared to exhibit hysterisis effects, with successive HgT reductions
for the three tests of 51, 39, and 9 percent.  The declining reductions in HgT capture were
mirrored by inlet reductions of HgT and Hg2+.

       The erratic nature and differences in capture for the CS-ESP units are probably due to
differences in the subbituminous coals being burned and the differences in the scrubber
operating conditions. Except for the Coronado tests, the test results on HS units were fairly
consistent. It is not known whether the sampling and analysis results from the Coronado unit
are incorrect or whether differences in the coal and operating conditions caused the lower HgT
capture results.
                                        6-47

-------
Errata Page 6-48, dated 3-21-02
                                  Table 6-19
             Wet FGD Scrubbers Burning Subbituminous Coal
Controls
and Test Unit
PS + Wet FGD
Boswell 4
Cholla 2
Col strip
Lawrence
Average
HS-ESP + Wet FGD
Coronado
Craig 1
Navajo
San Juan
Average
CS-ESP + Wet FGD
Jim Bridger
Laramie 1
Sam Seymour
Average
Reduction
in HgT, %
FGD
NA
NA
NA
NA


1
23
21
37
20

10
52
19
27
PM + FGD
-22
13
-7
-16
-8

11
31
29
44
29

14
54
19
29
      Two units, burning TX lignite and equipped with a CS-ESP, exhibited average HgT
captures of 46 percent (see Table 6-20). The SPF2+ for limestone was 0.65 and the SPF2+ for
Monticello 3 was 0.42, indicating moderately high relative concentrations of Hg2+ at the
scrubber inlets of these two units. TX lignites appear to have a higher oxidization and capture
potential than ND lignites.
                                      6-48

-------
Errata Page 6-49, dated 3-21-02
                                     Table 6-20
                       Wet FGD Scrubbers Burning TX Lignite
Controls and Test Unit
CS-ESP + Wet FGD
Limestone
Monticello 3
Average
Reduction in Hgx, %
FGD
51
36
44
PM + FGD*
51
36
44
*Estimated
6.6.6 Potential Effects of Post-combustion NOx Controls

       Post-combustion NOx controls convert NOx in the boiler flue gases to molecular
nitrogen and water using a catalytic process (selective catalytic reduction) or a noncatalytic
process (selective noncatalytic reduction). For both processes, a reducing agent (usually
ammonia) is injected into the boiler flue gas at a point upstream of any post-combustion PM
or SO2 control device. A limited amount of data is available in the ICR Hg emission database
regarding the potential effects of these post-combustion NOx controls on Hg capture.  These
data are presented in Table 6-21. Test results for pulverized-coal boilers burning bituminous
coal with either SNCR or SCR systems are compared to the results of tests on similarly
controlled units that do not use post-combustion NOx controls.

                                     Table 6-21
Potential Effects of Post-combustion NOx Control Technologies on Mercury Capture in
                      PC-fired Boilers Burning Bituminous Coal
Post-combustion
Controls
CS-ESP
SDA + FF
Post-
combustion
NOX Control
none
SNCR
none
SCR
Number of
Pulverized-
coal-fired
Boiler Units
Tested
6
1
2
1
Average Mercury
Capture by Control
Configuration
36%
91%
98%
98%
       Tests on the single pulverized-coal boiler unit using a CS-ESP with SNCR shows an
average Hg capture that is significantly higher than the six units tested with a CS-ESP using
no post-combustion NOx controls (91 percent with SNCR versus 36 percent without SNCR).
It was reported that the fly ash from the boiler unit using SNCR contained unusually high
levels of carbon. Because data are available only for this one test, it is not known whether
                                        6-49

-------
the high levels of Hg capture indicated by the test results are attributable to the high fly ash
carbon content, the use of an SNCR system, a combination of both, or some other site-
specific factor.

       A comparison of tests for pulverized-coal boiler units using an SDA with an FF
shows no discernable difference in Hg capture with or without the use of an SCR for post-
combustion NOx control. An average Hg capture of 98 percent was measured by the tests on
the one unit equipped with an SCR compared to 98 percent Hg capture for the two similar
units without SCR systems. Because  of the very high levels  of Hg capture by all of the tested
control configurations, it is not possible to determine the effect of SCR on Hg capture.

       Recent tests on a pilot-scale, pulverized-coal combustor, which was equipped with an
SCR and a CS-ESP, showed increased Hg capture when bituminous coals were burned but
not when a subbituminous coal was burned. Mercury emission reductions were observed
when the SCR system was operated normally with the injection of ammonia upstream of the
SCR catalyst.  Improvement of Hg capture was also noted when ammonia was injected, but
the SCR catalyst was bypassed.  These tests provide evidence that SNCR and SCR systems
may enhance Hg capture under some conditions.

6.7    COMBUSTION SYSTEM EFFECTS

       LNBs and combustion modification techniques are believed to increase the unburned
carbon in fly ash and increase the adsorption of Hg onto collectable fly ash. Since neither the
fly ash carbon content nor the LOI was measured during the  ICR field test, it is not possible
to evaluate Hg capture performance benefits that accrue from the use of NOx  control
combustion modification techniques.  The ICR field test program included tests on six
different unit classes using cyclone-fired boilers and six unit classes with FBCs. The results
of ICR tests on units with cyclone-fired boilers and FBCs are shown in Tables 6-22 and 6-23,
respectively.
                                       6-50

-------
Errata Page 6-51, dated 3-21-02
                            Table 6-22
                       Cyclone-fired Boilers
Hg Speciation at Inlet and Outlet (ug/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID
Run Hgpln Hg2* In
No. O-H O-H
Hg In HgTln
O-H O-H
HgT In Hgp Out
Coal O-H
Hg2*Out
O-H
Hg Out
O-H
HGTOut
O-H
%R HT
O-H
%R HgT
Coal
ND Lignite, Cyclone Boiler with CS-ESP
Leland Olds
Leland Olds
Leland Olds
Averac
1
2
3
e
0.56 0.23
0.26 0.46
2.85 0.81
0.41 0.34
3.30
8.80
4.77
6.05
4.09
9.51
8.43
6.80
5.63
10.18
7.94
7.90
0.00
0.00
0.00
0.00
0.82
1.09
1.60
0.95
4.04
5.26
LS
4.65
4.86
6.35
NA
5.60
-18.68
33.26
NA
7.29
13.66
37.64
NA
25.65
Subbituminous/Pet. Coke, Cyclone Boiler with HS-ESP
Nelson Dewey
Nelson Dewey
Nelson Dewey
Averac
1
2
3
e
0.01 0.49
0.01 0.24
0.01 0.12
0.01 0.28
3.20
2.19
2.06
2.48
3.69
2.43
2.18
2.77
6.62
6.47
6.09
6.39
0.10
0.04
0.04
0.06
0.26
0.16
0.25
0.22
3.33
2.40
2.44
2.72
3.69
2.60
2.73
3.00
0.13
-6.90
-24.95
-10.57
44.27
59.83
55.22
53.11
Lignite, Cyclone Boiler with Mechanical Collector
Bay Front
Bay Front
Bay Front
1
2
3
Average
0.76 0.78
1.08 0.67
0.09 0.77
0.64 0.74
2.17
1.94
1.74
1.95
3.70
3.69
2.60
3.33
3.58
3.01
3.36
3.32
1.19
0.86
0.48
0.84
0.60
2.75
3.57
2.30
1.91
1.80
1.78
1.83
3.69
5.40
5.84
4.98
0.34
-46.54
-125.00
-57.07
-2.95
-79.21
-73.79
-51.99
Lignite, Cyclone Boiler with SDA/FF
Coyote
Coyote
Coyote
1
2
3
Average
Bituminous,
Lacygne
Lacygne
Lacygne
Averac
Bituminous,
Bailly
Bailly
Bailly
Averac
Cyclone
1
2
3
e
Cyclone
1
2
3
e
0.69 1.62
1.18 2.98
1.69 3.07
1.19 2.34
Boiler with PS
6.70 3.99
6.52 3.34
5.98 0.59
6.40 2.64
13.68
13.90
14.91
14.29
and Wet FGD
1.30
0.60
0.61
0.84
Boiler with CS-ESP and wet
0.04 3.18
0.04 2.37
0.09 3.01

2.57
2.95
2.58

15.99
18.06
19.66
17.82
10.51
18.55
11.39
10.948803
0.08
0.14
0.08
0.08
0.04
0.24
0.44
0.24
13.97
LS
18.06
16.02
14.10
NA
18.58
16.34
11.81
NA
5.48
8.64
-34.23
NA
-63.12
-48.673625
Scrubbers
12.00
10.46
7.18
9.88
FGD
5.79
5.36
5.68

no inlet flow
no inlet flow
no inlet flow

Scrubber
4.41
5.20
4.08

0.04
0.05
0.09
0.06

0.00
0.00
0.00

0.44
0.43
0.41
0.43

0.36
0.31
0.39

8.74
7.41
5.10
7.08

2.85
2.62
2.78

9.22
7.89
5.59
7.57

3.22
2.93
3.17

23.18
24.53
22.17
23.29

54.24
54.95
54.11

no inlet flow
no inlet flow
no inlet flow


27.09
43.53
22.31

                              6-51

-------
Errata Page 6-52, dated 3-21-02
                                      Table 6-23
                              Fluidized Bed Combustors
Hg Speciation at Inlet and Outlet (ug/dscm(
Plant ID Run
No.
Lignite, FBC with
R.M. Heskett 1
R.M. Heskett 2
R.M. Heskett 3
Average
Anthracite Waste,
Kline Township 1
Kline Township 2
Kline Township 3
Average
Hgpln
O-H
CS-ESP
4.73
2.93
7.43
5.03
FBC with
44.54
43.12
44.97
44.21
Hg2* In
O-H

5.39
0.96
0.44
2.26
FF
0.12
0.06
0.06
0.08
Hgln
O-H

3.83
2.61
3.08
3.17

0.45
0.40
0.34
0.40
HgTln
O-H

13.95
6.50
10.94
10.46

45.11
43.58
45.37
44.69
 3%O2)
HgTln
Coal

13.54
12.68
11.11
12.44

148.68
212.95
153.77
171.80
: % Reduction for O-H Train and Coal Data
HgpOut
O-H

1.06
0.07
0.05
0.39

0.00
0.00
0.00
0.00
Hg2*0ut
O-H

1.44
0.41
0.18
0.68

0.06
0.06
0.06
0.06
Hg Out
O-H

4.57
5.31
4.74
4.87

0.06
0.06
0.06
0.06
HgTOut
O-H

7.07
5.78
4.98
5.95

0.12
0.12
0.12
0.12
%R HT
O-H

49.29
11.09
54.49
38.29

99.74
99.73
99.74
99.74
%R HgT
Coal

47.76
54.40
55.19
52.45

99.92
99.95
99.92
99.93
Bituminous Waste, FBC with FF
Scrubgrass 1
Scrubgrass 2
Scrubgrass 3
Average
184.04
124.11
76.68
128.28
Bituminous/Pet. Coke, FBC
Stockton Cogen 1
Stockton Cogen 2
Stockton Cogen 3
Average
2.71
1.56
2.08
2.12
0.68
0.42
0.22
0.44
with SNCR
0.06
0.07
0.06
0.07
0.19
0.09
0.07
0.12
and FF
0.06
0.06
0.06
0.06
184.91
124.62
76.97
128.83

2.83
1.69
2.20
2.24
100.09
101.35
100.25
100.56

1.68
1.44
1.66
1.59
0.00
0.00
0.00
0.00

0.02
0.03
0.03
0.03
0.07
0.05
0.04
0.05

0.04
0.05
0.05
0.05
0.08
0.07
0.07
0.07

0.05
0.05
0.05
0.05
0.15
0.12
0.11
0.13

0.11
0.13
0.12
0.12
99.92
99.91
99.85
99.89

96.09
92.16
94.48
94.25
99.85
99.89
99.89
99.88

93.39
90.80
92.67
92.29
Subbituminous, FBC with SCR and FF
AES Hawaii 1
AES Hawaii 2
AES Hawaii 3
Average
Lignite, FBC with
TNP 1
TNP 2
TNP 3
Average
0.26
0.35
0.36
0.32
CS-FF
21.65
10.65
28.12
20.14
0.04
0.17
0.11
0.10

8.68
4.51
13.78
8.99
1.29
1.38
1.18
1.28

7.42
6.09
7.04
6.85
1.59
1.90
1.64
1.71

37.74
21.25
48.94
35.98
3.77
3.72
2.51
3.33

63.81
44.22
95.04
67.69
0.00
0.00
0.00
0.00

0.04
0.03
0.04
0.04
0.02
0.02
0.02
0.02

12.13
6.78
13.54
10.82
0.68
0.90
0.55
0.71

4.74
2.94
5.07
4.25
0.70
0.92
0.58
0.73

16.91
9.76
18.66
15.11
55.84
51.35
64.91
57.37

55.20
54.07
61.88
57.05
81.39
75.16
77.06
77.87

73.50
77.93
80.37
77.27
6.7.1 Cyclone-fired Boilers

      Mercury capture and stack gas speciation for cyclone-fired boilers are shown in
Figures 6-20 and 6-21. The percentage of total Hg capture in these units appears to be similar
to the Hg captured in pulverized-coal-fired units burning similar fuels and equipped with
comparable air pollution control devices (see Table 6-24). Except for the unit equipped with
a mechanized collector, the Hg in flue gas consisted primarily of Hg.
                                       6-52

-------

d
5? ..
f) 14

(X)
1
a'
0
c
1 .
a 6
0
U
fe1 A

3 ,
S 2



DHg(p)Out QHg(2+)Out DHg(0)Out











r^
1  1


ps
1  |




ri


i^
i  1






	 rs^K-^
1 	 1











         Bit/CS-ESP/Wet-  Ligmte/CS-ESP  Lig/MC (-57)    Bit/PS (23)   Sub-Pet/HS-ESP Ligmte/SDA/FF
            FGD(54)        (7)                                  (-11)         (9)

                              Coal/APCD (Hg Reduction, %)

             Figure 6-20. Mercury speciation for cyclone-fired boilers.

   100

    90%
a  70%
o
U
sS
3
    60%


    50%


    40%


    30%


    20%


    10%
           Bit/CS-    Lignite/CS-ESP  Lig/MC (-57)   Bit/PS (23)  Sub-Pet/HS-ESP Lignite/SDA/FF
         ESP/Wet-FGD      (7)                                 (-11)         (9)

             (54)             Coal/APCD (Hg Reduction, %)

        Figure 6-21. Relative mercury speciation for cyclone-fired boilers.
                                           6-53

-------
                                     Table 6-24
                            Comparison of Class Average
                  HgT Reductions for PC- and Cyclone-fired Boilers
Unit Class
Lignite, CS-ESP
Subbituminous/Pet Coke, HS-ESP
Lignite, Multi cyclone
ND Lignite, SDA/FF
Bituminous, PM scrubber + wet FGD
Bituminous, CS-ESP + wet FGD
Reduction in HgT, %
Cyclone
9
0
0
7
23
54
PC-Fired
36
7
NA
2
12
81
6.7.2  Fluidized-bed Combustors

       Six fluidized-bed combustors were tested on the ICR program. Test results for the
fluidized-bed units are shown in Figures 6-22 and 6-23. All of the units injected limestone
into the FBC to control SC>2 emissions.  One unit was equipped with a CS-ESP while the
remaining five units were equipped with a FF. One of the FF units was also equipped with an
SNCR system. The unit equipped with the CS-ESP burned lignite.  The capture of HgT for
this unit averaged 38 percent. The reduction in HgT for units equipped with FF systems
depended primarily on the type of fuel that was burned. The one unit that burned
subbituminous coal was equipped with an SCR system and a FF.  Inlet and outlet HgT
concentrations for the two valid runs on this unit were 1.7 and 0.7 |ig/dscm, respectively,
resulting in a 57 percent capture efficiency. One unit that burned waste anthracite had an
average HgT reduction efficiency of 99.7 percent, while another unit burning bituminous coal
and petroleum coke had an average reduction of 94 percent.

       The best performance for any unit  tested during the Part in ICR program exhibited
average HgT inlet concentrations of  185 |ig/dscm,  outlet concentrations of 0.15 |ig/dscm, and
an average HgT reduction of 99.9 percent.
                                        6-54

-------
N
 ,0

8 s
u 8
^
a
i
c~
centratio
= t
o
U
=
b 2
i





r-ES





n
E3
1
1
1
1
1










n Hg(p) out
S Hg(2+) Out
DHg(0)Out

       Ligmte/CS-ESP  Ant (waste)TFF  Bit-Pet/FF (94)   Bit (waste)/FF  Sub/SCR/FF (57)  Lig/FF (57)
           (38)          (99)                      (99)
                           Coal/APCD (Hg Reduction, %)
                   Figure 6-22. Mercury speciation for FBCs.
  100%

   90%
   80%
CO
S  70%
I
g  60%
I
r?
1
   50%
40%
                                                                          DHg(0)Out
                                                                          QHg(2+)Out
                                                                          D Hg(p) Out
   30%
    0%
       ligtite/CS-ESP Ant (waste)/FF Bit-Pet/FF (94)  Bit (waste)'FF  Sub/SCR/FF    IigFF(57)
           (38)         (99)                     (99)         (57)
                          Coal/APCD (Hg Reduction, %)
              Figure 6-23. Relative mercury speciation for FBCs.
                                          6-55

-------
6.7.3  IGCC Facilities

        Table 6-25 summarizes the emission source test data and coal analysis data for the
Tampa Electric Company Polk Power Project and Wabash River Coal Gasification
Repowering Project.  A more detailed presentation of the test data is included in Appendix C
of this report.  Coal data were used to calculate inlet feed rates of total Hg to the coal-
gasification units.  The total Hg in the exhaust gas from the gas turbine was determined by
summing the three Hg species obtained using the OH Method during each test run (i.e.,  Hgp,
Hg2+,andHg).

                                            Table 6-25
       Calculated Mercury Removal in IGCC Power Plants Using Bituminous Coal
IGCC
Facility
Tampa
Electric
Company
Polk Power
Project
Wabash River
Coal
Gasification
Repowering
Project
Test
Run
Runl
Run 2
Run 3
3-Run
Average
Runl
Run 2
Run 3
3-Run
Average
Coal Fed to Gasifer
Coal
Flow
Rate
(kg/hr,
dry)
91,454
88,707
71,373
83,845
90,663
89,629
89,493
89,928
Total Hg
Content
(ppm,
dry)
NDa
NDa
NDa
	
0.064
0.068
0.070
0.067
Total Hg
Feed
Rate
(kg/hr)
0.0091 b
0.0089b
0.0071 b
0.0084
0.0058
0.0061
0.0063
0.0061
Gas Turbine Exhaust Gas Stream
Gas Stream
Flow Rate
(dscm/hr)
1,430,191
1,453,617
1,414,052
1,432,620
1,372,064
1,385,884
1,352,458
1,370,135
Total Hg
Content c
(pg/dscm)
3.94d
3.86d
3.68 d
3.83d
2.57 e
2.60 e
2.76 e
2.64 e
Total Hg
Emission
Rate
(kg/hr)
0.0056
0.0056
0.0052
0.0055
0.0035
0.0036
0.0037
0.0036
Overall
Mercury
Removal
(%)
38
37
27
34
40
41
41
41
(a)  No mercury was detected by the test method used to analyze the coal.
(b)  Feed rate calculated assuming total mercury content of the coal is at the detection limit for the analytical method (0.1 ppm).
(c)  Total mercury content of the gas turbine exhaust stream determined using Ontario-Hydro Method results.
(d)  No particle-bound mercury was detected by the Ontario-Hydro Method.  To calculate total Hg content, it is assumed that particle-bound
    mercury concentration in gas stream is negligible. It is reasonable to assume that, consistent with good IGCC operating practices, the
    total particle concentration of the syngas burned in the gas turbine needs to be very low in order to prevent premature wear of the gas
    turbine blades.
(e)  No particle-bound mercury or oxidized mercury was detected by the Ontario-Hydro Method. To calculate total Hg content it is assumed
    that particle-bound mercury and oxidized mercury concentrations in the gas stream are negligible. At the Wabash River facility the
    syngas is cleaned and conditioned before burning in the gas turbine by a barrier filter for particulate removal, a water scrubber for gas
    cooling, and an amine scrubber for removal of reduced-sulfur species. It is reasonable to assume that these filtration and scrubbing
    processes remove the particle-bound and oxidized mercury from the gas stream.


        The operating difficulties experienced at the Pinon Pine IGCC facility demonstrate that

good operating practices dictate the need for  the concentration of particulate matter in the

syngas to be continuously maintained at very low levels to prevent premature gas turbine

blade erosion.  The OH Method measurements obtained at both of the tested IGCC facilities

are consistent with this operating practice.  In both cases, the OH Method detected no particle-

bound Hg in the gas turbine exhaust gas.  With very low numbers of particles in the syngas
                                               6-56

-------
stream to begin with, the elemental Hg released during the coal gasification process has very
few opportunities to be adsorbed on solid particles to form particle-bound Hg.

      The OH Method test results show that elemental Hg is the predominant species in the
gas turbine exhaust gas. For the Polk IGCC facility, the measured distribution of gaseous Hg
species was approximately 90 percent elemental Hg and 10 percent Hg2+. For the Wabash
River IGCC facility, no Hg2+ was detected by the OH Method (i.e., 100 percent of the HgT in
the exhaust gas stream was in the form of Hg). One possible explanation for these results is
the different gas cleaning processes used at the two IGCC facilities.  The syngas from the
coal gasifier at the Wabash River IGCC facility is cleaned and conditioned using a system
that includes a water scrubber for gas cooling and an amine scrubber for removal of reduced-
sulfur species. Oxidized Hg is water-soluble and is readily absorbed by a wet scrubbing
system.  However, Hg is insoluble and passes through a wet scrubbing system.  Thus, it is
reasonable to  expect that the water and amine scrubbers used at the Wabash River IGCC
facility effectively remove the oxidized Hg in the syngas before it is burned in the gas
turbine.

      The Polk IGCC facility uses a hot gas-cleaning system.  There is no wet scrubbing
process to remove any Hg2+ from the syngas before it is burned.  The syngas is not cooled and
remains at elevated temperatures until it is fed to the gas turbine.  It cannot be determined
from the test data how the elevated syngas temperatures and combustion process in the gas
turbine combustors affect Hg speciation. However,  it is believed that any Hg2+ in the syngas
will be converted back to Hg when the syngas is burned.  The degree of oxidization will
probably be limited by the combustion gas composition and the rate at which it is cooled
before it is emitted to the atmosphere.

      The last column in Table 6-25 provides an estimate of the overall amount of Hg in the
coal removed by the IGCC process.  Based on these two tests, approximately one-third of the
Hg in the coal is removed. The Hg that remains in the combustion gas is primarily Hg.
 6.8  NATIONAL AND REGIONAL EMISSION ESTIMATES

       Estimates of the nationwide Hg emissions provide an indication of the overall level of
Hg capture being achieved by existing control systems used by coal-fired utility boilers in the
United States.  A number of different approaches can be used for these estimates. The EPA
evaluated four different methods for estimating nationwide Hg emissions using information
from the ICR database.  The method selected as being the best is outlined below:

         ICR Part II coal data were used to determine the  average Hg content and the
          amount of coal burned in each of 1143 units supplying data for 1999.

         Mercury in the flue gas from coal burned in each boiler unit in 1999 was
          calculated assuming that all of the Hg in the coal was in the flue gas leaving the
          furnace.

                                        6-57

-------
         Each unit was assigned a coal-boiler-control class that best met the characteristics
          of the unit.

         Total Hg in the boiler flue gas for each unit was multiplied by the class emission
          factors for speciated and total Hg that had been assigned to the unit.

         Total and speciated Hg emissions for each unit were added to provide estimates of
          national Hg emissions from coal-fired utility boilers in 1999.

       Computer runs using this procedure resulted in estimated national Hg emissions in
1999 of 43.5 tons.

       Using the EPA's ICR database, EPRI independently estimated the nationwide Hg
emissions from existing coal-fired utility boilers in the United States to be in the range of 45
to 48 tons in 1999. EPRI selected a different estimation methodology than the one used by
EPA.  EPRI's method is based on a model that correlates the level of Hg emissions with the
amount of chlorine in coal and the ratio of chlorine to sulfur in the coal for the case of units
with cold-side ESPs.  Both the EPA and EPRI estimate that approximately 75 tons of Hg was
in coals burned in!999.

       After EPA announced its decision to develop the NESHAP, the transfer of data from
the field test reports to the emission databases was rechecked for errors. It was found that
several test units had been assigned to the wrong coal-boiler-control classes. Also, the results
of a number of tests failed data quality requirements and were removed from the analysis set.
Subsequent computer evaluations resulted in the following estimates:

         48 tons of Hg was emitted to the atmosphere from coal-fired utility boilers in
          1999,  and

         27 tons of Hg was captured by existing flue gas cleaning devices.

       Nationwide, approximately 25  tons (52 percent) of Hg was emitted from the
combustion of bituminous coal, followed by 17 tons (36 percent) from the combustion of
subbituminous coals,  and 4 tons (8 percent) from the combustion  of lignite. The total
amounts of Hg emitted compared to the tonnage and types of coal burned in 1999 are
presented in Table 6-26.
                                        6-58

-------
                                      Table 6-26
                    Nationwide Coal Burned and Mercury Emitted
                 From Electric Utility Coal-fired Power Plants in 1999
Coal Type
Bituminous
Subbituminous
Lignite
Other
Total
Nationwide
Total Coal
Tonnage
Burned In 1999
(dry tons) (a)
427,572,000
279,227,000
50,932,000
10,756,000
768,487,000
Percent of
Total Coal
Burned
56
36
7
1
100
Nationwide
Total Mercury
Emitted in 1999
(tons)
25
17
4
2
48
Percent of
Total Mercury
Emitted
52
36
8
4
100
  (a) For wet tons (as received), nationwide total is 928,398,000 tons in 1999.
     Percentages for wet tons are 50% bituminous, 41% subbituminous, and 8% lignite.

6.9  SUMMARY AND CONCLUSIONS

       Previous research has shown that the capture of Hg by flue gas cleaning devices is
dependent on Hg speciation. Both Hg and Hg2+ are in a vapor phase at flue gas cleaning
temperatures.  Hg is insoluble in water and cannot be captured in wet scrubbers. The
predominant Hg2+compounds in coal flue gas are weakly-to-strongly soluble and can be
generally captured  in wet FGD scrubbers. Both Hg and Hg2+can be adsorbed onto porous
solids such as fly ash, PAC, or calcium-based acid gas sorbents for subsequent collection in a
PM control device.  Hg2+is generally easier to capture by adsorption than Hg. Hgp is attached
to solids that can be readily captured in ESPs and FFs.

       The evaluation of ICR data provides valuable insights into relationships between the
speciation and capture of Hg, the type of coal burned, the types of boilers used, and the types
of post-combustion technologies used for flue gas cleaning.  The evaluation of ICR data
indicates that the behavior of Hg in conventional PC-fired utility boilers is primarily
dependent on the type of coal burned and the control technologies used at each site. This
behavior is consistent with the ensuing interpretations.
                                         6-59

-------
Bituminous Coals

       The Hg in flue gas from the combustion of bituminous coal is readily oxidized and
converted to Hgp or Hg2+.  The best technologies for controlling corresponding Hg emissions
are dry or wet FGD scrubbers along with post-combustion PM controls. Dry scrubbing
systems that use a SDA/FF are superior in performance to those that use a SD A/ESP.  In
SDA/FF systems, Hg can be absorbed on PM in the SDA, and particulate- and gas-phase Hg
can be captured as it passes through the FF and its associated filter cake.  SD A/ESP systems
depend on the in-fight capture of Hg.

       A PM control device always precedes wet FGD scrubbers. Four types of PM control
devices are commonly used: FFs, CS-ESPs, HS-ESPs, and PM scrubbers. Units equipped
with a FF exhibit the best capture followed by units equipped with a CS-ESP, HS-ESP, and
PM scrubbers. Units that are equipped with FF + wet FGDs can capture Hg in FF and can
convert Hg to Hg2+ for subsequent capture in the scrubber. Hg capture in CS-ESP + wet FGD
systems depends on the degree of Hg capture and oxidization as the flue gas passes through
the CS-ESP. Hg capture in units equipped with HS-ESPs is generally lower than the capture
in CS-ESPs because HS-ESPs operate at temperatures where the oxidization and capture of
Hg is limited. The single test unit equipped with a PS + wet FGD system exhibited an average
HgT capture of 12 percent.

Subbituminous Coals

       Some subbituminous coals exhibit little, if any, Hg oxidization in PC-fired boilers.
Others display moderate amounts of Hg oxidization.  The use of low NOx burners tends to
increase the amount of unburned carbon and the potential for capturing gas-phase Hg.  The
ICR data  show that the oxidization of Hg can occur from gas-phase reactions or gas/solid
reactions  with fly ash or surface deposits in power plants. The unburned carbon in fly ash can
oxidize Hg or adsorb gas-phase Hg. Hg2+is believed to be more readily captured by
adsorption than Hg. Because of temperature considerations, the adsorption of Hg onto fly ash
in units equipped with CS-ESPs is believed to occur as the flue gas flows through the air
preheater and the ducting that leads to the ESP. Additional adsorption can also occur within
the ESP.

       Flue gas from the combustion of bituminous coal contains moderate to high levels of
Hg2+, primarily in the form of H
    The EPA ICR database provides a massive amount of information that can be mined for
additional information. However, its usefulness is limited by the uncertainty of some of the
measurements and by information that the data set does not contain.  Some of the uses and
limitations of the ICR data are summarized below. The data provide:

      Reasonable estimates of National and Regional emissions for Hgp, Hg2+, Hg, and
       HgT. They cannot be used to predict the total and speciated Hg emissions of
       individual plants.

                                       6-60

-------
       Information against which hypotheses and models of the speciation and capture of Hg
       in coal-fired boilers can be tested.  It cannot be used to identify or confirm specific
       mechanisms that control the speciation and capture of Hg.

       Information needed to guide the development of control technologies and identify
       effective strategies for the control of Hg emissions.
Cautions:
       Mercury speciation measurements made with the OH Method upstream of the PM
       control devices should be used with caution.  PM collected on the sampling train filter
       can result in physical and chemical transformations with the sampling train - with the
       result that OH Method speciation results do not accurately characterize the different
       forms of Hg in the flue gas where the samples were collected. The OH Method
       samples for HgT accurately reflect the concentration of HgT in the flue gas where the
       sample was collected. Also the samples collected at the inlet to air pollution control
       devices may not accurately represent the average Hg concentration because of flow
       stratifications near the sampling location.

       At low inlet and outlet concentrations, the precision of the OH Method can obscure
       real differences between these concentrations. When the capture across the control
       devices is being evaluated, the underlying imprecision of the measurements can show
       dramatic positive or negative reductions in emissions.

       It is believed that the positive variations in flue gas temperature can result in de-
       sorption of Hgp collected within PM control devices, resulting in flue gas
       concentrations of Hg that are higher at the outlet than at the inlet. Reentrainment of
       Hgp during rapping cycles of an ESP can also result in outlet concentrations that are
       higher than the inlet.

       There is uncertainty in the central values and  statistical characteristics of the OH
       measurements.  The samples represent a short snapshot in time, and the effects of
       long-term variations in coal properties and plant operating conditions are unknown.

       The adsorption of Hg onto fly ash is highly dependent on fly ash properties and
       particularly on the fly ash carbon content.  The lack of information on coal and fly
       ash properties limits the ability to evaluate the effects of LNBs on the capture of Hg.
                                         6-61

-------
6.10 REFERENCES

    1.  U.S. Environmental Protection Agency. Database of information collected in the
      Electric Utility Steam Generating Unit Mercury Emissions Information Collection
      Effort. OMB Control No. 2060-0396. Office of Air Quality Planning and Standards.
      Research Triangle Park, NC.  Available at:
      < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

    2.  Hargis, R., W. O'Dowd, and H. Pennline. "Sorbent Injection for Mercury Removal in
      a Pilot-Scale Coal Combustion Unit." Presented at the 93rd Annual Meeting of the
      Air & Waste Management Association, Salt Lake City, UT.  June 18-22, 2000.

    3.  Haythornthwaite, S., T. Hunt, M. Fox, J. Smith, G. Anderson, and C. Grover.
      "Investigation and Demonstration of Dry Carbon-Based Sorbent Injection for
      Mercury Control," Quarterly Report under DOE Contract No. DE-AC-22-
      96PC95256, December 1998.
                                       6-62

-------
                                      Chapter 7
                       Research and Development Status of
                Potential Retrofit Mercury Control Technologies
7.1 Introduction

       The Part III EPA ICR data show that ESP and FF control devices currently used to meet
PM emission standards do capture particle-bound mercury (Hgp) from coal-fired electric utility
boilers (see Chapter 6). The data also suggest that SDA and wet FGD scrubbers in place to meet
862 emission standards do capture oxidized mercury (Hg2+).  However, these data also show that
the air pollution control devices presently used at most electric utility power plants are not very
effective in capturing elemental mercury (Hg). Consequently, to achieve further reductions in
Hg emissions from existing coal-fired electric utility power plants, additional Hg reduction
strategies must be implemented.

       Potential Hg control strategies may be technology based (e.g., adding Hg emissions
control devices),  economics based (e.g., Hg emissions trading programs), or national energy
policy based (e.g., switching from coal to alternative energy sources for electrical power
production). This chapter discusses technology-based control strategies available for reducing
Hg emissions from existing coal-fired electric utility power plants (Section 7.2). Current
research and development is focused on retrofitting Hg control technologies to the coal-fired
electric utility power plant's existing air pollution control systems (Section 7.3). This retrofit
approach offers the potential for reduced costs of implementing Hg controls by enhancing the
capability of the air pollution control equipment already in place to capture more Hg.

       Building on the results of laboratory- and bench-scale research studies (discussed in
Chapter 5), additional studies have been, and currently are being, conducted using pilot-scale test
facilities to further investigate the more promising retrofit Hg control technologies (Section 7.4).
For the many existing coal-fired electric utility boilers that are equipped with only ESPs or FFs,
retrofit technologies under development are based on injecting sorbents into the flue gas
upstream of the control device (Section 7.5). Retrofit technologies to improve wet FGD scrubber
performance in capturing Hg are based on promoting oxidization of Hg to soluble species by the
addition of oxidizing agents or the installation of fixed oxidizing catalysts upstream of the
scrubber (Section 7.6).  The high levels of Hg control already achieved by the few existing
boilers using SDA for control of PM and SC>2 may be further enhanced by coinjection of a
second sorbent (Section 7.7). From a long-term perspective, the most cost-effective Hg controls
may be those implemented under a multipollutant emission control strategy. New
                                         7-1

-------
multipollutant control technologies, which potentially are effective in controlling Hg emissions,
are under development (Section 7.8).
7.2 Technology-based Mercury Control Strategies for Existing Coal-fired Electric Utility
     Boilers

7.2.1 Remove Mercury Prior to Burning by Coal Cleaning

       Reducing the amount of Hg in the coal burned in electric utility boilers would reduce the
level of Hg emissions from these boilers without the need for additional post-combustion Hg
controls.  Switching coal suppliers to obtain coals with lower Hg contents raises complex
economic and national energy policy issues that are beyond the scope of this report.

       Physical cleaning of coal (discussed in Chapter 2) has traditionally been used at coal
preparation plants to remove mineral matter (i.e., a source of coal combustion ash) and mineral-
bound sulfur (pyrite) from the mined coal. Mercury and other trace metals are also removed by
this cleaning depending on whether these metals are associated with the organic carbon structure
of coal or coal mineral inclusions. However, the existing commercially available coal-cleaning
methods remove only a portion of the Hg associated with the non-combustible mineral matter in
the coal and none of the Hg associated with the organic carbon structure of the coal.
Consequently, conventional physical coal cleaning can remove only a limited portion of the Hg
content of specific coals and may not be applicable to all coals.

       There is the potential for additional Hg reductions in the coal from several advanced
physical coal-cleaning processes using selective agglomeration or column froth flotation now
being developed. For example, Microcel is a type of column froth flotation available through
ICF Kaiser and Control International. The company is the exclusive licensee for use of the
technology for coal deposits east of the Mississippi River and has sold units for commercial
operation in Virginia, West Virginia, and Kentucky. Ken-Flote is another type of column
froth flotation cell coal-cleaning technology that is commercially available. Results of bench-
scale studies indicate that the combination of conventional with advanced coal-cleaning
techniques removes from 40 to 82 percent of the Hg contained in samples of raw coal.1'2

       Advanced coal-cleaning processes using naturally occurring microbes and mild chemical
treatments to  reduce the Hg content in coal have been investigated under DOE-funded bench-
scale studies.  The results of these tests indicate that these chemical and biological  coal-cleaning
processes have the potential for further reduction in the Hg content of coals.  However, DOE
viewed the processes as potentially high-cost control technologies, and DOE currently is not
funding development of these types of coal-cleaning technologies.3

       From  a near-term perspective, some reduction of the Hg content in certain coals burning
at existing coal-fired electric utility power plants can be achieved by physical coal-cleaning
processes. However, there are no easily identifiable coal deposits or coal types that will reliably
benefit from cleaning, with respect to reducing Hg content.  In addition, even with
                                         7-2

-------
implementation of widespread coal cleaning for Hg emissions control, significant quantities of
Hg will remain in the coal after cleaning; this will require that other control techniques be used
to achieve additional reductions in Hg emissions.

7.2.2 Retrofit Mercury Controls to Existing Air Pollution Control Systems

       In addition to reducing the amount of Hg in the coal before it is burned in a coal-fired
electric utility boiler, a second technology-based alternative is to remove more of the Hg in the
boiler flue gas before it is vented out the stack. One strategy is to retrofit or adapt control
technologies to the facility's existing air pollution control systems to increase the amount of Hg
captured by these systems rather than install new, separate Hg control  devices. The strategy
offers the potential advantage of reducing the costs of implementing the Hg controls by
enhancing the capability of the air pollution control equipment already in place to capture more
Hg.

       The existing air pollution controls used for a given coal-fired electric utility boiler
depends on site-specific factors including the properties of the coal burned, age and size of the
boilers, the geographic location of the  facility, individual state regulatory requirements, and
preferences of the facility owner or operator.  For approximately 70 percent of the existing coal-
fired electric utility boilers in the United States, the control device used is an ESP (see Table 3-6
in Chapter 3).  These power plants typically burn low-sulfur coals to control SC>2 emissions  and
use combustion modifications for NOx emissions control. Most boilers use a "cold-side" ESP
where the control device is installed downstream of the boiler air heater (discussed in Section
3.4.1). Some of the boilers use a "hot-side" ESP where the control device is installed upstream
of the boiler air heater. A small number of existing boilers (7 percent) that fire low-sulfur coal
use FFs instead of ESPs.  In general, FFs are being used at these coal-fired electric utility power
plants to obtain better  PM control or to solve ESP performance problems associated with high-
resistivity fly ash. A FF can be used only downstream of the boiler  air heater because of
temperature limitations of the fabric filter bags.

       Post-combustion 862 emissions controls are used at approximately 27 percent of existing
coal-fired electric utility boilers.  The SC>2 control technology most commonly used for these
boilers is a wet FGD scrubber. In all cases, a PM control device, usually an ESP, precedes a
scrubber. Wet FGD scrubbers remove gaseous 862 from flue gas by absorption. In absorption,
gaseous species are contacted with a liquid in which they are soluble.  For 862 absorption,
gaseous SC>2 is contacted with a caustic slurry, typically water and limestone or water and lime.
The newer semi-dry SC>2 scrubber technologies currently are used at small number of the existing
coal-fired utility boilers (about 5 percent). However, for retrofit Hg control, these semi-dry
scrubbers have the advantage of an existing sorbent delivery system coupled with, in most cases,
a downstream FF to collect the reacted sorbent already in place. A detailed discussion of
potential retrofit options and current research and development status is presented in following
sections.
                                         7-3

-------
7.2.3  Integrate Mercury Control Under a Multipollutant Control Strategy

       The most cost-effective, long-term Hg controls may be those implemented as part of a
multipollutant control strategy. Selection and deployment of new SC>2, NOx, and fine PM
controls, which also control or contribute to the control of gaseous Hg in coal combustion flue
gas, may reduce or eliminate the need for Hg-specific controls.  For example, installation of a
wet or semi-dry FGD unit should reduce oxidized Hg emissions by 90 to 95 percent over
previous levels; adding upstream NOx controls, which assist in oxidation of Hg, would further
reduce total Hg emissions.  Replacing or supplementing existing ESPs with FFs will likely
remove  additional Hg, especially for bituminous coal applications.

       The remaining majority, Sections 7.3 through 7.7,  discusses control technologies that
reduce Hg emissions by improving the performance of existing air pollution control devices to
capture the Hg in coal combustion flue gas. Section 7.8 discusses new multipollutant control
technologies (other than serial SOx-NOx-PM control devices), which are under development and
are potentially applicable to electric utility coal-fired electric utility power plants.

7.3 Post-combustion Mercury Control Retrofit

      Retrofits that reduce Hg emissions from existing electric utility coal-fired electric utility
power plants are implemented by modifying existing post-combustion emission control
techniques already in place. Potential retrofit options are identified and discussed below.
Options that are discussed may not be technically feasible or economically practical to install  and
operate  at all facilities.

7.3.1  Cold-side ESP Retrofit Options

      Add Flue Gas Cooling. Lowering the flue gas temperature entering the ESP assists
natural fly ash sorption of Hg as well as improves the performance of any sorbents injected
upstream for Hg control. However, the acid dew point temperature limits gas cooling when the
flue gas has significant HC1 or H2SO4 formation potential.

      Add Sorbent Injection. Gaseous Hg can be converted to Hgp by adsorption onto  solid
particles in flue gas. Injecting sorbents into the flue gas upstream of the ESP increases the
amount  of Hg captured in the form of Hgp.  This modification may require  adding ducting
between the sorbent injection point and the ESP, and adding a gas absorber/humidifier upstream
of the ESP. This approach also may be limited by the ability of the ESP  to collect high-
resistivity sorbents.  For coal-fired electric utility boilers with marginally performing ESPs that
have difficulty meeting opacity requirements and may not be candidates  for a sorbent injection
retrofit,  the following option may be preferred.

      Add Downstream FFwith Sorbent Injection. Adding a FF downstream of the existing
ESP, while a more expensive retrofit option, allows a significant portion of the fly ash to be
collected without reacted sorbent and enhances overall PM control efficiency where ESP
performance is marginal. Further, because the FF would have a much lower particulate  loading,
                                         7-4

-------
the collecting surface can be smaller (higher air-to-cloth ratio) or have longer cleaning cycles
(good for sorbent performance and bag life).

       ESP Modifications. Potential ESP modifications include converting the last field of the
ESP to a wet ESP or a very compact pulsejet FF. These conversions would likely be made
because of PM collection improvements needed, rather than Hg control considerations;
nonetheless, associated Hg control benefits would also be likely.

7.3.2  Hot-side ESP Retrofit Options

       Convert to Cold-side ESP with Sorbent Injection. Adding flue gas cooling is  not an
option for a hot-side ESP because of its location upstream of the air preheater. The only
potential retrofit option for Hg capture without adding a new downstream PM control device is
to convert the existing ESP from a hot-side configuration to a cold-side configuration.
Depending on the plant layout and ESP design, this may be possible by reconfiguring the ducting
and retuning the ESP to operate at the lower temperature.  Adding sorbent injection with the
modification would further improve Hg capture.  The lower flue gas temperature entering the
ESP enhances the adsorption of gaseous Hg onto fly ash or sorbent (if injected upstream) and
subsequent collection of the particulate Hg in the ESP.

       Add Downstream FFwith Sorbent Injection. The same as for a cold-side ESP, adding a
FF downstream of the existing ESP, while a more expensive retrofit option, allows a  significant
portion of the fly ash to be collected without reacted sorbent.

7.3.3  Fabric Filter Retrofit Options

       Add Flue Gas Cooling. As is the case for ESPs, lowering the flue gas temperatures
entering the FF  enhances the adsorption of gaseous Hg onto fly ash or sorbent (if injected
upstream). Again, the acid dew point temperature limits gas cooling when the flue gas has
significant HC1  or H2SO4 formation potential.

       Add Sorbent Injection.  Use of sorbent injection may require some internal FF
modifications to ensure good sorbent performance.  In general, existing FFs were not designed as
adsorbers, so some modifications may be in order to ensure that sorbent particles stay entrained
and become part of the filter cake. This may be accomplished by removing baffles, changing the
point of gas entry, increasing gas velocity, or using smaller sorbent particles.  Operating
requirements of the FF may require more frequent cleaning with the additional sorbent loading.

       FF Modifications. Potential FF retrofit options include replacing fabric bags  with
catalytic bags that oxidize Hg to Hg++ and Hgp or adding electrostatic augmentation  to increase
the bag cleaning cycle interval time and hence increase sorbent/gas contact time.  This last
improvement would be especially beneficial with higher-cost, high-capacity sorbents.
                                         7-5

-------
7.3.4 Spray Dryer Absorber Retrofit Options

       Use Oxidation Additives. Existing SDA systems already achieve very high Hg removal
on certain coals but show poor performance on other coals. Possible causes are low oxidation
potential resulting from high alkaline fly ash content as well as low effective carbon content in
fly ash.  Therefore possible performance improvements include producing a higher carbon
content fly ash by NOx combustion control modifications, direct addition of activated carbon to
the absorber with lime, and addition of oxidants to the absorber.

      Replace Existing ESP with FF Control Device.  Where the PM control device used for
the absorber is an ESP, replacement of the unit with a FF would likely improve Hg removal as a
result of enhanced PM control as well as greater conversion of Hg2+ to Hgp.

7.3.5 Wet FGD Scrubber Retrofit Options

       Use Oxidation Additives. Oxidation of the gaseous Hg to gaseous Hg2+ can potentially
increase the total Hg removed by wet scrubbing since gaseous Hg + is more readily captured by
these systems than gaseous Hg. Several flue gas additives and scrubbing additives are being
developed to increase the conversion of Hg to Hg++ prior to the scrubber inlet. Flue gas and
scrubber additives are also being developed for use in preventing the conversion of absorbed
Hg2+ to gaseous Hg in wet FGD systems.  The one caution is that increasing oxidants upstream
or within the scrubber may also oxidize other species such as  SO2 and NO/NO2 to sulfuric and
nitric acid aerosols.

      Add Fixed Oxidizing Catalysts Upstream of Scrubber. Improvements in wet scrubber
performance in capturing Hg may be accomplished by installation of fixed oxidizing catalysts
upstream of the scrubber to promote oxidization of Hg to soluble species.  Potential catalysts
currently are being tested.

       Wet FGD Scrubber Modifications.  Several studies of pilot-scale wet FGD systems
suggest that modifying the scrubber operation and design (as well as the control and design of
upstream ESPs) may improve the capture of gaseous Hg2+ and reduce the conversion of absorbed
Hg2+ to Hg.  Specifically, these studies have found that the liquid-to-gas ratio and tower design
of a wet FGD unit affect the absorption of gaseous Hg2+, while the oxidation air influences the
conversion of absorbed Hg2+ back to Hg which is then emitted to the atmosphere in the scrubber
exhaust gas.

7.3.6 Particle Scrubber Retrofit Options

      A few existing power plants use wet scrubbers exclusively for control of PM emissions.
Knowledge gained in the enhancing control of Hg emissions from wet FGD scrubbers by
operating modifications also may be useful in improving the Hg removal performance of these
particle scrubbers.
                                         7-6

-------
7.4 Retrofit Control Technology Research and Development Programs

       None of the retrofit options discussed in Section 7.3 are routinely being used by the
electric utility industry at this time. In addition, the Hg emissions control technologies that are
successfully used for municipal waste combustors (MWCs) in the United States and Europe
cannot be directly retrofitted to existing coal-fired electric utility boilers.  Differences in flue gas
properties, combustion unit design, and other factors (discussed in Section 7.4.1) prevent the Hg
control devices now used for MWCs to be directly installed at coal-fired electric utility power
plants. Consequently, development of effective retrofit control technologies for coal-fired
electric utility boilers is the subject of bench-scale, pilot-scale, and full-scale test programs.
Chapter 5 discusses laboratory studies investigating potential Hg control techniques for coal-
fired electric utility boilers. To further develop the most promising of these control techniques
for full-scale application to coal-fired electric utility boilers, pilot-scale and full-scale research
studies are being funded by the EPA, DOE, EPRI, state agencies, and private companies.
Section 7.4.2 describes several pilot-scale test units that are being used for research and
development programs. Building upon the results obtain using these test facilities, a number of
full-scale test programs currently are being conducted to provide a more thorough
characterization of the performance and potential for widespread commercial application of
specific retrofit Hg control technologies.

7.4.1  MWC Mercury  Control Technology

       Injection of activated carbon into the flue gas from a MWC and collecting the reacted
sorbent in a downstream FF is one Hg control method widely used for MWCs.4'5 Mercury
removal levels in excess of 90 percent are achieved.  However, the level of Hg control achieved
by adding sorbents into the flue gas from a particular combustion unit is influenced by the
particular characteristics of the flue gas from that unit including flue gas temperature, flow rate,
Hg content, and chloride Hg content.  Table 7-1 compares selected properties of the flue gas
from a coal-fired utility boiler with those for a MWC flue gas.  As shown in this table, Hg
concentrations in MWC flue gas streams may be up to several orders of magnitude greater than
those seen in utility flue gas streams.  In addition, MWC flue gas contains mostly Hg2+, while
flue gas from coal-fired electric utility boilers can  have substantial amounts of Hg, which
generally is less likely to be adsorbed. Additionally the flue gas ductwork for a coal-fired utility
boiler is substantially larger and more complex (multiple passes) than for a MWC, therefore duct
injection of a sorbent is more complicated and its performance more difficult to predict for a
coal-fired utility boiler due to variations in temperatures, residence time, and other factors.

       Similarly, the wet scrubber technology used by European MWCs is not directly
applicable to controlling emissions from coal combustion. European MWCs typically have two-
stage scrubbers consisting of a low-pH water scrubber to control hydrochloric acid (HC1)
emissions, produced as a result of the large quantities of plastics in the garbage burned, followed
by an alkaline scrubber to control 862 emissions.  In contrast, wet scrubbing systems typically
used by the electric utility industry in the United States to control 862 emissions resulting from
burning high sulfur coal consist of a single-stage wet scrubber using a limestone or lime
scrubbing agent.  As a consequence, there are significant differences in the underlying chemistry
                                          7-7

-------
Table 7-1.  Comparisons of typical uncontrolled flue gas parameters for coal-fired
utility boiler versus municipal waste combustor (MWC).
              Flue Gas
              Parameter
Coal-fired Electric Utility Boiler
Municipal Waste Combustor
            Temperature
                                         121 to 177
                                    177 to 299"
             Hg Content
             (|ig/dscm)
          1 to 25
      400 to 1,400'
           Chloride Content
              (|j,g/dscm)
      1,000 to 140,000
    200,000 to 400,00"
              Flow Rate
             (dscm/min)
     11,000 to 4,000,000
    80,000 to 200,000"
      (a) Temperature, chloride content, and flow rate data taken or determined from Reference 6
      (b) Mercury content data taken from Reference 4.
                                          7-8

-------
of the scrubbing systems used for MWCs compared to those currently in use at coal-fired electric
utility power plants.

7.4.2 Pilot-scale Coal-fired Test Facilities

       To date, most of the retrofit control technology development has been conducted using
pilot-scale test units that simulate full-scale coal-fired electric utility boiler combustion
conditions. The DOE Federal Energy Technology Center, the Ohio Coal Development Office
(OCDO), and McDermott Technology, Inc., jointly funded one program titled the Advanced
Emissions Control Development Program (AECDP).  This test program was conducted in three
phases using a 10 MW coal-fired test faculty.7'8'9 The test facility is capable of testing a full-flow
ESP, a partial-flow pulsejet FF, and a wet FGD scrubber.  All testing under the AECDP was
performed firing Ohio bituminous coals. Figure 7-1 shows a schematic of the test facility.
Specific AECDP test results related to specific retrofit options are discussed later in this chapter
under the relevant topic headings.

       For a DOE cooperative agreement test program, the project team of Public Service
Company of Colorado (PSCO), ADA Technologies, and EPRI fabricated a pilot-scale
particulate control module (PCM) to investigate Hg control in actual coal combustion flue gas by
different dry sorbents.10  Figure 7-2 shows  a schematic of the PCM. The PCM draws a
slipstream of flue gas (600 actual cubic feet per minute) from the 350-MWe coal-fired electric
utility boiler (Unit 2) at PSCO's Comanche Station power plant.  This boiler is an opposed-fired,
pulverized-coal boiler firing Powder River Basin (PRB) subbituminous coal.  Flue gas can be
drawn either from the inlet (high particulate loading) or the outlet (essentially particle free) of the
full-size Unit 2 reverse-gas FF.  In addition,  the PCM can be  configured as an ESP, a reverse-gas
or pulse-jet FF, and as EPRI's TOXICON pulse-jet FF. Gaseous Hg is injected into the flue gas
to the PCM along with recycled fly ash and/or sorbent; the solids can be injected at various
locations upstream  of the PCM  to investigate the effects of Hg adsorption at different in-flight
residence times (0.5 to 3 seconds). The PCM is  also equipped with in-duct heating and water
spraying to investigate the effects of Hg adsorption at different temperatures.  Specific results
from testing using the PCM are discussed later in this chapter under the relevant topic headings.

       The DOE National Energy Technology Laboratory (NETL) is conducting in-house
research studies using a 500-lb/hr coal combustion unit to simulate a pulverized-coal-fired
electric utility boiler.  U2  Figure 7-3 shows a schematic of the DOE/NETL coal combustion test
facility. The system consists of a wall-fired, pulverized-coal furnace equipped with a water-
cooled convection system, a recuperative air heater, spray dryer, sorbent injection duct (SID) test
section, and  FF. Sorbent can be injected at numerous locations along the SID test section; this
allows for a wide range of sorbent in-duct residence times relative to the FF and to the SID flue-
gas sampling locations.
                                         7-9

-------
                                                               r	i
                                                        jtiwtft*
                      C^SPBlV

                      ~"-;aoi  j	"~>

                         i  bin       i
          BfSfl
                                   .,--'  C*R
     i        F.u, >



     4	,____.___J?-;-^*-; Si__Jv?
     I        *v!h-".wi; r*-ap-
     t        e'.'JMSfi' i *S;
             v-^>


 Nuqil     KP
	  ftt^*3^|        '




          "*   j A
                                                   Ofy !%v.
                                                                    Wei
                                                                       Slack
                                 U.I' I'ii


                                iJIft|lMll
                                                            A*hto
          Q



          B A PC OP TwSI FiJ iiipiMiM
  . Slip
Figure 7-1.  Schematic of 10-MWe coal-fired Babcock & Wilcox (B&W) Clean

Environment Development Facility (CEDF) as used for Advanced Emissions

Control Development Program (AECDP) (source: Reference 9).
                                     7-10

-------
                      Ash

rv
.gfl

t
y
-0 
inlet
Sample^^Duct Heater
\ *
Hg
Doping
i
., Carbon
*r injection

OD
X"
Particulate Control Modu!
x^

1
Outlet
Sample
                                            V
Figure 7-2. Schematic of Participate Control Module (PCM) at Public Service
Company of Colorado (PSCO) Comanche Station (source: Reference 10).
                                7-11

-------
      nil
            *.!
                            V
                             .pt
                                            -
                                                      e| - 
                                                          , .
Figure 7-3. Schematic of DOE/NETL in-house 500-1 b/hr coal combustion test
facility (source: Reference 12).
                                 7-12

-------
7.5 Mercury Control Retrofits for Existing Coal-fired Electric Utility Boilers Using ESP or
    FF Only

       The focus of research and development for existing coal-fired electric utility boilers
equipped only with an ESP or FF has been the use of dry sorbent injection.  As discussed in
Chapter 5, gaseous Hg can be adsorbed onto solid particles in the flue gas. A solid particle that
absorbs gaseous species is called a "sorbent."  The flue gas from every electric utility boiler that
directly burns coal (i.e.,  all boilers except for IGCC units) contains fly ash particles that adsorb
gaseous Hg in the flue gas to various degrees.  Other types of solid particles can be injected into
the flue gas for the purpose of adsorbing gaseous Hg. Materials being investigated as possible
sorbents for Hg control include activated carbon, calcium-based and sodium-based (trona)
sorbents, various clays and zeolites, alkaline-earth sulfides, and lime and lime-silica
multipollutant sorbents.  An alternative sorbent-based Hg control approach that has been
investigated is passing the flue gas through a fixed bed of a noble-metal-based sorbent.

7.5.1  Sorbent Injection Configurations

       In general, four retrofit configurations are possible for injecting dry sorbent particles into
the flue gas from a coal-fired utility boiler.  It may not be technically feasible to implement one
or more of these configurations at a given existing coal-fired power plant because of site-specific
factors such as the existing flue gas duct configuration, availability of space to add additional
ducting or new control device, use of a wet FGD scrubber, or other plant layout and operation
considerations.

       Configuration A  - Sorbent injection into the flue gas duct upstream of existing ESP or FF.
       Cooling of the flue gas upstream of the sorbent injection point or modifications to the
       ducting may be needed.

       Configuration B  - Sorbent injection into the flue gas duct downstream of the existing PM
       control device followed by a new FF (to collect the reacted sorbent), with or without flue
       gas cooling upstream of the injection point.  This configuration requires higher capital
       costs but reduces sorbent costs compared to Configuration A. The configuration also
       allows the fly ash collected by the upstream PM control device to  be sold without being
       mixed with the injected sorbent.

       Configuration C - Sorbent injection into a circulating fluidized-bed absorber (CFA)
       upstream of the existing ESP or FF, with or without flue gas cooling upstream of the
       CFA.  The advantage to using a CFA is that it recirculates reacted materials with fresh
       sorbent to create an entrained bed with a high number of reaction sites resulting in higher
       sorbent utilization and enhanced capture of Hg and other pollutants.

       Configuration D - Sorbent injection into a CFA downstream  of the existing PM control
       device and followed by a new FF (to collect the reacted sorbent).  Like Configuration B,
       this configuration allows the fly ash collected by the upstream PM control device to be
       sold without being mixed with the injected sorbent.
                                         7-13

-------
       The level of Hg capture using sorbent injection with a downstream ESP depends on in-
flight adsorption of Hg by entrained sorbent particles. Mercury capture in a downstream FF
occurs by this same in-flight adsorption process as well as a second mechanism when flue gas
must pass through the filter cake collected on the FF bags.  This filter cake contains a mixture of
previously captured fly ash  and sorbent particles, and provides good contact between gaseous Hg
and captured particles. Filter cake retention times between bag cleaning cycles may be as long
as 60  minutes, greatly increasing the adsorption of Hg on the sorbent particles. This compares
with the relatively short time that in-flight adsorption occurs upstream of the control device
(nominal times for in-flight adsorption are 0.5 to 1.5 seconds). In addition, FFs generally are
more  efficient than ESPs in collecting fine particles and any associated Hgp (see Table 3-3).  The
extra  contact time and higher collection efficiency provided by a FF reduces the amount of
sorbent needed for adsorption compared to what is needed for an ESP to achieve a given level of
control.

       Cooling the flue gas before the sorbent injection point can improve Hg adsorption by the
sorbent, which in turn may reduce the amount of sorbent needed for a given level of control.
However, the temperature to which the flue gas may be cooled is limited because sulfuric acid
(and perhaps hydrochloric acid) mists may be formed if the flue gas temperature drops below the
acid dew point(s) of the flue gas. For all four configurations, sorbent capacity may be
maximized by recycling and reinjecting sorbent and fly ash collected in the PM control device(s)
located downstream of the injection point.

7.5.2  Sorbent Adsorption Theory

       Gas-phase adsorption occurs when a gaseous specie contacts the surface of a solid and is
held there by attractive forces between the gaseous specie and the solid.  In adsorption
terminology, the gaseous specie being adsorbed is called the "adsorbate," and the solid is called
the "adsorbent" or "sorbent." While all solids have the potential to adsorb gaseous species,
adsorption is not very pronounced unless a solid has a large surface area. As a result, most solids
for gas-phase adsorption are highly porous and in the form of particles or granules. The porosity
of the solids provides large  amounts of internal surface area where most adsorption takes place.
When a gaseous specie is adsorbed onto the surface of a solid particle, the gaseous specie
becomes a particle-bound specie.

       Gas-phase adsorption may be classified as chemisorption or physical adsorption
depending on the nature of the attractive force between the adsorbate and sorbent.  In
chemisorption, the adsorbate reacts with the surface of the sorbent, thus, the attractive force
between the adsorbate and sorbent is similar to a chemical bond. Chemisorption often involves
the use of sorbents impregnated with compounds that are reactive with the adsorbate. In physical
adsorption, the attractive force between an adsorbate and sorbent is electrostatic in nature
(similar to the attraction between metal filings and a magnet, where the metal filings are
analogous to the adsorbate and the magnet is analogous to the sorbent).  Different adsorbates
have different attractive forces for a given sorbent due to differences in molecular weight,
normal boiling point (or vapor pressure), degree of unsaturation, polarity, and structural
configuration. When a sorbent is exposed to more than one adsorbate, preferential adsorption
                                         7-14

-------
tends to take place due to differences in the attractive forces between the different adsorbates and
the sorbent particles.

       Equilibrium adsorption capacity is the maximum amount of adsorbate a given mass of
sorbent can hold at a given temperature and adsorbate gas concentration. Generally, the
adsorption capacity of a sorbent for a given adsorbate increases with increased adsorbate
concentration and decreases with increases adsorption temperature.

       In a dynamic adsorption system (i.e., an adsorption system involving a moving gas
stream), a gas stream containing one or more adsorbates is passed through a fixed or fluidized
bed of sorbent particles or the sorbent particles are injected directly into the gas stream.  In
dynamic adsorption systems, the contact time between the sorbent particles and the adsorbate in
the gas stream is critical.  While contact time does not affect the equilibrium adsorption capacity
of the sorbent, it directly affects the sorbent's ability to capture the adsorbate from the gas
stream. Maximum capture of adsorbate from the gas stream will not take place unless the
adsorbate has sufficient time to contact the sorbent and diffuse into its pores. Thus, increasing
the contact time increases Hg capture by the sorbent.

7.5.3 Pilot-scale and Full-scale Research and Development Status

       The laboratory studies of using dry sorbents for Hg control based on bench-scale reactor
testing are discussed in Section 5.4.  This section discusses the results from field studies testing
different sorbents in pilot-scale or full-scale systems.

7.5.3.1 Coal Fly Ash Reinjection

       As discussed in Chapter 5, fly ash generated naturally when burning certain coals in a
utility boiler adsorbs some of the gaseous Hg in the flue gas.  The adsorption of gaseous Hg by
the fly  ash vented in the flue gas from the boiler, referred to by some researchers as "native fly
ash," is believed to occur at active sites on the ash surface similar to those on sorbent (e.g., fly
ash carbon analogous  to activated carbon or fly ash alkaline species akin to injected lime). As
part of the DOE cooperative agreement test program to investigate dry sorbents, the project team
of PSCO, ADA Technologies, and EPRI evaluated Hg removal rates by the fly ash in the flue
gas from burning two  types of Western coals and the potential for Hg removal by reinjection of
low levels of collected fly ash back into the flue gas upstream of the particulate control device.10
The use of reinjected fly ash for Hg control avoids the potentially adverse impact on the
commercial viability of selling the fly ash collected in the downstream particulate control
devices.  The use of activated carbon as a Hg sorbent may increase the level of carbon in the
collected fly ash/activated carbon mixture above allowable maximum levels for some
commercial fly  ash applications (e.g., sale of fly ash for use as a concrete additive).

       Full-scale testing was conducted at three PSCO coal-fired electric utility power plants to
characterize gaseous Hg removal by the native fly ash in flue  gas at each facility; a boiler using a
FF for  PM control was tested.  At one facility, a second boiler using an ESP was also tested.
Two of the three power plants burned subbituminous coal from the Powder River Basin (PRB),
                                         7-15

-------
and the other burned a Colorado-mined bituminous coal. Flue gas measurements were taken
concurrently at the inlet and outlet of each particulate control device. At two of the power
plants, testing was conducted in both the summer and winter in order to investigate the effect of
ambient temperature on the adsorption of Hg on the fly ash.

      Results of the full-scale tests are summarized in Table 7-2. Mercury removal measured
across the three FFs ranged from 61 to 99 percent.  Mercury removal across the ESP was
significantly lower at 28 percent. The two boilers units demonstrating Hg removals above
80 percent (Arapahoe 4 and Cherokee 3) were equipped with low-NOx burner retrofits.  The use
of these burners often causes elevated levels of unburned carbon in the fly ash. Measuring
unburned carbon by the "loss-on-ignition" (LOT) test, the fly ashes from Arapahoe 4 and
Cherokee 3 had LOT contents approximately 7 to 14 times higher than the fly ashes from the
other two boilers. The Hg levels measured for the Cherokee 3 unit was essentially the same in
both summer and winter, indicating no adverse temperature effects on adsorption.  In contrast,
the Arapahoe 4 tests showed better adsorption at cooler test conditions (i.e., winter versus
summer).

      To examine the use of fly ash reinjection for Hg emissions controls, a series of pilot-scale
tests were conducted by collecting the fly ash samples from the three power plants and injecting
the collected fly ash into the PCM located at the Comanche Station (discussed in Section 7.4.2).
For the recycled fly ash tests, the PCM was configured as a reverse-gas FF and drew fly-ash-free
flue gas from the outlet  side of the FF serving the coal-fired boiler. The flue gas was spiked with
gaseous Hg to produce a Hg concentration of approximately 10 |ig/Nm3. The gaseous Hg
concentration was sampled at the inlet and outlet of the PCM using a Hg continuous emissions
monitor (Perkin Elmer MERCEM). Recycled fly ash was injected into the flue gas just
downstream  of the inlet sampling port. Except during one test, the injected fly ash samples were
not treated in any way to enhance their Hg-adsorbing properties.  For one test, a sample of fly
ash from the Comanche 2 unit  was treated with a hot nitrogen purge in an attempt to desorb any
Hg on the ash particles.

      Table 7-3 summarizes Hg removal data for the fly ashes tested. Reinjected
subbituminous coal fly ash removed 84 to 86 percent of the gaseous Hg across the PCM. In
contrast, reinjecting fly ash from the boiler burning bituminous coal showed only a 10 percent
removal of gaseous Hg. The removal efficiency for bituminous coal fly ash was increased to 31
percent when this ash was thermally pretreated to desorb Hg before injection into the PCM. The
results in Table 7-3 show that the recycled fly ashes from the Cherokee and Arapahoe boilers had
additional capacity to adsorb gaseous Hg (beyond what they had adsorbed from their source flue
gas), while the untreated recycled fly ash from the Comanche 2 boiler appeared to be saturated or
no longer reactive. The LOI contents of the Cherokee 3 and Arapahoe 4 fly ash samples were 8
and 14 percent, respectively. The LOI contents of the Comanche 2 fly ash samples were 0.3 to
0.4 percent.  As was observed during the full-scale testing,  fly ashes with the highest LOI
contents (those from the Arapahoe  4 and Cherokee 3 boilers) adsorbed more Hg than fly ashes
with lower LOI contents (those from the Comanche 2 boiler).
                                        7-16

-------
Table 7-2. Hg removal by native fly ashes measured across PM control devices at
PSCO power plants burning selected western coals (source: Reference 10).
Power Plant
PSCO"
Cherokee
PSCO
Arapahoe
PSCO
Comanche
Type of Coal
Burned
Bituminous
(Colorado)
Subbituminous
(Powder River
Basin)
Subbituminous
(Powder River
Basin)
PM Control
Device
Reverse-gas FF
(Boiler Unit #3)
ESP
(Boiler Unit #1)
Reverse-gas FF
(Boiler Unit #4)
Reverse-gas FF
(Boiler Unit #2)
Ash Carbon
Content
(% L0lb)
7.6
<1
14.4
0.4
Gaseous Hg
Removal
(%)
98 (summer)
99 (winter)
28
62 (summer)
82 (winter)
61
       (a) PSCO = Public Service Company of Colorado
       (b) LOI = Loss on ignition
                                   7-17

-------
Table 7-3. Hg removals by fly ash reinjection measured across PCM at PSCO
Comanche power plant for selected western coals (source: Reference 10).
Reinjected Fly Ash
Coal Source
(PSCO power plant)
PRB Subbituminous coal
(Arapahoe 4)
PRB Subbituminous coal
(Cherokee 3)

Colorado
Bituminous coal
(Comanche 2)

Flue Gas
Temperature
(F)
320
320

280




280
Ash
Reinjection
Rate
(grains/acf)
0.13
033

1.13




1.21
Ash
Carbon
Content
(% L0la)
14.4
76

0.42




0.26
Gaseous Hg
Removal
(%)
84
86

10




31 b
      (a) LOI = Loss on ignition
      (b) Deadsorbed ash.
                                  7-18

-------
       In addition to evaluating the adsorption capacity of recycled fly ashes, several tests were
made using the PCM to evaluate the effects of temperature on fly ash adsorption. For the
temperature tests, fly-ash-laden flue gas was extracted from the inlet of the FF serving the
Comanche 2 boiler and passed through the PCM; gaseous Hg was injected upstream of the PCM.
Hg adsorption across the PCM was monitored as the temperature of the flue gas through the
PCM was varied. Table 7-4 summarizes the results of the temperature tests.  For the baseline
tests (no heating or cooling), the temperature of the flue gas through the PCM was in the range of
135 C (275 F); at this temperature, the Comanche 2 fly ash removed 20 to 40 percent of the
gaseous Hg present in the flue gas. When the flue gas was heated to around 152 C (305 F), the
fly-ash Hg removal dropped to zero, while spray cooling to reduce the flue gas temperature to
about 110 C (230 F) increased the Hg removal to  around 60 percent. As expected, the data from
these tests show that adsorption is greatly affected by temperature, with adsorption increasing
with decreasing flue gas temperature.

7.5.3.2 Activated Carbon Sorbent Injection

       The most frequently tested activated carbon for Hg removal from coal combustion gases
has been a commercially available carbon manufactured by Norit Americas, Inc. (trade name
Darco FGD). The Darco FGD carbon is produced from lignite specifically for the removal
of heavy metals and other contaminants from MWC flue gas streams.  Other commercially
available activated carbons and experimental carbons also have been tested.

       A full-scale test program jointly funded by EPRI and Public Service Electric and Gas
(PSE&G) evaluated the potential of activated carbon injection for Hg control.13  The tests were
performed at the PSE&G Hudson Generating Station, which fires low-sulfur bituminous coal and
uses an ESP for PM control. Two types of activated carbon were tested, the Darco FGD
carbon and  an experimental carbon identified as AC-1.  Results from these tests are shown in
Table 7-5. The data indicate a distinct reduction in total Hg removal efficiency with increased
temperature.  The maximum Hg removal measured was 83 percent using the  Darco FGD
carbon at a  C:Hg ratio  of 45,000:1 and an ESP operating temperature of 221 F.  Full-scale ESP
operation at this low temperature is not practical, however, due to potential problems with acid
condensation.

       Sorbent injection using Darco FGD carbon and an ESP was also tested as part of the
AECDP Phase III studies.9  For this test, the coal burned was an Ohio bituminous coal. The
carbon was injected upstream of the ESP, with an approximate in-flight particle residence time
of 1 second. The injection temperature was approximately 204 C (400 F) and the ESP inlet
temperature was about 174  C (345 F). The carbon flow rate was approximately 14 Ib/hr, which
is equivalent to a C:Hg mass ratio of 9,000:1. Both paniculate and gaseous Hg species were
measured at the inlet and outlet of the ESP during the carbon injection test. The test results are
presented in Figure 7-4. Also shown in this figure  are baseline Hg concentrations measured
before any injection tests.  Compared to the baseline condition,  injection of the activated carbon
resulted in a total Hg removal of 53 percent.  Carbon injection at the test conditions had no effect
on the removal of gaseous Hg, suggesting that Hg removal appears to be a result of the capture
                                        7-19

-------
Table 7-4. Effect of flue gas temperature on fly ash Hg adsorption measured
across PCM at PSCO Comanche power plant burning PRB subbituminous coal
(source: Reference 10).
Test Condition
Baseline
Heated flue gas
Cooled flue gas
Flue Gas
Temperature
(C)
135
152
110
Gaseous
Hg Removal
(%)
20 to 40
0
60
                                7-20

-------
Table 7-5.  Hg removal by activated carbon injection measured at PSE&G Hudson
Station burning low-sulfur bituminous coal and using ESP (source: Reference
13).
Sorbent
Tested
Baseline
(no sorbent injection)
Darco FGD
Activated Carbon
Experimental
Activated Carbon
AC-1
ESP Operating
Temperature
(F)
255
268 -278
240-255
240-255
220 -235
275 -280
270 -275
240 -250
240-250
280
Sorbent Injection
Ratio
(C:Hg)
0
0
11,500:1
20,000:1
45,000:1
27,000:1
45,000:1
18,000:1
45,000:1
29,000:1
Total Hg
Removal Range
(%)
3
0
13to17
41 to 42
76 to 83
14 to 38
28 to 45
33 to 45
56 to 58
28
                                 7-21

-------
  ~  25."
  E
                                                     Parikiikite


                                                        "'J	*
  =
  i  i IN,
  g   *

      0.0
           FurmicL E.vlt Before       I unuicc
             Sorbent Injection    After Sortwnf li^Jcclkm

Figure 7-4. Hg removal by activated carbon injection measured at AECDP test
facility burning Ohio bituminous coal and using ESP (source: Reference 9).
                                   7-22

-------
             9-1-
of gaseous Hg  (onto or into the particulate phase) and then the subsequent removal of the
particulate in the ESP.

       The DOE/NETL also tested injecting Darco FGD carbon for Hg control using the
DOE/NETL in-house coal combustion test facility.11  For these tests, low-sulfur bituminous coal
was burned based on the rationale that this is a coal-type likely be burned in utility power plants
that do not have flue gas desulfurization systems. Throughout testing, the furnace was operated
to achieve high combustion efficiency with low levels of unburned carbon in the fly ash.
Unburned carbon levels in the fly ash under baseline conditions were generally less than two
percent. Flue gas measurements of Hg were conducted at the FF inlet using the OH Method, and
a Modified  Ontario-Hydro Method (MOH Method). The modified method samples the flue gas
non-isokinetically whereas the former samples the flue gas isokinetically.  Stack measurements
downstream of the FF were made for speciated Hg using the OH Method and total Hg using EPA
Method 101 A. Analysis of coal and ash deposits was made using ASTM D3684. The MOH
Method was used at the inlet to minimize PM collection during sampling.  Eliminating entrained
PM in the sample flue gas allowed researchers to determine in-duct Hg removals. In addition,
the effect of filtered solids on Hg speciation was deduced by comparison with the Hg speciation
measured with the OH Method.

       Test results measured using the DOE/NETL test facility for sorbent injection upstream of
a FF using the Darco FGD carbon are presented in Table 7-6. Total Hg removals measured
ranged from 39 to 86 percent at injection C:Hg ratios of 2,  600:1 to 10, 300:1. The test results
show a general trend where the total Hg removal increased with increasing C:Hg ratios. A
second commercially available activated carbon has also been tested for possible Hg control
using the NETL test facility.12  Mercury removals of 30 to  40 percent were measured injecting
Calgon FluePac  activated carbon at C:Hg injection ratios of 2,500:1 to 5,100:1. The
DOE/NETL in-house research also shows no significant in-duct removals of Hg under the test
conditions,  and Hg appears to be oxidized by the filter cake. On-going research on activated
carbon injection using the DOE/NETL test facility includes tests to quantify the effects of
humidification and FF pressure drop on Hg removal, evaluating novel sorbents, determining
sorbent effectiveness downstream of a FF with and without recycle, and comparing Hg removals
using sorbent injection with ESP versus FF.12

       A multiple-site, full-scale field test program is currently being conducted under a
DOE/NETL cooperative agreement to obtain performance  and cost data for using activated
carbon injection to reduce Hg emissions from existing coal-fired electric utility power plants
equipped only with an ESP or FF for post-combustion air pollution controls.14 The DOE/NETL
is working in partnership with ADA-ES, PG&E National Energy Group (NEG), Wisconsin
Electric, a subsidiary of Wisconsin Energy Corp., Alabama Power Company, a subsidiary of
Southern Company, EPRI, and Ontario Power Generation on a field evaluation program at four
power plant facilities. Other organizations participating in this test program as team members
include EPRI, Apogee  Scientific, URS Radian, Energy & Environmental Strategies, Physical
Sciences, Inc., Southern Research Institute, Ham on Research-Cottrell, Environmental Elements
Corporation, Norit Americas, and EnviroCare International. The first test site is a boiler unit at
the Alabama Power Gaston facility that burns various low-sulfur bituminous coals and is
                                        7-23

-------
Table 7-6.  Hg removal by activated carbon injection measured at DOE/NETL in-
house test facility burning low-sulfur bituminous coal and using FF (Source:
Reference 11).
Test
Run ID
9907-1
(baseline)
9907-2
9907-3
9907-4
9908-1
(baseline)
9908-2
9908-3
9908-4
Fabric Filter
Temperature
(F)
294
294
265
268
296
296
296
270
Sorbent
Injection
Ratio
(C:Hg)
0
9,500:1
10,300:1
6,200:1
0
2,600:1
5,400:1
2,900:1
Total
Hg
Removal
(%)
2.7
86.0
82.3
75.1
35.0
38.8
64.0
54.2
Mass Balance (%)
Fabric Filter
103.2
77.4
130.1
80.0
84.4
100.6
94.7
103.2
Overall
79.4
78.6
76.7
98.1
67.1
90.8
89.1
86.8
                                 7-24

-------
equipped with a hot-side ESP followed by a COHPAC FF. Testing at this site was conducted in
the spring of 2001.15 The next test site being tested is a boiler unit at the Wisconsin Electric
Pleasant Prairie facility that burns PRB subbituminous coal and uses a cold-side ESP for PM
control. The other two sites are scheduled to be tested in 2002, and are the PG&E NEG Salem
Harbor and Brayton Point facilities that burn low-sulfur bituminous coals and are equipped with
cold-side ESPs.

7.5.3.3 Calcium-based Sorbent Injection

       An alternative to using activated carbon is to use a calcium-based sorbent. Laboratory
studies conducted by the EPA and Acurex Environmental Corporation (funded by the State of
Illinois, ICCI) indicated that the injection of calcium-based sorbents into flue gas could result in
significant removal of Hg (discussed in Section 5.3).  Other benefits associated with the use of
limestone injection for Hg control include an incremental amount of SC>2 removal and a high
probability  for 863 removal.  Flue gas Hg removal using furnace limestone injection was
evaluated as part of a study conducted by McDermott Technology, Inc. titled Combustion 2000
Project/Low Emission Boiler System Program.16  In this study, limestone was injected into the
upper furnace firing Ohio bituminous coal at a temperature of about 1,204 C (2,200 F). The
Ca:S ratio was set at 1.40 mol/mol.  An 80 percent efficient cyclone was then used to collect the
fly ash and  calcined lime.  At this location the flue gas temperature was approximately 163 C
(325 F).  The Hg concentration in the flue gas was measured downstream of the cyclone using
the OH Method.  The measured Hg concentrations for the baseline (no limestone injection) and
the six limestone injection tests are shown in Figure 7-5. The  data show that the Hg
concentration in the flue gas was significant lower when limestone was injected compared to the
baseline.  The overall average Hg reduction for the six limestone injection runs was 82 percent.
The researchers note that using more efficient ESP or FF PM control devices with collection
efficiencies of greater than 99 percent in place of a cyclone (see  Table 3-3) is expected to
provide an additional increase in Hg removal.

       Based on the test results from the EPA/Acurex ICCI studies and the Combustion 2000
Project/Low Emission Boiler System Program, McDermott Technology, Inc. conducted
additional limestone injection tests during Phase  III of the AECDP.9 The same limestone
previously tested in the Combustion 2000 program was used for the Phase III tests.  Two
limestone flow rates were tested. The flow rates chosen for the limestone injection tests were
200 Ib/hr (Ca:S = 0.35 mol/mol) and 25 Ib/hr (Ca:S = 0.04 mol/mol).  An injection temperature
target of 1,149 C to 1260 C  (2,100 F to 2,300 F) was chosen  as the optimum range to
calcine the limestone (CaCOs) into lime (CaO).  It was assumed that CaO would be more
reactive with Hg, as it is with SO2, because of the increased surface area and reactivity.
Limestone was injected upstream of an ESP.  The ESP inlet flue gas temperature was 177 C
(350 F).  Mercury  concentrations were determined at the inlet and outlet of the ESP with
triplicate Ontario Hydro measurements. One set of triplicate measurements was performed prior
to sorbent injection to provide a baseline set of comparison data.

       Figure 7-6 shows the Hg partitioning and speciation for three sets of Hg measurement
locations: 1) at the ESP inlet without limestone injection (baseline); 2) at the ESP inlet with
                                        7-25

-------
   I



   f
   =
   L.
   i





   
14
>* .
1
1 ii
111
(l
3




n











^^^H
















n
. D





















G Baseline Without
Limestone Injection
 Limestone Injection
iv ith Cyclone
PurticulHlf Removal
D Avernpewilh
Injection



n -










           Unse
#2
  43      #4


Injection "lests
#6  Average
Figure 7-5. Hg removal by limestone injection measured in Combustion 2000

furnace using mechanical cyclone separator (source: Reference 9).
                                  7-26

-------
    15.0
    10.0
     s.o
 s  -
                                                        Parti culuk'
                                                        EJeneriuil
                                                        OxiJiicd
              h iicn:ir^ Exit
             Uefare Sorbent
                Injeclhin
    Furnace Exit
VIK-r SHI-IK-nt InjeclHin
ISP f
Figure 7-6.  Hg removal by limestone injection measured at AECDP test facility
burning Ohio bituminous coal and using ESP (Source: Reference 9)
                                    7-27

-------
 limestone injection of 200 Ib/hr; and 3) at the ESP outlet with limestone injection of 200 Ib/hr.
 As shown in Figure 7-6, the total Hg in the flue gas at the ESP inlet with and without limestone
 injection is about the same. Limestone injection substantially increases the Hgp, thereby
 substantially reducing gaseous Hg +.  The Hgp is then removed by the ESP, providing an overall
 Hg removal of 53 percent compared to the baseline condition. Reducing the limestone feed rate
 to 25 Ib/hr showed the same Hg partitioning trends observed for 200 Ib/hr but with a reduction in
 total Hg removal. An overall Hg removal of 41 percent compared to the baseline condition was
 measured.  The increased removal provided by limestone injection compared to the baseline
 appears to be a result of the capture of Hg2+ by the CaO particulate (onto or into the particulate
 phase) and the subsequent removal of the paniculate  in the ESP.  Limestone injection had no
 apparent effect on the Hg.

        Table 7-7 presents a summary comparison of limestone sorbent injection test results with
 the activated carbon injection results from the AECDP Phase III studies (discussed in
 Section 7.5.3.2).  The table shows that limestone sorbent injection at 200 Ib/hr achieved an
 equivalent level of total Hg removal with activated carbon injection. The difference in sorbent-
 to-Hg ratios for these two tests is about a factor of 15. Based on the test results, the researchers
 concluded that activated carbon is a more effective sorbent than limestone on  a mass basis;
 however, because the  cost of activated carbon typically is an order of magnitude more than the
 cost for limestone, limestone is more effective on a sorbent cost basis.

 7.5.3.4 Multipollutant Sorbent Injection

        The EPRI/PSE&G Hudson sorbent injection study discussed in section 7.5.3.2 included
measurement of Hg removal by coinjection of activated carbon with calcium-based sorbents for
SC>2 control.13 The calcium-based sorbents tested were sodium bicarbonate and hydrated lime.
With the coinjection of either of the calcium-based sorbents, the researchers reported
improvement in the adsorption of gaseous Hg by the activated carbon.

        A study of the coinjection of a sodium-based  sorbent with activated carbon showed that
 the removal of gaseous Hg by the native fly ash and the activated carbon was  impeded
 when the sodium sesquicarbonate was coinjected. As part of the AECDP Phase III studies using
 the PCM at the PSCO Comanche Station, tests were conducted to investigate whether any
 synergistic removal of Hg or impairment of 862 removal  occurs when injecting both activated
 carbon for Hg control and sodium sesquicarbonate for 862 control into the flue gas and collected
 in a FF.17'18 The activated carbon tested was Darco FGD.

        When no sorbent  (carbon or sodium) was injected into the flue gas, the measured Hg
 removal across the PCM  by the native fly ash ranged from 41 to 76 percent at the respective
 temperatures of 162 C (324 F) and 138 C (280 F). When activated carbon  was injected into
 the flue gas with no sodium sesquicarbonate, measured Hg removal across the PCM was
 74 percent at 162 C (324 F).  When sodium sesquicarbonate was injected into the flue gas with
 no activated  carbon injection, gaseous Hg removal percentages were in the negative range (i.e.,
 test measurements indicated an increase in Hg concentrations at the PCM outlet compared to the
 inlet).  When both activated carbon and sodium sesquicarbonate were injected into the flue gas,
                                         7-28

-------
Table 7-7.  Comparison of Hg removals for activated carbon injection versus
limestone injection measured at AECDP test facility burning Ohio bituminous
coal and using ESP ( Source: Reference 9).
Parameter
Sorbent injection rate
SorbentHg mass ratio
Sorbent injection
temperature (F)
ESP operating
temperature (F)
Total Hg removal (%)
Sorbent Injected Upstream of ESP
Activated
Carbon
14 Ib/hr
9,000:1
400
345
53
Limestone
0.35 Ca:S mass ratio
200 Ib/hr
125,000:1
2,200
350
53
0.04 Ca:S mass ratio
25 Ib/hr
16,000:1
2,200
350
41
                                 7-29

-------
Hg removal percentages ranged from -104 to 22 percent. The SC>2 removal percentages did not
appear to be either impeded or improved with the coinjection of the activated carbon.

       Based on the limited data, the researchers speculated that the impediment of Hg capture
occurred either because of inhibition of the sorbent mechanism  or because the addition of sodium
increased the level of NC>2 in the flue gas.  During the sodium sesquicarbonate tests, NC>2 in the
flue gas increased from 5  to 41  ppmv, with the higher values associated with the higher
temperatures tested. If the increase in the NC>2 levels was real, researchers are questioning
whether NC>2 had a negative impact on Hg removal and subsequent Hg desorption in the flue gas.
Nitrogen dioxide is a strong oxidizer, which may have stripped  Hg from the internal surfaces of
the PCM, resulting in higher Hg measured at the outlet than the inlet (thus explaining the
negative removal efficiencies for Hg).  If this were the case, the effect would diminish over time
as the Hg on the walls of the pilot unit came into equilibrium with the flue gas. No tests were
run with sufficient time to observe this effect, and credible Hg data were not available in real
time.

       The negative impact of the sodium sesquicarbonate injection on Hg removal by activated
carbon injection is contrary to the results reported for the Hudson Station power plant tests where
injecting either sodium bicarbonate or hydrated lime with activated carbon improved the
activated carbon=s Hg adsorption capability. The Hudson data were taken over a single test day,
and the two power plants  tested burned different coal types with different fly ash properties and
flue gas compositions (eastern bituminous coal at Hudson versus PRB subbituminous at
Comanche). Drawing any definite conclusions regarding coinjection of alkaline materials and
activated carbon based on these two tests would be conjecture.

7.5.3.5 Noble-metal-based Sorbent in Fixed-bed Configuration

       ADA Technologies Inc. (ADA) has patented a  sorbent process for Hg control in coal
combustion flue gas, trade name Mercu-RE M. Unlike the dry sorbent injection processes
previously discussed, the  Mercu-RE process is based on the adsorption of the Hg by noble
metals in a fixed-bed, regeneration of the sorbents by thermal means, and recovering the
desorbed Hg for commercial recycle or disposal.19' 20 Laboratory testing of the noble-metal
sorbent showed that the sorbent captured virtually all of the Hg and mercuric chloride injected
into a simulated coal combustion flue gas. During 1999, the noble-metal  sorbent was tested for
6 months using a flue gas slipstream from the PSE&G Hudson Station. The acid gases in the
flue gas degraded the performance of the noble-metal sorbent. The field data suggested that
there are limitations on the commercial application of using noble-metal sorbents for removal of
Hg from coal combustion flue gas without upstream acid gas controls  installed.  Laboratory
testing indicated that sorbent capacity can be recovered by scrubbing acid gases from flue gas
prior to the sorbent bed.  Additional testing is being conducted to determine if noble-metal
sorbents can be used effectively on scrubbed flue gas.
                                        7-30

-------
7.6 Mercury Control Retrofits for Existing Coal-fired Electric Utility Boilers Using Semi-
Dry Absorbers

7.6.1  Retrofit Options

       Spray dryer absorber systems are the most common semi-dry scrubbers currently being
used at electric utility coal-fired electric utility power plants.  With this control technology, a
slurry of hydrated or slaked lime is sprayed into an absorber vessel where the flue gas reacts with
the drying slurry droplets.  The resulting particle-laden dry flue gas then flows to an ESP or an
FF where fly ash and 862 reaction products are collected. In  some cases, water-soluble sodium-
based sorbents are used instead of calcium-based sorbents.  SDA systems can  also provide
opportunities for injection of other dry sorbents for Hg or multipollutant control  schemes.

       In a dry sorbent injection (DSI) system, a sorbent is injected  into a flue gas duct upstream
of the PM collector. In many cases water is injected upstream of the sorbent injection location to
increase flue gas moisture content. This water spray, called spray humidification, reduces the
flue gas temperature and increases the sorbent reactivity. DSI systems can also provide
opportunities for injection of Hg or multipollutant sorbents. A circulating fluid-bed absorber
(CFA) is effectively a Avertical duct absorber that allow simultaneous gas cooling, sorbent
injection and recycle, and gas sorption by flash drying of wet lime reagents. It is believed that
CFAs can potentially control Hg emissions at costs lower than those associated with use of spray
dryers. With these absorbers, opportunities for use of advanced sorbents appear to be more
favorable than for DSI,  due to the improved sorbent utilization by re-circulation, recycle, and
flash evaporative cooling.

7.6.2  Pilot-scale and Full-scale Research and Development Status

      Full-scale tests on eastern bituminous coals (i.e., a 180 MWe boiler with a SDA-FF
control system and a 55 MWe boiler with CFA-FF controls) were conducted in September
2000.21  The EPA Method 101A was used for absorber inlet Hg measurements and the OH
Method for the boiler stacks. Both units averaged over 97 percent Hg removal in the respective
control systems based on outlet and inlet flue gas measurements. Using the raw coal analysis
and the stack OH Method measurements, each system removed about 95 percent of total Hg.
Further Hg/multipollutant testing of SDA and CFA units are planned in DOE-EPRI-EPA pilot
and field test programs.

7.7 Mercury Control Retrofits for Existing Coal-fired Electric Utility Boilers Using Wet
     FGD Scrubbers

7.7.1 Retrofit Options

       Wet FGD scrubbers are typically installed downstream of an ESP or FF.  Removal of PM
from the flue gas before it enters the wet scrubber reduces solids in the scrubbing solution and
avoids chemistry problems that may be associated with fly ash. In the United  States, plants that
use wet limestone scrubbers for SO2 control generally capture more  than 90 percent of the Hg +
                                        7-31

-------
in the flue gas entering the scrubber.  Consequently these FGD scrubbers may lower Hg
emissions by about 20 to more than 80 percent, depending on the speciation of Hg in the inlet
flue gas.

       Improvements in wet scrubber performance in capturing Hg depend primarily on the
oxidation of Hg to Hg2+.  This may be accomplished by the injection of appropriate oxidizing
agents or installation of fixed oxidizing catalysts to promote oxidization of Hg to soluble
species. Oxidation of gaseous Hg to gaseous Hg2+ can potentially increase the total Hg removed
by wet scrubbing and sorbent systems since gaseous Hg2+ is more readily captured by these
systems than gaseous Hg. Several flue gas additives and scrubbing additives are being
developed for this purpose.  Flue gas and scrubber additives are also being developed for use in
preventing the conversion of absorbed Hg2+ to gaseous Hg in wet FGD systems.

       An alternative strategy for controlling Hg emissions from wet FGD scrubbing systems is
to inject sorbents upstream of the PM control device. In units equipped with FFs this allows for
increased Hg capture and oxidization of Hg as the flue gas flows through the filter cake.
Increased oxidization afforded by FFs results in increased Hg removal in the downstream
scrubber.  In FGD units equipped with ESPs, performance gains are limited by sorbent injection
and Hg adsorption rates.

7.7.2 Mercury Absorption Theory

       Gaseous Hg is insoluble in water and therefore does not absorb in the aqueous  slurry of a
wet FGD system. Gaseous compounds of Hg + are water-soluble and do absorb in such slurries.
When gaseous compounds of Hg2+ are absorbed in the  liquid slurry of a wet FGD system, the
dissolved species are believed to react with dissolved sulfides to form mercuric sulfide  (HgS);
the mercuric sulfide precipitates from the liquid solution as a sludge.  In the absence of sufficient
sulfides in the liquid solution, a competing reaction that reduces/converts dissolved Hg2+ to Hg
is believed to take place. When this conversion takes place, the newly formed (insoluble) Hg is
transferred to the flue gas passing through the wet FGD unit.  The transferred Hg increases the
concentration of Hg in the flue gas passing through the wet FGD unit (since the incoming Hg is
not absorbed) giving rise to a higher concentration of gaseous Hg in the flue gas exiting the wet
FGD than entering it. Transition metals in the slurry (originating from the flue gas) are
suspected to play an active role in the conversion reaction since they can act as catalysts and/or
reactants for reducing oxidized species

7.7.3 Pilot-scale and Full-scale Research and Development Status

7.7.3.1 Oxidation Additives

       As part of the AECDP Phase  III studies, tests were conducted to investigate two potential
chemical additives for controlling the conversion of oxidized Hg to the elemental  form, and
enhancing the control of Hg in a pilot-scale wet FGD system.9 The first additive was gaseous
H2S.  The selection of H2S as a potential additive was based on the possibility that a sulfide-
donating species could assist in capturing Hg +. A H2S gas stream at a concentration of about 2
                                        7-32

-------
                                                    	                                 9-1-
ppm was injected into the flue gas entering the scrubber. The Hg concentrations of gaseous Hg
and gaseous Hg measured at the wet scrubber inlet and outlet for the baseline and H2S injection
tests are shown in Figure 7-7.  Gaseous Hg removal by the wet scrubber increased with the
addition of H^S (at about 2 ppm) from 46 to 71 percent. This increase was attributed mainly to a
                              94-              n
decrease in the conversion of Hg  to gaseous Hg .

       The second additive tested was ethylenediaminetetraacetic acid (abbreviated EDTA).
This chemical was  selected because EDTA is strong chelating agent. Chelating agents react with
metallic ions to form soluble nonionic compounds.  Because, transition metals may act as a
catalyst in the conversion of Hg2+ to gaseous Hg in wet FGD scrubbers, their chemical binding
may reduce the conversion.  For the test, EDTA was added to the scrubbing slurry. The Hg
concentration of gaseous Hg2+ and gaseous Hg measured at the wet scrubber inlet and outlet for
the ESP baseline and EDTA additive tests is shown in Figure 7-8.  The total Hg removal
increased to 73 percent with the addition of EDTA. Under a new cooperative agreement with
DOE/NERL, McDermott Technologies, Inc. is conducting a full-scale test program of using
scrubber additives to achieve increased Hg removal at two power plants burning high-sulfur
Ohio bituminous coal: 1) Michigan South Central Power Agency's (MSCPA) 55-MWe Endicott
Station located in Litchfield, MI, and 2) Cinergy's 1300-MWe Zimmer Station located near
Cincinnati, OH.22

7.7.3.2 Mercury Oxidation Catalysts

       Under a DOE/NETL cooperative agreement, laboratory and field tests were conducted to
investigate catalytic oxidation of gaseous Hg in coal-fired electric utility boiler flue gas.23  The
project tested the actual rate to convert gaseous Hg to a soluble form using different candidate
catalysts under simulated and actual  coal combustion flue gas conditions. The results of the
bench-scale studies are discussed in Chapter 5.  Additional extended tests with the most-active
catalysts and fly ash were conducted in the field to assess their adsorption and/or oxidation of Hg
in an actual coal-fired boiler flue gas.24 These tests were conducted in a fixed-catalyst-bed test
rig using a flue gas slipstream from a electric utility boiler firing a Texas lignite. Total Hg
concentrations in the flue gas slipstream varied from 7 to 35 |ig/Nm3, with the gaseous Hg
                                       O  	                  n,
concentrations varying from 4 to 18 |ig/Nm . The inlet gaseous Hg  also was variable,  ranging
from 30 to 80 percent of the total, and the concentrations of SO2 and NOx varied considerably
during the testing period.  The catalysts and fly ash were exposed to flue over periods ranging
from 3,480 to 3,490 hours. Table 7-8 presents the oxidation results over the 5-month-plus period
of testing. For the values of the catalyst field measurements shown in the table, the Hg
oxidation measured across the sand "blank" was subtracted from the actual measured Hg
oxidation for each catalyst.  In general, the field test results indicate that while the  initial Hg
oxidation percentages achieved by the catalysts matched the percentages measured in the
laboratory tests, the metal-based and some carbon-based catalysts were deactivated after a
relatively short time exposure to the  actual coal combustion flue gas. The researchers identified
sulfur trioxide and  selenium (or selenium compounds) as possible flue gas constituents that
rapidly deactivate the iron-based and other metal catalysts. Additional bench-scale laboratory
tests conducted as part of the study indicate that regeneration of spent catalysts should be
possible.
                                        7-33

-------
    2U-
 T
     '
Removal
                                                    S IriUl
Figure 7-7.  Effect of using H2S as an oxidation additive on wet FGD scrubber Hg
removal measured at AECDP test facility burning Ohio bituminous coal (source:
Reference 9).
                                  7-34

-------
E
a
at
-_-
c"
   ls
fl
01
             KSf Baseline Test
                       -If-,',
                                                        EDTA Addition
                                                                3*r Rcnuwal
           WS
                                                   WB InUu
WiOmln
Figure 7-8. Effect of using EDTA as an oxidation additive on wet FGD scrubber
Hg removal measured at AECDP test facility burning Ohio bituminous coal
(source: Reference 9).
                                  7-35

-------
Table 7-8.  Comparison of field test results using flue gas from electric utility
boiler firing Texas lignite versus bench-scale results using simulated flue gas for
selected candidate Hg oxidation catalysts (Source: adapted from Reference 24).
Test Parameters
Catalyst Type
Sand (non-catalyst blank)
Activated carbon #1 (1st Bed)
Activated carbon #1(2nd Bed)
Activated carbon #2
Pd#1
SB #5 (fly ash)
Laboratory
Bench-Scale
Results
Field Test Results ""
at hour
24
at hour
1,000
at hour
2,400
at hour
3,055
at hour
3,477
Percent of Hg Oxidized Across Catalyst Bed
3%
100%
100%
96%
91 %
4/70 % c
3.3-8.1%
100%
100%
97%
90%
100%
7%
66%
81 %
not
recorded
not
recorded
36%
9-12%
45%
42 - 59 %
76%
82%
82%
23%
0%
0%
0%
0%
73%
0%
89%
0%
76%
0%
0%
Test Conditions
Catalyst Bed Temp. C (F)
Inlet Hg(ng/Nm3)
Total Hg (ng/Nm3)
149 (300)
50
50
149(300) 149(300) 149(300) 104(220) 149(300)
3.7-16.2 5.4 8.3-9.3 17.8 3.7
7.0-26.1 9.8 15-27 31-35 27
   "All catalyst oxidation values corrected for the sand blank oxidation values.
   b Number of hours passing flue gas through the catalyst materials
   c Laboratory tests using SB#5 (fly ash) were conducted in a simulated flue gas with HCI (70 percent oxidation
     with 1 ppmv of HCI) and without HCI (4 percent oxidation).
                                        7-36

-------
       A pilot-scale field test program is currently being conducted under a DOE/NETL
cooperative agreement to obtain addition data on the potential commercial application of Hg
oxidation catalysts to enhance Hg capture by an existing wet FGD system downstream of high-
efficiency ESP.25  This study is testing selected catalysts previously identified as being effective
by the DOE-sponsored studies in a commercial form in larger pilot-scale units for longer periods.
The DOE/NETL is working in partnership with URS Group, Inc., EPRI, and two electric utility
companies, Great River Energy and City Public Service of San Antonio, TX. The first test site is
the Great River Energy Coal Creek Station, which fires North Dakota lignite. The second site
the City Public Service of San Antonio's J.K. Spruce Plant, which fires a PRB subbituminous
coal. The pilot-scale tests will continue for over a year at each of two sites.

 7.7.3.3 Wet FGD Scrubber Design and Operating Modifications

       Several studies of pilot-scale wet FGD scrubbers suggest that modifying the operation
and design of the scrubber unit as well as the upstream ESP may improve the capture of gaseous
Hg2+ and reduce the  conversion of absorbed Hg2+ to Hg. Specifically, these studies have found
that the liquid-to-gas ratio and tower design of a wet FGD unit affect the absorption of gaseous
Hg2+, while the oxidation air influences the conversion of absorbed Hg2+.  The operating voltage
of ESPs upstream  of wet FGD systems has also been shown to influence the latter. The
remainder of this section summarizes these findings.

       Scrubber Liquid-to-gas Ratio. The liquid-to-gas ratio (L/G ratio) of a wet FGD system is
dictated by the desired removal efficiency to control SO2 emissions.  The selected L/G ratio also
can impact the removal efficiency of gaseous Hg2+. In general, high efficiency FGD systems
(95+ percent SO2 removal) are designed with L/G ratios in the range of 120 to 150 gallons (gal.)
of aqueous slurry per 1,000 actual cubic feet (acf) of gas flow. In two separate pilot-scale
studies26 increasing the L/G ratio from approximately 40 to 125 gal./l,000 acf increased the
removal efficiency of gaseous Hg2+ from 90 to 99 percent. However, increasing the L/G ratio
did not affect the removal of gaseous Hg, which was close to zero percent. Similar studies were
                                                        01 0"7
conducted prior to these studies and produced similar findings. '

       Scrubber Tower Design. Most of the existing wet FGD systems in the United States use
either an open-spray tower or tray tower design.  In one study of wet FGD systems, where the
composition of the flue gas was mostly gaseous Hg2+, the tray tower design removed from 85 to
95 percent of the total Hg, whereas the open spray tower design removed from 70 to 85 percent
of the total Hg. 28  This study suggests that a tray tower design is more effective in removing
gaseous Hg + from boiler flue gas than an open spray tower design for a given SO2 removal
level.

       Scrubber Oxidation Air. When SO2 is absorbed in the scrubbing slurry of a wet FGD
system, the dissolved SO2 reacts with lime or limestone to form insoluble sulfate/sulfite sludge;
the sulfate reaction consumes oxygen, which is present in the flue gas. Some wet FGD systems
add air to the system to increase the amount of oxygen available for the reaction; the additional
oxygen accelerates the reaction between SO2 and lime or limestone.
                                        7-37

-------
       The effect of oxidation air on FGD Hg removal was investigated as part of the AECDP
Phase III studies by conducting test runs at baseline, intermediate and low levels of oxidation
air.9 Figure 7-9 compares wet scrubber inlet and outlet Hg concentration measured for the base
case and the runs at a mid- and low-level of oxidation air. The bars include the elemental and
oxidized fractions of the total gaseous Hg. The relative amounts of Hg at the inlet and outlet did
not change significantly for the three tests. However, the amount of absorbed Hg2+ converted to
Hg decreased as the oxidation air decreased.  This point is further illustrated in Figure 7-10 that
shows only the gaseous Hg for the three tests. For the baseline test, gaseous Hg increased by
265 percent across the wet scrubber.  This improved to a 76 percent increase for the second test,
and only two percent for the low oxidation air test. Total gaseous-phase Hg removal improved
from 46 percent for the  base case to 80 percent for the low oxidation air case. These normalized
oxidation air stoichiometry results show a strong relationship between oxidation air and wet
scrubber Hg removal for a wet FGD system. The researchers of this study hypothesize that low
oxidation air must somehow inhibit the reduction of absorbed Hg2+, or provide a species needed
to sequester the absorbed Hg2+ in the slurry. The researchers also note that the level to which the
scrubber oxidation air can be reduced at a given coal-fired electric utility power plant is highly
site-specific specific and depends on several factors such as scaling considerations and gypsum
purity requirements.

       Voltage of ESP Upstream of Scrubber. The effect of ESP operating power on wet
scrubber Hg removal was  investigated as part  of the AECDP Phase III studies.9 Concentrations
of gaseous Hg2+ and gaseous Hg were measured at the inlet and outlet of the wet FGD system
for three different ESP operating conditions. For the first operating condition (the baseline
operation), the pilot-scale  ESP was operated with three of its four fields in service, and the power
was set to maintain an outlet particulate loading of 0.02 to 0.03 Ib/MBtu (below the PM limit of
the New Source Performance Standard for utility boilers). In the second operating condition, the
ESP voltage was increased by 60 percent above the baseline voltage.  In the third operating
condition, the ESP power  was turned off and an FF was used for PM control upstream of the wet
FGD system.  For all three operating conditions, triplicate measurements of Hg were made at the
inlet and outlet of the pilot-scale  wet FGD system.
                                                         94-              n
       Figure 7-11 compares the concentrations of gaseous Hg and gaseous Hg  measured at
the inlet and outlet of the wet FGD system for the three different ESP operating conditions.
Since the Hg measurements were taken downstream of the ESP and FF, very little Hgp was
measured; thus, Hgp measurements are not shown in the Figure 7-11.  Figure 7-12 presents only
gaseous Hg for the same three ESP conditions as those in Figure 7-11.  The figures clearly show
that the operating voltage  of the ESP has a direct, negative impact  on the wet scrubber Hg
control performance.  The proportion of gaseous Hg2+ and gaseous Hg at the wet scrubber inlet
is the same for all three tests. However, for the high-power test, the amount of gaseous Hg
significantly increased across the wet scrubber. The gaseous Hg remains constant for the
no-power test, which is the observed behavior when the scrubber is preceded by the FF. This
indicates that the electric field affects some component of the flue  gas, which, in turn, has a
negative impact on wet scrubber chemistry.
                                        7-38

-------
...
1
"-f
*
,s 15
_
*
K | (1
C"
' 5
*g?













I










latHim* C)v -




"' ^



 	 1

lir











D









OH-''
vAiFMid




"*  , ::\ -.*: j. -.' i:


D'


0










\-Mrl.i.w





M.-;R.,: :,=..,=
j


Figure 7-9. Effect of oxidation air on wet FGD scrubber Hg removal as measured
at AECDP test facility burning Ohio bituminous coal (source: Reference 9).
                                 7-39

-------
    12
  S  J
        Biiji'jljjjj- O.y \\r
                                                          CK -%ir I ,fit
Figure 7-10. Effect of oxidation air on Hg in wet FGD scrubber flue gas as
measured at AECDP test facility burning Ohio bituminous coal (source:
Reference 9).
                                   7-40

-------
    Ill
 j
 I  15
 7S
    W
                                 ESP
               4>v, Ri.nii>v.tl
                                        Jll1. Hviiii". ;,l
F.SI" So Pnwrr
         S E-iI.,1   V. -, -: i r .1
                                                         V, s ;-iU-i    vs s f i i:
Figure 7-11.  Effect of ESP operating voltage on wet FGD scrubber Hg removal as
measured at AECDP test facility burning Ohio bituminous coal (source:
Reference 9).
                                    7-41

-------
     1ft

     y
Figure 7-12.  Effect of ESP operating voltage on Hg in wet FGD scrubber flue gas
as measured at AECDP test facility burning Ohio bituminous coal (Source:
Reference 9).
                                 7-42

-------
7.8 Multipollutant Control Technologies

       This section presents  a summary of control systems being commercially offered or
developed for multipollutant emissions control.  The current status of many systems is based
upon reports that targeted one or two pollutants. A caution here is that, when evaluating the best
system for a specific application, it is important to consider both: 1) how a given system affects
the emissions of all pollutants, and 2) how that system affects the long-term performance,
operation, and cost of other downstream systems, including ductwork, heat exchangers, stacks,
and other emission control equipment.  To date no comprehensive long-term evaluations of the
multipollutant systems described below have been conducted.

7.8.1  Corona Discharge

       Generation of an intense corona discharge (ionization of air by a high voltage electrical
discharge) in the boiler flue gas upstream of an ESP and wet scrubber is being investigated with
respect to improving PM control by oxidation of a portion of the entering SO2 to SO3. 9  The
corona discharge creates oxygen-carrying reactive species, which, in turn, oxidize the Hg in the
flue gas (i.e., convert Hg to Hg2+).  The increased SO3 both improves ESP collection of PM and
acts to convert Hg to Hg2+which may then be captured by an alkaline FGD  scrubber
downstream.  Representative reactions for SO2 oxidation by corona discharge include:

                          O2 + e- ->2O  + e-
                          O2 + O -> O3
                        SO2 + O3 -> SO3 + O2
                        SO3 +H2O->H2SO4

Similarly, for NO,

                          NO + e- -> NO-
                        NO + NO-  ->NO2 + N + e-
                            O2 + e-  ->2O + e-
                            O2 +O  ->O3
                          H2O + O3  ->2OH + O2
                        NO2 + OH  ->HNO3

       Environmental Elements  Corporation is developing a process based on corona discharge
that recovers the oxidized sulfur  and nitrogen compounds as marketable sulfuric and nitric acids
in wet ESP sections and or/absorbers.  A slipstream pilot plant has been installed at Alabama
Power Miller Plant (Unit 3).  Initial tests indicated 80 percent Hg removal and complete
oxidation of Hg at 10 and 20 W/cfm, respectively.

       Powerspan Corporation is developing a single, integrated pollution control device that
uses a proprietary technology called Electro-Catalytic Oxidation or ECO to control SO2,
NOx,  Hg, and fine PM in coal-fired boiler flue gas.30 The first stage of the device uses a
dielectric barrier discharge to convert NOx and SO2 to acids and to oxidize Hg. A condensing,
                                        7-43

-------
wet ESP is used to collect acid mists, fine PM, and Hg.  The effluent from the wet ESP is
processed to produce salable byproducts (e.g., concentrated acids, gypsum for wallboard
manufacture, and ammonia for fertilizer). Before entering the ECO unit, flue gas passes
through a  conventional ESP to remove the majority of the ash particles. In partnership with
FirstEnergy Corporation, Powerspan has built a pilot-scale ECO test facility at FirstEnergy's R.E.
Burger Plant near Shadyside, OH.31 This test facility processes a slipstream of flue gas from a
150-MW boiler unit burning high-sulfur eastern bituminous coal.  The test results showed a Hg
emission reduction of 68 percent. Under a new DOE cooperative agreement, Powerspan and
FirstEnergy are conducting a research project using the ECO pilot test facility to optimize the
technology's Hg removal capability while maintaining the performance of the ECO unit for
removal of nitrogen  oxides,  sulfur dioxide, and fine PM.32 In addition, Powerspan and
FirstEnergy are currently constructing an $11.9 million ECO commercial demonstration unit at
FirstEnergy's Eastlake Plant near Cleveland,  OH.  The project is being cofunded by a $3.5
million grant from the Ohio Coal Development Office.

7.8.2  Electron Beam Irradiation

       The E-Beam Process has been offered commercially since the 1980s and is now used in
Japan and China.33 The chemical reactions are identical to corona discharge, except that the
power source is a battery of irradiating electron "guns" and the oxidation products then enter a
semi-dry absorption system  with ammonia reagent and are converted to ammonium sulfate and
nitrate salts suitable  for use as a fertilizer. It is presumed that the Hg solids would also be
present in the fertilizer as contaminants.  The polishing reactions for E-Beam are:

                      NH4OH + HNO3 ->NH4NO3 + H2O

                     2 NH4OH + H2SO4 -> (NH4)2SO4 + 2 H2O
7.8.3  Oxidant Injection in Flue Gas

       A number of proposed schemes would add an oxidant such as chlorine, peroxide, or
ozone to the flue gas upstream of an absorber. Again the reaction products would be similar to
corona or electron beam, and the recovered products could range from weak acids to
sulfate/nitrate fertilizers or lower-value soil amendments; trace Hg salts would likely be
contained within these products. An example of ozone injection is the Lo-TOx.34 The ISCA is a
chlorine-based system producing byproduct acids. Hydrogen peroxide and other chlorine-based
oxidation schemes have been investigated but have not been proposed for commercial use.35
Typical oxidation reactions are:
                                        7-44

-------
          Hydrogen Peroxide:                       Ozone:

           H2O2 -->  2 OH                   NO + O3 ->NO2 + O2
      H2O2 + OH ->  HO2 + H2O          2 NO2  +  O3 -> N2O5 + O2
       NO + OH  -> NO2 + H             SO2 + O3  -> SO3 + O2
       NO + OH  -> HNO2                N2O5 + H2O  -> 2 HNO3
      NO + HO2 -> HNO3                 SO2 + N2O5  -> SO3 + 2 NO2
      NO2 + OH ->HNO3

7.8.4  Catalytic Oxidation

       Catalysts can be employed in higher temperature regimes to speed up oxidation of SO2
and NOx, but not Hg.  However, increasing the SO3  and NO2/N2O4/N2Os concentrations will
likely result in increased conversion of Hg to Hg2+ downstream, as acid gases and PM are
removed in control devices.  Lower temperature catalysis (less than 500 F) would likely directly
          n      94-  	
oxidize Hg to Hg . Thus, any number of catalytic oxidation schemes that produce byproduct
acids would likely remove a substantial portion of total Hg with the acids as a Hg salt  chloride,
sulfate, or nitrate.  A number of catalytic technologies are under commercial development; an
example of this class -  SNOx - has been evaluated under DOE's Clean Coal Technology
Program. At least one  current DOE-sponsored project is examining the effectiveness of an
oxidation catalyst upstream of wet FGD scrubber to decrease total Hg emissions.36

7.8.5   Oxidant Addition to Scrubber

       One current DOE test program is measuring the effectiveness of a Hg oxidant added to
the liquor of commercial wet scrubbers.  The EPA is  sponsoring similar research, which will
culminate in a pilot-scale slipstream evaluation of oxidant addition.37  Another DOE-sponsored
project is investigating the use of oxidated-lime and lime-silica sorbents to a semi-dry circulating
bed absorber for combined SO2, NOx, and Hg control.38 Other combinations of sorbents injected
upstream of an efficient PM collector such as the EPRI Toxecon process may be used for a
multiple pollutant control strategy centered around PM control.

7.8.6  Catalytic Fabric Filters

       Some pilot-scale efforts have reported substantial oxidation of Hg within a FF,
presumably by catalytic action of certain fibers or residual fly ash imbedded within the fabric.39
Several investigations are being made into woven carbon fibers or other catalytic materials
integrated into the bag  filters for a combined Hg/PM  control device.

7.8.7   Carbon-fiber FFs and ESPs

       Carbon-fiber FFs are commercially available.  Carbon-fiber ESP plates are being
investigated under a study sponsored the Ohio Coal Development Office. While combined
Hg/PM control using this approach would be initially effective, the Hg capacity would be
                                        7-45

-------
realized in a relatively short time period; therefore, means of regenerating the carbon active sites
without replacing the fabric filter bags or ESP plates have to be devised.

7.9  Summary

       A practical approach to controlling Hg emissions at existing utility plants is to minimize
capital costs by adapting or retrofitting existing equipment to capture Hg. Potential retrofit
options for control of Hg were investigated for units that currently use the following post
combustion emission control methods: (1) ESPs or FFs for control of PM, (2) dry FGD scrubbers
for control of PM and 862, and (3) wet FGD scrubbers for the control of PM and 862.

Hg Control Retrofits for ESP andFF

       ESPs and FFs are either cold-side or hot-side devices. Hot-side devices are installed
upstream of the air heater while cold-side devices are installed downstream.  Flue gas
temperatures in hot-side devices typically range from 350 to 450 C while cold-side devices
typically operate at temperatures ranging from 140 to 160 C. Based on current information, it
appears that little Hg can be captured in hot-side ESPs or FFs.

       Least-cost retrofit options for the control of Hg emissions from units with ESP or FF are
believed to include:

         Injection of a sorbent upstream of the ESP or FF. Cooling of the  stack gas or
          modifications to the ducting may be needed to keep sorbent requirements at
          acceptable levels.

         Injection of a sorbent between the ESP and a pulsejet FF retrofitted downstream of
          the ESP. This approach will increase capital costs but reduce sorbent costs.

         Installation of a semi-dry CFA upstream of an existing ESP used  in conjunction with
          sorbent injection.  The CFA recirculates both fly ash and sorbent to create an
          entrained bed with a large number of reaction sites. This leads to higher sorbent
          utilization and enhanced  fly ash capture of Hg and other pollutants.

       Units equipped with a FF require less sorbent than units equipped with an ESP. ESP
systems depend on in-flight adsorption of Hg by entrained fly ash or sorbent  particles. The FFs
obtain in-flight capture and capture as the flue gas passes through the FF.

       In general, the successful  application of cost-effective sorbent injection technologies for
ESP and FF units will depend on: (1) the development of lower cost and/or higher performing
sorbents, and (2) appropriate modifications to the operating conditions or equipment being
currently used to control emission of PM, NOx, and SC>2.
                                         7-46

-------
Mercury Control Retrofits for Wet FGD Scrubbers

       Wet FGD scrubbers are typically installed downstream of an ESP or FF. Wet limestone
FGD scrubbers are the most commonly used scrubbers on coal-fired electric utility boilers.
These FGD units generally capture more than 90 percent of the Hg2+ in the flue gas entering the
scrubber.  Consequently, existing wet FGD scrubbers may lower Hg emissions by about 20
percent to more than 80 percent, depending on the speciation of Hg in the inlet flue gas.

       Improvements in wet scrubber performance in capturing mercury depend primarily on the
oxidation of Hg to Hg2+.  This may be accomplished by 1) the injection of appropriate oxidizing
agents, or 2) the installation of fixed oxidizing catalysts upstream of the scrubber to promote
oxidization of Hg to soluble species.

       An alternative strategy for controlling Hg emissions from wet FGD scrubbers is to inject
sorbents upstream of the PM control device.  In wet FGD systems equipped with ESPs,
performance gains  are limited by the in-flight oxidization of Hg, and the in-flight capture of
Hg2+ and Hg. In systems equipped with FFs, increased oxidization and capture of Hg can be
achieved as the flue gas flows through the FF. Increased oxidization of Hg in the FF will result
in increased Hg removal in the downstream scrubber.

Mercury Control Retrofits for Semi-dry FGD Systems

       SDA systems that use calcium-based sorbents are the most common dry FGD systems
used in the utility industry.  An aqueous slurry containing the sorbent is sprayed into an absorber
vessel where the flue gas reacts with the drying slurry droplets.  The resulting, particle-laden, dry
flue gas then flows to an ESP or a FF where fly ash and 862 reaction products are collected.

       CFAs are "vertical duct absorbers" that allow simultaneous gas cooling, sorbent injection
and recycle, and gas absorption by flash drying of wet lime reagents. It is believed that CFAs
can potentially control Hg emissions at costs lower than those associated with use of spray
dryers.

       Dry FGD systems are already equipped to control emissions of SC>2 and PM.  The
modification of these units  by the use of appropriate  sorbents for the capture of Hg and other air
toxics is considered to be the easiest retrofit problem to solve.
7.10 References
1.  Smit, F. J., G.L. Shields,  and CJ. Mahesh. "Reduction of Toxic Trace Elements in Coal By
    Advanced Cleaning." Presented at the Thirteenth Annual International Pittsburgh Coal
    Conference, September 3-7, 1996.
                                        7-47

-------
2.  Topical Report No. 5 Trace Element Removal Study." Prepared for U.S. Department of
   Energy's Pittsburgh Technology Center by ICF Kaiser Engineers, Fairfax, VA.  March 1995.

3.  Brown, T. D., D.N. Smith, R.A. Hargis, Jr., and WJ. O'Dowd.  "1999 Critical Review:
   Mercury Measurement and Its Control: What We Know, Have Learned, and Need to Further
   Investigate.,'" Journal of the Air & Waste Management Association, June 1999. pp. 1-97.

4.  Nebel, K. L., D.M. White, W.H. Stevenson, and M.G. Johnston. A Summary of Mercury
   Emissions and Applicable Control Technologies for Municipal Solid Waste Combustors.
   U.S. EPA, Office of Air Quality Planning and Standards, Research Triangle Park, NC.
   September 1991.

5.  Getz, N. P., B.T. Ian, and C.K. Amos. "Demonstrated and Innovative Control Technologies
   for Lead, Cadmium and Mercury for Municipal Waste Combustors," Proceedings of the Air
   & Waste Management Association 85th Annual Meeting and Exhibition, Kansas City, MO.
   1992.

6.  Brown, B., and K. Felsvang. "Control of Mercury and Dioxin Emissions from United States
   and European Municipal Waste Incinerators by Spray Dryer Absorption Systems," in
   Proceedings of the Municipal Waste Combustion International Specialty Conference, Air
   and Waste Management Association, VIP-19, Tampa, FL, pp 685-705, April 1991.

7.  Babcock & Wilcox Alliance Research Center. Advanced Emissions Control Development
   Program Phase I'- Approved Final Report prepared for the U.S. Department of Energy
   (U.S. DOE-FETC contract DE-FC22-94PC94251) and Ohio Coal Development Office
   (grant agreement CDO/D-922-13), July 1996.

8.  McDermott Technologies, Inc. Advanced Emissions Control Development Program
   Phase II - Approved Final Report, prepared for the U.S. Department of Energy (U.S. DOE-
   FETC contract DE-FC22-94PC94251) and Ohio Coal Development Office (grant agreement
   CDO/D-922-13), RDD:98:43509-500-200:01R, April 1998. Available at:
   .

9.  McDermott Technologies, Inc. Advanced Emissions Control Development Program
   Phase III'- Approved Final Report, prepared for the U.S. Department of Energy (U.S. DOE-
   FETC contract DE-FC22-94PC9425122) and Ohio Coal Development Office (grant
   agreement CDO/D-922-13).  July 1999. Available at:
   < http://www.osti.gov/dublincore/servlets/purl/756595-LACvcL/webviewable/756595.pdf>.

10. Grover, C., J.  Butz, S. Haythornthwaite, J. Smith, M. Fox, T. Hunt, R. Chang, T. Brown, and
   E. Prestbo. "Mercury Measurements Across Particulate Collectors of PSCO Coal-fired
   Electric Utility Boilers," EPRI/DOE/EPA Mega-Symposium, Atlanta, GA. August 1999.
                                      7-48

-------
11. Hargis, R. A., WJ. O'Dowd, and H.W. Pennline.  "Sorbent Injection for Mercury Removal
   in a Pilot-scale Coal Combustion Unit," presented at the 93th Annual Meeting & Exhibition
   of the Air & Waste Management Association, Salt Lake City, UT. June 2000.

12. U.S. Department of Energy, National Energy Technology Laboratory. In-House Research
   on Mercury Measurement and Control at NETL. Pittsburgh, PA . November 2001. Available
   at: < http://www.fetc.doe.gov/coalpower/environment/mercury/pubs/poster.pdf >.

13. Waugh, E.G., B.K. Jensen,  L.N. Lapatnick, F.X. Gibbons, S. Sjostrom, J. Ruhl, R. Slye, and
   R. Chang. "Mercury control in utility ESPs and baghouses through dry carbon-based sorbent
   injection pilot-scale demonstration," In Proceedings of the EPRI/DOE/EPA Combined
   Utility Air Pollutant Control Symposium, EPRI TR-108683-V3; Washington, DC, August
   25-29, 1997.

14. Durham, M.D, C.J. Bustard, R. Schlager, C. Martin, S. Johnson, and S. Renninger. "Field
   Test Program to Develop Comprehensive Design, Operating and Cost Data for Mercury
   Control Systems on Non-Scrubbed Coal-Fired Boilers," presented at the Air & Waste
   Management Association 2001 Annual Conference and Exhibition, Orlando, FL. June 24-28,
   2001.

15. Bustard, C.  J., M. Durham, C. Lindsey, T. Starns, K. Baldrey, C. Martin, S. Sjostrom, R.
   Slye, S. Renninger, and L. Monroe, "Full-Scale Evaluation of Mercury Control with Sorbent
   Injection and COHPAC at Alabama Power E.G. Gaston," presented at the A&WMA
   Specialty Conference on Mercury Emissions: Fate, Effects, and Control and the U.S.
   EPA/DOE/EPRI Combined Power Plant Air Pollutant Control Symposium: The Mega
   Symposium, Chicago, IL. August 20-23, 2001.

16. Madden, D.A., and M.J. Holmes. "B&W's E-LIDS TM Process - Advanced SOx,
   Paniculate, and Air Toxics  Control for the Year 2000," presented at the 1998 EPRI-DOE-
   EPA Combined Utility Air Pollutant Control Symposium, Washington, DC. August 25-29,
   1997.

17. Sjostrom, S., J. Smith, T. Hunt, R. Chang, and T. D. Brown. "Demonstration of Dry Carbon-
   Based Sorbent Injection for Mercury Control in Utility ESPs and Baghouses."  Presented at
   the Air & Waste Management Association's 90th Annual Meeting & Exhibition, Toronto,
   Ontario, Canada. June 8-13, 1997.

18. Haythornthwaite, S., S. Sjostrom, T. Ebner, J. Ruhl, R. Slye, J. Smith, T. Hunt, R. Chang,
   and T.D. Brown. "Demonstration of Dry Carbon-Based Sorbent Injection for Mercury
   Control in Utility ESPs and FFs," in Proceedings of the EPRI/DOE/EPA Combined Utility
   Air Pollutant Control  Symposium; Washington, DC; EPRI TR-108683-V3. August 25-29,
   1997.
                                      7-49

-------
19. Roberts, D.L., J. Albiston, T. Broderick, C. Greenwell, and R. Stewart. Novel Process for
   Removal and Recovery of Vapor Phase Mercury, Phase I Final Report under Contract DE-
   AC22-95PC95257 to DOE Federal Energy Technology Center, Pittsburgh, PA. September
    1997.

20. Turchi, C.S., J. Albiston, T.E. Broderick, and R.M. Stewart. "Removal of Mercury from
   Coal-Combustion Flue Gas Using Regenerable Sorbents," presented at the 92nd Annual
   Meeting of the Air & Waste Management Association, St. Louis, MO. June 1999.

21.  ARC ADIS Geraghty & Miller. Roanoke Valley Energy Facility Mercury Testing. Research
   Triangle Park, NC. November 6, 2000.

22. U.S. Department of Energy, National Energy Technology Laboratory. "Full-Scale Testing of
   Enhanced Mercury Control in Wet FGD,"  November 2001. Available at
   < http://www.fetc.doe.gov/coalpower/environment/mercury/index.html >.

23. Hargrove, O.W., Jr., T.R. Carey, C.F. Richardson, R.C.  Skarupa, F.B. Meserole, R.G. Rhudy,
   and T.D. Brown. "Factors Affecting Control of Mercury by Wet FGD," Presented at the
   EPRI/DOE/EPA Combined Utility Air Pollutant Control Symposium, Washington, DC.
   August 1997.

24. Blythe, G.M, T.R.Carey, C.F. Richardson , F.B. Meserole, R.G. Rhudy,  and T.D. Brown.
   "Enhanced Control of Mercury by Wet Flue Gas Desulfurization Systems," Presented at the
   92nd Annual Meeting & Exhibition of the Air & Waste Management Association, St. Louis,
   MO. June 1999.

25. U.S. Department of Energy, National Energy Technology Laboratory. "Pilot Testing of
   Mercury Oxidation Catalysts," Pittsburgh, PA. November 2001.  Available at:
   < http://www.fetc.doe.gov/coalpower/environment/mercury/index.html >.

26. Redinger, K. E., A. P. Evans, R. T.  Bailey, and P. S. Nolan. "Mercury Emissions Control in
   FGD Systems," presented at the EPRI/DOE/EPA Combined Air Pollutant Control
    Symposium, Washington, DC. August 25-29, 1997.

27. Hargrove, O.W., Jr., J.R. Peterson, D.M. Seeger, R.C. Skarupa, and R.E. Moser.  "Update of
   EPRI Wet FGD Pilot-Scale Mercury Emissions Control Research," presented at the
   EPRI/DOE International Conference on Managing Hazardous and Particulate Pollutants,
   Toronto, Canada.  August 15-17, 1995.

28.  Electric Power Research Institute.  Electric Utility Trace Substances Synthesis Report -
    Volume 3: Appendix O, Mercury in the Environment. EPRI TR-104614-V3, Project
   3081,3297. November 1994.
                                       7-50

-------
29. Helfritch, D.J., and P.L. Feldman.  "Flue Gas Mercury Control by Means of Corona
   Discharge," Paper 99-157, Air & Waste Management Association 92nd Annual Meeting,
   St. Louis, MO.  June 20-24, 1999.

30. McLarnon, C. R., M. L. Horvath, and P. D. Boyle. "Electro-Catalytic Oxidation Technology
   Applied to Mercury and Trace Elements Removal from Flue Gas," presented at Conference
   on Air Quality II, McLean, VA. September 20, 2000.

3 1 . McLarnon, C. R, and M. D. Jones. "Electro-Catalytic Oxidation Process for Multi-Pollutant
   Control at FirstEnergy 's R.E.  Burger Generating Station," presented at Electric Power 2000,
   Cincinnati, OH.  April 5, 2000.

32. U.S. Department of Energy, National Energy Technology Laboratory. "Non-thermal Plasma
   Based Removal of Mercury,"  November 2001. Available at
   < http://www.fetc.doe.gov/coalpower/environment/mercury/index.html >.
33. Hirona, S. "Simultaneous SO2, SOs andNOx Removal by Commercial Application of the
   EBA Process," presented at the EPRI/DOE/EPA Combined Utility Air Pollution Control
   Symposium, Atlanta, GA, EPRI TR-113187-V2, pp 8-1 through 8-14. August 1999.

34. Anderson, M.H., A.P. Skelley, E. Goren, and J. Cavello.  "A Low Temperature Oxidation
   System for the Control of NOx Emissions Using Ozone Injection," presented at the Institute
   of Clean Air Companies Forum 98: Cutting NOx Emissions, Durham, NC. March 18-20,
   1998.

35. Livengood, C.D., and M.H. Mendelsohn.  "Process for Combined Control of Mercury and
   Nitric Oxide," presented at the EPRI/DOE/EPA Combined Utility Air Pollution Control
   Symposium, Atlanta, GA, EPRI TR-113187-V2, pp 19-30 through 19-41. August 1999.

36. Richardson, C.F., G.M. Blythe, T.R. Carey, R.G. Rhudy,  and T.D. Brown. "Enhanced
   Control of Mercury by Wet FGD Systems," EPRI/DOE/EPA Combined Utility Air Pollution
   Control Symposium, Atlanta, Georgia, EPRI TR-113187-V3, pp 20-41 through 20-54,
   August 1999.

37. Roy, S.,  and G.T. Rochelle.  Chlorine Absorption in S (IV) Solutions .  EPA-600/R-0 1-054
   (NTIS PB2001-107826), National Risk Management Research Laboratory, Research
   Triangle Park, NC .  August 2001 .
38.  Ghorishi, S.B.,  C.F. Singer, W.S. Jozewicz, R.K. Srivastava, and C.B. Sedman.
    "Simultaneous Control of Hg, SO2, and NOx by Novel Oxidized Calcium-Based Sor
    Paper # 243, presented at the 94th AWMA Annual Meeting, Orlando, FL. June 2001
                                      7-51

-------
39.  McManus, T.J., R.O. Agbede, and R.P. Khosah.  "Conversion of Elemental Mercury to the
    Oxidized Form in a Baghouse," Paper 98-WP79A.07, presented at the A&WMA 91st
    Annual Meeting, San Diego, CA. June 14-18, 1998.
                                       7-52

-------
                                       Chapter 8
                 Cost Evaluation of Retrofit Mercury Controls for
                          Coal-fired Electric Utility Boilers
8.1 Introduction

       A practical approach to controlling Hg emissions at existing coal-fired electric utility
power plants is to minimize control costs by adapting or retrofitting existing air pollution control
equipment to capture Hg. As discussed in Chapter 3, coal-fired electric utility power plants
currently use a wide variety of technologies to control the emission of criteria air pollutants (e.g.,
PM, SO2, and NOX emissions).  Generally, the air pollution control methods and configurations
used for a given coal-fired electric utility boiler depend on the type of coal burned, age and size
of the boiler unit, and the power plant location.

       Potential retrofit technologies for the control of Hg emissions from existing coal-fired
electric utility boilers are discussed in Chapter 7.  Control technologies using injection of
powdered activated carbon (PAC) into the flue gas have been applied successfully on municipal
waste combustors to reduce Hg emissions. Pilot-scale testing indicates that these technologies
offer the potential to provide significant Hg removal from the flue gas of coal-fired electric utility
boilers. This chapter discusses an initial evaluation of annual Hg control costs based on the
retrofit of PAC injection-based control technologies to a series of model plant scenarios (not
actual full-scale applications) representative of the coal-fired electric utility power plants
operating in the United States.  It is worth noting that, while performance and cost of only PAC-
related technologies were evaluated, other non-P AC-based Hg  control technologies are expected
to be available in the future.  For example, enhanced Hg oxidation using oxidants or catalysts
followed by wet scrubbing may become available.  Also, the role of an SCR-FGD combination
may become more cost effective and attractive. The information presented in this chapter was
used in the EPA's recent regulatory determination regarding Hg and other air toxics.

       The cost estimates of the PAC injection-based Hg control technologies presented in this
chapter are based on relatively few data points from pilot-scale tests and, therefore, are
considered to be preliminary estimates.  As discussed in Section 8.2, factors that are known to
affect adsorption of Hg on activated carbon include speciation of Hg in the flue gas, flue gas and
ash characteristics, and the degree of mixing between the flue gas and activated carbon.  The
effects of these factors may not be entirely accounted for in the relatively few pilot-scale data
points available for this evaluation. Successful testing of a control approach at small pilot plants
                                          3-1

-------
does not necessarily guarantee successful implementation of the approach in full-scale systems.
Temporary wall effects at small scale will generally not be realized at full scale. Appropriate
mass transfer associated with mixing and the number, placement, and design of reagent and
sorbent injection equipment may also need to be determined.  Further, potential longer-term
problems such as deposits, fouling, and corrosion of the control equipment are frequently not
addressed by pilot-scale tests because of shorter-term, non-continuous operation.  Ongoing
research is expected to address these issues to improve the potential of using sorbents for Hg
control in coal-fired boilers.

       Coal-fired electric utility power plants are currently required to reduce emissions of NOx,
SC>2, and PM.  The EPA has also revised the National Ambient Air Quality Standards (NAAQS)
for PM and ozone. These revisions may require electric utility sources to adopt control measures
aimed at reducing concentrations of fine PM in the atmosphere. In addition, as discussed above,
the EPA has recently expressed its intent to regulate Hg emissions from these sources. Adding to
these environmental requirements and activities, Congress is introducing bills aimed at
developing legislation requiring simultaneous reductions in emissions of multiple emissions.
Improved sorbents and other methods for controlling Hg and multipollutant (e.g., Hg and NOx)
emissions are also under development by DOE, EPA, EPRI, the electric industry, and equipment
vendors. These development activities include large demonstration programs that are underway
under the sponsorship of DOE/NETL and industrial participants. The demonstrations are
focused on full-scale testing of powdered activated carbon injection and  modifications to flue gas
cleaning systems aimed at improving Hg capture.

       It is expected that, when the research and development activities being conducted by
DOE, EPA, EPRI, and others are completed, there will be many more control options for Hg and
multipollutants with attendant benefits in improved cost effectiveness.
8.2 Cost Estimate Methodology

       The methodology used for the Hg control cost evaluation consists of the following six
steps:
          First step, a set of model plant and Hg control scenarios was defined;
          Second step, cost estimates were made for selected scenarios using a cost model
          developed collaboratively by the DOE and the EPA;
          Third step, the cost impacts of selected variables were examined;
          Fourth step, the cost model results were used to develop indications of costs for those
          model plant scenarios for which data on PAC use are currently not available;
          Fifth step, potential future  improvements in the cost estimates were examined; and
          Sixth step, in order to place Hg control costs in perspective, these costs were
          compared to current costs of applying NOx controls to coal-fired electric utility
          boilers.
                                         8-2

-------
8.2.1  Mercury Control Technologies Evaluated

       The cost evaluation is based primarily on the application of potential PAC injection-based
control technologies. These technologies were selected because sufficient pilot-scale data are
available to make reasonable estimates of the Hg capture efficiency of the technologies. Mercury
capture performance data are currently not available for other potential Hg control technologies
(e.g., use of catalysts to oxidize Hg in wet scrubber systems) that conceivably could be applied
to coal-fired electric utility boilers at this time. Table 8-1 lists the PAC injection-based Hg control
technologies defined for this study.  Pilot-scale applications of most of these technologies have
been reported in published literature.l!2'3'4'5'6

       PAC injection-based retrofit control technologies ESP-1, ESP-3, ESP-4, ESP-6, and
ESP-7 are applicable to coal-fired electric utility boilers equipped with a cold-side ESP.

       In ESP-1, PAC is injected between the air preheater and the cold-side ESP (CS-ESP, i.e.,
an ESP located downstream of the boiler=s air preheater). This configuration is the simplest to
install, requiring only PAC injection equipment upstream of the ESP. Activated carbon
consumption is expected to be relatively high because the high temperature of the flue gas would
inhibit adsorption of Hg onto PAC.

       In ESP-3, PAC is injected downstream of the CS-ESP and is collected using a polishing
fabric filter (PFF).  This technology permits recycling of the PAC sorbent to increase its
utilization. Typically, this recycling is achieved by transferring a portion of used sorbent from
the PM control device (e.g., PFF) to the sorbent injection location using a chain or a belt
conveyor, mixing the used sorbent with fresh sorbent, and injecting the  resulting sorbent mixture
into the flue gas. Further, the technology provides a contact bed (i.e., filter cake on PFF) for
increased adsorption of Hg.

       ESP-4 is similar to ESP-1, but adds spray cooling (SC) upstream of the  PAC injection
location.  Cooling the flue gas aids adsorption and reduces PAC injection requirements.
However, adding too much water to the flue gas could cause acid condensation, which  would
corrode ductwork and equipment. In the cost modeling conducted for this work, flue-gas
temperatures are not allowed to reach the acid dewpoint (i.e., the temperature at which  the acidic
components in the flue gas would condense).

       ESP-6 is similar to ESP-3, but provides SC upstream of PAC injection.  Cooling the flue
gas aids adsorption and reduces PAC injection requirements. Also, use of PFF  permits sorbent
recycling, leading to improved sorbent utilization.

       ESP-7 is the same as ESP-6 except for the  addition of a second sorbent, lime. In addition
to Hg removal, this technology would remove acid gases  from the flue gas. Pilot-scale results
have indicated that this may result in significant lowering of PAC injection rates.

-------
Table 8-1.  Mercury control technologies.
Existing Post-combustion
Control Devices
Used for
Coal-fired Boiler Unit ab
CS-ESP
HS-ESP
FF
SDA + FF
SDA + CS-ESP
Mercury Control Technologies b
Identification
Code
ESP-1
ESP-3
ESP-4
ESP-6
ESP-7
HESP-1
FF-1
FF-2
SD/FF-1
SD/ESP-1
Additional Control Equipment Installed
PAC injection
PAC injection + PFF
SC + PAC injection
SC + PAC injection + PFF
SC + PAC injection + lime injection + PFF
SC + PAC injection + PFF
PAC injection
SC + PAC injection
PAC injection
PAC injection
 (a) Existing controls may include wet FGD scrubber system or post-combustion NOxcontrols such as selective
   catalytic reduction (SCR) and selective noncatalytic reduction (SNCR).
 (b) CS-ESP = cold-side electrostatic precipitator
   HS-ESP = hot-side electrostatic precipitator
   FF = fabric filter
   PAC = powdered activated carbon
   PFF = polishing fabric filter
   SC = spray cooling
   SDA = spray dryer adsorber system
                                          8-4

-------
       In HESP-1, SC, PAC injection, and a PFF are added downstream from a hot-side ESP (an
ESP located upstream of the boiler=s air preheater). This configuration is identical to ESP-6, only
the location of the ESP is different.

       Two PAC injection-based retrofit controls are applicable to coal-fired electric utility
boilers equipped with a fabric filter.  FF-1 is the fabric filter analogue of ESP-1. However, Hg
collection should be better than that in ESP-1 because the FF provides added residence time and
a contact bed (filter cake on the bags) for increased adsorption of Hg. FF-2 is the fabric filter
analogue of ESP-4; spray cooling and PAC injection are installed upstream of an existing fabric
filter. As with ESP-4, cooling reduces PAC requirements, which reduces total annual PAC costs
for FF-2 compared to FF-1.

       Finally, use of a PAC injection in combination with an existing spray dryer adsorber
system for 862 control was evaluated.  In SD/FF-1, PAC is injected into the flue gas of a boiler
that uses a SDA + FF combination. In this configuration, only PAC injection equipment is added
to the existing air pollution control system, with the SDA providing flue gas cooling. SD/ESP-1
is similar to SD/FF-1 except that an ESP is used in place of an FF for particulate collection.  The
advantages are similar to those of SD/FF-1; however, larger amounts of PAC may be needed to
achieve performance levels comparable to those achieved by SD/FF-1.

8.2.2 Model Plant Descriptions

       Costs for installing and operating the Hg control technologies described in Table 8-1 are
estimated by combining these control configurations with appropriate model plant descriptions
representing plants firing different types of coal on varying boiler sizes. Eighteen different
model plant descriptions or Ascenarios@ were defined for the cost evaluation. Table 8-2 lists these
scenarios.

       Approximately 75 percent of the existing coal-fired electric utility boilers in the United
States are equipped with an ESP for the control of PM.7  The remaining boilers employ fabric
filters, particulate scrubbers, or other equipment for control of PM. Additionally, units firing
medium-to-high  sulfur coals may use FGD technologies to meet their SC>2 control requirements.
Generally, larger units firing high-sulfur coals employ wet FGD, and smaller units firing
medium-sulfur coals use SDAs.  While developing the model plant scenarios, these PM and SCh
control possibilities were taken into account.  It may be worth noting that,  since the majority of
boilers use an ESP  for PM control, most Hg control technology applications would likely take
place on such boilers and would reflect pertinent performance and costs.

       The two coal-fired boiler sizes (expressed as gross electricity output), used for the model
plant scenarios listed in Table 8-2, were selected to approximately span the range of typical
electric utility boiler sizes, and to be consistent with the model plant sizes used in previous cost
studies.1   It was  also envisioned that the use of post-combustion NOx controls (i.e.,  SCR  or
SNCR) may enhance oxidation of Hg in flue gas and result in the Acobenefit@ of
                                         3-5

-------
             Table 8-2. Matrix of model plant scenarios.
oo
Model
Plant
Scenario
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Gross
Electricity
Output
975 MWe
975 MWe
975 MWe
975 MWe
975 MWe
975 MWe
975 MWe
975 MWe
975 MWe
100 MWe
100 MWe
100 MWe
100 MWe
100 MWe
100 MWe
100 MWe
100 MWe
100 MWe
Coal Burned
Type
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Subbituminous
Subbituminous
Subbituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Subbituminous
Subbituminous
Subbituminous
Sulfur content
3%
3%
3%
0.6 %
0.6 %
0.6 %
0.5 %
0.5 %
0.5 %
3%
3%
3%
0.6 %
0.6 %
0.6 %
0.5 %
0.5 %
0.5 %
Existing Post-combustion
PM and SO2
Control Devices
CS-ESP + wet FGD
FF + wet FGD
HS-ESP + wet FGD
CS-ESP
FF
HS-ESP
CS-ESP
FF
HS-ESP
SDA + CS-ESP
SDA + FF
HS-ESP + wet FGD
CS-ESP
FF
HS-ESP
CS-ESP
FF
HS-ESP
Applicable
Mercury Control
Retrofit
Configuration(s)
(see Table 8-1)
ESP-1, ESP-3
FF-1
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
SD/ESP-1
SD/FF-1
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
Co-benefit
Cases
with SCR
SCR
SCR
SCR
















-------
increased Hg removal in wet FGD systems. This is especially relevant since many SCR
applications are expected to take place in the next few years and, in response to 862 reduction
requirements, more wet FGD systems may be installed.  However, at the time of this study, some
data on this co-benefit were available for SCR applications only.  Since SCR is a capital-
intensive technology, generally its use is more cost-effective for larger boilers.  Accordingly, in
this work, the Hg co-benefit resulting from SCR use was evaluated for model plant scenarios 1,
2, and 3, utilizing large (975 MWe) boilers and wet FGD.

8.2.3  Computer Cost Model

       The DOE/NETL developed a cost model for estimating the costs of Hg control options
for coal-fired electric utility boilers. This cost model, called the NETL Mercury Control Cost
Model, can provide capital and operating and maintenance (O&M) costs estimated in year 2000
constant dollars for the application of selected Hg control configurations to coal-fired electric
utility boilers.  The model has been used for other studies conducted to characterize the costs
associated with using PAC injection on coal-fired electric utility boilers.8 For this evaluation, the
EPA collaborated with the DOE to modify this cost model to incorporate the PAC injection rate
algorithms described in the following section. An overview of the modified version of the NETL
Mercury Control Cost Model used for this cost evaluation is presented in Appendix D to this
report. This model is hereafter referred to simply as the cost model.

8.2.4 PAC Injection Rate Algorithms
                                                                                       o
       The current understanding is that Hgp is well collected in PM or SC>2 control systems, Hg'
is not so well collected, and Hg2+ is collected to a greater or lesser degree depending on
characteristics of the control device and conditions within it.  Therefore, for a specified Hg
removal requirement, the rate of PAC injection needed will depend, in part, on the ability of
existing controls to remove the three forms of Hg. The major factor affecting the cost of PAC
injection-based technologies is the rate of PAC injection needed for the required Hg removal
efficiency.  In general, this rate depends on the time of contact between carbon particle and flue
gas, the properties of the carbon (particle size,  micropore surface area, pore size distribution, and
Hg adsorption capacity), the temperature of the flue gas, and the type of coal-fired in the boiler.
For this work, PAC injection rates at specific flue gas temperatures and Hg removal efficiencies
achieved in pilot-scale tests were fitted to the form of Equation  (8-1) with curve-fit parameters a,
b, and c (see Attachment 2 in Appendix D). For each technology for which pilot-scale test data
are available, separate correlations of Hg removal efficiency and PAC injection rate were
determined for bituminous and subbituminous coals.  These coals are predominantly used at
electric utility boilers and, therefore, were chosen for this work.


  Mercury Removal Efficiency (%) = 100	(Eq. 8-1)
                                         [PAC Injection Rate  (ib / W6 acf)+b]
                                         3-7

-------
       Equation 8-1 can be used to calculate the PAC injection rate (lb/106 acf) needed to
achieve a specified Hg removal efficiency (percent) for the control technology of interest. Note
that Hg removal efficiency (percent) is based on total Hg (the sum of Hg, Hg2+, and Hgp)
removed from the flue gas and is defined as
                                          fjr           77       \            (Eq.8-2)
    *s       r>       i rjr       fn,\  i  (Emission  -Emission t)
   Mercury Removal Efficiency (%) = lOOx-	(^LJ-
                                                Emissionin
   where:     Emission^ = total flue gas Hg concentration at the inlet to the first air pollution
              control device; and
              Emissionout = total flue gas Hg concentration at the outlet of the last air pollution
              control device.

       Preliminary analysis of the Pat in EPA ICR data 9 reflected that, at boilers firing
bituminous coals and using a CS-ESP for PM capture, higher levels (more than 50 percent) of Hg
were being removed with fly ash than were found in earlier pilot-scale tests (see Attachment 2 in
Appendix D).  Accordingly, for each of technologies ESP-1, ESP-3, ESP-4, and ESP-6, two
separate sets of correlations, relating PAC injection rate (lb/106 acf) to Hg removal efficiency
(percent), were created for use with bituminous-coal-fired boilers. The first of these sets,
hereafter referred to as the pilot-scale PAC injection rate, was derived using presently available
pilot-scale test data. The other set, hereafter referred to as the ICR/pilot-scale PAC injection rate,
was derived using preliminary ICR results for fly ash capture of Hg (i.e., no PAC injection) and
pilot-scale results for PAC injection.

       Note that the above data-fitting procedure resulted in  correlations of PAC injection rate
(lb/106 acf) versus Hg removal efficiency (percent), as a function of flue gas temperature, for all
of the technologies except:  (1) FF-1, FF-2, and SD/FF-1, applied on boilers firing bituminous
coals, for which no data  are available; (2) HESP-1, applied on boilers  firing either bituminous or
subbituminous coals, for which no data are available; and (3) ESP-7, applied on boilers firing
either bituminous or subbituminous coals. The only available data on ESP-7 are from a pilot-
scale application on a boiler firing a bituminous coal.10 Since these data reflect that more than 90
percent of the Hg can be removed by injecting relatively small amounts of PAC with lime, in this
work, application of ESP-7 was evaluated at 90 percent Hg removal efficiency in a sensitivity
analysis.

       The algorithms describing sorbent injection rates for various technologies can be found  in
Attachment 2 in Appendix D. The PAC injection rate algorithms could not be determined for the
retrofit configurations defined for model plant scenarios 2, 3, 5, 6, 9, 11, 12, 14, 15, and 18.  As
such, costs  for these model plant configurations cannot be estimated using the cost model.

-------
8.2.5  Cost Estimate Assumptions

       To estimate the costs for the model plant configurations using the cost model, the
following specifications were used.

   (1) Mercury concentration in the flue gas for each model plant scenario is 10 jig/Nm3.  This
       concentration has been used in previous cost studies1'8 and is in the range of mean
       concentrations (1.7-50.1 |ig/dscm) determined from ICR data for pulverized-coal-fired
       electric utility boilers equipped with different air pollution controls.9 Note also that the
       corresponding median and mean concentrations are 9.1 and 11.4 |ig/dscm, respectively.

   (2) For each of retrofit configurations ESP-1, ESP-3, ESP-4, and ESP-6, two separate sets of
       correlations, relating PAC injection rate (lb/106  acf) to Hg removal efficiency (percent),
       were created for use with bituminous-coal-fired boilers.  The first of these sets, hereafter
       referred to as the pilot-scale PAC injection rate, was derived using presently available
       pilot-scale test data.  The other set, hereafter referred to as the ICR/pilot-scale PAC
       injection rate, was derived using preliminary EPA ICR results for fly ash capture of Hg
       (i.e.,  no PAC injection) and pilot-scale results for PAC injection. Accordingly, two sets
       of cost estimates for applying retrofit configurations ESP-1, ESP-3, ESP-4, and ESP-6
       were made: one estimate used the pilot-scale PAC injection rate, and the other used the
       ICR/pilot-scale PAC injection rate.

   (3) PAC injection rate correlations generally reflect that PAC injection requirements increase
       nonlinearly with increases in Hg removal efficiency.  To characterize the impact of this
       behavior, wherever possible,  model plant costs were estimated for Hg removal
       efficiencies of 60, 70, 80, and 90 percent.

   (4) In general, for any given Hg removal requirement, the PAC injection rate decreases if the
       temperature of the flue gas is lowered.  For this  reason, the flue gas is cooled by water
       injection in some of the retrofit configurations (see Table 8-1).  However, injecting water
       into an acidic flue gas can lead potentially to corrosion of downstream equipment.  To
       avoid this corrosion, an approach to acid dew point (ADP) of 18 F was used for the
       retrofit configurations with spray cooling (i.e., ESP-4, ESP-6, ESP-7, and FF-2).11  For
       these retrofit configurations, the extent of SC provided was determined based on the
       temperature of the flue gas before cooling  and the temperature nearest to the above
       approach to ADP for which a PAC injection rate correlation was available.  Note that, in
       the high-sulfur coal applications with relatively  high ADPs, this constraint resulted in no
       SC if the SO2 control technology was wet FGD.  However, in applications using SDAs
       for SC>2 control, SC is inherent and acid gases are removed prior to PAC injection;
       therefore, this constraint was not applied.

   (5) No data are currently available for recycling of sorbent in technology applications
       utilizing PAC injection and PFF.  Accordingly,  no sorbent recycle was used in retrofit
       configurations ESP-3 and ESP-6.
                                         8-9

-------
   (6) Mercury speciation in the flue gas from bituminous-coal-fired boilers is assumed to be 70
       percent of the total Hg being oxidized, with 30 percent being Hg. The corresponding
       assumption for boilers firing subbituminous coals is 25 percent oxidized with 75 percent
       Hg.  These Hg speciation percentages were determined from a preliminary analysis of
       ICR data (see Attachment 2 in Appendix D).

   (7) Wet FGD systems are assumed to remove 100 percent of Hg2+ and no Hg.  This is based
       on the fact that mercuric chloride (the assumed major oxidized species) is soluble in
       water, while Hg is insoluble. It is anticipated that ongoing research on wet scrubbers
       will result in improved performance through the use of reagents or catalysts to convert Hg
       to chemical compounds that are soluble in aqueous-based scrubbers.

   (8)  Use of SCR is assumed to increase Hg2+ content in flue gas by 35 percent for both
       bituminous- and subbituminous-coal-fired boilers.  This increase in mercury oxidation
       was determined from a preliminary analysis of ICR data as follows. As explained above,
       oxidized mercury content in flue gas from bituminous-coal-fired boilers is assumed to be
       70 percent. Also, ICR data revealed that SCR application with SDA at one  plant firing
       bituminous coal resulted in greater than 95 percent mercury removal. It is hypothesized
       that virtually all of the mercury removed at this plant was oxidized mercury. Based on
       these considerations, it is assumed that SCR increases oxidized mercury content by
       35 percent (also see Attachment 2 in Appendix D). Currently, research and development
       efforts are underway to investigate the effects of SCR on Hg oxidation. A more mature
       set of findings regarding SCR impacts are expected from these efforts.

   (9) For each of the model plant scenarios, a plant capacity factor of 65 percent was used.

   (10) The cost of PAC is assumed to be $1.00 per kilogram.12

   Other specifications are described in Attachments 1, 2, and 3 in Appendix D.
8.3 Estimated Costs of Reducing Mercury Emissions

       This section describes the estimates of total annual cost determined using the cost model
for application of Hg controls to those model plant scenarios for which PAC injection rate
algorithms could be determined (i.e., model plant scenarios 1, 4, 7, 8, 10, 13,  16, and 17). It is
important to note that cost estimates presented in this section are based on currently available
data and, as explained later, may be improved with R&D efforts and as long-term operating data
from full-scale demonstrations become available.

       In general, capital costs of PAC injection-based Hg control technologies comprise a
relatively minor fraction of the total annual costs of these technologies; the major fraction is
associated with the costs related to the use of PAC.12 As an example, for application of SC+PAC
injection (ESP-4) to achieve 80 percent Hg reduction on a 975-MWe boiler firing bituminous
                                        8-10

-------
coal and using an ESP, the capital cost contributes about 23 percent of the total annual cost.
Therefore, for such technologies, the cost assessment should be based on total annual costs.
Accordingly, total annual costs of controlling Hg emissions from coal-fired electric utility boilers
are examined in this section. These costs include  annualized capital charge, annual fixed
operation and maintenance (O&M) costs, and annual variable O&M costs.  Note that Reference
12 provides an examination of the contribution of various cost elements, including cost of PAC,
to total annual cost of Hg controls.

8.3.1 Bituminous-coal-fired Boiler Using CS-ESP

       Several of the Hg control technologies listed in Table 8-1 are potential options for
reducing Hg emissions from a electric utility boiler that fires bituminous coal and  already is
using an ESP for PM control.  For boilers firing low-sulfur bituminous coals, these options
include  configurations ESP-4 (SC + PAC injection) and ESP-6 (SC + PAC injection + PFF). For
large boilers firing high-sulfur bituminous coals, the options include configurations ESP-1 (PAC
injection + wet FGD) and ESP-3 (PAC injection + PFF + wet FGD). For smaller boilers
(typically less than 300 MW), these options include configuration SD/ESP-1 (SDA + PAC
injection + ESP). For each of these cases, cost estimates were determined using the cost model.

       Table 8-3 presents the estimated total annual Hg control  costs for a bituminous-coal-fired
boiler with existing CS-ESP. The table presents two sets of cost estimates. The first set of
estimates was made based on levels of Hg capture on fly ash using PAC injection  rates derived
from the available pilot-scale test data.  A subsequent review of the Part in EPA ICR data
(discussed in Section 6.2), however, suggests that levels of Hg capture higher than those
measured in the pilot-scale tests may be occurring. Consequently, the cost estimates based solely
on pilot test data for Hg control technologies applied to bituminous-coal-fired boilers using ESP
may be  overstating the costs. Therefore, a second set of estimates is presented based on the
preliminary ICR results for fly ash capture of Hg (i.e., no PAC injection) in combination with the
pilot-scale results for PAC injection.

       For ESP-4 applied to low-sulfur (0.6 percent) bituminous coal and using pilot-scale PAC
injection rates, the estimated total annual cost ranges from 2.81 mills/kWh for a 100-MWe boiler
removing 90 percent of the total Hg to 0.53 mill/kWh for a 975-MWe boiler removing 60 percent
of the total Hg.  The corresponding costs with ICR/pilot-scale PAC injection rates are 1.65
mills/kWh for the 100-MWe boiler and 0.24 mill/kWh for the 975-MWe boiler.

       In general, these results reflect that, for a given boiler, the total annual cost increases non-
linearly with increases in the Hg reduction requirement in concert with the behavior of the PAC
injection rate algorithms (see Attachment 2 in Appendix D). A comparison of results obtained
with pilot-scale and ICR/pilot-scale PAC injection rates also indicates that research and
development efforts aimed at ensuring broad availability of relatively high levels of fly ash
capture  of Hg have the potential of providing significant reductions in Hg control  costs.
                                         8-11

-------
Table 8-3.  Estimated total annual mercury control costs for bituminous-coal-fired
boiler with existing CS-ESP.
Model
Plant
Scenario
1
4
10
13
Model Plant
Size
975 MWe
975 MWe
100 MWe
100 MWe
Coal
Sulfur
Content
high sulfur
(3 %)
low sulfur
(0.6 %)
high sulfur
(3 %)
low sulfur
(0.6 %)
Mercury
Control
Retrofit
Configuration
(see Table 8-1)
ESP-1
ESP-3
ESP-4
ESP-6
SD/ESP-1
ESP-4
ESP-6
Mercury
Capture
Efficiency
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
Total Annual Mercury Control
Costs
' Ml/kWh
Pilot-scale data1
2.594
0.727
0.006
0.006
2.086
1.501
1.273
0.006
1.966
1.017
0.696
0.533
2.381
1.817
1.625
1.528
1.925
1.197
0.945
0.815
2.810
1.793
1.442
1.262
4.966
3.783
3.170
2.957
i aj
ICR/pilot data b
0.427
0.006 d
0.006 d
0.006 d
1.416
0.006 d
0.006 d
0.006 d
0.883
0.464
0.319
0.240
1.735
1.485
1.397
1.353
1.094
0.759
0.637
0.008
1.647
1.184
1.018
0.922
3.080
2.798
2.695
2.637
  (a)   Mercury capture efficiency of Hg controls based on fly ash using PAC injection rates derived from the available pilot-scale
        test data.
  (b)   Mercury capture efficiency of mercury controls based on preliminary EPA ICR results for fly ash capture of Hg (i.e., no PAC
        injection) in combination with the pilot-scale results for PAC injection.
  (c)   The cost of monitoring Hg emissions is 0.006 mill/kWh.  Based on 70% of total Hg being oxidized, 0% Hg removal with fly
        ash, and all Hg2* being removed in wet FGD scrubber system, a minimum of 70% of total Hg is captured by existing control
        system(s).  Therefore, add-on of PAC injection is not needed to meet target Hg control efficiency of 70% or lower.
  (d)   The cost of monitoring Hg emissions is 0.006 mill/kWh.  Based on 70% of total Hg being oxidized, 58% mercury removal
        with fly ash, and all Hg2* being removed in wet FGD scrubber system, approximately 87% of total Hg  is captured by
        existing control system(s). Therefore, Add-on of PAC injection is not needed to meet target mercury control efficiency.
                                                  8-12

-------
       Another option for boilers firing low-sulfur bituminous coals is to utilize ESP-6 for Hg
control. For this option, using the pilot-scale PAC injection rates, the estimated total annual cost
ranges from 4.966 mills/kWh for the 100-MWe boiler removing 90 percent of total Hg to 1.528
mills/kWh for the 975-MWe boiler removing 60 percent of total Hg. The corresponding costs
with ICR/pilot-scale PAC injection rates are 3.08 mills/kWh for the 100-MWe boiler and 1.353
mills/kWh for the 975-MWe boiler. In general, these results reflect that the ESP-6 control option
is more expensive than ESP-4 because of the capital cost associated with the PFF. To make this
control option more cost-effective, R&D efforts are needed to develop less expensive PFF
designs and high capacity sorbents, which may be recycled sufficiently to improve sorbent
utilization.

       As seen in Table 8-3, for ESP-1 application on a large (975-MW) high-sulfur bituminous-
coal-fired boiler that uses wet FGD for 862 control, using pilot-scale PAC injection rates, the
estimated total annual cost ranges from 2.594 mills/kWh for removing 90 percent of the total Hg
to 0.006 mill/kWh (cost of monitoring of Hg emissions) for removing 70 percent of the total Hg.
The costs with ICR/pilot-scale PAC injection rates are 0.427 mill/kWh for 90 percent removal
and 0.006 mill/kWh for about 87 percent removal.  Note that, with the assumptions of this work,
a minimum of 70 percent of total Hg is removed in wet  FGD systems if no Hg is removed with
fly ash (pilot-scale test results) and a minimum of about 87 percent is removed if about
58 percent of Hg is removed with fly ash (preliminary ICR data analyses results). These results
reflect that, if significant amounts of Hg can be captured along with fly ash in ESPs and in wet
FGD systems, costs of achieving high levels of Hg removal would be quite low. Considering
these results, R&D efforts are needed to ensure that these Hg capture mechanisms are broadly
available.

       Another option for large boilers firing high-sulfur bituminous coals and using wet FGDs
is to utilize ESP-3 for Hg control.  Using this option on  a 975-MWe boiler, with pilot-scale PAC
injection rates, the estimated total annual cost ranges from 2.086 mills/kWh for removing
90 percent of the total Hg to 1.273 mills/kWh for removing 70 percent of the total Hg.  The costs
with ICR/pilot-scale PAC injection rates are 1.416 mills/kWh for removing 90 percent of the
total Hg and 0.006 mill/kWh for about 87 percent removal.  Interestingly, this control option is
more cost-effective than the one using PAC injection (ESP-1) at 90 percent Hg removal.
However, at or below 80 percent removal, this option is more expensive because the PAC
injection rate decreases more rapidly than the capital cost of PFF.  It may be possible to make this
option competitive across a  wide range of Hg removal efficiencies by conducting R&D  efforts
directed towards reducing both PFF capital cost and operating cost through sorbent recycling.

       Finally, as seen in Table 8-3,  for ESP-1 application on a relatively small boiler (100-MW)
that fires a high-sulfur bituminous coal and uses an SDA for SC>2 control, with pilot-scale PAC
injection rates, the estimated total annual cost ranges from  1.925 mills/kWh for removing
90 percent of the total Hg to 0.815 mill/kWh for removing 60 percent of the total Hg. The
corresponding costs with ICR/pilot-scale PAC  injection rates are 1.094 and 0.008 mills/kWh,
respectively.  A significant increase in costs is observed on increasing the Hg control requirement
from 80 to 90 percent. Again, considering the  differences in total annual costs obtained using
                                        8-13

-------
ICR/pilot-scale and pilot-scale PAC injection rates, R&D efforts are needed, aimed at providing
broad availability of relatively high levels of fly ash capture of Hg.

8.3.2  Subbituminous-coal-fired Boiler Using CS-ESP

       Shown in Table 8-4 are two potential options to reduce total Hg emissions from boilers
that fire subbituminous coals and use ESPs for PM control. These options include SC + PAC
injection (ESP-4) and SC + PAC injection + PFF (ESP-6).

       For ESP-4 application on boilers firing subbituminous coals, estimated total annual costs
range from 3.232 mills/kWh for a 100-MWe boiler removing 90 percent of the total Hg to 0.473
mill/kWh for the 975-MWe boiler removing 60 percent of the total Hg. Further, total annual cost
appears to drop sharply as the Hg removal requirement is reduced from 90 to 80 percent due to
the nonlinear nature of the PAC injection rate algorithms.

       For ESP-6 application on boilers firing subbituminous coals, the estimated total annual
cost ranges from 2.754 mills/kWh for a 100-MWe boiler removing 90 percent of the total Hg to
1.405 mills/kWh for the 975-MWe boiler removing 60 percent of the total Hg.  Interestingly, this
control option is more cost-effective than the one using SC + PAC injection (ESP-4)  at 90
percent Hg removal. However, at or below 80 percent removal, this option is more expensive
because the PAC injection rate decreases more rapidly than capital costs of PFF. These results
again indicate possibilities of making this option competitive by reducing both the PFF capital
cost and operating cost through sorbent recycling.

       A comparison of the results shown in Tables 8-4 and 8-3 reveals that applications  of
SC+PAC injection (ESP-4) to achieve high Hg reductions could cost more for boilers firing
subbituminous coals compared to boilers firing bituminous coals.  Further, in general, relatively
few FGDs would be used on subbituminous-coal-fired boilers. Considering these factors,
research and development efforts are needed to ensure that cost-effective control of Hg is
achieved at these boilers.

8.3.3 Subbituminous-coal-fired Boilers Using FF

       As seen in Table 8-5,  for boilers firing subbituminous coals and utilizing SC + PAC
injection (FF-2) for Hg control, the estimated total annual cost ranges from 1.120 mills/kWh for
a 100-MWe boiler removing 90 percent of the total Hg to 0.219 mill/kWh for the 975-MWe
boiler removing 60 percent of the total Hg.  These cost estimates reflect that the combination of
SC + PAC injection + FF is very efficient in removing Hg from the boiler flue gas.
                                        8-14

-------
Table 8-4. Estimated total annual mercury control costs for
subbituminous-coal-fired boiler with existing CS-ESP.
Model
Plant
I.D.
7
16
Model
Plant
Size
975 MWe
100MWe
Coal
Sulfur
Content
0.5 %
0.5 %
Mercury
Control
Retrofit
Configuration
(see Table 8-1)
ESP-4
ESP-6
ESP-4
ESP-6
Mercury
Capture
Efficiency
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
Total Annual Mercury
Control Costs
(mills/kWh generated)
2.384
1.150
0.731
0.473
1.444
1.419
1.410
1.405
3.232
1.915
1.460
1.174
2.754
2.723
2.712
2.703
                                  8-15

-------
Table 8-5. Estimated total annual mercury control costs for
subbituminous-coal-fired boiler with existing FF.
Model
Plant
I.D.
8
17
Model
Plant
Size
975 MWe
100MWe
Coal
Sulfur
Content
0.5 %
0.5 %
Mercury
Control
Retrofit
Configuration
(see Table 8-1)
FF-2
FF-2
Assumed
Mercury
Capture
Efficiency
90%
80%
70%
60%
90%
80%
70%
60%
Total Annual Mercury
Control Costs
(mills/kWh generated)
0.423
0.299
0.226
0.219
1.120
0.977
0.888
0.879
                               8-16

-------
8.3.4  Coal-fired Boilers Using SCR for NOX Control

       As mentioned before, this work assumes that flue gas resulting from bituminous coal
combustion has an oxidized-Hg content of 70 percent, and SCR augments this by 35 percent.
This leads to a total of 94.5 percent of total Hg being Hg2+ after SCR. Using the results of ICR
data analysis, about 58 percent of the total Hg is captured along with fly ash in an ESP, and all of
the remaining Hg2+ is captured in a wet FGD system.  Thus, a total Hg capture of 97.6 percent is
achieved.  The cost of this removal is 0.006 mill/kWh, which is simply the cost of monitoring the
Hg emissions.  On the other hand, using pilot-scale test results, no Hg is captured along with fly
ash in ESPs and all of the Hg2+ is captured in wet FGDs. Therefore, a total Hg capture of 94.5
percent is achieved, and the cost of this removal is again 0.006 mill/kWh; i.e., the cost of
monitoring the Hg emissions.

       It may be mentioned that, since the majority of boilers use cold-side ESPs, the most
frequently occurring costs would be those related to technology applications on such boilers.
Based on this work, these costs would range between 0.727 and 1.915 mills/kWh.
8.4 Impacts of Selected Variables on Mercury Control Costs

       The impacts of certain selected variables on Hg control costs were examined by
performing sensitivity analyses using the cost model with the pilot-scale PAC injection rates.
These analyses are summarized below with a more detailed discussion of the analyses presented
in Attachment 3 in Appendix D.  A model plant size of 500 MWe was used in these analyses.
This size was selected because it is approximately the midpoint of the range of 100- and 975-
MWe boiler sizes used for the model plant cost estimates.

       Note that, in general, the cost of sorbent constitutes an important component of the total
annual  cost. This cost is likely to fall in the future because of the many active research programs
aimed at producing low-cost sorbents. While a sensitivity analysis with respect to this cost was
not conducted in this work, Reference 12 provides an examination of the contribution of various
cost elements, including the cost of PAC, to the total annual cost of Hg controls.

8.4.1 A cid Dew Point Approach Setting

       Adsorption of Hg on PAC is dependent on the temperature of the flue gas at the point
where the PAC is injected. The acid dew point (ADP) is the temperature of the flue gas at which
acidic components will  condense. It is important to keep the flue gas temperature above the ADP
because below the ADP, acid mist will form in the flue gas and corrode the downstream ducting
and control equipment.  In determining the estimates of Hg control costs for the model plant
scenarios presented in Section 8.2, the approach to ADP was kept at 10 C (18 F), resulting in a
flue gas temperature of ADP + 10 C. Some investigators have expressed concern that, in some
cases, this approach may be too low to prevent corrosion of downstream equipment, especially
for those Hg control retrofit configurations where spray cooling is used.  To test the sensitivity of
                                        8-17

-------
the cost estimate results to the ADP approach setting, annual Hg control costs were computed
using the cost model for a set of Hg control applications with SC using a boiler size of 500 MWe
and burning low-sulfur coals for the nominal flue gas temperature of ADP + 10 C (ADP + 18 F)
and a higher temperature of ADP + 22.2 C (ADP + 40 F).

       As seen in Table 8-6, for a 500-MWe boiler firing low-sulfur bituminous coal and using
ESP-4, the total annual cost increase ranges from 126.3 to 38.2 percent. Again for the same
boiler using ESP-6, the cost increase ranges from 18.8 to 2 percent.  Interestingly, the results for
subbituminous coal presented in Table 8-7 reflect that the total annual  cost decreases with an
increase in approach to ADP.  This is due to a significant decrease in water injection
requirements, while PAC injection does not increase much to provide the required Hg removal.
These results indicate that, for a bituminous-coal-fired boiler using a CS-ESP, changes in the
ADP approach can influence estimated costs significantly. However, the same is not true for
subbituminous-coal-fired boilers.

8.4.2   PAC Recycle

       As discussed in Section 8.2, estimates of Hg control costs for model plant scenarios using
PFF obtained using no sorbent recycle, are, in general, higher than those of other options. A
sensitivity analysis was conducted to examine the impact of increasing PAC utilization in ESP-3
and ESP-6 retrofit configurations on associated costs. Specifically, cost estimates were obtained
with 20 percent of the PAC being recycled in the following applications evaluated with a 500-
MWe boiler: model plant  1 retrofitted with ESP-3; model plant 4 retrofitted with ESP-6; and
model plant 7 retrofitted with ESP-6.

       The results shown in Table  8-8 reflect that a recycle rate of 20 percent does not have
much of an impact on total annual costs estimated by the cost model. This is because the capital
cost of the new PFF is the dominant cost component. To utilize the benefits of increased PAC
utilization, higher recycle rates would  be needed, but such rates would require that sorbents used
have relatively high adsorption capacities.

8.4.3  Increased Flue Gas Residence Time

       Adsorption of Hg on PAC is dependent on the time of contact between the flue gas and
the PAC. In general, about half of the existing electric utility boilers have a flue gas residence
time in the duct of 1 second, and about 30 percent have a time of 2 seconds.13 Although it is not
entirely clear at this time as to how much time is needed for particular  levels of Hg capture, in
this sensitivity analysis the impact of adding ductwork to increase the flue gas residence time by
1 second on the cost of Hg control was evaluated as a conservative measure. This analysis was
conducted using the model plant 4 with a 500-MWe  boiler retrofitted with ESP-4.

       The results shown in Figure 8-1 reflect that the impact of adding ductwork on the total
annual cost is quite small. The increase in cost ranges from 16.4 percent at the lowest cost
                                        8-18

-------
              Table 8-6.  Impact of acid dew point setting on annual mercury control costs for a 500-MWe electric
              utility boiler burning bituminous coal.
Coal Burned in Boiler Unit
Type
Bituminous
Bituminous
Sulfur
Content
low sulfur
(0.6 %)
low sulfur
(0.6 %)
Existing
Post-combustion
PM and SO2
Control Devices
CS-ESP
CS-ESP
Mercury Control
Retrofit
Configuration
(see Table 8-1)
ESP-4
ESP-6
Mercury
Capture
Efficiency
90%
80%
70%
60%
90%
80%
70%
60%
Total Annual Mercury Control Costs
(mills/kWh generated)
Acid dew point
+ 18F
2.095
1.132
0.804
0.637
2.650
2.075
1.879
1.779
Acid dew point
+ 40F
4.741
2.282
1.451
1.030
3.263
2.307
1.982
1.816
oo

VO

-------
              Table 8-7. Impact of acid dew point setting on annual mercury control costs for a 500-MWe electric

              utility boiler burning subbituminous coal.
Coal Burned in Boiler Unit
Type
Subbituminous
Subbituminous
Subbituminous
Sulfur
Content
low sulfur
(0.5 %)
low sulfur
(0.5 %)
low sulfur
(0.5 %)
Existing
Post-combustion
PM and SO2
Control Devices
CS-ESP
CS-ESP
FF
Mercury Control
Retrofit
Configuration
(see Table 8-1)
ESP-4
ESP-6
FF-2
Mercury
Capture
Efficiency
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
Total Annual Mercury Control Costs
(mills/kWh generated)
Acid dew point
+18 F
2.513
1.261
0.835
0.571
1.693
1.667
1.658
1.652
0.520
0.392
0.315
0.308
Acid dew point
+ 40F
2.392
1.140
0.714
0.478
1.683
1.597
1.567
1.550
0.399
0.0271
0.216
0.197
oo
to
o

-------
                Table 8-8.  Effect of PAC recycle on annual mercury control costs for a 500-MWe electric utility boiler
                burning bituminous coal.
Coal Burned in Boiler Unit
Type
Bituminous
Subbituminous
Sulfur
Content
high sulfur
(3 %)
low sulfur
(0.6 %)
low sulfur
(0.5 %)
Existing
Post-combustion
PM and SO2
Control Devices
CS-ESP + wet FGD
CS-ESP
CS-ESP
Mercury Control
Retrofit
Configuration
(see Table 8-1)
ESP-3
ESP-6
ESP-6
Assumed
Mercury
Capture
Efficiency
90%
80%
70%
60%
90%
80%
70%
60%
90%
80%
70%
60%
Total Annual Mercury Control Costs
(mills/kWh generated)
No Recycle
2.324
1.727
0.006"
0.006 a
2.650
2.075
1.879
1.779
1.693
1.667
1.658
1.652
20% Recycle
2.173
1.686
0.006 a
0.006 a
2.457
1.989
1.829
1.747
1.686
1.664
1.657
1.652
oo
I
to
                    (a) The cost of monitoring of Hg emissions is 0.006 mill/kWh. Add-on of PAC injection is not needed to meet target Hg control efficiency.
                    Based on 70% of total Hg being oxidized, 0% Hg removal with fly ash, and all Hg2+ being removed in wet FGD scrubber system, a minimum
                    of 70% of total Hg is captured by existing control system.

-------
MODEL PLANT 4, ESP-4, 500 MW, Bituminous Coal, 0.6% Sulfur,
With and Without Added Ductwork
i cn
Total Annual Cost, mills/kWh
2OC
.23
9 on
1 7S
1 'SO -
1 ?>
i on
o 7s
o so
n T

9
//
1
J
Jf
^3&^
ICZZ^' 


 * W/O Ductwork
  W/ Ductwork

40 50 60 70 80 90 100
Total Hg Removed, %
Figure 8-1.  Change in total annual cost resulting from addition of ductwork to
provide additional residence time.
                                8-22

-------
0.535 mill/kWh to 4.3 percent at the highest cost of 2.095 mills/kWh. Based on this analysis, it
appears that addition of ductwork is not a sensitive cost parameter.

8.4.4 Use of Composite PA C and Lime Sorbent

       As discussed above, high levels of Hg have been removed in pilot-scale tests using lime
and PAC with PFF.10 To assess the potential economic impact, this analysis was based on
removing 90 percent of Hg from model plant 4 with a 500-MWe boiler retrofitted with ESP-7
and using a composite PAC-lime sorbent, with a PAC-to-lime mass ratio of 2:19. The results of
this analysis shown in Figure 8-2 reflect that use of the composite sorbent lowers the total annual
cost by 34.7 to 38.1 percent.
8.5 Cost Indications for Other Model Plant Scenarios

       As discussed in Section 8.2.4, since data are not available on Hg control technology
applications involving HS-ESPs or boilers firing bituminous coals and using FFs, PAC injection
rate algorithms could not be developed for these applications.  Consequently, cost estimates for
these applications (i.e., model plant scenarios 2, 3, 5, 6, 9, 11,  12, 14,  15, and 18) could not be
obtained using the cost model. In this section, estimates of cost for these latter applications are
developed using the estimates described in previous sections.

       Cooling the flue gas after the air preheater, injecting PAC, and collecting the spent PAC
in a downwind PFF may achieve Hg control on boilers equipped with HS-ESPs. This
configuration is identical to ESP-6, with only the location of the ESP being different. Therefore,
Hg reduction performance and costs should be similar to those found for ESP-6. However, on
boilers equipped with HS-ESPs and firing high-sulfur bituminous coals, application of SC may
not be possible due to corrosion concerns; for such boilers, Hg control may be achieved using
ESP-3.  With these considerations, cost of Hg control technology applications involving
HS-ESPs are: model plant 3 costs are the  same  as those for model plant 1 with ESP-3; model
plant 6 costs are the same as those for model plant 4 with ESP-6;  model plant 9 costs are the
same as those for model plant 7 with ESP-6; model plant 12 costs are the same as those for
model plant 12 with ESP-3; model plant 15 costs are the same as  those for model plant 13 with
ESP-6; and model plant 18 costs are the same as those for model  plant 16 with ESP-6.

       The combination of PAC injection and FF provides better sorbent utilization than the
corresponding PAC injection and ESP combination because FF provides added residence time
and a contact bed for increased adsorption of Hg.  This superior performance of FF has been
validated in full-scale tests on MWCs and pilot-scale tests on coal-fired combustors. Field tests
have shown that it takes 2 to 3 times more PAC to achieve the same performance on MWCs
equipped with SDAs and ESPs than with  SDAs and FFs.14 As a result of increased sorbent
utilization, the total annual cost of a PAC  injection and FF application would be lower than that
of the corresponding PAC injection and ESP combination. An analysis  of cost data for ESP-4
applications on Model plant scenarios 7 and 16 and FF-2 applications on Model plant scenarios 8
and 17 (see Tables 8- 4 and 8-5) reveals that, in reducing Hg emissions between
                                         8-23

-------
           MODEL PLANT 4, ESP-6 & 7, 500 MW, Bituminous Coal, 0.6% Sulfur,

                    Comparison of PAC and Lime-PAC Sorbents
       1
       In
          2.75
          1.50
       =  2.25
O
O

"ra
3
c
c


"ra

5  1
          !.00
           ,75
          1.50
             40     50     60     70     80

                        Total Hg Removed, %
                                           T
                                                    - ESP-6, AC sorbent

                                                    ESP-7, Lime-AC Sorbent
                                            90     100
Figure 8-2. Change in total annual cost resulting from use of a composite

PAC-lime sorbent instead of PAC.
                                    8-24

-------
     60 and 90 percent using FFs instead of ESPs, the total annual cost decreases by an average
     of about 70 percent for the 975-MWe boiler and 45 percent for the 100-MWe boiler.
     Considering these numbers, an average about 58 percent decrease in total annual cost may
     be expected if FFs are used in place of ESPs for Hg removal.
8.6 Projection of Future Mercury Control Costs

       Shown in Table 8-9 is a summary of costs of Hg control technology applications
developed in previous sections.  This summary presents current estimates of costs developed
using the pilot-scale PAC injection rates and projections based on use of potentially more
effective sorbents. The following assumptions were used in developing these estimates.

   (1) A Hg capture of 80 percent is obtained in technologies using ESPs and 90 percent in
       technologies using FFs.  This assumption is based on the consideration that it is more
       cost-effective to remove Hg on boilers equipped with FFs.

   (2) For technology applications on bituminous-coal-fired boilers using ESPs, current
       estimates are based on levels of Hg capture on fly  ash derived from pilot-scale test data.
       ICR data, however, reflect that levels of capture higher than those seen in pilot-scale tests
       may be occurring. In this light, these cost estimates may be conservative.

   (3) Current estimates for boilers using HS-ESPs, as well as boilers firing bituminous coals
       and using FFs, are based on the information presented  in Section 8.4.  For other cases,
       these estimates are based on the results obtained with the cost model.

   (4) Results of sensitivity analyses presented in Section 8.3, especially impacts of increase in
       approach-to-ADP at boilers firing bituminous coals and using ESP-4, are not included in
       the current estimates because the estimates are preliminary in nature and because it is not
       clear whether such an increase is broadly applicable. Generally an approach of ADP +
       18 F is considered to be optimum.11 Where a higher approach is desired, use of ESP-6
       may be less expensive.

   (5) Finally, sensitivity analyses reflect that using a potentially more effective sorbents (e.g., a
       composite PAC + lime) may remove Hg cost effectively.  Although some data are
       available for applications using a PAC + lime sorbent with PFF,  there does not appear to
       be any significant technical constraint to using such sorbents in other applications.
       Consequently, projected Hg control costs are based on using such more-effective
       sorbents. Specifically, sensitivity analyses reflected that a 35 to 40 percent decrease in
       total annual cost might be experienced if a PAC + lime sorbent is used. Since these
       indications are based on using PFF, the capital cost of which is a dominant component of
       the corresponding total annual cost, in applications without PFF  greater benefits may be
       possible. Considering these factors, a 40 percent reduction in total annual cost is used to
       arrive at the cost projections shown in Table 8-9. Note that current research to develop
       more effective sorbents should in the near future provide such sorbents.
                                           8-25

-------
                    Table 8-9.  Projected future mercury control costs.
oo
to
Coal Burned
Type
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Subbituminous
Subbituminous
Subbituminous
Sulfur
Content
3%
3%
3%
0.6 %
0.6 %
0.6 %
0.5 %
0.5 %
0.5 %
Existing
PM and SO2
Control Devices (a)
CS-ESP + wet FGD
FF + wet FGD
HS-ESP + wet FGD
CS-ESP
FF
HS-ESP
CS-ESP
FF
HS-ESP
Applicable
Mercury Control(s)
(see Table 8-1)
ESP-1, SD/ESP-1
FF-1
ESP-3
ESP-4
FF-2
HESP-1
ESP-4
FF-2
HESP-1
Total Annual Mercury Control Cost Range
(mills/kWh generated) (b)
Current Estimate
(from Tables 8-3 and 8-4)
0.727-1.197
0.305-0.502
1.501 (b)
1.017-1.793
0.427-0.753
1.817-3.783
1.150-1.915
0.423-1.120
1.419-2.723
Projected Estimate
0.436-0.718
0.183-0.301
0.901 -NA(c)
0.610-1.076
0.256-0.452
1.090-2.270
0.690-1.149
0.254-0.672
0.851 -1.634
                      (a)  CS-ESP = cold-side electrostatic precipitator
                          HS-ESP = hot-side electrostatic precipitator
                          FF = fabric filter
                          wet FGD = wet flue gas desulfurization system
                      (b)  Boiler size range is 975-100 MW.
                      (c)  NA = not available

-------
The 40 percent reduction in cost described above simply indicates the potential cost savings that
may be possible once more effective sorbents are available.

       Earlier, EPA=s Office of Air and Radiation (OAR) conducted preliminary analyses examining
potential pollution control options for the electric utility power industry to lower the emissions of its
most significant air pollutants, including Hg.15 These analyses were conducted using the Integrated
Planning Model (IPM),16 which was supplemented with previously developed estimates of
performance and cost of Hg emission control technologies. These estimates were based on using lime
with PAC injection.  In these previous estimates, Hg control costs ranged from 0.17 to 1.76
mills/kWh for boilers ranging in size from 100 to 1000 MW.12 As seen from Table 8-10, the range
of projected cost estimates (i.e., 0.183 to 2.27 mills/kWh) is comparable to the range of previously
developed estimates.

       Finally, it is noted that, in the wake of recent NOx control regulations, many plants are
planning to install  SCRs. As discussed in Section 8.3.2, Hg control costs may be negligible at
bituminous-coal-fired plants using SCR and wet FGD where Hg2+ content in the flue gas is 95 percent
and higher as a result of fuel and combustion conditions and an increase in Hg oxidation due to SCR.
8.7 Comparison of Mercury and NOx Control Costs

       An understanding of Hg control costs may be gained by comparing them with costs of
currently used controls for NOx. In the U.S., commercial NOx control technologies are being used to
comply with emission reduction requirements. Therefore, the costs associated with these NOx
controls are being experienced at full-scale applications. A comparison of Hg control costs with costs
of currently used NOx controls provides insight into how far or near the Hg control costs are from
costs that are presently being experienced at full-scale applications to control another pollutant.

       Table 8-10 presents the ranges of total annual costs in 2000 constant dollars for the Hg
controls examined in this work and for two currently used NOx control technologies; i.e., low NOx
burner (LNB) and SCR.  The LNB and SCR costs were derived from the information in Reference
16. The NOx control costs presented are for applications on dry-bottom, wall-fired pulverized-coal
boilers ranging in size from 100 to 1000 MWe and being operated at a capacity factor of 0.65. In
general, costs associated with LNB and SCR are expected to span the costs of currently used NOx
controls; therefore, these costs were chosen for comparison with Hg control costs.

       As seen from Table 8-10, total annual costs for Hg controls lie mostly between applicable
costs for LNB and SCR.  However, Table 8-9 shows total  annual costs of Hg controls to be higher for
the minority of plants using HS-ESPs. Excluding these costs, both currently estimated and projected
Hg control costs are in the spectrum of LNB and SCR costs.
                                         8-27

-------
Table 8-10.  Comparison of mercury control costs with NOX control costs.
Air Pollutant
Controlled
Hg
NOX
Control Technology
PAC injection
Low-NOx burners
Selective catalytic reduction
Total Annual Control Cost
Range
(mills/kWh generated)
0.305B3.783(a)
0.183 to 2.270 (b)
0.210 B 0.827 (c)
1. 846 B 3.619 (c)
             (a) Current estimate of costs from Tables 8-3 and 8-4.
             (b) Projected costs from Table 8-9.
             (c) Actual costs from Reference 16.
                                        8-28

-------
8.8 Summary

       Preliminary estimates of costs of PAC injection-based Hg control technologies for coal-fired
electric utility boilers have been determined.  These estimates include those based on currently
available data from pilot-scale PAC injection tests, as well as projections for future applications of
more effective sorbents. Estimates based on currently available data range from 0.305 to 3.783
mills/kWh. However, the higher costs are associated with the minority of plants using HS-ESPs.  If
these costs are excluded, the estimates range from 0.305 to 1.915 mills/kWh. Cost projections,
developed based on using a composite lime-PAC sorbent for Hg removal, range from 0.183 to 2.270
mills/kWh with the higher costs being associated with the minority of plants using HS-ESPs.

       For technology applications on bituminous-coal-fired boilers using ESPs, current estimates
are based on levels of Hg capture on fly ash derived from pilot-scale test data. The EPA ICR data,
however, reflect that levels of capture higher than those seen in pilot-scale tests may be occurring. In
this light, the cost estimates for technology applications on bituminous-coal-fired boilers using ESPs
may be conservative.

       Results of sensitivity analyses conducted on the total annual cost of Hg controls reflect that:
(1) addition of ductwork to increase residence time does not have a significantjmpact on cost, (2) a
sorbent recycle rate of 20 percent is not adequate to reflect significant improvement in sorbent
utilization, (3) increasing the approach to ADP from ADP + 10 C to ADP + 22.2 C can have a
significant impact on total annual costs of Hg controls applicable to bituminous-coal-fired boilers,
and (4) a composite sorbent containing a mixture of PAC and lime offers great promise of
significantly reduced control costs.

       A comparison of Hg control costs with those of NOx controls reveals that total annual costs
for Hg controls lie mostly between applicable costs for LNB and SCR. As mentioned above,
estimates of total annual cost are higher where applicable to the minority of plants using HS-ESPs.
Excluding these costs, both currently estimated and projected Hg control costs are in the spectrum of
LNB and SCR costs.

       The performance and cost estimates of the PAC injection-based Hg control technologies
presented in this paper are based on relatively few data points from pilot-scale tests and, therefore, are
considered to be preliminary. Factors that are known to affect the adsorption of Hg  on PAC or other
sorbent include the speciation of Hg in flue gas, the effect of flue gas and ash characteristics, and the
degree of mixing between the flue gas and the sorbent. This mixing may be especially important
where the sorbent has to be injected in relatively large ducts. The effect of these factors may not be
entirely accounted for in the relatively few pilot-scale data points that comprised the basis for this
work. Ongoing research is expected to address these issues and to improve the cost effectiveness of
using sorbents for Hg control.  Research is also needed on ash and sorbent residue to evaluate Hg
retention and the potential for release back into the environment.
                                          8-29

-------
8.9 References
1.  Keating, M.H., K.R. Mahaffey, R. Schoeny, G.E. Rice, O.R. Bullock, R.B. Ambrose, J. Swartout,
   and J.W. Nichols.  Mercury Study Report to Congress, Volumes I-VIII. EPA-452/R-97-003
   through 010. Office of Air Quality Planning and Standards and Office of Research and
   Development, Research Triangle Park, NC. December 1997. Available at:
   < http://www.epa.gov/airprogm/oar/mercury.html >.

2.   French, C.L., W.H. Maxwell, W.D. Peters, G.E. Rice, O.R. Bullock, A.B. Vasu, R. Hetes,
    A. Colli, C. Nelson, andB.F. Lyons. Study of Hazardous Air Pollutant Emissions from Electric
    Utility Steam Generating Units  Final Report to Congress, Volume 1. EPA-453/R-98-004a
    (NTIS PB98-131774). Office of Air Quality Planning and Standards, Research Triangle Park,
    NC. February 1998.  Also available at:
   < http://www.epa.gov/ttn/atw/combust/utiltox/utoxpg.html >.

3.   Waugh, E.G., B.K. Jensen, L.N. Lapatnick, F.X. Gibbons,  S. Sjostrom, J. Ruhl, R. Slye, and R.
    Chang. "Mercury Control in Utility ESPs and Fabric Filters Through Dry Carbon-Based Sorbent
    Injection Pilot-Scale Demonstration." EPRI/DOE/EPA Combined Air Pollutant Control
    Symposium, Particulates and Air Toxics. Volume 3, EPRITR-108683-V3. Electric Power
    Research Institute, Palo Alto,  CA. August 1997.  pp. 1-15.

4.   Haythornthwaite,  S.M., J. Smith, G. Anderson, T. Hunt, M. Fox, R. Chang, and T. Brown. APilot-
    Scale Carbon Injection for Mercury Control at Comanche Station. Presented at the A&WMA
    92nd Annual Meeting & Exhibition. St. Louis, MO. June 1999.

5.   Haythornthwaite,  S.M., S. Sjostrom, T. Ebner, J. Ruhl, R. Slye, J. Smith, T. Hunt, R. Chang, and
    T.D. Brown. "Demonstration  of Dry Carbon-Based Sorbent Injection for Mercury Control in
    Utility ESPs and Fabric Filters."  EPRI/DOE/EPA Combined Air Pollutant Control Symposium,
    Particulates and Air Toxics, Volume 3. EPRI TR-108683-V3, Electric Power Research Institute,
    Palo Alto, CA.  August 1997.

6.   Redinger, K.E., A.P. Evans, R.T. Bailey, and P.S. Nolan.  "Mercury Emissions Control in FGD
    Systems." EPRI/DOE/EPA Combined Air Pollutant Control Symposium, Particulates and Air
    Toxics, Volume 3. EPRI TR-108683-V3, Electric Power Research Institute, Palo Alto, CA.
    August 1997.

7.   Brown, T.D., D.N. Smith, R.A Hargis, and W.J. O=Dowd. A1999 Critical Review, Mercury
    Measurement and Its Control: What We Know, Have Learned, and Need to Further Investigate,
    Journal of the Air & Waste Management Association, June 1999. pp. 1-97.

8.   Brown, T.D., W. O'Dowd, R. Reuther, and D.  Smith. "Control of Mercury Emissions from
    Coal-Fired Power Plants: A Preliminary Cost Assessment," in Proceedings of the Conference on
    Air Quality, Mercury, Trace Elements, and Particulate Matter,  Energy & Environmental Research
    Center, McLean, VA. December 1998.  pp. 1-18.
                                        8-30

-------
9.   U.S. Environmental Protection Agency. Electric Utility Steam Generating Unit Mercury
    Emissions Information Collection Effort, OMB Control No. 2060-0396. Office of Air Quality
    Planning and Standards, Research Triangle Park, NC.  Available at:
    < http://www.epa.gov/ttn/atw/combust/utitox/utoxpg.html >.

10. Bute, J.R., R. Chang, E.G. Waugh, B.K. Jensen, and L.N. Lapatnick.  "Use of Sorbents for Air
    Toxics Control in a Pilot-Scale COHPAC Baghouse," presented at the A&WMA 92nd Annual
    Meeting & Exhibition, St. Louis, MO.  June 1999.

11. Dotson, R.L., F.A Sodhoff, and T.A. Burnett. Lime Spray Dryer Flue Gas Desulfurization
    Computer Model UsersManual. EPA-600/8-86-016 (NTIS PB87-140968). U.S. Environmental
    Protection Agency. Air and Energy Engineering Research Laboratory, Research Triangle Park,
    NC. June 1986. p. 15.

12. Srivastava, R.K., C.B. Sedman, and J.D. Kilgroe. "Preliminary Performance and Cost Estimates
    of Mercury Emission Control Options for Electric Utility Boilers." Presented at the A&WMA
    93rd Annual Conference & Exhibition, Salt Lake City, UT.  June 2000.

13. U.S. Department of Energy. DOE=s Duct Injection Survey Results. Pittsburgh Energy
    Technology Center, Pittsburgh, PA. August 1988.

14. Kilgroe, J.D. "Control of Dioxin, Furan, and Mercury Emissions from Municipal Waste
    Combustors," Journal of Hazardous Materials, 47,  1996. pp. 163-194.

15. U.S. Environmental Protection Agency. Analysis of Emissions Reduction Options for the
    Electric Power Industry. Office of Air and Radiation, Washington, DC, March 1999. Available
    at: < http://www.epa.gov/capi/multipol/mercury.htm >.

16. U.S. Environmental Protection Agency. Analyzing Electric Power Generation Under CAAA.
    Office of Air and Radiation. Washington, DC. March 1998. Available at:
    < http://www.epa.gov/capi >.
                                         8-31

-------
                                      Chapter 9
                Coal Combustion Residues and Mercury Control
9.1 Introduction

       The burning of coal in electric utility boilers generates residual materials including fly
ash, bottom ash, boiler slag, and wet FGD scrubber solids/sludges. These residual materials are
collectively referred to as "coal combustion residues" (CCRs).  Currently, about 70 percent of the
CCRs are land-disposed and the other 30 percent are reused or recycled for commercial uses such
as production of wallboard, cement, and asphalt. Use of Hg emission control technologies on
coal-fired electric utility boilers will probably increase the amount of Hg in certain types of
CCRs, and could also change the composition and physical properties of these materials, possibly
impacting their suitability for commercial reuse and recycling applications. Many of the
potential retrofit Hg control technologies for coal-fired electric utility power plants discussed in
Chapter 7 remove Hg from the flue gas and concentrate the captured Hg into CCRs (i.e., fly ash
collected by PM control devices or solids/sludges generated by wet FGD  scrubbers).  Concern
has been raised as to whether the Hg in the CCRs may later be re-released back to the
environment.

       A life-cycle evaluation is being conducted by NRMRL to help evaluate any potential
environmental trade-offs and to ensure that there is not an increased environmental risk for the
management of CCRs resulting from  Hg control technologies. In support of this evaluation, the
NRMRL is gathering data and information to assess future increases in Hg concentrations in
CCRs resulting from application of Hg emissions control requirements to coal-fired electric
utility boilers.  This chapter summarizes some of the CCR information gathered by NRMRL to
date and identifies the major data gaps and priorities of EPA's research to ensure that Hg
controlled at the coal-fired electric utility power plant stack is not later released from CCRs in an
amount that is problematic for the environment.
9.2 CCR Types

       The coal combustion process generates many different types of residues. At a given
power plant, CCRs can be grouped as those generated on a continuous basis in high-volume
quantities and those generated either continuously or intermittently in low-volume quantities.
These low-volume CCRs include those resulting from maintenance and coal cleaning.  However,

                                          9-1

-------
the focus of this chapter is on high-volume CCRs. High-volume CCRs include the bottom ash or
slag removed directly from the boiler furnace and the fly ash collected by downstream PM
control devices. For those coal-fired electric utility boilers using wet FGD scrubbers for SC>2
emissions control, large quantities of scrubber solid wastes and sludges are generated.

       Nationwide quantities of high-volume CCRs generated in 1999 from coal combustion are
available from data prepared by the American Coal Ash Association (ACAA).1 Table 9-1
summarizes the characteristics  and nationwide generation quantities for the major types of CCRs
resulting from combustion of coal in power plants.
9.3 CCR Mercury Concentrations

       An initial review by NRMRL indicated that limited laboratory data were available on Hg
concentrations in CCRs. Therefore, a nationwide Hg mass balance approach was taken to
estimate Hg concentrations in CCRs.  This Hg mass balance approach used data from the EPA
Parts n and HI ICR data bases on coal Hg concentrations and control device Hg capture
efficiencies. The EPA ICR data were used with additional ACAA data on CCR generation rates,
to estimate Hg concentrations in various CCRs. The Hg concentrations estimated with the
nationwide  mass balance approach are shown in Table 9-2. Table 9-2 shows calculated mean, 5th
percentile, and 95th percentile values for Hg concentrations in CCRs. Mercury concentrations are
projected to be highest in fly ash, with a mean value of 0.33 ppm, and a 95th percentile value of
1.2 ppm.  Mercury concentrations in wet FGD scrubber  solids/sludges are calculated to have a
mean value of 0.20 ppm, and a 95th percentile value of 0.72 ppm. Mercury concentrations in
bottom ash  and boiler slag were calculated to be much lower, with mean values of 0.067  ppm,
and 0.042 ppm, respectively.

       Subsequent to performing the nationwide Hg mass balance to determine Hg
concentrations in CCRs, more extensive laboratory data became available from the Electric
Power Research Institute (EPRI)  and the University of North Dakota Environmental and  Energy
Research Center (UND/EERC). A summary of available laboratory measurements of Hg in
CCRs is shown in Table 9-3.  The laboratory measurements in Table 9-3 generally show  good
correlation with the nationwide mercury mass balance predictions in Table 9-2. For example, the
EPRI fly ash data (382 samples) have a mean mercury concentration of 0.44 ppm, with a 95th
percentile value of 1.13 ppm, and the UND/EERC data (20 samples) have a mean Hg
concentration of 0.22 ppm, and a 95th  percentile value of 1.03 ppm. Both these sets of data
correlate well with fly ash calculations obtained by the nationwide Hg mass balance, which
indicates  a mean concentration of 0.33 ppm, and 95th percentile value of 1.2 ppm.
9.4 Nationwide Management Practices

       A summary comparison of the quantities and management techniques for various CCRs is
presented in Figure 9-1. The CCRs are either land-disposed (in a monofill or surface

                                          9-2

-------
 Table 9-1.  Coal combustion residues.
Coal
Combustion
Residue
Fly ash
Bottom ash
Boiler slag
Wet FGD
scrubber
solids/sludges
Description
Fine, powdery non-combustible mineral matter
in the boiler flue gas and collected by
electrostatic precipitator or fabric filter
Dark gray, granular, porous non-combustible
mineral matter heavier than fly ash and
collected in bottom of the boiler furnace.
Coarse, black, glassy mineral matter
that forms when molten bottom ash contacts
quenching waters in wet-bottom furnaces.
Solid material or sludge generated by
scrubbing processes used to
remove sulfur from the flue gases.
Average Quantity
Generated Per
Ton of Coal Burned a
160 Ib/ton
40 Ib/ton
100 Ib/ton
350 Ib/ton
Total Nationwide
Quantity Generated
in 1999"
63, 000, 000 tons
17, 000, 000 tons
3,000,000 tons
25,000,000 tons
    (a) Source: Reference 2.
    (b) Source: Reference 1.
Table 9-2.  Calculated Hg concentrations in CCRs using EPA ICR data.
Coal
Combustion
Residue
Fly ash
Bottom ash
Boiler slag
Wet FGD scrubber
solids/sludges
Hg Concentration (ppm)'1
5th Percent! le
0.062
0.019
0.012
0.038
Mean
0.33
0.067
0.042
0.20
95th Percentile
1.2
0.16
0.10
0.72
     (a) Changes in Hg control technology requirements for coal-fired electric utility power plants will cause changes
        in the Hg concentration in fly ash and wet FGD scrubber solids/sludges.
                                            9-3

-------
     Table 9-3.  Summary of available test data on Hg concentrations in major types of CCRs.
Coal
Combustion
Residue
Fly ash
Bottom ash
Boiler slag
Wet FGD solids/sludges
Test Data Source
(Reference)
EPRI (References)
UND/EERC (Reference 4)
EPA (Reference 5)
(fine fly ash)
EPA (Reference 5)
(mechanical hopper)
EPA (Reference 5)
(1993 data)
EPA (Reference 5)
EPA (Reference 6)
(combined bottom ash and slag)
EPA (Reference 5)
EPA (Reference 5)
Number
of
Samples
382
20
n.r. a
n.r.
n.r.
12
n.r.
12
15
Hg Concentration (ppm)
Min.
0.0002
<0.002
0.005
0.008
0.013
0.003
0.005
0.005
0.073
5th
Percent! le
0.0002
0.002
n.r.
n.r.
n.r.
n.r.
n.r.
n.r
n.r
Median
0.09
0.076
0.10
0.073
0.10
0.009
0.023
0.023
4.8
Mean
0.44
0.22
n.r.
n.r
n.r
n.r
n.r
n.r
n.r
95th
Percentile
1.13
1.03
n.r
n.r
n.r
n.r
n.r
n.r
n.r
Max.
27.7
1.24
2.50
n.r
n.r
0.040
4.2
4.2
39.0
VO
       (a) n.r. = not reported.

-------
                                               Commercial use

                                              D Land-disposed
                 fly ash
bottom ash
boiler slag
FGD material
Figure 9-1. Nationwide CCR management practices in the year 1999 (source:
graph prepared using data from Reference 1).
                                     9-5

-------
impoundment) or are being used for commercial applications. In the United States in the year
1999, approximately 110 million tons of CCRs were generated.  Approximately one-third (31
percent) of these materials were reused or recycled in various commercial uses, with the
remainder being land-disposed.

9.4.1  Reuse and Recycling of CCRs

       The primary commercial uses of CCRs are listed in Table 9-4. The table presents how
each of four types of high-volume CCRs were used for commercial application in 1999. The use
of fly ash as a replacement ingredient for concrete or grout is the most common use for any CCR.
In this application, the fly ash can serve as a replacement for sand or as a partial replacement for
Portland cement in the concrete mix. Significant amounts of fly and bottom ash are used for
structural fills (e.g.,  creation of highway embankments). The addition of CCR to form a road
base allows for greater long-term strength development than conventional materials. Bottom ash
is used as a substitute for salt for road de-icing operations. Almost all of the boiler slag
generated in 1999 was used as blasting grit or roofing granules.  Wet FGD scrubber solid wastes
and sludges that do not contain high levels of fly ash can be used either directly or, with
additional processing, in the production of gypsum wallboard. The substitution of wet FGD
scrubber solids/sludges for natural gypsum in wallboard manufacturing has been growing rapidly.

       For some commercial uses of CCRs, there is concern regarding the potential re-release of
Hg, particularly for those uses  involving high-temperature processes.  In cement manufacturing,
for example, the high temperatures in the cement kiln will revolatilize the Hg contained in the
coal fly ash that is used as a material substitute. Questions exist regarding the fraction of Hg in
the fly ash that may be emitted when fed to a cement kiln.  Other commercial processes that
expose CCRs to elevated temperatures include wallboard manufacturing (during the drying
process) and when CCRs are used as fillers in asphalt.

       For some of the other commercial uses, it appears unlikely that significant Hg in CCRs
would be re-introduced into the environment. For example, Hg  is unlikely to be re-volatilized or
leached from concrete, flowable fill, or structural fill. However, the various  commercial uses
will be evaluated to  determine if there is any significant increase in environmental risk as a result
of changes occurring to CCRs.

9.4.2  Land-disposal of CCRs

       There are currently approximately 600 waste disposal units (monofills or surface
impoundments) being used for disposal of CCRs from electric utility coal-fired electric utility
power plants in the United States.5 The monofills used for these residues may be located either
on-site at the power plant or off-site. Surface impoundments are almost exclusively located at
the power plant site. While the distribution of units presently is  about equal  between monofills
and surface impoundments, there is an increased trend to use monofills as the primary disposal
method.
                                          9-6

-------
           Table 9-4. Commercial uses for CCRs generated in 1999 (source:  data from Reference 1).
Commercial
Application or Use
Concrete/grout
Waste stabilization/solidification
Structural fill
Mining applications
Raw feed for cement clinker
Road base/subbase
Flow/able fill
Other
Mineral filler
Soil modification
Agriculture
Snow and ice control
Blasting grit/roofing granules
Wallboard
Nationwide Total a
Coal Combustion Residue
Fly ash
tons
10,000,000
1,900,000
3,200,000
1,500,000
1,300,000
1,200,000
850,000
460,000
160,000
78,000
78,000
3,200
0
0
21,000,000
%
49
9.3
15
7.3
6.1
5.9
4.1
2.2
0.8
0.4
0.4
0.1
0
0
100
Bottom ash
tons
700,000
69,000
1,400,000
150,000
160,000
1,100,000
13,000
450,000
63,000
17,000
43,000
1,100,000
160,000
0
5,400,000
%
13
1.3
26
2.8
2.9
20
0.2
8.3
1.2
0.3
0.8
20
2.9
0
100
Boiler slag
tons
11,000
0
52,000
10,000
0
5,500
0
76,000
12,000
13,000
0
51,000
2,100,000
0
2,300,000
%
0.5
0
2.2
0.4
0
0.2
0
3.2
0.5
0.5
0
2.2
90
0
100
Wet FGD solids/sludges
tons
290,000
16,000
580,000
230,000
0
17,000
0
180,000
0
2,100
80,000
0
0
3,100,000
4,500,000
%
6.5
0.4
13
5.2
0
0.4
0
4.1
0
<0.1
1.8
0
0
69
100
Nationwide
Total
(tons)
11,000,000
2,000,000
5,200,000
1,900,000
1,500,000
2,300,000
860,000
1,200,000
240,000
110,000
200,000
1,200,000
2,300,000
3,100,000
33,000,000
VO
            (a) Sum of individual values may not equal total due to rounding.

-------
       On May 22, 2000, the EPA made the regulatory determination that the disposal of CCRs
does not warrant regulation under subtitle C of RCRA and retained the hazardous waste
exemption for these materials provided under RCRA section 3001(b)(3)(C).  However,  the
EPA also determined that national regulations under subtitle D of RCRA are warranted for CCRs
when they are disposed of in landfills or surface impoundments, and that regulations under
subtitle D of RCRA [and/or possibly modifications to existing regulations established under
authority of the Surface Mining Control and Reclamation Act (SMCRA)] are warranted when
these materials are managed in surface or underground mines.  The national regulations will
apply to disposal of coal combustion wastes that are generated at electric utility and independent
power producing facilities and managed in surface impoundments, landfills, and mines.

       The EPA will re-evaluate the risk posed by managing coal combustion residues if levels
of Hg or other hazardous constituents change due to any future Clean Air Act air pollution
control requirements for coal burning utilities. When any rulemaking under the Clean Air Act
proceeds to the point where an assessment of the likely changes to the character of CCRs is
completed, EPA will evaluate the implications of these changes relative to existing or planned
national RCRA regulations governing these materials and take appropriate action.
9.5 Current Status of CCR Research Activities

       The EPA/NRMRL is preparing a report on characterization and management of CCRs
from coal-fired electric utility power plants. The report examines changes in the Hg content of
CCRs that potentially could occur as the result of implementing different control technologies to
reduce stack emissions of Hg from coal-fired electric utility power plants. This report is
scheduled to be published in the near future.

       Test methods to characterize CCRs and to determine Hg volatilization and leaching from
CCRs in various management practices are being reviewed by EPA/NRMRL.  The goal of this
review is to ensure that leaching and volatilization testing conducted by all parties, inside and
outside of the EPA, is uniform and appropriate.

       Multiple-site, full-scale field test programs are currently being conducted under a
DOE/NETL cooperative agreement to obtain performance and cost data for using different Hg
control technologies to reduce Hg emissions from existing coal-fired electric utility power plants
(discussed in Chapter 7). As  part of these test programs, field data are being collected that will
help determine changes in the Hg content of CCRs as a result of implementing these Hg controls
technologies. In addition, CCR commercial applications requiring elevated temperature
processes, such as cement manufacturing and wallboard production, are being evaluated to
determine the amount of Hg revolatilization that occurs, and the impacts of this revolatilization
on the environment.

       The EPA/NRMRL is planning to prepare a report, scheduled for publication in 2003,
presenting data and other information relating to changes to CCRs as a result of implementing
                                        9-8

-------
different Hg control technologies. This report will also help identify any potential concerns due
to increased environmental risk from the management of CCRs resulting from Hg control
measures.
9.6 Future CCR Research Activities and Needs

       Coordination with industry and others will continue to identify available data and
information that will help to characterize any changes to CCRs as a result of Hg control measures.
Different methods are being used to characterize CCRs which result in data of questionable value.
The EPA ORD/NRMRL is working closely with EPA/OSW to identify methods for
characterizing CCRs to identify potential changes to CCRs as a result of Hg control measures.

       Samples of the resulting CCRs from the on-going full-scale field test programs of different
Hg control technologies will be collected to characterize the resulting CCRs and to identify any
changes occurring to CCRs that would increase environmental risk from waste management and
potential commercial applications.

       Questions regarding the potential release of Hg from land-disposal result in the need to
conduct field test measurements to ensure that Hg is not being emitted through either biological
processes or leaching.  Opportunities will be identified to help address questions regarding any
increased environmental risk due to changes occurring to CCRs.

       Questions also  exist relating to CCRs being used in high-temperature processes such as
cement manufacturing and wallboard production. Effort is needed to  determine the amount of Hg
that may be released during the manufacturing process and other life-cycle stages, including final
disposal in a landfill.
9.7 References
1.  American Coal Ash Association. Coal Combustion Products (CCPs) Production and Use
   Survey - 2000.  Syracuse, NY. Available at:
   < http://216.22.250.39/CCP%20Survev/CCP%20Survev.htm >.

2.  Butalia,  T., W. Wolfe, W. Dick, D. Limes, and R. Stowell. Coal Combustion Products  Ohio
   State University Fact Sheet.  1999.  Available at: <
   http://www.ag.ohio-state.edu/~ohioline/aex-fact/0330.html. >

3.  Ladwig, K. Electric Power Research Institute, Palo Alto, CA. Personal communication
   presenting total composition of mercury in fly ash based on EPRI's PISCES database, May
   29,2001.
                                        9-9

-------
4.  Hassett, D. J., D.F. Pflughoeft-Hassett, D.L. Laudal, and J.H. Pavlish. Mercury Release from
   Coal Combustion By-Products to the Environment.  Energy & Environmental Research
   Center, University of North Dakota, Grand Forks, ND, 2000.

5.  U.S. Environmental Protection Agency. Report to Congress: Wastes from the Combustion of
   Fossil Fuels: Volume 2B. Methods, Findings and Recommendations.  EPA-530-R-99-010.
   Office of Solid Waste and Emergency Response, Washington, DC. 1999.

6.  U.S. Environmental Protection Agency. Report to Congress: Wastes from the Combustion of
   Coal by Electric Utility Power Plants.  EPA/530-SW-88-002 (NTIS PB88-177977). Office
   of Solid Waste and Emergency Response, Washington, DC. 1988.

7.  U.S. Environmental Protection Agency. Regulatory Determination on Wastes from the
   Combustion of Fossil Fuels; Final Rule. Federal Register. Vol. 65, No. 99.  pp.32213-32237.
   May 22, 2000.

8.  U.S. Environmental Protection Agency. Technical Background Document for the Report to
   Congress on Remaining Wastes from Fossil Fuel Combustion:  Industry Statistics and Waste
   Management Practices. Office of Solid Waste and Emergency Response, Washington, DC.
   1999. Available at: < http://www.epa.gov/epaoswer/other/fossil/index.htm >.
                                       9-10

-------
                                      Chapter 10
                          Conclusions and Recommendations
10.1 Electric Utility Coal Combustion and Air Pollution Control Technologies

      The diversity of coals and combustion technologies used for electrical utility steam
generating units (i.e., coal-fired boilers) is reflected in data gathered from Phases I and II of the
EPA's formal information collection request (ICR) to the electric utility industry. In 1999,
electric utility coal-fired boilers in the United States burned 786 million tons of coal of which
about 52 percent was bituminous coal, 37 percent was subbituminous coal, and 9 percent was
lignite.  Other fuels burned in electric utility coal-fired boilers included mixtures of bituminous
and subbituminous coal, mixtures of coal and petroleum coke (pet-coke), reclaimed coal wastes,
and mixtures of coal and tire-derived fuel (TDF).

       There were 1,140 coal-fired boiler units that burned coal by conventional methods and
three units that used gasification to produce a fuel gas. Pulverized-coal-fired (PC) boilers, by far
the largest group of coal-fired boiler units, represent approximately 86 percent of the total
number of units and 90 percent of the total utility boiler capacity. Based on capacity, other types
of boilers include cyclone-fired boilers (7.6 percent), fluidized-bed combustors (1.3 percent), and
stoker-fired boilers (1.0 percent).

       The Part II EPA ICR responses indicate that a variety of air emission control technologies
are employed to meet requirements for control  of sulfur dioxide (802), nitrogen oxides (NOx),
and paniculate matter (PM). Most utilities control NOx by combustion modification techniques
and control SO2by the use of compliance coals. For post-combustion controls, 77.4 percent of
units by number have PM control only, 18.6 percent have both PM and SO2 controls, 2.5 percent
have PM and post-combustion NOx controls, and 1.3 percent have three post-combustion control
devices.

       For PM emissions control of electric utility coal-fired boilers, electrostatic precipitators
(ESPs) are used on 84 percent of the units and fabric filters (FFs) on  14 percent. Post-
combustion SO2 controls are less common. Wet flue gas desulfurization (FGD) scrubbers are
used on 15.1 percent of the units, and spray dryer absorbers (SDA) are used on 4.6 percent of the
units surveyed.  In  1999, while the application  of post-combustion NOx controls was becoming
more prevalent, only 3.8 percent of the units used either selective non-catalytic reduction
(SNCR) or selective catalytic reduction (SCR)  systems.
                                          10-1

-------
 10.2 Mercury Measurement Methods

       Reliable and valid speciated and total Hg measurements, by either manual test methods or
continuous emission monitors (CEMs), are critical to the characterization and future reduction of
Hg emissions from coal-fired powered plants. Although viable measurement techniques exist for
certain measurement scenarios, acceptable measurement techniques are not available to meet all
measurement needs. Additional research and development is needed to enable quality
measurements from various measurement environments.

       The Ontario-Hydro (OH) Method is the only manual test method that currently is
recognized in the United States for the collection of speciated Hg emissions data from the
combustion of coal. The OH Method appears to provide valid speciation results at sampling
locations downstream of PM control devices where most of the fly ash has been removed.
However, measurements made upstream of PM control devices are susceptible to measurement
artifacts that bias speciation measurements, causing significant uncertainty in results.

       A limited number of CEMs exist (both commercial and prototype) for the measurement
of total gas-phase Hg and, to a lesser extent, speciated gas-phase Hg.  Also, demonstration of
acceptable measurement performance under field applications is limited.  Because of the
diversity and severity of associated measurement environments, numerous measurement
obstacles exist (e.g., PM artifacts, interferences, Hg2+ conversion systems, sample
conditioning/delivery) that have not been adequately addressed,  particularly with respect to
speciated measurements. While experts use Hg CEMs as a research tool, the Hg CEMs are not
currently suitable for routine use on power plants in the United States.

       Improved methods for the sampling and analysis are critical for:  the development and
evaluation of Hg emission control technologies; use as Hg control technology process controls;
and potential use as compliance tools.  Research is specifically needed to:

         Develop and verify a manual test method that is suitable for measuring total and
          speciated Hg at sampling locations upstream of PM control devices,

         Develop and verify a manual test method that can simultaneously measure speciated
          Hg and other hazardous trace metals,

         Develop and demonstrate measurement techniques that are capable of directly
          identifying and quantifying trace levels of individual ionic species of Hg [e.g., HgCb,
          HgCl, HgO, HgS,  HgS04, and Hg(NO3)2],

         Develop and demonstrate improved Hg CEM measurement techniques that address
          known and potential measurement obstacles (e.g., PM artifacts, interferences/biases,
          conversion systems, and sample conditioning/delivery),
                                          10-2

-------
          Verify the ability of Hg CEMs to accurately measure total gas-phase Hg and
          speciated gas-phase Hg at diverse stack conditions representative of fuel type and
          pollution control device configurations (e.g., downstream of PM control devices, dry
          FGD scrubbers, and wet FGD scrubbers),

          Verify the ability of Hg CEMs to accurately measure total gas-phase Hg and
          speciated gas-phase Hg upstream of PM control devices,

          Demonstrate Hg CEM long-term monitoring performance and operational
          requirements,

          Identify and evaluate CEMs capable of measuring 863 and other hazardous air
          pollutant emissions,

          Identify and evaluate alternative, cost-effective semi-continuous methods for
          measuring the stack emission of total Hg, and

          Demonstrate the use of Hg CEMs and semi-continuous monitoring methods as
          potential Hg emission compliance tools.
10.3 Mercury Speciation and Capture

10.3.1  Mercury Speciation

       When the coal is burned in an electric utility boiler, the resulting high combustion
temperatures vaporize the Hg in the coal to form gaseous elemental mercury (Hg).  Subsequent
cooling of the combustion gases and interaction of the gaseous Hg with other combustion
products result in a portion of the Hg being converted to gaseous oxidized forms of mercury
(Hg +) and particle-bound mercury (Hgp).  The term speciation is used to denote the relative
amounts of these three forms of Hg  in the flue gas of the boiler. It is important to understand
how Hg speciates in the boiler flue gas because the  overall effectiveness of different control
strategies for capturing  Hg often depends on the concentrations of the different forms of Hg
species present in the boiler flue gas.

       In general, Hg speciation is dependent on:  1) coal properties, 2) combustion conditions,
3) the flue gas composition, 4) fly ash properties, 5) the time/temperature profile between the
boiler and air pollution  control devices, and 6) the flue gas cleaning methods, if any, in use.  The
mechanisms by which Hg is oxidized in flue gas are believed to include gas-phase reactions, fly
ash or deposit-mediated reactions, and oxidation reactions in post-combustion NOx control
systems. Data reveal that gas-phase oxidation is kinetically limited and occurs due to reactions
of Hg with oxidizers such as Cl and C^. Research also suggests that gas-phase oxidation may be
inhibited by  the presence of NO, SO2, and water vapor.
                                          10-3

-------
       Certain fly ashes have been shown to promote oxidation of Hg more than others.  The
differences in oxidation appear to be attributable to the composition of the fly ash and the
presence of certain flue gas constituents.  The results of bench-scale research conducted at EPA
indicate that the presence of HC1 and NOx in flue gas and iron in fly ash assists in oxidation.
                                                    9-1-
Other research indicates that y-Fe2O3 may be causing Hg  formation and that surface area may
be a dominant factor in this regard. Also, there are indications that HC1, NC>2, and 862 in the
flue gas may contribute to Hg oxidation, while the presence of NO may suppress Hg oxidation.

10.3.2 Development and Evaluation ofSorbents

       Mercury can be captured and removed from a flue gas stream by injection of a sorbent
into the exhaust stream with  subsequent collection in a particulate matter (PM) control device
such as an ESP or a FF.  However, adsorptive capture of Hg from flue gas is a complex process
that involves many variables. These include the temperature and composition of the flue gas, the
concentration of Hg in the exhaust stream, and the physical and chemical characteristics of the
sorbent (and associated functional groups). The implementation of an effective and efficient Hg
control strategy using sorbent injection requires the development of low-cost and efficient Hg or
multipollutant sorbents.  Of the known Hg sorbents, activated  carbons and calcium-based
sorbents have been the most  actively studied.

       Oxidized mercury is readily absorbed by alkaline solutes/slurries or adsorbed by alkaline
particulate matter (or by sorbents). Flue gas desulfurization systems, which use alkaline
materials to neutralize the acidic SO2 gas, remove Hg + effectively in the flue gas. Current
research is focusing on optimization of the existing desulfurization systems as a retrofit
technology for controlling Hg2+ emissions and on development of new multipollutant control
technologies for simultaneously controlling both SO2 and Hg emissions. Sorbents containing
oxidizing agents are also being developed for the oxidization and capture of Hg.
10.4 Evaluation of EPA ICR Mercury Emission Test Data

       The air pollution control technologies used on coal-fired utility boilers exhibit levels of Hg
control that range from 0 to 99 percent.  The best levels of control are generally obtained by
emission control systems that use FFs. Since Hg emission control technologies are not currently
used by the utility industry, the capture of Hg by existing controls results from:  1) adsorption of
Hg onto fly ash with subsequent capture of Hgp in a PM control device; 2) adsorption of Hg by
the alkaline sorbents used in dry scrubbers; or 3) the capture of Hg 2+ in wet scrubbers.

       The amount of Hg captured by a given control technology is better for bituminous coal
than for either subbituminous coal or lignite. The lower levels of Hg capture in plants firing sub-
bituminous coal and lignite are attributed to low fly ash carbon content, and the higher relative
amounts of Hg in the flue gas are due to the combustion of these fuels.

       Combinations of coal, boiler, and control technologies that are expected to behave in a
similar manner with respect to speciation and capture of mercury can be grouped into data sets
                                          10-4

-------
called coal-boiler-control technology classes. Accordingly, the ICR Part III emission data were
sorted into appropriate coal-boiler-control classes. Next, the data in each class were evaluated for
consistency, and the data between classes were evaluated based on the current understanding of
speciation and capture of mercury. With few exceptions, the differences in data between the
different classes were consistent with this understanding.

       Plants that employ only post-combustion PM controls display Hg emission reductions
ranging from 0 to 93 percent.  The lower levels of control were observed for units with FFs.
Decreasing levels of control were shown for units with ESPs, PM scrubbers, and mechanical
collectors.

       Units equipped with dry scrubbers (SDA/ESP or SDA/FF systems) exhibited  average Hg
captures ranging from 98 percent for units burning bituminous coals to 3 percent for units burning
subbituminous coal. The poor Hg capture in units firing subbituminous coal is attributed to the
predominance of Hg in the flue gas from these units.

       The capture of Hg in units equipped with wet FGD scrubbers is primarily dependent on
the relative amount of Hg2+ in the inlet flue gas.  Average Hg  captures in wet FGD scrubbers
ranged from 33  percent, for one PC-fired ESP + FGD unit burning subbituminous coal, to 96
percent in a PC-fired FF  + FGD unit burning bituminous coal. The high Hg capture in the FF +
FGD unit is attributed to the increased oxidization and  capture of Hg in the FF.

     The EPA ICR data base provides a massive amount of data that can be mined for additional
information. However, the usefulness of these data is limited by the uncertainty of some of the
measurements and/or the information that the data set does not contain. Some of the  uses and
limitations of the EPA ICR data are summarized below.

    The EPA ICR data provide:

           Reasonable estimates of national and regional emissions for total Hg, Hg, Hg2+,
            and Hgp.  The data cannot be used to predict the  total and speciated Hg emissions of
            individual coal-fired power plants.

           Data for testing hypotheses and models that predict speciation and capture of Hg in
            coal-fired boilers. The data cannot be used to identify or confirm specific
            mechanisms that control the speciation and capture of Hg.

           Information needed to guide the development of control technologies and identify
            effective strategies for the control of Hg emissions.

    Caution should be used in interpreting the EPA ICR data since:

           Adsorption  of Hg onto fly ash is highly dependent on fly ash properties, particularly
            on the fly ash carbon content. The lack of information  on coal and fly ash
            properties greatly limits the usefulness of the ICR data.
                                          10-5

-------
           Results of Hg speciation measurements made with the OH Method upstream of PM
            control devices should be used with great caution. PM collected on the sampling
            train filter can result in physical and chemical transformations (e.g., sample
            artifacts) within the sample train.

           Because of the limited number of samples, there is a great deal of uncertainty in the
            central values and statistical characteristics of the ICR data. The flue gas Hg tests
            represent a short snapshot in time, and the effects of long-term variations in coal
            properties and plant operating conditions are unknown.

           At low inlet and outlet concentrations, the imprecision of the OH Method can
            obscure real differences between inlet and outlet concentrations.
10.5  Potential Retrofit Mercury Control Technologies

       A practical approach to controlling Hg emissions at existing utility plants is to minimize
capital costs by adapting or retrofitting the existing equipment to capture Hg. For units that
currently use only PM controls, dry FGD scrubbers, or wet FGD scrubbers, there are three
potential retrofit options: 1) Cold-side ESP, hot-side ESP, and FF systems; 2) Dry FGD systems;
and 3) Wet FGD systems.

10.5.1 Cold-side ESP, Hot-side ESP, and FF Systems

       The most cost-effective retrofit options for the control of Hg emissions from units,
currently equipped with  only an ESP or FF, include:

           Injection of a sorbent upstream of a cold-side ESP or FF.  Cooling of the stack gas
            and/or modifications to the ducting may be needed to keep sorbent requirements at
            acceptable levels.

           Injection of a sorbent between a cold- or hot-side ESP and a pulsejet FF retrofitted
            downstream of the ESP.  This approach will increase capital costs but reduce
            sorbent costs.

           Installation of a semi-dry CFA upstream of an existing cold-ESP used in
            conjunction with sorbent injection.  The CFA recirculates both fly ash and sorbent
            to create an entrained bed with a high number of reaction sites. This leads to higher
            sorbent utilization and enhanced fly ash capture of Hg and other pollutants.

       Units equipped with a FF  require less sorbent than units equipped with an ESP.  ESP
systems depend on in-flight adsorption of Hg by entrained fly ash or sorbent particles. FFs
obtain both in-flight and fixed-bed capture as the flue gas passes through the FF.
                                          10-6

-------
       In general, cost-effective sorbent injection technologies for cold-side ESP units will
depend on: (1) the development of lower cost and/or higher performing sorbents, and (2)
appropriate modifications to the operating conditions or equipment being currently used to
control emissions of PM, NOx, and  SC>2.

       It is believed that the above technologies will be available for use on boiler units that
must comply with the Clean Air Act hazardous air pollutant (HAP) maximum achievable control
technology (MACT) requirements for electric utility steam-generating units.  The performance
and cost of these technologies are yet to be determined.

10.5.2 Semi-dry FGD scrubbers

       Semi-dry FGD scrubbers are already equipped to control emissions of SC>2 and PM. The
modification of these units by the use of appropriate sorbents for the capture of Hg and other air
toxics is considered to be the easiest retrofit problem to solve.  SDA/FF systems are capable of
higher levels of Hg control than SD A/ESP systems.

10.5.3 Wet FGD Scrubbers

       Improvements in wet FGD scrubber performance in capturing Hg depend primarily on
                  n      94-   	
the oxidation of Hg to Hg  .  This may be accomplished by 1) the injection of appropriate
oxidizing agents, or 2) the installation of fixed oxidizing catalysts upstream of the scrubber to
promote oxidization of Hg to soluble Hg compounds.

       An alternative strategy for controlling Hg emissions from wet FGD scrubbing systems is
to inject sorbents upstream of the PM control device. Better performance can be expected for
units equipped with FFs than those equipped with ESPs. SCR systems used with wet FGD
scrubbers may enhance Hg capture in the scrubber.

       Additional research is needed on the oxidization of Hg and the removal and
sequestration of Hg collected in the  scrubbing liquid.
10.6 Costs of Retrofit Mercury Control Technologies

       Preliminary annualized costs of Hg controls using powdered activated carbon (PAC)
injection have been estimated based on recent pilot-scale evaluations with commercially
available sorbents.  These control costs range from 0.305 to 3.783 mill/kWh, with the highest
costs associated with plants having hot-side electrostatic precipitators (HS-ESPs). For plants
representing approximately 89 percent of current plant capacity and using controls other than
HS-ESPs, the costs range from 0.305 to 1.915 mill/kWh. Assuming a 40 percent reduction in
sorbent costs by use of a composite lime-PAC sorbent for Hg removal, cost projections range
from 0.183  to 2.27 mill/kWh, with higher costs again being associated with the plants using
HS-ESPs.
                                          10-7

-------
       In comparison, the estimated annual costs of Hg controls, as a function of plant size, lie
mostly between the costs for low-NOx burners (LNBs) and selective catalytic reduction (SCR)
systems.  The costs of Hg control will dramatically diminish if retrofit hardware and sorbents are
employed for control of other pollutants such as NOx, SC>2, or fine PM.

       The performance and cost estimates of PAC injection-based Hg control technologies
presented in this report are based on relatively few data points from pilot-scale tests and are
considered to be preliminary. However, based on data from pilot-scale tests and the results of
ICR data evaluations, it is expected that better sorbents and technologies now being developed
will reduce the costs of Hg controls beyond current estimates.

       Within the next 2-3 years, it is expected that the evaluation of retrofit technologies at
plants where co-control is being practiced will lead to a more thorough characterization of the
performance and costs of Hg control.  Future cost studies will focus on the development of
performance and cost information needed to: 1) refine cost estimates for  sorbent injection based
controls, 2) develop cost estimates for wet scrubbing systems that employ methods for oxidizing
Hg, and 3) determine the costs of various multipollutant control options.

       The issue of Hg in many power plant residues will also be examined to address concerns
related to the release of captured Hg species into the environment.  These evaluations will be
conducted in conjunction with the development and evaluation of air pollution emission control
technologies.

10.7 Coal Combustion Residues  and Mercury Control

       Power plant operations result in solid discharges including fly ash, bottom ash, boiler
slag, and FGD residues. These residues already contain Hg, presumably bound Hg that is
relatively insoluble and non-leachable. In  1998, approximately  108 million tons of coal
combustion residues (CCRs) were generated. Of this amount, about 77 million tons were land-
filled and about 31 million tons were utilized for beneficial uses.

       Increased control of Hg emissions from coal-fired power plants may change the amount
and composition of CCRs. Such changes may increase the potential for release of Hg to the
environment from either land filling (approximately 70 percent) or  commercial uses
(approximately 30 percent) of CCRs.  Mercury volatilization from CCRs in landfills and/or
surface impoundments is expected to be low due to the low temperatures involved and minimal
gas-to-solid interfaces within impacted wastes. For Hg control retrofits involving dry or wet
FGD scrubbers, the residues are typically alkaline and the acid leaching potential of Hg from
these residues is expected to be minimal.

       There are several commercial uses of CCR where available  data on which to characterize
the Hg emission potential are lacking. The following CCR uses are given a priority for
developing additional data in order to  characterize the ultimate fate of Hg:

            Use of fly ash in cement production,


                                          10-8

-------
           Volatilization and leaching of residues used for structural fills,

           Leaching of residues exposed to acidic conditions during mining applications,

           Volatilization of Hg during the production of wallboard from gypsum in wet
            scrubber residues,

           Mercury volatilization during the production and application of asphalt with fly ash
            fillers, and

           Leaching or plant uptake of Hg from fly ash, bottom ash, and FGD sludge that are
            used as soil additives.


10.8  Current and Future Research

      It is important to continue collaborative Hg research efforts between DOE, EPA, EPRI,
and the utility industry. The focus of these efforts should be to provide scientific and
engineering data that support the Administration's Energy Plan and that can be used to:

           Develop HAP MACT requirements for coal-fired electric utility steam generating
            units.

           Optimize control of Hg emissions from units that must comply with more stringent
            NOx emission requirements under the NOx state implementation plan (SIP) call.

           Develop technologies that can be used to control emissions under multipollutant
            control legislation that is under consideration by the Congress.

     Current and future research is needed to:

           Control Hg emissions for units now equipped only with ESPs.

           Develop cost-effective sorbents to control emissions from  subbituminous coals and
            lignite.

           Control Hg emissions from subbituminous  coals and lignite.

           Determine the effects of coal  blending on Hg speciation and capture.

           Evaluate the enhancement of fly-ash capture by combustion modification
            techniques.
                                          10-9

-------
    Develop Hg oxidizing methods for wet FGD systems.

    Optimize NOx controls for Hg control.

    Evaluate controls for non-PC fired units.

    Control Hg, SC>3, and other air toxic emissions from units equipped with SCR and
     wet FGD scrubbers.

    Demonstrate Hg control for units with SD/ESP and SD/FF.

    Demonstrate Hg control in wet FGD systems.

    Determine the effects of cyclone, stoker, and fluidized-bed combustion systems on
     Hg control.

    Minimize the effects of Hg controls on power plant operability.

    Conduct tests with CEMs to study the variability of Hg emissions.

    Characterize and control the stability of Hg and other hazardous trace metals in
     CCRs and byproducts.
                                  10-10

-------
                                   Appendix A
                       Summary of Part II EPA ICR Data

                Mercury Content and Selected Fuel Properties of
                     As-fired Coals and Supplemental Fuels
                  Burned in Coal-fired Electric Utility Boilers
                               Nationwide in 1999
Table A-l.  Selected properties of anthracite coal burned in 1999	 A-2
Table A-2.  Selected properties of bituminous coal burned in 1999	A-2
Table A-3.  Selected properties of South American bituminous coal burned in 1999	A-3
Table A-4.  Selected properties of subbituminous coal burned in 1999	A-3
Table A-5.  Selected properties of Indonesian  subbituminous coal burned in 1999	A-4
Table A-6.  Selected properties of lignite burned in 1999	A-4
Table A-7.  Selected properties of waste anthracite coal burned in 1999	A-5
Table A-8.  Selected properties of waste bituminous coal burned in 1999	A-5
Table A-9.  Selected properties of waste subbituminous coal burned in 1999	A-6
Table A-10. Selected properties of petroleum coke burned in 1999	A-6
Table A-ll. Selected properties of tire-derived fuel burned in 1999	A-7
Table A-12. Comparison of mercury content by fuel type as burned in 1999	A-8
Table A-13. Comparison of mercury content with fuel heat content as burned in 1999	A-9
                                        A-l

-------
Table A-l. Selected properties of anthracite coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry
Range
5.02-35.19
0.06-0.31
0.44-2.01
5,721 -14,376
9.86-56.07
100-807
Mean
15.28
0.16
0.67
10,944
24.41
384.04
Median
13.37
0.16
0.60
11,598
20.26
382.5
Standard deviation
6.23
0.05
0.22
1,660
10.43
169.97
Table A-2. Selected properties of bituminous coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry"
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dryb
Range
0.04-103.81
0.0-1.3
0.07-13.92
8,068-21,503
0.11 -40.37
2.5-11,000
Mean
8.59
0.11
1.70
13,196
11.13
1,031
Median
7.05
0.091
1.22
13,181
10.86
905
Standard deviation
6.69
0.09
1.14
712
3.33
879
    (a) Includes 112 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
    (b) Includes 940 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
       Four analyses did not report chlorine content.
                                              A-2

-------
Table A-3. Selected properties of South American bituminous coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry Mercury,
ID/"IU bl a IVI ID/'IU
Mercury, ppm dry
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry"
Range
0.70-66.81
0.01 -0.95
0.45-1.44
11,515-17,195
2.03-16.32
45-1,521
Mean
5.94
0.08
0.77
13,678
6.81
286.71
Median
4.91
0.0642
0.75
13,701
6.66
264.5
Standard deviation
5.28
0.07
0.14
490
1.90
180.91
    (a)  Includes 7 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
Table A-4. Selected properties of subbituminous coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry"
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dryb
Range
0.39-71.08
0.005-0.9
0.02-3.12
7,636-14,401
0.43-31.75
1 -5,800
Mean
5.74
0.07
0.48
11,969
7.90
108.51
Median
5.00
0.06
0.45
12,060
7.15
50
Standard deviation
3.59
0.04
0.25
672
3.63
235.27
    (a)  Includes 193 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
        Two analyses did not report mercury content due to lost samples.
    (b)  Includes 3,961 values reported at the non-detect level;  2 the non-detect level has been used in the
        analyses. One analysis did not report chlorine content.
                                              A-3

-------
Table A-5.  Selected properties of Indonesian subbituminous coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry"
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry"
Range
0.79-4.61
0.01 -0.05
0.059-0.68
10,840-13,157
1.15-12.44
25-400
Mean
2.51
0.03
0.31
12,501
5.26
90.79
Median
2.39
0.03
0.30
12,532
6.46
100
Standard deviation
0.86
0.01
0.17
337
3.79
68.65
    (a)  Includes 4 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
    (b)  Includes 19 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
Table A-6.  Selected properties of lignite burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry"
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry"
Range
0.93-75.06
0.01 -0.75
0.35-3.47
7,022-11,943
8.93-39.10
25-1,380
Mean
10.54
0.10
1.30
10,026
19.45
166.09
Median
7.94
0.08
1.21
10,205
18.75
101.00
Standard deviation
9.05
0.09
0.44
784
6.42
199.60
    (a)  Includes 133 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
    (b)  Includes 361 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
                                               A-4

-------
Table A-7. Selected properties of waste anthracite coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry"
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry"
Range
2.49-73.02
0.02-0.54
0.25-1.96
2,998-10,941
22.34-72.41
11.6-1,855
Mean
29.31
0.19
0.51
6,687
49.13
231.29
Median
27.77
0.17
0.44
6,889
47.77
102.20
Standard deviation
11.94
0.08
0.20
2,002
12.54
241.36
    (a) Includes two values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
    (b) Includes 21 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
Table A-8. Selected properties of waste bituminous coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry
Range
2.47-172.92
0.03-1.18
0.74-7.73
5,194-13,370
7.89-61.49
29.77-4,277
Mean
60.50
0.46
2.38
8,753
37.54
847.92
Median
53.32
0.45
2.38
7,929
41.43
848.00
Standard deviation
44.35
0.30
0.88
2,153
13.03
493.75
                                             A-5

-------
Table A-9. Selected properties of waste subbituminous coal burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry"
Range
5.81 -30.35
0.07-0.35
0.83-3.59
7,849-11,801
11.17-38.52
50-100
Mean
11.42
0.12
1.95
10,506
20.53
53.77
Median
10.79
0.11
1.82
10,704
19.35
50
Standard deviation
4.66
0.05
0.69
927
6.40
13.33
    (a) Includes 49 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
Table A-10. Selected properties of petroleum coke burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry"
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry"
Range
0.06-32.16
0.0009-0.5
0.54-7.91
10,892-16,463
0.04-27.53
7-3,000
Mean
23.18
0.045
4.88
15,233
0.64
203.70
Median
2.16
0.03
5.18
15,319
0.40
110
Standard deviation
3.18
0.05
1.58
439.29
1.17
269.93
    (a) Includes 131 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
    (b) Includes 169 values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
                                              A-6

-------
Table A-ll.  Selected properties of tire-derived fuel burned in 1999.
Fuel Property
Mercury, lb/1012 Btu dry
Mercury, ppm dry"
Sulfur, % dry
Heat content, Btu/lb dry
Ash, % dry
Chlorine, ppm dry"
Range
0.38-19.89
0.006-0.33
0.86-2.8
11,457-17,035
0.75-23.14
100-6,483
Mean
3.58
0.05
1.56
15,261
7.50
1,059
Median
2.79
0.04
1.58
15,890
5.91
800
Standard deviation
2.78
0.04
0.26
1,384
4.49
933
    (a)  Includes four values reported at the non-detect level; 2 the non-detect level has been used in the analyses.
    (b)  Includes three values reported at the non-detect level; 2 the non-detect level has been used in the
        analyses.
                                               A-7

-------
Table A-12. Comparison of mercury content by fuel type as burned in 1999.
Fuel Type
Anthracite Coal
Bituminous Coal
South American
Bituminous Coal
Subbituminous Coal
Indonesian
Subbituminous Coal
Lignite Coal
Waste
Anthracite Coal
Waste
Bituminous Coal
Waste
Subbituminous Coal
Petroleum Coke
Tire-derived fuel
Number of
analyses
114
27,883
269
8,190
78
1,047
377
575
53
1,149
149
Mercury Concentration in Fuel (ppm)
Range
0.06-0.31
0.0-1.3
0.01 -0.95
0.005-0.9
0.01 -0.05
0.01 -0.75
0.02-0.54
0.03-1.18
0.07-0.35
0.0009-0.05
0.006-0.33
Mean
0.16
0.11
0.08
0.07
0.03
0.10
0.19
0.46
0.12
0.049
0.054
Median
0.16
0.09
0.06
0.06
0.03
0.08
0.17
0.45
0.11
0.03
0.04
Standard
deviation
0.05
0.09
0.07
0.04
0.01
0.09
0.08
0.30
0.05
0.05
0.04
                                       A-8

-------
Table A-13. Comparison of mercury content with fuel heat content as burned in 1999.
Fuel Type
Anthracite Coal
Bituminous Coal
South American
Bituminous Coal
Subbituminous Coal
Indonesian
Subbituminous Coal
Lignite Coal
Waste
Anthracite Coal
Waste
Bituminous Coal
Waste
Subbituminous Coal
Petroleum Coke
Tire-derived fuel
Number of
analyses
114
27,883
269
8,190
78
1,047
377
575
53
1,149
149
Ratio of Mercury to Fuel Heat Content (Ib Hg/trillion Btu)
Range
5.02-35.19
0.04-103.81
0.70-66.81
0.39-71.08
0.79-4.61
0.93-75.06
2.49-73.02
2.47-172.92
5.81 -30.35
0.06-32.16
0.38-19.89
Mean
15.28
8.59
5.94
5.74
2.51
10.54
29.31
60.50
11.42
23.18
3.58
Median
13.37
7.05
4.91
5.00
2.39
7.94
27.77
53.32
10.79
2.16
2.79
Standard
deviation
6.23
6.69
5.28
3.59
0.86
9.05
11.94
44.35
4.66
3.18
2.78
                                       A-9

-------
              Appendix B

Background Material of Methodology Used
   To Estimate 1999 National Mercury
    Emissions from Coal-fired Electric
             Utility Boilers
                  B-l

-------
                                                           September 15.2000


MEMORANDUM


To:          William Maxwell, EPA/OAQPS/ESD/CG


From:        Jeffrey Cole, RTI


Subject:       Draft Interim Report on Data Analyses


PURPOSE

The purpose of this memorandum is to discuss RTI's data analyses after the delivery of the Study
 of Hazardous Air Pollutant Emissions from Electric Utility Steam Generating Units Final Report
 to Congress (issued on February 24,1998)

 BACKGROUND

 Section 112(a)(8) of the Clean Air Act, as amended (CAA), defines an "electric utility
 steam-generating unit" as "any fossil-fuel-fired combustion unit of more than 25 megawatts
 electric (MWe) that serves a generator that produces electricity for sale."  A unit that cogenerates
 steam and electricity and supplies more than one-third of its potential electric output capacity and
 more than 25 MWe output to any utility power distribution system  for sale is also considered an
 electric utility (EU) steam-generating unit (i.e., utility unit).

 Section 112{n)(!)(A) also requires that:

         The EPA develop and describe alternative control strategies for hazardous air pollutants
         (HAPs) that may warrant regulation under section 112; and

         The EPA proceed with rolemaking activities under section 112 to control HAP emissions
         from utilities if EPA finds such regulation is appropriate and necessary after considering
         the results of the study.
                                        B-2

-------
Based on available information and current analyses, the EPA concluded that: mercury from
coal-fired utilities is the HAP of greatest potential concern and merits additional research and
monitoring; and, further research and evaluation are needed to gain a better understanding of the
risks and impacts of utility mercury emissions.

Two of the potential areas identified for further study included: (l)additional data on the mercury
content of various types of coal fired in U.S. utility units; and (2) additional data on mercury
emissions (e.g., how much is emitted from various types of units, how much is divalent vs
elemental mercury, and how do factors such as control device, fuel type, and plant configuration
affect emissions and speciation).

DATA COLLECTION

Following the issuance of the Report to Congress, EPA initiated, under the authority of
section 114 of the CAA, the Electric Utility Steam Generating Unit Mercury Emissions
Information Collection Effort (EU/ICE). As a part of this effort, RTI provided support to EPA in
 the EU/ICE development, distribution and processing. The EU/ICE has three basic
 sections: Part 1: General Facility Information; Part 0: Coal/Fuel Analysis; and Part ffl: Speciatcd
 Mercury Emissions Testing Data. Parts I, n, and ffl of the EU/ICE; were mailed out to all
 facilities that met the CAA definition.  As the completed Part I data forms were returned to EPA,
 they had their data extracted to create a unit configuration database.

 From January through December 1999, Part fl of the EU/ICE resulted in the collection of data for
 over 152,000 coal shipments from 1.143 units at 464 coal-fired facilities, a total of approximately
 40,500 individual mercury and chlorine analyses. To achieve this compilation of data, a
 web-based data collection system was developed  to allow the affected electric utilities to submit
 their Part II data over the Internet. The website provided a high-end technical solution to a large
 data collection effort and reduced the time required for data submission and the potential for
 data entry errors.  The website ended up collecting greater than 8 million pieces of data.

 Under Part HI of the EU/ICE, stack tests were conducted at 85 separate EU units to measure
 speciated mercury emissions. These data were reported by the companies to EPA and were
 entered into a database. These data were used in the emission program (see below) developed to
 estimate nationwide mercury emissions as well as to estimate mercury collection across different
 control devices, boilers, and fuel types.

  WEBSITE

  A web-based, interactive data collection system was developed to gather data on the mercury and
  chlorine content of the coals fired (including tire-derived fuel and petroleum coke) in U.S. utility
  units, as well as the fuel consumption data, directly for each electric utility company  or facility.
  The data were collected through the development and use of user-friendly web pages. The
  website had two different portions, a nonsecure and a secure portion, each with different
                                         B-3

-------
purposes. The nonsecure portion of the website was developed to serve as a conduit for
information from EPA in response to questions from the electric utility industry. The nonsecure
website also served as an entry point to the secure website.

The secure website was developed as an online collection point for EU/ICE Part 0 shipment,
analysis, and fuel usage data by unit and by month from 464 coal-fired facilities. To familiarize
EU/ICEindustry contacts with the on-line data entry procedures, a website user guide was created
in Adobe Acrobat format.

A toll-free customer service hotline (with email access) was manned to answer questions and
disseminate information to the participants of the EU/ICE. If a question was thought to be of
interest to more than just the facility/company that submitted it, the question and answer were
posted in the frequently-askcd-question (FAQ) webpage on the nonsecure website.  The process
was initiated by requesting e-mails from the EU/ICE contacts explaining their problems.  The
EPA would then answer the questioning facility by e-mail with a copy being posted on the
 website's FAQ pages.

 This method of answering EU/ICE questions had the additional benefit of creating a paper trail of
 EU/ICE modifications that could be recorded in the Electric Utility Hazardous Air Pollutant
 Emission Study Docket No. A-92-55.

 QUALITY CONTROL

 Quality control and customer service were important parts of the EU/ICE. To enhance quality
 control and customer service during the EU/ICE, the following actions were initiated.

 Early in the planning of the EU/ICE, EPA determined that the data from each facility should be
 able to pass a statistical confidence test on their coal/fuel analyses data. EPA statisticians
 established what confidence intervals were necessary to determine if the industry were sampling
 and analyzing at a sufficient frequency to obtain statistically reliable data. Programmers created
 a statistical evaluation portion of the website to determine if incoming data could pass the
 EPA-imposed confidence interval test for data quality.  The EU/ICE website was programmed to
 advise contacts on how to increase or decrease the frequency of analyses through a set of flow
 tables.

  For those EU/ICE contacts with numerous facilities who found webpage data entry too
  time-consuming, a data-upload capability was developed through the use of data entry
  spreadsheets (saved as text files). To familiarize EU/ICE contacts with these data entry
  spreadsheets, a spreadsheet user guide was created in Adobe Acrobat format.

  Before a data entry spreadsheet-based, tab-delimited, text file upload was allowed, the website
  checked the data for completeness. If there were data missing from the upload, loading would
  not occur, and an error message would be displayed. The website was also designed to flag
                                         B-4

-------
errors in data and to stop all incomplete or incorrect data loading. This error checking was
accomplished by implementing online quality control checks. These online checks would not
allow the entry of obviously incorrect data (e.g., zeros, text in numeric fields).

Programmers also created a duplicate development website (a mirror image of the live website,
but not available to the public) where new features and changes could be tried without affecting
the live website. This tactic proved to be a valuable tool because complex procedures could be
fully developed before being seen by the EU/ICE contacts.

DATA ANALYSIS

To estimate total mercury emissions from coal-fired electric utility units,  an emission factor
program (EFP) was developed. The EFP was built to accept data from the three data sources in
the EU/ICE. The first is a data input file containing plant configurations created from responses
from the ICE (Part I). The second source is a database containing detailed mercury analyses and
fuel consumption data, by unit, and by month for all of 1999 (Part D). The third data file is the
emission modification factor (EMF) database. This database contains results from the
 85 speciated mercury emission tests conducted by the electric utility industry under authority of
 the ICE (Part HI),  Eight of the 85 tests were done previously under a DOE study between
 1996 and 1998. The use of the eight tests was permitted by EPA because the test methods were
 the same as or similar to the EU/ICE Part ffl units (tested in 1999 and 2000) and were tested for
 speciated mercury.

 The program first categorizes each coal-fired electric utility unit by its fuel type/boiler
 type/emission control system(s) and assigns them to bins with other common units.  The program
 then categorizes the 85 units that were subject to stack testing under Part ffl by fuel type/boiler
 type/emission control system(s) and places them into similar combinations of bins. The program
 then computes unit-by-unit the mercury loading by analyzing the fuel burned by the unit for
 1999 and the concentration of mercury in the facilities fuel in 1999 (from Part H).  The mercury
 removals are then averaged from all the units in the stack-tested bins and that removal is applied
 to the mercury loading for each individual unit.  This procedure results in a kg/yr mercury output
 for each unit and  thus, when all units emissions are totaled, estimates nationwide mercury
 emissions.

 Because stack testing did not analyze every configuration of fuel type/emission control/boiler
 type system, the units that did not match perfectly had to be assigned a stack-tested bin in order
 for their emissions to be quantified. A hierarchy (fuel type/boiler type/emission control system)
 was used, as well as engineering judgment, to assign bins to those units that did not perfectly fit
 into a stack-tested bin.
                                         B-5

-------
RESULTS

The results of the EFP model are the following. The estimated national total of mercury emitted
from all coal-fired electric utility steam-generating units in 1999 is 43.4 tons. This amount of
mercury was emitted from 1,143 units at 464 facilities.
                                       B-6

-------
                                                           June 19,2001


MEMORANDUM


To:          William Maxwell, EPA/OAQPS/ESD/CG


From:        Jeffrey Cole, RTI


Subject:      Updated Draft Interim Report on Data Analyses


PURPOSE

The purpose of this memorandum is to discuss changes to RTFs data analyses data that were
made after my 9/15/00 memorandum to you entitled, "Draft Interim Report on Data Analyses."
In February-March 2001, EPA conducted a thorough quality assurance/quality-control (QA/QC)
review of the data extracted from the Hg emissions stack test reports. Changes to the national Hg
emissions total as a result of updated data and the initial presentation of the estimated specialed
 Hg split are presented in this memorandum.

 BACKGROUND

 Since the "Draft Interim Report on Data Analyses" was written, several changes have been made.
 RTI found three of four units that had reported their type of fuel burned incorrectly.  Five or six
 plants had incorrect Department of Energy's, Energy Information Administration (DOE/EIA)
 Office of the Regulatory Information System (ORIS) codes. These and other minor changes were
 made to the unit configuration, fuel usage, and mercury (Hg)/chlorine (Cl) analyses databases.
 The Brayton Point facility, was mis-located in the previous configuration output file. It was
 listed as being in Maryland, when it's actual location is Massachusetts.

 In February 2001, RTI began a thorough review of the data extracted from the 80 Electric
 Utility/Information Collection Effort (EU/ICE) speciated mercury emissions stack test reports.
 RTI developed and implemented a QA/QC study of the reports to determine the quality of the
 original data extraction and to correct any errors.  As part of this effort, RTI developed a
 sophisticated spreadsheet data entry form that allowed all test reports to be analyzed in a uniform
 manner using a single standard QC method.
                                       B-7

-------
RTI found several types of errors in data extracted from different test reports. Occasionally the
data errors were a result of incorrectly transferring data to EPA's national utility mercury model
or in misreporting the fuel a specific unit burned.  However, more data errors occurred when the
testing contractor or the report writing contractor incorrectly reported details about the testing
and process data. RTI reviewed the original Ontario Hydro Method data sheets, in each test
report, and examined each data point to determine its validity. RTI had to call several plant
representatives either to confirm items reported or to extract crucial  data that was not included in
the test report.

The individual data corrections inside a test report had some effect on the average EMF for that
test. However the more significant change occurred when a tested plant's fuel type was
incorrectly reported. This error could cause the test data to reside in a different database segment
(bin) than originally modeled. Once these errors were found and corrected RTI proceeded with
updating the Electric Utility National Mercury Emissions Model with the newly
quality-controlled data from the emissions test reports.

RESULTS

Because of this further QA/QC examination of data extracted from  the Hg emissions stack test
reports, the estimated national total of mercury emitted from all coal-fired electric utility
steam-generating units in 1999 is now approximately 48 tons. This amount of mercury was
emitted from 1,143 units at 461 facilities. This total  is composed of 1.48 tons/yr of particle-
 bound mercury (Kg"), 20.41 tons/yr of oxidized mercury(Hg2*), and 26.10 tons/yr of elemental
 mercury (Hg).
                                         B-8

-------
                                                           January 17,2001


MEMORANDUM


To:          William Maxwell, EPA/OAQPS/ESD/CG


From:        Jeffrey Cole, RTI
Subject:       Detailed Overview of How the Electric Utility Mercury National Emissions
              Model Estimated Nationwide Emissions

PURPOSE

       The purpose of this memorandum is to provide a detailed explanation of the mercury
national emissions model addressing specific inquiries about how the emissions modification
factors (EMFs) were developed and implemented in the model.  This memorandum is intended to
be an addendum to the memorandum "Draft Interim Report on Data Analyses" dated 9/15/00
available on EPA's website http://www.epa.gov/ttn/uatw/combust/utiltox/atoxpg.htinl.  Also
included in this memorandum are tables showing the EMFs used in the national emissions
 model.

 INTRODUCTION

        The national emissions model first categorizes each coal-fired electric utility unit by its
 fuel type/boiler type/emission control system(s) and classifies each unit under a simplified
 nomenclature. An example of this nomenclature would be a CYCJjONE/NONOX/WET
 w/ESP-CS. This unit would be a cyclone-fired boiler/furnace with a wet bottom (ash removed in
 a molten state) without NO, control followed by a cold-side electrostatic precipitator.

        The program then categorizes the 80 units were stack test data were available by fuel
 type/boiler type/emission control system(s) and classifies them into similar nomenclatures and
 places them in their similar categories (stack-tested bins). The program then computes,
 unit-by-unit, the mercury (Hg) loading in the boiler flue-gas by analyzing the fuel burned by the
 unit for 1999 and the concentration of Hg in the facility's fuel in 1999 (from Part E of the
 Electric Utility/Information Collection Effort {EU/ICE]). The Hg removals (obtained from
 EU/ICE Part JJI data) are then averaged from all the units in each of the stack-tested bins and
 those removals are applied to the Hg loading for each individual unit according to its bin
 configuration. This procedure results in a Hg output in kg/yr for each unit and thus, when all unit
 emissions are totaled, estimates  nationwide Hg emissions.  A detailed explanation follows.
                                       B-9

-------
DETAILED EXPLANATION

       Because stack testing did not analyze every configuration of fuel type/emission
control/boiler type system, the units that did not match perfectly had to be assigned a stack-tested
bin in order for their emissions to be quantified.  A hierarchy chosen was fuel type/boiler
type/emission control system.  Some engineering judgment was also used to assign bins to those
units that did not perfectly fit into a stack-tested bin. The speciated Hg concentration data,
extracted from the test reports and segregated into similar stack-tested bins, were then analyzed
by the national emissions model.

       Typically, each emissions test report contained data from three emission stack test runs
which included speciated Hg information.  The national emissions model sums the speciated Hg
concentrations at the inlet and outlet of the final control device of each unit.  Thus, an EMF is
determined for each run. The EMF is a fraction of the amount of total Hg exiting an air pollution
control device (APCD) divided by the amount of the total Hg entering that device. The EMF can
 also be defined as one minus the total Hg removal fraction. For example, an EMF of 0.68 is
 equal to a Hg removal of 0.32 (or 32 %). The equation used to compute the EMF also contained
 a correction made to the EU/ICE Part ffl test data to account for flowrate differences at the
 emissions test sampling inlet  and outlet (see Equation 1).

        The total EMF's of each run were then averaged in each bin (sec Table 1). This average
 EMF was multiplied by each  Hg loading (in kilograms) for each unit that matched the bin
 configuration. As a rule, when  a non-detect was encountered in the national emissions model,
 one-half the non-detection value was used. This procedure is consistent with the method EPA
 has used to deal with non-detect values since  the electric utility study was begun.

         A different method was used to average fuel type/boiler type/emission control system(s)
 from all dual controlled units (units having both a PM and an SO2 control device). Since stack
 test flue-gas speciated Hg was analyzed at the inlet and outlet of the last control device, the effect
 of the PM control on Hg removal on these dual controlled units was not clear. Thus:

        EPA decided that it would be more realistic to add the PM control device removal of Hg
         to the S02 control device removal of Hg for dual controlled units.

        The PM control device average EMF was taken from the bin of a unit with a similar fuel
         type/boiler type/PM emission control system to the dual controlled bin it was modifying.

        The average EMF of a tested unit with a single PM control device was multiplied by each
         individual  run EMF from a similarly configured dual controlled unit.

        These modified EMFs  were averaged.  This average was used to compute the Hg removal
         of a dual controlled unit in the national emissions model.

  Table 2 (the accompanying spreadsheet) shows how  the modified EMFs were calculated.
                                       B-10

-------
       Coal gasification units could not have their EMF's determined by either of the previous
methods.  Mercury testing was done in the coal feed, and emissions testing was done after the
boiler/furnace. These plants do not have conventional APCD's.  Thus, EMF's were determined
across two points, the Hg in the coal and the total Hg exiting the boiler/furnace for each run.

       It should be noted that not all EMF averages (single or dual control) showed removal of
Hg.  Some EMF averages showed a generation of Hg by the tested unit.  This discrepancy could
be accounted for by a number of factors, such as, an inaccurate mercury-in-coal analysis or an
inaccurate flue-gas Hg test before and after the last control device. None of the stack tests used
in these computations were found to have any documentation as to errors or problems found
during or after the stack testing. EPA decided that if no errors or problems were presented in the
stack tests that the data would be used.  EPA's main purpose in this model was to determine the
1999 Hg emissions total nationwide from all coal-fired electric utility power plants in the U.S.
Because of the combination of test inaccuracies and bin assignments, it was inevitable that some
units would be modeled as emitting less Hg than they actually emitted and some would be
modeled as emitting more. In some cases the computed emission rate was greater than the inlet
amount. Although physically improbable, these cases were used as computed to balance cases
that were modeled as emitting too little mercury.

       Two small discrepancies were found in the EMF portion of the nationwide emission
model. A single emission test run outlet speciated Hg content was entered incorrectly in the
thousandth place. This error produced an EMF of 0.238499 during Run 1 from the stack testing
on the Intermountain facility instead of the correct EMF of 0.23448. This difference would have
had the effect of changing the modified dual controlled unit combined EMF (Bin 12) from
0.11071 (used in  the existing model) to 0.110446. EPA feels that this change would have a
negligible effect on the national Hg emissions estimate. The second discrepancy was the
 incorrect use of the modified, dual controlled EMF for Bin 36 (0.58277) for Bin 32 instead of the
 correct modified  EMF (0.19477).  Bin 32 is used on only 3 units at one  plant and has a negligible
 effect on national Hg emissions estimate. Both  discrepancies are highlighted in Table 1 and 2
 and will be corrected in future versions of the national emissions model.
                                     B-ll

-------
Table  1.
A LISTING  OF  TEST  RUNS  WITH EMISSION  MODIFICATION  FACTORS
Test Report
Emission control device
Polk Power
Polk Power
Polk Power
Wabash River Qen Sta
Wabash River Qen Sta
Wabash River Gen Sta
Run * EKF
bin/ type 0
1 .55622
2 .56521
3 .64844
1 .71294
2 1.08624
3 .69517
EMF Analysis

coal to stack
coal to stack
coal to stack
coal to stack
coal to stack
coal to stack
Average emission factor .7107
Emission control device
Brayton Point 1
Brayton Point 1
Brayton Point 1
Brayton Point 3
_, Brayton Point 3
 Brayton Point 3
^ Meramec
Meramec
Meramec
George Neal South
George Neal South
George Neal South
Jack Watson
Jack Watson
Jack Watson
Widows Creek
Widows Creek
Widows Creek
Presque Isle 5
Presque Isle 5
Presque Isle S
bin/ type 1
1 .85267
2 .63886
3 .6781
1 .61791
2 .71529
3 .82287
1 .18986
2 .3003
3 .28044
1 .84823
2 .91106
3 1.52286
1 .78292
2 .623
3 .70962
1 .55725
2 .50002
3 .34892
1 .34713
2 .3813
3 .29159

across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
Average emission factor .61525
Emission control device
Presque Isle 6
Presoue Isle 6
bin/ type 2
1 .45459
2 .4971

across control
across control
                                                             Unit Configuration
                                                             COAL GAS
                                                             COAL GAS
                                                             COAL GAS
                                                             COAL GAS
                                                             COAL GAS
                                                             COAL GAS
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/ PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/ PC / NOX / DRY
                                                             CONV/ PC /NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX / WET
                                                             CONV/ PC /NONOX/WET
                                                             CONV/PC/NONOX/WET
                                                              CONV/PC/NONOX/WET
                                                              CONV/PC/NONOX/WET
                                                                                   PM Control
                                                                                                  SO2 Control
                                                                                                                 Fuel
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
COAL GAS
Bituminous
Bituminous
Bituminous
Subbituminous
Subbi tuminous
Subbi tuminous
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
COMP COAL
COMP COAL
COHP COAL
COMP COAL
COMP COAL
COMP COAL
NONE
NONE
NONE
COMP COAL
COMP COAL
COMP COAL
NONE
NONE
NONE
COMP COAL
COMP COAL
COMP COAL
COMP COAL
COMP COAL
COMP COAL
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bi tuminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
                                                                     ESP- CS
                                                                     ESP- CS
COMP COAL
COMP COAL
Bituminous/Pet Coke
Bituminous/Pet Coke

-------
        EMF
1-1-
                          Hg species concentration the outlet
                         ZHg species concentration at the inlet
                                                                [21 - Ot concentration at the inlet) \
                                                                [21 - Oj concentration at the outlet]]
                   Equation (1) Individual run EMF calculation with flow correction
Emission control
Salem Harbor
Salem Harbor
Salem Harbor
Average emission
Emission control
Cholla 3
Cholla 3
Cholla 3
Cliffside
Cliffside
Cliffside
Dunkirk
Dunkirk
Dunkirk
Gaston
Gaston
Gaston
device bin/ type 3
1 .12978
2 .07435
3 .06945
factor
.09119
across control
across control
across control

device bin/type 4
1
2
3
1
2
3
1
2
3
1
2
3
1.01304
.79577
1.12281
.99144
.61999
.57491
.93916
.6867
.82925
1.00192
1.41373
1.1
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
Average emission  factor
                             .92406
Emission control device bin/type   5
 Cholla 2                1    .70701
 Cholla 2                2    1.07508
 Cholla 2                3    .81712
 Bruce Mansfield         1    .85193
 Bruce Mansfield         2    .85059
 Bruce Mansfield         3    .92579
Average emission  factor
                             .87125
Emission control  device  bin/type   6
 Port Washington          1    .59319
 Port Washington          2    .47394
 Port Washington          3    .58628

 Average emission factor     .55114

 Emission  control  device  bin/type   7
across control
across control
across control
across control
across control
across control
across control
across control
across control
                                                             CONV/PC/NOX/SNCR/DRY
                                                             CONV/ PC/NOX/SNCR/DRY
                                                             CONV/PC/NOX/SNCR/DRY
                                                             CONV/PC/NONOX/DRY
                                                             COKV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             COHV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                              CONV/PC/NONOX/DRY
                                                              CONV/PC/NONOX/DRY
                                                              CONV/PC/NONOX/DRY
                                                                           ESP- CS
                                                                           ESP- CS
                                                                           ESP- CS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                           ESP- HS
                                                                    MECH/PARTSCRUB
                                                                    MECH/PARTSCRUB
                                                                    MECH / PARTSCRUB
                                                                    PARTSCRUB
                                                                    PARTSCRUB
                                                                    PARTSCRUB
                                                                    ESP- CS
                                                                    ESP- CS
                                                                    ESP- CS
                                                                                             Presque Isle  6           3    .41997
                                                                                      across control         CONV/PC/NONOX/WET
                                                                                      CS     COMP  COAL      Bituminous /Pet Coke
                                                                                                                                                    ESP-
Average emission factor .45722
COMP COAL
COMP COAL
COMP COAL
NONE
NONE
NONE
NONE
NONE
NONE
COMP COAL
COMP COAL
COMP COAL
NONE
NONE
NONE
WETSCRUB
WETSCRUB
WETSCRUB
NONE
NONE
NONE
IBENT INJECTION
IBENT INJECTION
IBENT INJECTION
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous

-------
W. H. Samnis
W. H. Sanrnis
W . H . Santni s
Shavmee
Shawnee
Shawnee
Valley
Valley
Valley
Valroont
Valroont
Valroont
1
2
3
1
2
3
1
2
3
1
2
3
.08586
.07033
.06992
.33271
.29648
.28837
.94333
1.00699
1.25146
.13019
.10475
.15846
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
 Average emission  factor
                              .3949
Emission control device bin/type 8
Logan Gen Plant
Logan Gen Plant
Logan Gen Plant
Mecklenburg Cogen
Mecklenburg Cogen
Mecklenburg Cogen
1
2
3
1
2
3
.01292
.01834
.01273
.00773
.03096
.02409

across control
across control
across control
across control
across control
across control
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                              CONV/PC/NOX/DRY
                                                              CONV/PC/NOX/DRY
                                                              CONV/PC/NOX/DRY
                                                              CONV/PC/NOX/DRY
                                                              CONV/PC/NOX/DRY
                                                              CONV/PC/NOX/DRY
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
                                                      BAGHOUSE
_v^^___B^P^_B^^_l__
 Average emission  factor
.01779
Emission control device bin/type 9
SET Birchwood Power Facility
SET Birchwood Power Facility
SET Birchwood Power Facility
1
2
3
.03439
.02211
.01613
across control
across control
across control
 Average emission  factor

Emission control device bin/type   10
 AES Cayuga  (NY)  (formerly Milliken)
 AES Cayuga  !NY)  (formerly Milliken)
 AES Cayuga  (NY)  (formerly Milliken!
                                             02421
             .3811?
             .32086
             .38365
last control
last control
last control
                                                                             CONV/PC/NOX/SCR/DRY
                                                                             CONV/PC/NOX/SCR/DRY
                                                                             CONV/PC/NOX/SCR/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
 Second unit  Average
 First unit correction factor
 Combined  emission factor

Emission control device bin/type  11
 R. D. Morrow            1      .46544
 R. D. Morrow            2      .50824
 R. D. Morrow            3      -553
             .36189
             .61525
             .22266
             last control
             last control
             last control
          CONV/PC/NOX/DRY
          CONV/PC/NOX/DRY
          CONV/PC/NOX/DRY
        ESP- HS
        ESP- HS
        ESP- HS
 Second unit Average            .5089
 First unit correction factor   .92406
 Combined emission factor       .47025
                                              NONE
                                              NONE
                                              NONE
                                              COMP COAL
                                              COMP COAL
                                              COMP COAL
                                              NONE
                                              NONE
                                              NONE
                                              COMP COAL
                                              COMP COAL
                                              COMP COAL
                                              SDA
                                              SDA
                                              SDA
                                              SDA
                                              SDA
                                              SDA
                                                                     BAGHOUSE
                                                                     BAGHOUSE
                                                                     BAGHOUSE
ESP- CS
ESP- CS
ESP- CS
WETSCRUB
WETSCRUB
WETSCRUB
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                     Bituminous
                                                              SDA    Bituminous
                                                              SDA    Bituminous
                                                              SDA    Bituminous
WETSCRUB
WETSCRUB
WETSCRUB
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous

-------
      Emission control device bin/type  12
03
tn
Intel-mountain
Intermountain
Intermountain
Clover
Clover
Clover
Second unit Average
First unit correction
1 .23448
2 .40334
3 .31322
1 .41849
2 .17405
3 .1345
.27968
factor .39585
last control
last control
last control
last control
last control
last control


Combined emission factor .11071
Emission control device
Gibson (03/00)
Gibson (03/00)
Gibson (03/00)
Gibson (10/99)
Gibson (10/99)
Gibson (10/99)
Montrose
Montrose
Montrose
Newton
Newton
Newton
St Clair
St Clair
St Clair
bin/ type 13
1 1.03854
2 .90019
3 .98339
1 .61081
2 .41305
3 .91315
1 .82536
2 1.02306
3 .87465
1 1.00101
2 .8367
3 .91535
1 .64369
2 .80355
3 .91291

across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
Average emission factor .84636
Emission control device
Clifty Creek
Clifty Creek
Clifty Creek
Columbia
Columbia
Columbia
Platte
Platte
Platte
Presque isle 9
Presque isle 9
Presque isle 9
bin/ type 14
1 .64416
2 .70486
3 .63012
1 .94982
2 .72657
3 1.01475
1 .73075
2 1.33203
3 1.02404
1 .99901
2 1.05232
3 1.05765

across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
                                                                   CONV/PC/NOX/DRY
                                                                   CONV/PC/NOX/DRY
                                                                   CONV/PC/NOX/DRY
                                                                   CONV/PC/NOX/DRY
                                                                   CONV/PC/NOX/DRY
                                                                   CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NONOX/DRY
CONV/PC/NONOX/DRY
CONV/PC/NONOX/DRY
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
                                                                    CONV/ PC/NOX/WET
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
                       ESP- HS
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
Bituminous
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
NONE
NONE
NONE
NONE
NONE
NONE
COMP
COM?
COMP
COMP
COMP
COMP
COMP
COMP
COMP






COAL
COAL
COAL
COAL
COAL
COAL
COAL
COAL
COAL
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous-
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
COMP COAL      Subbituminous
       Average emission  factor
                                    .90551

-------
Emission control
Clay Boswell 2
Clay Boswell 2
Clay Boswell 2
Comanche
Comanche
Comanche
Average emission
Emission control
Clay Boswell 3
Clay Boswell 3
Clay Boswell 3
Clay Boswell 4
Clay Boswell 4
Clay Boswell 4
Colstrip
Colstrip
Colstrip
Lawrence
OB Lawrence
H* Lawrence
Os
Average emission
Emission control
GRDA
GRDA
GRDA
Laramie River 3
Laramie River 3
Laramie River 3
Wyodak
Wyodak
Wyodak
device bin/ type 15
1 .2394
2 .12551
3
1
2
3
factor
.15683
.3142
.47328
.34481
.27567
device bin/ type 16
1
2
3
1
2
3
1
2
3
1
2
3

iractor
.92841
.86191
.94752
1.0308
1.18251
1.44396
1.86536
1.07793
.25733
1.00071
1.31139
1.14729

1.08793
device bin/ type 17
1
2
3
1
2
3
1
2
3
.57991
1.24231
.44283
4.55004
.42652
.37468
.56046
.60129
.60014
across control
across control
across control
across control
across control
across control


across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control
across control



across control
across control
across control
across control
across control
across control
across control
across control
across control
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
CONV/PC/NOX/DRY
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
COMP COAL
COMP COAL
COMP COAL
COMP COAL
COMP COAL
COMP COAL
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NONOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
                                                             CONV/PC/NOX/DRY
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
PARTSCRUB
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
WETSCRUB/COMP COAL
WETSCRUB/COMP COAL
WETSCRUB/COMP COAL
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
        Subbi tuminous
SDA
SDA
SDA
SDA
SDA
SDA
SDA
SDA
SDA
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Subbi tuminous
Average emission liactor
                            1.04202

-------
w
Emission control device bin/type 18
Craig 3
Craig 3
Craig 3
Rawhide
Rawhide
Rawhide
Sherburne
Sherburne
Sherburne






County
County
County
1
2
3
1
2
3
1
2
3
,6621
.58165
.68334
.82842
.67975
.52691
.7619
1.10936
.98954
across
across
across
across
across
across
across
across
across
control
control
control
control
control
control
control
control
control
        Average emission factor      .75811

       Emission control device bin/type   19
Jim Bridger
Jim Bridger
Jim Bridger
I/aramie River
I/aramie River
I/aramie River
Sam Seymour
Sam Seymour
Sam Seymour



1
1
1



1
2
3
1
2
3
1
2
3
.89679
.84355
.9694
.47432
.55464
.42468
1 .03425
.79918
.71422
last control
last control
last control
last control
last control
last control
last control
last control
last control
        Second unit Average            .74567
        First unit correction  factor   .84636
        Combined emission  factor       .63111
Emission control device
Charles R. I/owman
Charles R. I/owman
Charles R. I/owman
Coronado
Coronado
Coronado
Craig 1
Craig 1
Craig 1
Nava jo
Nava jo
Nava jo
San Juan
San Juan
San Juan
Second unit Average
First unit correction
bin /type
1
2
3
20
63558
70694
58821
1 1.12946
2
73182
3 1.11298
1
2
3
1
2
3
1
2
3
.
factor
Combined emission factor
56966
7707
97558
58001
87346
91752
6326
68646
57686
76586
9055
69348

last control
last control
last control
last control
last control
last control
last control
last control
last control
last control
last control
last control
last control
last control
last control



                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/ PC / NONOX / DRY
                                                                     CONV/ PC/NONOX/DRY
                                                                     CONV/ PC/NONOX/DRY
                                                                     CONV/ PC/NONOX / DRY
                                                                     CONV/ PC /NONOX /DRY
                                                                     CONV/PC/NONOX/DRY
                                                                     CONV/PC/NOX/WET
                                                                     CONV/PC/NOX/WET
                                                                     CONV/PC/NOX/WET
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/PC/NOX/DRY
                                                                     CONV/ PC/NONOX/DRY
                                                                     CONV/ PC /NONOX /DRY
                                                                     CONV/PC/NONOX/DRY
                                                                     CONV/PC/NONOX/DRY
                                                                     CONV/ PC /NONOX / DRY
                                                                     CONV/ PC /NONOX / DRY
                                            BAGHOUSE
                                            BAGHOUSE
                                            BAGHOUSE
                                            BAGHOUSE
                                            BAGHOUSE
                                            BAGHOUSE
                                            BAGHOUSE
                                            BAGHOUSE
                                            BAGHOUSE
                                            ESP- CS
                                            ESP- CS
                                            ESP- CS
                                            ESP- CS
                                            ESP- CS
                                            ESP- CS
                                            ESP- CS
                                            ESP- CS
                                            ESP- CS
SDA
SDA
SDA
SDA
SDA
SDA
SDA
SDA
SDA
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
Subbi tuminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
ESP- HS
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
Subbituminous
        Emission control device  bin/type  21
         Stanton Station 1       1   .95581
across  control
                      CONV/PC/NOX/DRY
                                             ESP- CS
                                                            NONE
                                                                           Lignite

-------
      Stanton Station 1
      Stanton Station 1
1.09703
1.05413
across control
across control
      Average emission  factor
                                  1.03566
     Emission control  device  bin/type  22
      La Cygne                l    .7682
      La Cygne                2    .75471
      La Cygne                3    .7783
          across control
          across control
          across control
     Average  emission  factor
                                  .76707
     Emission control  device bin/type  23
     Nelson  Dewey            l    .99871
     Nelson  Dewey            2    1.06895
     Nelson  Dewey            3    1.24952
          across control
          across control
          across control
     Average  emission  factor
                                  1.10573
    Emission control device bin/type  24
     Leland  Olds Station     1    1.18696
i    Leland  Olds Station     2    .66739
OB   Leland  Olds Station     3    .73521
          across control
          across control
          across control
     Average emission factor
                                  .86319
    Emission control device bin/type  25
     Stockton Cogen Company  1   .03689
     Stockton Cogen Company  2   .07829
     Stockton Cogen Company  3   .05571
         across control
         across control
         across control
CONWPC/NOX/DRY
CONWPC/NOX/DRY
                      CYCLONE/NOX/WET
                      CYCLONE/NOX/WET
                      CYCLONE/NOX/WET
                      CYCLONE/NONOX/WET
                      CYCLONE/NONOX/WET
                      CYCLONE/NONOX/WET
                      CYCLONE/NONOX/WET
                      CYCLONE/NONOX/WET
                      CYCLONE/NONOX/WET
                      FBC/SNCR
                      FBC/SNCR
                      FBC/SNCR
     Average emission factor     .05696

    Emission control device bin/type  26


     Bin/type 26 was not used
ESP- CS
ESP- CS
                      PARTSCRUB
                      PARTSCRUB
                      PARTSCRUB
                      ESP- HS
                      ESP- HS
                      ESP- HS
                      ESP-  CS
                      ESP-  CS
                      ESP-  CS
                     BAGHOUSE
                     BAGHOUSE
                     BAGHOUSE
NONE
NONE
               WETSCRUB
               WETSCRUB
               WETSCRUB
               COKP COAL
               COKP COAL
               COKP COAL
              NONE
              NONE
              NONE
              FBC
              FBC
              FBC
Emission control device bin/type 27
Scrubgrass Generating Company L.P.
Scrubgrass Generating Company L.P.
Scrubgrass Generating Company L.P.
1
2
3
.0008
.00101
.00148
across control
across control
across control
FBC/NONOX
FBC/NONOX
FBC/NONOX
BAGHOUSE
BAGHOUSE
BAGHOUSE
FBC
FBC
FBC
Lignite
Lignite
               Bituminous
               Bituminous
               Bituminous
               Subbi tuminous
               Subbi tuminous
               Subbi tuminous
               Lignite
               Lignite
               Lignite
               Bituminous
               Bituminous
               Bituminous
                                                                                                                     Waste Bituminous
                                                                                                                     Waste Bituminous
                                                                                                                     Waste Bituminous
     Average  emission  factor
                                             .0011

-------
Emission control  device bin/type  28
 R. M. Heskett            1    .50707     across control
 R. M. Heskett            2    .88912     across control
 R. M. Heskett            3    .45514     across control
 Average  emission  factor
                              .61711
Emission  control  device bin/type   29
 TNP-One                  1    .44798
 TNP-One                  2    .45928
 TNP-One                  3    .38122
         across  control
         across  control
         across  control
 Average  emission  factor
                              .42949
Emission  control device  bin/type   30
 Kline  Township Cogen    1    .00259
 Kline  Township Cogen    2    .00268
 Kline  Township Cogen    3    .00257
         across control
         across control
         across control
                               FBC/NONOX
                               FBC/NONOX
                               FBC/NONOX
FBC/NONOX
FBC/NONOX
FBC/NONOX
FBC/NONOX
FBC/NONOX
FBC/NONOX
                      ESP- CS
                      ESP- CS
                      ESP- CS
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
               FBC
               FBC
               FBC
                                                                                                         FBC
                                                                                                         FBC
                                                                                                         FBC
                                                                                                         FBC
                                                                                                         FBC
                                                                                                         FBC
                                                                                                                       Lignite
                                                                                                                       Lignite
                                                                                                                       Lignite
Lignite
Lignite
Lignite
Waste Anthracite
Waste Anthracite
Waste Anthracite
 Average  emission factor
                              .00261
CO
Dwayne Collier Battle Cogen
Dwayne Collier Battle Cogen
Dwayne Collier Battle Cogen
1
2
3
.07072
.05519
.06285
across control
across control
across control
 Average  emission factor
                                     .06292
Emission  control  device  bin/type   32
 Big Bend                1      .33301    last control
 Big Bend                2      .25633    last control
 Big Bend                3      .35991    last control

 Second unit Average            .31642
 First unit  correction factor   .61525
 Combined  emission factor      .58277

Emission  control  device  bin/type   33
 Big Brown
 Big Brown
 Big Brown
 Monticello  1-2
 Monticello  1-2
 Monticello  1-2
1
2
3
1
2
.99082
1 .12876
1 .12202
.71314
1 .51459
last control
last control
last control
last control
last control
                                                                      STOKER/NOX/DRY
                                                                      STOKER/NOX/DRY
                                                                      STOKER/NOX/DRY
                               STOKER/NOX/WET
                               STOKER/NOX/WET
                               STOKER/NOX/WET
1.40836   last control
                               CONV/PC/NONOX/DRY
                               CONV/PC/NONOX/DRY
                               CONV/PC/NONOX/DRY
                               CONV/PC/NONOX/DRY
                               CONV/PC/NONOX/DRY
                               CONV/PC/NONOX/DRY
                                                             BAGHOUSE
                                                             BAGHOUSE
                                                             BAGHOUSE
                                             SDA
                                             SDA
                                             SDA
                      ESP- CS
                      ESP- CS
                      ESP- CS
               WETSCRUB
               WETSCRUB
               WETSCRUB
                       ESP-  CS/BAGHOUSE  NONE
                       ESP-  CS/BAGHOUSE  NONE
                       ESP-  CS/BAGHOUSE  NONE
                       ESP-  CS/BAGHOUSE  NONE
                       ESP-  CS/BAGHOUSE  NONE
                       ESP-  CS/BAGHOUSE  NONE
                              Bituminous
                              Bituminous
                              Bituminous
                                                                                                                        Bituminous
                                                                                                                        Bituminous
                                                                                                                        Bituminous
                                      Lignite
                                      Lignite
                                      Lignite
                                      Lignite
                                      Lignite
                                      Lignite
 Second  unit  Average           1.14628
 First  unit correction factor 1.03566
 Combined emission factor     1.18716
 Emission control device bin/type  34

-------
03
Antelope Valley 1
Antelope Valley 2 1
Antelope Valley 3
Stanton Station 10 1
Stanton Station 10 21
Stanton Station 10 31
Average emission factor
05466 across control
.06273 across control
91505 across control
98757 across control
.0254 across control
.01769 across control
84385
Emission control device bin/type 35
Lewis w Clark 1
Lewis w Clark 2
Lewis w Clark 3
Average emission factor
49252 across control
61262 across control
91187 across control
67234
Emission control device bin/type 36
Monticello 3 1
Monticello 3 2
Monticello 3 3
Limestone 1
Limestone 2
Limestone 3
Second unit Average
First unit correction factor
Combined emission factor
.78694 last control
.54432 last control
.57559 last control
.50597 last control
.52407 last control
.43934 last control
.5627
1.03566
.58277
      Emission control device bin/type  37
       Bay Front               1    .99664    across control
       Bay Front               2    1.46539   across control
       Bay Front               3    2.24996   across control
Average
Emission
Bailly
Bailly
Bailly
emission factor 1.57066
control device bin/type 38
1 .55564
2 .54693
3 .55714
last control
last control
last control
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NOX/DRY
                                                                    CONV/PC/NONOX/DRY
                                                                    CONV/PC/NONOX/DRY
                                                                    CONV/PC/NONOX/DRY
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
                                                                    CONV/PC/NOX/WET
CYCLONE/NONOX/WET
CYCLONE / NONOX / WET
CYCLONE / NONOX / WET
                                                                    CYCLONE / NONOX / WET
                                                                    CYCLONE/NONOX/WET
                                                                    CYCLONE/NONOX/WET
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                      BAGHOUSE
                      PARTSCRUB
                      PARTSCRUB
                      PARTSCRUB
MECH
MECH
MECH
                       ESP-  CS
                       ESP-  CS
                       ESP-  CS
               SDA
               SDA
               SDA
               SDA
               SDA
               SDA
               NONE
               NONE
               NONE
COMP COAL
COMP COAL
COMP COAL
               WETSCRUB
               WETSCRUB
               WETSCRUB
               Lignite
               Lignite
               Lignite
               Lignite
               Lignite
               Lignite
               Lignite
               Lignite
               Lignite
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
ESP- CS
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
WETSCRUB
Lignite
Lignite
Lignite
Lignite
Lignite
Lignite
Bituminous
Bituminous
Bituminous
               Bituminous
               Bituminous
               Bituminous
       Second unit Average            .55324
       First unit correction  factor   .86319
       Combined emission factor       .47755

-------
      Emission  control device bin/type  39
      Coyote                   1   .88207
      Coyote                   2   .8932
      Coyote                   3   .94525
across control
across control
across control
       Average  emission factor
                                   .90684
      Emission  control  device bin/type  40
      AES  Hawaii              1   .44101
      AES  Hawaii              2   .4865
      AES  Hawaii              3   .35088
across control
across control
across control
CYCLONE/NONOX/WET
CYCLONE/NONOX/WET
CYCLONE/NONOX/WET
FBC/SNCR
FBC/SNCR
FBC/SNCR
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
BAGHOUSE
SDA
SDA
SDA
FBC
FBC
FBC
Lignite
Lignite
Lignite
Subbi tuminous
Subbi tuminous
Subbi tuminous
      Average  emission  factor
                                   .42613
w

-------
                                Appendix C
                   Summary of Part II EPA ICR Data

               Mercury Capture Efficiencies of Existing
                        Post-combustion Controls
                            Used for Coal-fired
                          Electric Utility Boilers
Table C-l.  Post-combustion controls: cold-side ESPs	C-2
Table C-2.  Post-combustion controls: hot-side ESPs	C-4
Table C-3.  Post-combustion controls: fabric filters	C-5
Table C-4.  Post-combustion controls: miscellaneous PM controls	C-6
Table C-5.  Post-combustion controls: dry FGD scrubbers using ESP	C-7
Table C-6.  Post-combustion controls: dry FGD scrubbers using FF	C-8
Table C-7.  Post-combustion controls: wetFGD scrubbers	C-10
Table C-8.  Cyclone-fired boilers	C-13
Table C-9.  Fluidized-bed combustion	C-14
                                     C-l

-------
Table C-1.  Post-combustion controls: cold-side ESPs.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In HgT In HgT In HgpOut Hg2+Out Hg Out HGTOut %R HT %R HgT
No. O-H O-H O-H O-H Coal O-H O-H O-H O-H O-H Coal
Bituminous Coal, PC Boiler with CS-ESP
Brayton Point 1
Brayton Point 1
Brayton Point 1
Average
Brayton Point 3
Brayton Point 3
Brayton Point 3
Average
Gibson 0300
Gibson 0300
Gibson 0300
Average
Gibson 1099
Gibson 1099
Gibson 1099
Average
Meramec
Meramec
Meramec
Average
Jack Watson
Jack Watson
Jack Watson
Average
Widows Creek
Widows Creek
Widows Creek
Average
1
2
3

1
2
3

1
2
3

1
2
3

1
2
3

1
2
3

1
2
3

2.01
2.61
2.17
2.27
3.14
1.83
1.40
2.12
1.94
1.25
1.75
1.65
5.53
27.57
4.60
12.57
7.61
9.34
5.65
7.53
3.60
4.91
4.64
4.38
3.36
2.98
2.87
3.07
3.34
3.69
3.50
3.51
3.67
3.14
3.26
3.36
31.74
38.06
44.44
38.08
10.33
3.78
11.02
8.38
0.49
1.36
1.93
1.26
1.22
1.16
0.60
0.99
0.44
0.45
0.47
0.45
0.32
0.25
0.26
0.28
0.36
0.34
1.60
0.77
4.39
2.92
1.65
2.99
2.34
1.25
1.58
1.72
0.14
0.44
0.62
0.40
0.92
0.25
0.23
0.47
0.54
0.51
0.50
0.51

5.68
6.55
5.93
6.05
7.17
5.31
6.26
6.25
38.08
42.23
47.85
42.72
18.20
32.60
17.20
22.67
8.23
11.15
8.19
9.19
5.74
6.32
5.46
5.84
4.34
3.94
3.83
4.04

6.80
4.21
5.01
5.34
8.55
5.30
5.58
6.48
13.69
13.33
13.53
13.52
14.00
15.09
14.69
14.59
8.46
10.72
5.89
8.36
4.70
5.67
6.20
5.52
3.11
2.67
2.15
2.64

0.77
0.75
0.77
0.76
0.78
0.96
0.01
0.59
0.00
0.01
0.01
0.01
0.03
0.05
0.03
0.04
0.00
0.01
0.00
0.00
0.05
0.05
0.06
0.05
0.14
0.01
0.01
0.06

3.83
3.19
3.02
3.35
3.18
2.47
3.43
3.03
32.03
32.21
42.87
35.70
6.06
8.41
11.03
8.50
0.76
2.20
1.51
1.49
2.57
2.99
2.92
2.83
1.48
1.28
0.65
1.14

0.23
0.25
0.24
0.24
0.46
0.37
1.70
0.85
7.51
5.80
4.17
5.83
5.03
5.00
4.65
4.90
0.80
1.13
0.79
0.91
1.87
0.89
0.89
1.22
0.78
0.68
0.67
0.71

4.84
4.18
4.02
4.35
4.43
3.80
5.15
4.46
39.54
38.01
47.05
41.54
11.12
13.46
15.71
13.43
1.56
3.35
2.30
2.40
4.49
3.94
3.88
4.10
2.40
1.97
1.34
1.90

14.73
36.11
32.19
27.68
38.21
28.47
17.70
28.13
-3.85
9.98
1.66
2.60
38.92
58.69
8.68
35.43
81.01
69.97
71.96
74.32
21.71
37.70
29.04
29.48
44.75
50.00
65.11
53.29

28.86
0.68
19.64
16.39
48.20
28.27
7.71
28.06
-188.83
-185.13
-247.76
20.62
10.76
-6.93
8.15
81.54
68.77
60.99
70.43
4.39
30.53
37.45
24.13
22.95
26.25
37.81
29.00
Average 4.80 8.00 1.02 13.82 8.06 0.22 8.00 2.09 10.31 35.85 -4.44
Minimum
Maximum
STDEV



1.25
27.57
5.62
0.44
44.44
13.05
0.14
4.39
1.09
3.83
47.85
13.86
2.15
15.09
4.35
Bituminous Coal and Pet Coke, PC Boiler with CS-ESP
Presque Isle 5
Presque Isle 5
Presque Isle 5
Average
Presque Isle 6
Presque Isle 6
Presque Isle 6
Average
1
2
3

1
2
3

4.56
3.60
5.06
4.40
2.73
2.97
2.96
2.89
0.48
0.66
0.45
0.53
0.63
0.72
0.62
0.65
0.14
0.57
0.12
0.27
0.17
0.25
0.17
0.20
5.17
4.82
5.63
5.21
3.52
3.94
3.75
3.74
4.27
3.48
3.93
3.89
2.29
4.34
3.85
3.49
0.00
0.96
0.34

0.01
0.00
0.02
0.01
0.06
0.03
0.03
0.04
0.65
42.87
12.01

0.72
0.82
0.71
0.75
0.84
1.00
0.73
0.86
0.23
7.51
2.24

1.06
1.02
0.91
1.00
0.70
0.93
0.81
0.81
1.34
47.05
13.67

1.80
1.84
1.64
1.76
1.60
1.96
1.57
1.71
-3.85
81.01
23.90

65.29
61.87
70.84
66.00
54.54
50.29
58.00
54.28
-247.76
81.54
88.31

57.92
47.14
58.19
54.42
30.10
54.87
59.17
48.05
Average 3.65 0.59 0.24 4.47 3.69 0.02 0.81 0.90 1.73 60.14 51.23
Minimum
Maximum
STDEV



2.73
5.06
0.96
0.45
0.72
0.11
0.12
0.57
0.17
3.52
5.63
0.86
2.29
4.34
0.75
0.00
0.06
0.02
0.71
1.00
0.11
0.70
1.06
0.13
1.57
1.96
0.15
50.29
70.84
7.44
30.10
59.17
11.25
                                                                    (continued)
                                     C-2

-------
Table C-1.  (continued).
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In HgT In
No. O-H O-H O-H O-H
Bituminous Coal, PC Boiler with SNCRand CS-ESP
Salem Harbor
Salem Harbor
Salem Harbor
Average
1
2
3

4.12
4.09
3.96
4.06
0.32
0.04
0.06
0.14
0.32
0.16
0.15
0.21
Subbituminous Coal, PC Boiler with CS-ESP
Montrose
Montrose
Montrose
Average
George Neal So.
George Neal So.
George Neal So.
Average
Newton
Newton
Newton
Average
1
2
3

1
2
3

1
2
3

1.94
0.91
1.63
1.49
0.17
0.07
0.02
0.09
0.04
0.04
0.08
0.05
1.85
2.52
2.85
2.41
4.78
4.35
3.53
4.22
0.58
0.63
1.65
0.95
6.00
4.93
4.68
5.20
6.34
8.24
3.77
6.12
9.70
9.85
9.26
9.61
4.76
4.29
4.17
4.41
9.79
8.36
9.16
9.10
11.29
12.66
7.32
10.42
10.32
10.52
11.00
10.61
HgTln
Coal
3.44
2.35
3.27
3.02
44.90
51.99
47.76
48.21
8.96
7.82
10.19
8.99
9.07
8.05
9.34
8.82
HgpOut
O-H
0.07
0.10
0.08
0.08
0.03
0.02
0.02
0.02
0.03
0.06
0.02
0.04
0.00
0.00
0.00
0.00
Hg^Out
O-H
0.28
0.07
0.08
0.14
2.57
2.60
2.30
2.49
4.07
4.60
4.74
4.47
2.26
1.66
2.04
1.99
Hg Out
O-H
0.27
0.15
0.14
0.19
5.48
5.94
5.69
5.70
5.47
6.87
6.39
6.24
8.07
7.13
8.03
7.74
HGT Out
O-H
0.62
0.32
0.29
0.41
8.08
8.56
8.01
8.22
9.58
11.53
11.15
10.75
10.33
8.80
10.07
9.73
%R HT
O-H
87.07
92.57
93.06
90.90
17.46
-2.31
12.54
9.23
15.18
8.89
-52.29
12.04
-0.10
16.33
8.46
8.23
%R HgT
Coal
82.11
86.40
91.16
86.55
82.01
83.54
83.22
82.92
-6.90
-47.37
-9.36
-21.21
-14.00
-9.28

-10.36
Average 0.54 2.53 6.98 10.05 22.01 0.02 2.98 6.56 9.57 2.69 17.12
Minimum
Maximum
STDEV
SPF




0.02
1.94
0.76
0.05
0.58
4.78
1.50
0.25
3.77
9.85
2.34
0.69
7.32
12.66
1.61
1.00
7.82
51.99
19.75
1.00
Subbituminous/ Bituminous Coal, PC Boiler with CS-ESP
St Glair
St Clair
St Clair
Average
1
2
3

2.53
2.87
0.98
2.13
2.29
2.13
1.94
2.12
Lignite, PC Boiler with CS-ESP
Stanton 1
Stanton 1
Stanton 1
Average
1
2
3
0.04
0.13
0.08
0.15
0.13
0.05
1.97
1.40
4.28
2.55
11.96
10.81
11.66
6.79
6.39
7.20
6.79
12.15
11.06
11.79
16.26
14.36
17.71
16.11
31.51
41.24
19.94
0.00
0.06
0.02
0.00

0.01
0.01
0.01
0.01
0.04
0.02
0.01
1.66
4.74
1.16
0.31

1.35
1.39
1.33
1.35
0.42
0.43
0.45
5.47
8.07
1.02
0.69

3.01
3.74
5.24
4.00
11.16
11.68
11.97
8.01
11.53
1.30
1.00

4.37
5.14
6.57
5.36
11.62
12.14
12.43
-52.29
17.46
21.75


35.63
19.65
8.71
21.33
4.42
-9.70
-5.41
-47.37
83.54
50.89


73.13
64.24
62.89
66.75
63.13
70.56
37.67
0.08 0.11 11.48 11.67 30.89 0.02 0.44 11.60 12.06 -3.57 57.12
                                     c-:

-------
Table C-2. Post-combustion controls: hot-side ESPs.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg" In Hg In
No. O-H O-H O-H
Bituminous Coal, PC Boiler with HS-ESP
Cliffside
Cliffside
Cliffside
Average
Gaston
Gaston
Gaston
Average
Dunkirk
Dunkirk
Dunkirk
Average
1
2
3

1
2
3

1
2
3

0.17
0.09
0.08
0.11
4.28
2.57
0.43
2.42
0.09
0.01
0.01
0.04
3.72
3.54
4.15
3.80
0.86
0.71
3.94
1.84
8.56
8.91
9.15
8.87
3.31
3.33
7.27
4.63
2.64
3.56
2.83
3.01
2.82
1.43
3.20
2.48
HgTln
O-H
7.20
6.95
11.49
8.55
111
6.84
7.20
7.27
11.47
10.36
12.36
11.40
HgTln
Coal
5.43
3.84
8.80
6.02
5.20
6.27
4.70
5.39
10.06
10.30
9.65
10.00
HgpOut
O-H
0.41
0.10
0.10
0.20
0.74
0.40
1.15
0.76
0.21
0.08
0.03
0.11
HgpOut
O-H
2.79
2.27
3.97
3.01
4.70
5.80
4.73
5.08
6.89
4.57
6.40
5.95
Hg Out
O-H
3.95
1.95
2.54
2.81
2.34
3.47
2.04
2.62
3.67
2.46
3.82
3.32
HGT Out
O-H
7.14
4.31
6.61
6.02
7.78
9.66
7.92
8.46
10.77
7.12
10.25
9.38
%R HT
O-H
0.86
38.00
42.51
27.12
-0.19
-41.37
-10.00
6.08
31.27
17.08
18.14
%R HgT
Coal
-31.58
-12.17
24.94
-6.27
-49.56
-54.19
-68.41
-7.09
30.90
-6.26
5.85
Average 0.86 4.84 3.38 9.07 7.14 0.36 4.68 2.92 7.95 9.36 -19.27
Minimum
Maximum
STDEV



0.01
4.28
1.52
0.71
9.15
3.28
1.43
7.27
1.59
6.84
12.36
2.30
3.84
10.30
2.55
0.03
1.15
0.37
2.27
6.89
1.54
1.95
3.95
0.80
4.31
10.77
2.02
-41.37
42.51
26.41
-68.41
30.90
34.54
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg"' In Hg In HgT In HgT In
No. O-H O-H O-H O-H Coal
Subbituminous Coal, PC Boiler (Dry Bottom) with HS-ESP
ChollaS
ChollaS
ChollaS
Average
Columbia
Columbia
Columbia
Average
1
2
3

1
2
3

0.07
0.51
0.45
0.34
0.01
0.01
0.01
0.01
0.37
0.32
0.43
0.37
0.93
5.82
0.46
2.41
1.93
0.46
0.61
1.00
14.27
13.40
14.65
14.11
2.37
1.28
1.49
1.71
15.22
19.24
15.12
16.52
51.98
54.43
40.48
48.96
9.85
10.30
10.35
10.17
HgpOut
O-H
0.01
0.01
0.01
0.01
0.00
0.00
0.00
0.00
HgpOut
O-H
0.51
0.01
0.39
0.30
2.74
2.16
2.65
2.51
Hg Out
O-H
1.87
1.00
1.27
1.38
11.71
11.82
12.68
12.07
HGT Out
O-H
2.40
1.02
1.67
1.70
14.45
13.98
15.34
14.59
%R HT
O-H
-1.30
20.42
-12.28
2.28
5.02
27.34
-1.47
10.30
%R HgT
Coal
95.39
98.12
95.87
96.46
-46.78
-35.71
-48.18
-43.56
Average 0.18 1.39 7.55 9.12 29.57 0.01 1.41 6.73 8.14 6.29 26.45
Minimum
Maximum
STDEV



0.01
0.51
0.23
0.32
5.82
2.18
0.46
14.65
7.21
1.28
19.24
8.26
9.85
54.43
21.77
Subbituminous Coal, PC Boiler (Wet Bottom) with HS-ESP
Platte
Platte
Platte
Average
Presque Isle 9
Presque Isle 9
Presque Isle 9
Average
1
2
3

1
2
3

0.03
0.02
0.03
0.03
0.04
0.01
0.01
0.02
4.15
1.92
4.39
3.48
0.14
0.14
0.10
0.13
9.82
11.31
11.63
10.92
6.70
6.89
6.43
6.68
14.00
13.25
16.04
14.43
6.89
7.05
6.55
6.83
11.10
9.65
6.05
8.93
9.86
8.92
9.91
9.56
0.00
0.01
0.00

0.01
0.01
0.01
0.01
0.00
0.00
0.00
0.00
0.01
2.74
1.24

1.45
0.78
1.51
1.25
0.57
0.67
0.54
0.59
1.00
12.68
5.87

8.76
16.86
14.90
13.51
6.30
6.74
6.38
6.47
1.02
15.34
7.09

10.23
17.65
16.43
14.77
6.88
7.41
6.92
7.07
-12.28
27.34
14.88

26.93
-33.20

-2.89
0.10
-5.23

-3.63
-48.18
98.12
76.82

7.88
-82.85

-82.18
30.22
16.87
30.11
25.73
Average 0.02 1.80 8.80 10.63 9.25 0.01 0.92 9.99 10.92 -3.26 -28.22
Minimum
Maximum
STDEV



0.01
0.04
0.01
0.10
4.39
2.03
6.43
11.63
2.41
6.55
16.04
4.27
6.05
11.10
1.72
Subbituminous/Bituminous Coal, PC Boiler with HS-ESP
Clifty
Clifty
Clifty
Average
3
1
2

0.01
0.40
0.02
0.14
3.41
2.35
3.58
3.11
11.46
11.17
11.13
11.25
14.87
13.92
14.73
14.51
7.84
8.02
7.66
7.84
0.00
0.01
0.01

0.07
0.70
0.01
0.26
0.54
1.51
0.44

5.50
3.60
5.04
4.71
6.30
16.86
4.69

3.80
4.67
5.34
4.60
6.88
17.65
4.91

9.37
8.96
10.39
9.57
-33.20
26.93
19.13

36.99
35.58
29.51
34.03
-171.57
30.22
82.08

-19.53
-11.78

-22.29
                                   C-4

-------
Table C-3.  Post-combustion controls: fabric filters.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In
No. O-H O-H
Bituminous Coal, PC Boiler with FF
Sammis
Sammis
Sammis
Average
Valmont
Valmont
Valmont
Average
1
2
3

1
2
3

11.78
15.35
14.62
13.92
0.92
0.92
1.23
1.02
0.48
0.50
0.51
0.50
0.12
0.07
0.10
0.10
Hgln
O-H
0.61
0.54
0.52
0.55
0.18
0.14
0.17
0.17
HgTln
O-H
12.86
16.38
15.65
14.97
1.22
1.12
1.51
1.29
HgTln
Coal
6.64
9.54
9.55
8.58
0.80
0.44
0.60
0.61
HgpOut
O-H
0.01
0.01
0.02
0.01
0.00
0.00
0.00
0.00
Hg^Out
O-H
0.49
0.58
0.51
0.53
0.12
0.10
0.21
0.14
Hg Out
O-H
0.61
0.55
0.57
0.57
0.04
0.02
0.03
0.03
HGT Out
O-H
1.11
1.14
1.10
1.12
0.16
0.12
0.24
0.17
%R HT
O-H
91.37
93.04
92.97
92.46
86.98
89.53
84.15
86.89
%R HgT
Coal
83.28
88.05
88.48
86.60
80.04
73.26
60.16
71.16
Average 7.47 0.30 0.36 8.13 4.59 0.01 0.34 0.30 0.64 89.67 78.88
Minimum
Maximum
STDEV



0.92
15.35
7.16
0.07
0.51
0.22
0.14
0.61
0.22
1.12
16.38
7.59
0.44
9.55
4.49
0.00
0.02
0.01
0.10
0.58
0.22
0.02
0.61
0.30
Bituminous Coal/Pet. Coke, PC Boiler with FF (Measurements not valid, disregard)
Valley
Valley
Valley
Average
1
2
3

0.04
0.05
0.04
0.04
1.44
1.49
1.22
1.38
1.21
0.45
0.67
0.78
2.69
1.99
1.92
2.20
Bituminous/Subbituminous Coal, PC Boiler with FF
Shawnee
Shawnee
Shawnee
Average
1
2
3

3.18
3.01
3.44
3.21
0.58
0.98
0.57
0.71
0.72
0.66
0.67
0.68
Subbituminous Coal, PC Boiler with FF
Boswell 2
Boswell 2
Boswell 2
Average
Comanche
Comanche
Comanche
Average
2
3
1

1
3
2

1.99
0.83
2.75
1.85
1.81
5.27
2.59
3.23
1.26
1.15
1.81
1.41
3.93
1.28
1.45
2.22
1.46
2.49
1.60
1.85
5.71
3.67
5.77
5.05
4.48
4.65
4.68
4.61
4.71
4.46
6.16
5.11
11.46
10.22
9.82
10.50
0.95
1.33
1.52
1.27
2.39
4.29
2.66
3.11
4.35
5.20
8.35
5.97
15.91
14.24
17.08
15.74
0.11
0.04
0.00
0.05
0.01
0.02
0.01
0.01
0.00
0.00
0.07
0.03
0.00
0.00
0.00
0.00
2.02
1.55
1.89
1.82
0.63
0.61
0.60
0.61
0.35
0.58
1.26
0.73
3.33
3.20
3.99
3.51
0.41
0.42
0.52
0.45
0.84
0.75
0.68
0.76
0.23
0.12
0.14
0.16
0.27
0.33
0.65
0.42
0.12
1.14
0.52

2.54
2.00
2.41
2.31
1.49
1.37
1.28
1.38
0.59
0.70
1.47
0.92
3.60
3.52
4.65
3.92
84.15
93.04
3.54

5.67
-0.70

-6.73
66.73
70.51
72.62
69.95
87.45
84.32
76.06
82.61
68.58
65.52
52.67
62.26
60.16
88.48
10.76

-165.84
-50.84

-91.81
37.66
68.03
51.82
52.50
86.43
86.54
82.34
85.10
77.37
75.26
72.80
75.14
Average 2.54 1.81 3.45 7.80 10.86 0.01 2.12 0.29 2.42 72.43 80.12
Minimum
Maximum
STDEV



0.83
5.27
1.50
1.15
3.93
1.06
1.46
5.77
1.94
4.46
11.46
3.06
4.35
17.08
5.59
0.00
0.07
0.03
0.35
3.99
1.57
0.12
0.65
0.20
0.59
4.65
1.72
52.67
87.45
12.91
72.80
86.54
5.85
                                     C-5

-------
Table C-4. Post-combustion controls: miscellaneous PM controls.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In HgT In
No. O-H O-H O-H O-H
TX Lignite, PC Boiler with CS-ESP and FF (COHPAC)
Bigbrown
Bigbrown
Bigbrown
Average
Monticello 1-2
Monticello 1-2
Monticello 1-2
Average
1
2
3

1
2
3

2.59
0.54
0.14
1.09
15.97
0.37
7.97
8.10
8.35
10.37
14.14
10.95
22.54
14.82
22.74
20.03
31.24
27.31
21.93
26.83
8.82
46.29
44.19
33.10
42.18
38.21
36.21
38.87
47.34
61.48
74.90
61.24
HgTln
Coal
50.86
49.95
46.92
49.24
53.79
54.09
84.65
64.18
HgpOut
O-H
0.01
0.01
0.01
0.01
0.17
0.11
0.08
0.12
Hg^Out
O-H
16.58
17.66
18.49
17.58
32.01
78.08
86.89
65.66
Hg Out
O-H
25.20
25.47
22.12
24.26
1.58
14.93
18.51
11.67
HGT Out
O-H
41.80
43.13
40.62
41.85
33.76
93.11
105.48
77.45
%R HT
O-H
0.92
-12.88

-7.68
28.69
-51.46
-40.84
-21.20
%R HgT
Coal
17.82
13.65
13.42
14.96
37.23
-72.13
-24.61
-19.83
Average 4.60 15.49 29.96 50.05 56.71 0.07 41.62 17.97 59.65 -13.63 -2.44
Minimum
Maximum
STDEV



0.14
15.97
6.31
8.35
22.74
6.03
8.82
46.29
14.07
36.21
74.90
15.16
Subbituminous Coal, PC Boiler with PM Scrubber
Boswell 3
Boswell 3
Boswell 3
Average
1
2
3

0.01
0.01
0.06
0.03
0.25
0.31
0.62
0.39
6.06
6.00
5.21
5.76
6.32
6.32
5.89
6.18
46.92
84.65
13.94

5.00
6.38
5.79
5.72
0.01
0.17
0.07

0.00
0.00
0.00
0.00
16.58
86.89
32.27

0.05
0.06
0.06
0.06
1.58
25.47
8.99

5.82
5.39
5.51
5.57
33.76
105.48
31.13

5.87
5.45
5.58
5.63
-51.46
28.69
26.49

7.16
13.81
5.25
8.74
-72.13
37.23
39.61

-17.51
14.54
3.60
0.21
                                    C-6

-------
Table C-5. Post-combustion controls: dry FGD scrubbers using ESP.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In
No. O-H O-H O-H
Bituminous Coal, PC Boiler with DSI/CS-ESP
Washington
Washington
Washington
Average
1
2
3

0.00
0.00
0.00
0.00
4.33
7.75
6.40
6.16
11.63
11.02
9.95
10.86
HgTln
O-H
15.97
18.77
16.36
17.03
Subbituminous Coal, PC Boiler with SDA/CS-ESP
GRDA
GRDA
GRDA
Average
Laramie 3
Laramie 3
Laramie 3
Average
Wyodak
Wyodak
Wyodak
Average
1
2
3

1
2
3

1
2
3

0.13
0.53
0.51
0.39
0.03
1.69
4.55
2.09
2.49
3.05
2.25
2.60
4.42
2.97
8.78
5.39
0.22
0.52
0.44
0.39
3.88
4.71
3.57
4.05
7.77
6.50
3.71
5.99
0.63
8.53
9.28
6.15
11.63
9.42
11.51
10.85
12.31
9.99
13.01
11.77
0.88
10.75
14.27
8.63
18.00
17.17
17.34
17.50
HgTln
Coal
13.01
13.36
13.33
13.24
11.22
10.73
12.24
11.40
15.03
17.67
14.94
15.88
4.46
6.41
8.17
6.34
HgpOut
O-H
0.00
0.00
0.04
0.02
0.01
0.01
0.01
0.01
0.03
0.03
0.03
0.03
0.05
0.05
0.05
0.05
Hg^Out
O-H
6.41
5.84
6.52
6.26
1.55
1.28
0.34
1.06
0.10
0.04
0.04
0.06
0.07
0.17
0.25
0.16
Hg Out
O-H
3.06
3.05
3.04
3.05
5.58
11.12
5.41
7.37
3.87
4.52
5.27
4.56
9.97
10.11
10.11
10.06
HGT Out
O-H
9.47
8.90
9.59
9.32
7.13
12.42
5.76
8.44
4.00
4.58
5.35
4.64
10.09
10.32
10.41
10.27
%R HT
O-H
40.68
52.61
41.37
44.89
42.06
-24.23
55.72
24.51
0.00
57.35
62.53
39.96
43.95
39.87
39.99
41.27
%R HgT
Coal
27.22
33.42
28.06
29.56
36.42
-15.70
52.94
24.55
73.40
74.05
64.22
70.56
-126.38
-61.16

-71.65
Average 1.69 3.28 7.67 12.64 11.21 0.03 0.43 7.33 7.78 35.25 7.82
Minimum
Maximum
STDEV



0.03
4.55
1.54
0.22
8.78
2.72
0.63
11.63
3.60
0.88
18.00
5.28
4.46
17.67
4.32
0.01
0.05
0.02
0.04
1.55
0.57
3.87
11.12
2.91
4.00
12.42
3.06
-24.23
62.53
28.72
-126.38
74.05
70.07
                                  C-7

-------
Table C-6. Post-combustion controls: dry FGD scrubbers using FF.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In
No. O-H O-H O-H
Bituminous Coal, PC Boiler with SDA/FF
Mecklenburg
Mecklenburg
Mecklenburg
Average
1
2
3

11.34
5.66
6.90
7.97
3.40
4.21
3.04
3.55
6.16
0.02
0.02
2.07
HgTln
O-H
20.91
9.89
9.96
13.59
Bituminous Coal, PC Boiler with SCR and SDA/FF
Logan
Logan
Logan
Average
SEI
SEI
SEI
Average
1
2
3

1
2
3

12.87
12.74
12.45
12.69
13.48
9.47
12.01
11.66
7.22
4.36
4.59
5.39
0.30
0.25
0.25
0.26
0.21
0.35
0.25
0.27
0.14
0.18
0.16
0.16
20.31
17.46
17.29
18.35
13.92
9.90
12.42
12.08
HgTln
Coal
11.52
13.28
11.50
12.10

18.28
18.14
17.51
17.98
11.79
11.74
11.97
11.83
HgpOut
O-H
0.00
0.00
0.00
0.00

0.02
0.02
0.01
0.02
0.01
0.01
0.02
0.01
Hg^Out
O-H
0.07
0.07
0.01
0.05

0.08
0.13
0.04
0.09
0.34
0.09
0.08
0.17
Hg Out
O-H
0.09
0.23
0.23
0.18

0.16
0.17
0.17
0.16
0.13
0.12
0.11
0.12
HGT Out
O-H
0.16
0.31
0.24
0.24

0.26
0.32
0.22
0.27
0.48
0.22
0.21
0.30
%R HT
O-H
99.23
96.91
97.60
97.91

98.71
98.16
98.72
98.53
96.56
97.79
98.34
97.56
%R HgT
Coal
98.60
97.70
97.92
98.07

98.57
98.23
98.74
98.51
95.94
98.13
98.28
97.45
Average 12.17 2.83 0.22 15.22 14.90 0.02 0.13 0.14 0.28 98.05 97.98
Minimum
Maximum
STDEV



9.47
13.48
1.41
0.25
7.22
2.98
0.14
0.35
0.08
Subbituminous Coal, PC Boiler with SDA/FF
Craig 3
Craig 3
Craig 3
Average
Rawhide
Rawhide
Rawhide
Average
NSPSherburne
NSPSherburne
NSPSherburne
Average
1
2
3

1
2
3

1
2
3

0.57
0.92
0.90
0.80
0.25
1.92
3.76
1.98
0.03
0.03
0.03
0.03
0.65
0.50
0.23
0.46
1.38
0.83
0.46
0.89
0.53
0.23
0.19
0.32
0.20
0.17
0.12
0.16
12.46
12.85
14.79
13.37
10.92
10.92
10.24
10.69
9.90
20.31
3.82

1.42
1.60
1.25
1.42
14.09
15.59
19.01
16.23
11.48
11.18
10.46
11.04
11.74
18.28
3.38

1.20
1.06
0.93
1.06
8.09
7.33
9.24
8.22
8.29
8.27
7.73
8.10
0.01
0.02
0.00

0.00
0.00
0.00
0.00
0.12
0.01
0.03
0.05
0.12
0.14
0.27
0.18
0.04
0.34
0.11

0.04
0.04
0.03
0.04
0.76
0.69
0.98
0.81
0.20
0.18
0.24
0.20
0.11
0.17
0.03

0.90
0.89
0.82
0.87
10.80
9.91
9.00
9.90
8.42
12.09
9.84
10.12
0.21
0.48
0.10

0.94
0.93
0.86
0.91
11.68
10.60
10.01
10.76
8.74
12.40
10.35
10.50
96.56
98.72
0.81

33.79
41.83
31.67
35.76
17.16
32.03
47.31
32.16
23.81
-10.94
1.05
4.64
95.94
98.74
1.03

21.49
11.88
7.38
13.58
-44.36
-44.61
-8.41
-5.43
-49.92
-34.01
-29.78
Average 0.93 0.56 8.07 9.56 5.79 0.08 0.35 6.96 7.39 24.19 -16.22
Minimum
Maximum
STDEV



0.03
3.76
1.23
0.19
1.38
0.37
0.12
14.79
6.08
1.25
19.01
6.63
0.93
9.24
3.59
0.00
0.27
0.09
0.03
0.98
0.36
0.82
12.09
4.68
0.86
12.40
4.97
-10.94
47.31
18.96
-49.92
21.49
27.38
                                                                   (continued)

-------
Table C-6.  (continued).
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In
No. O-H O-H
ND Lignite, PC Boiler with SDA/FF
Antelope Valley
Antelope Valley
Antelope Valley
Average
Stanton 10
Stanton 10
Stanton 10
Average
1
2
3

1
2
3

0.16
0.21
0.16
0.18
0.22
0.27
0.50
0.33
0.38
0.42
0.16
0.29
0.24
0.36
0.69
0.43
Hgln
O-H
7.80
7.82
7.67
7.75
10.23
9.86
9.45
9.85
HgTln
O-H
8.34
8.45
8.00
8.22
10.70
10.49
10.64
10.61
HgTln
Coal
13.85
16.03
12.50
14.27
12.82
15.63
9.45
12.63
HgpOut
O-H
0.01
0.02
0.02
0.02
0.02
0.01
0.01
0.01
Hg^Out
O-H
0.25
0.79
0.33
0.56
0.40
0.17
0.01
0.19
Hg Out
O-H
omit
8.16
6.97
7.56
10.14
10.58
10.81
10.51
HGT Out
O-H
NA
8.98
7.32
8.15
10.56
10.76
10.83
10.72
%R HT
O-H
NA
-6.27
8.49
1.11
1.24
-2.54
-1.77
-1.02
%R HgT
Coal
NA
44.01
41.45
42.73
17.63
31.15
-14.61
11.39
Average 0.27 0.38 9.01 9.65 13.29 0.02 0.34 9.33 9.69 -0.17 23.93
Minimum
Maximum
STDEV



0.16
0.50
0.13
0.16
0.69
0.20
Bituminous, Stoker with SDA/FF
Dwayne Collier
Dwayne Collier
Dwayne Collier
Average
1
2
3

2.19
2.14
1.99
2.11
0.03
0.18
0.03
0.08
7.67
10.23
1.18

0.06
0.42
0.11
0.20
8.00
10.70
1.32

2.28
2.75
2.13
2.39
9.45
16.03
2.67

3.37
3.48
3.29
3.38
0.01
0.02
0.01

0.06
0.03
0.01
0.03
0.01
0.79
0.29

0.02
0.03
0.03
0.03
6.97
10.81
1.69

0.08
0.09
0.06
0.08
7.32
10.83
1.53

0.16
0.15
0.10
0.14
-6.27
8.49
5.53

92.84
94.48
95.43
94.25
-14.61
44.01
23.91

95.16
95.64
97.04
95.95
                                     C-9

-------
Table C-7. Post-combustion controls: wet FGD scrubbers.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In HgT In HgT In
No. O-H O-H O-H O-H Coal
Bituminous Coal, PC Boiler with PS and Wet FGD Scrubber
Bruce Mansfield
Bruce Mansfield
Bruce Mansfield
Average
1
2
3

0.27
0.73
0.27
0.42
8.65
9.84
8.34
8.94
1.58
2.08
1.70
1.79
10.50
12.65
10.31
11.15
10.93
8.93
11.82
10.56
Subbituminous Coal, PC Boiler with PS and Wet FGD Scrubber
Boswell 4
Boswell 4
Boswell 4
Average
Cholla2
Cholla2
Cholla2
Average
Colstrip
Colstrip
Colstrip
Average
Lawrence
Lawrence
Lawrence
Average
1
2
3

1
2
3

1
2
3

1
2
3

0.11
2.98
2.75
1.95
0.42
1.11
0.41
0.65
1.78
1.94
1.63
1.78
0.23
0.53
0.24
0.33
0.33
1.07
0.55
0.65
0.97
0.93
2.06
1.32
2.29
2.37
2.86
2.51
1.65
0.63
0.65
0.98
5.05
1.47
1.16
2.56
4.68
2.62
2.99
3.43
1.08
6.37
5.39
4.28
4.99
4.41
4.96
4.79
5.48
5.53
4.45
5.15
6.07
4.66
5.46
5.40
5.15
10.68
9.88
8.57
6.86
5.58
5.86
6.10
6.98
6.63
7.93
7.18
6.99
6.37
5.09
6.15
7.63
7.98
7.93
7.85
6.24
5.47
6.03
5.91
HgpOut
O-H
0.04
0.06
0.04
0.05
0.02
0.20
0.28
0.17
0.15
0.19
0.11
0.15
0.05
0.02
0.02
0.03
0.01
0.08
0.09
0.06
Hg^Out
O-H
1.89
2.73
1.22
1.95
0.10
0.44
0.59
0.38
0.21
0.14
0.14
0.16
0.42
0.45
0.39
0.42
0.49
0.53
0.51
0.51
Hg Out
O-H
7.01
7.96
8.29
7.76
5.53
5.89
5.57
5.66
3.93
4.67
4.22
4.27
9.13
11.03
2.13
7.43
6.37
6.71
6.20
6.42
HGT Out
O-H
8.95
10.76
9.55
9.75
5.65
6.53
6.43
6.21
4.29
5.01
4.46
4.59
9.60
11.51
2.54
7.88
6.87
7.32
6.81
7.00
%R HT
O-H
14.81
14.94
7.42
12.39
-3.08
-18.25

-21.91
29.30
-7.51
18.29
13.36
-86.54
-7.74
74.27
-6.67
-0.07
-31.14

-15.81
%R HgT
Coal
18.11
-20.57
19.25
5.60
19.00
1.41
18.91
13.11
38.59
21.38
12.27
24.08
-25.89
-44.19
67.94
-0.71
-10.01
-33.75

-18.91
Average 1.18 1.36 3.76 6.30 6.77 0.10 0.37 5.95 6.42 -7.76 4.39
Minimum
Maximum
STDEV



0.11
2.98
1.02
0.33
2.86
0.85
1.08
6.37
1.82
4.45
10.68
1.96
ND Lignite, PC Boiler with PS and Wet FGD Scrubber
Lewis and Clark
Lewis and Clark
Lewis and Clark
Average
1
2
3

1.15
1.68
1.41
1.41
16.47
13.64
6.28
12.13
11.65
8.43
10.20
10.09
29.27
23.75
17.89
23.64
5.09
7.98
0.98

15.33
15.54
18.96
16.61
0.01
0.28
0.09

0.06
0.00
0.00
0.02
0.10
0.59
0.17

0.50
0.35
0.50
0.45
2.13
11.03
2.34

13.86
14.19
15.81
14.62
2.54
11.51
2.40

14.42
14.55
16.31
15.09
-86.54
74.27
39.47

50.75
38.74
8.81
32.77
-44.19
67.94
31.96

5.98
6.41
13.94
8.78
                                                                   (continued)
                                   C-10

-------
Table C-7. (continued).
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In HgT In HgT In HgpOut
No. O-H O-H O-H O-H Coal O-H
Bituminous Coal, PC Boiler with CS-ESP and wet FGD Scrubber
AES Cayuga
AES Cayuga
AES Cayuga
Average
Big Bend
Big Bend
Big Bend
Average
2
1
3

1
2
3

0.00
0.00
0.00
0.00
0.09
0.05
0.02
0.05
6.40
5.87
5.55
5.94
4.86
4.92
4.26
4.68
2.58
2.24
2.95
2.59
2.40
2.31
2.13
2.28
8.98
8.11
8.50
8.53
7.34
7.29
6.41
7.01
11.87
10.70
10.80
11.12
17.52
11.25
12.01
13.59
0.00
0.00
0.00
0.00
0.05
0.00
0.03
0.03
Hg^Out
O-H
0.18
0.36
0.18
0.24
0.21
0.12
0.23
0.19
Hg Out
O-H
2.70
2.73
3.08
2.83
2.18
1.75
2.05
1.99
HGT Out
O-H
2.88
3.09
3.26
3.08
2.44
1.87
2.31
2.21
%R HT
Wet FGD
67.91
61.88
61.63
63.81
66.70
74.37
64.01
68.36
%R HgT
PM+FGD
76.06
71.56
71.38
73.00
75.16
80.88
73.15
76.39
Average 0.03 5.31 2.43 7.77 12.36 0.01 0.22 2.41 2.64 66.08 74.70
Minimum
Maximum
STDEV



0.00
0.09
0.03
4.26
6.40
0.78
2.13
2.95
0.30
6.41
8.98
0.94
10.70
17.52
2.59
0.00
0.05
0.02
Subbituminous Coal, PC Boiler with CS-ESP and wet FGD Scrubber
Jim Bridger
Jim Bridger
Jim Bridger
Average
Laramie River 1
Laramie River 1
Laramie River 1
Average
Sam Seymour
Sam Seymour
Sam Seymour
Average
1
2
3

1
2
3

1
2
3

0.05
0.44
0.07
0.19
0.25
0.04
0.02
0.10
0.03
0.01
0.01
0.01
2.49
2.04
1.78
2.10
3.14
2.16
3.08
2.79
3.00
4.08
5.39
4.16
5.21
5.64
4.50
5.12
7.52
8.35
7.53
7.80
9.10
13.10
11.96
11.38
7.74
8.12
6.35
7.41
10.91
10.55
10.63
10.70
12.13
17.19
17.35
15.56
no coal flow
no coal flow
no coal flow
not includec
13.52
15.45
15.71
14.90
60.48
43.20
51.04
51.58
0.06
0.05
0.03
0.05
0.02
0.00
0.01
0.01
0.06
0.11
0.06
0.07
0.12
0.36
0.08

0.25
0.29
0.20
0.25
0.29
0.12
0.03
0.15
0.24
0.29
0.35
0.29
1.75
3.08
0.50

6.63
6.51
5.92
6.36
4.86
5.73
4.48
5.02
12.25
13.33
11.99
12.53
1.87
3.26
0.53

6.95
6.85
6.15
6.65
5.18
5.85
4.52
5.18
12.54
13.74
12.39
12.89
61.63
74.37
4.78

10.32
15.64
3.06
9.68
52.57
44.54
57.53
51.55
1.51
23.90
31.99
19.13
71.38
80.88
3.56

14.60
19.67
7.69
13.99
54.83
47.18
59.56
53.86
1.51
23.90
31.99
19.13
Average 0.10 3.02 8.10 11.22 33.24 0.04 0.23 7.97 8.24 26.78 28.99
Minimum
Maximum
STDEV



0.01
0.44
0.15
1.78
5.39
1.13
4.50
13.10
2.94
6.35
17.35
3.88
13.52
60.48
20.84
TX Lignite, PC Boiler with CS-ESP and wet FGD Scrubber
Monticello 3
Monticello 3
Monticello 3
Average
Limestone
Limestone
Limestone
Average
1
2
3

1
2
3

0.19
0.11
0.13
0.14
0.01
0.01
0.02
0.02
16.49
19.77
25.83
20.70
23.55
24.55
28.15
25.42
29.39
28.15
27.21
28.25
13.38
13.11
14.11
13.54
46.07
48.03
53.16
49.09
36.94
37.68
42.29
38.97
61.96
63.13
76.52
67.20
14.49
20.84
15.29
16.87
0.00
0.11
0.03

0.31
0.18
0.24
0.24
0.04
0.33
0.12
0.17
0.03
0.35
0.10

6.50
0.44
7.26
4.73
2.69
3.18
1.27
2.38
4.48
13.33
3.50

29.45
25.52
23.10
26.02
15.96
16.23
17.18
16.46
4.52
13.74
3.59

36.25
26.14
30.60
31.00
18.69
19.74
18.58
19.01
1.51
57.53
21.09

21.31
45.57
42.44
36.44
49.40
47.59
56.07
51.02
1.51
59.56
20.83

21.31
45.57
42.44
36.44
49.40
47.59
56.07
51.02
Average 0.08 23.06 20.89 44.03 42.04 0.20 3.56 21.24 25.00 43.73 43.73
Minimum
Maximum
STDEV



0.01
0.19
0.07
16.49
28.15
4.24
13.11
29.39
8.09
36.94
53.16
6.28
14.49
76.52
28.12
0.04
0.33
0.11
0.44
7.26
2.76
15.96
29.45
5.63
18.58
36.25
7.32
21.31
56.07
11.89
21.31
56.07
11.89
                                                                         (continued)
                                       C-ll

-------
Table C-7.  (continued).
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In HgT In HgT In HgpOut
No. O-H O-H O-H O-H Coal O-H
Bituminous Coal, PC Boiler with HS-ESP and wet FGD Scrubber
Charles Lowman
Charles Lowman
Charles Lowman
Average
Morrow
Morrow
Morrow
Average
1
2
3

1
2
3

2.64
1.55
3.45
2.55
0.05
0.01
0.03
0.03
3.33
3.98
3.55
3.62
10.80
8.31
6.98
8.70
2.09
2.17
2.02
2.09
4.41
4.10
3.32
3.94
8.06
7.69
9.01
8.26
15.27
12.42
10.33
12.67
23.49
21.50
23.94
22.98
5.48
5.42
5.38
5.43
0.06
0.07
0.05
0.06
0.05
0.03
0.04
0.04
HgpOut
O-H
1.68
1.86
2.06
1.87
2.06
1.79
1.12
1.65
Hg Out
O-H
3.39
3.50
3.19
3.36
5.00
4.50
4.55
4.68
HGT Out
O-H
5.13
5.44
5.30
5.29
7.11
6.31
5.71
6.38
%R HT
Wet FGD
36.44
29.31
41.18
35.64
53.46
49.18
44.70
49.11
%R HgT
PM+FGD
44.29
38.03
48.44
43.58
59.20
55.45
51.52
55.39
Average 1.29 6.16 3.02 10.46 14.20 0.05 1.76 4.02 5.83 42.38 49.49
Minimum
Maximum
STDEV



0.01
3.45
1.50
3.33
10.80
3.05
2.02
4.41
1.08
7.69
15.27
2.91
5.38
23.94
9.65
0.03
0.07
0.02
Subbituminous Coal, PC Boiler with HS-ESP and wet FGD Scrubber
Coronado
Coronado
Coronado
Average
Craig 1
Craig 1
Craig 1
Average
Navajo
Navajo
Navajo
Average
San Juan
San Juan
San Juan
Average
1
2
3

1
2
3

1
2
3

1
2
3

0.03
0.03
0.03
0.03
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.03
0.02
0.08
0.02
0.04
0.99
0.82
1.09
0.96
0.33
0.29
0.16
0.26
2.91
0.45
0.62
1.33
6.25
3.31
5.07
4.87
2.19
1.86
1.87
1.97
3.61
2.52
1.99
2.71
3.55
3.93
3.50
3.66
5.81
4.26
3.62
4.56
3.20
2.71
2.99
2.96
3.97
2.85
2.19
3.01
6.49
4.41
4.16
5.02
12.08
7.65
8.70
9.47
4.45
4.76
3.86
4.36
2.45
2.79
2.30
2.51
4.37
2.63
2.63
3.21
7.94
8.69
11.00
9.21
0.02
0.08
0.11
0.07
0.00
0.00
0.01
0.01
0.05
0.02
0.01
0.03
0.05
0.08
0.05
0.06
1.12
2.06
0.35

0.04
0.07
0.13
0.08
0.13
0.11
0.09
0.11
0.04
0.04
0.04
0.04
0.45
0.38
0.31
0.38
3.19
5.00
0.75

3.56
1.83
3.08
2.82
2.13
2.09
2.03
2.08
3.67
3.79
3.77
3.75
7.14
4.79
4.66
5.53
5.13
7.11
0.75

3.61
1.98
3.32
2.97
2.26
2.20
2.14
2.20
3.76
3.85
3.82
3.81
7.64
5.25
5.02
5.97
29.31
53.46
8.74

-12.95
26.82

0.86
43.05
22.93
2.44
22.81
42.00
12.65
8.25
20.97
36.74
31.35
42.31
36.80
38.03
59.20
7.66

-0.87
34.64
0.60
11.46
49.14
31.17
12.87
31.06
48.20
21.99
18.06
29.42
43.50
38.69
48.48
43.56
Average 0.03 1.86 3.23 5.12 4.82 0.04 0.15 3.54 3.74 20.36 28.87
Minimum
Maximum
STDEV



0.02
0.08
0.02
0.16
6.25
2.05
1.86
5.81
1.19
2.19
12.08
3.02
2.30
11.00
2.86
0.00
0.11
0.04
0.04
0.45
0.14
1.83
7.14
1.52
1.98
7.64
1.64
-12.95
43.05
20.32
-0.87
49.14
18.15
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In HgT In HgT In
No. O-H O-H O-H O-H Coal
Bituminous Coal, PC Boiler with FF and wet FGD scrubber
Clover
Clover
Clover
Average
Intermountain
Intermountain
Intermountain
Average
1
2
3

1
2
3

0.06
0.03
0.08
0.06
0.01
0.01
0.01
0.01
1.00
1.11
1.16
1.09
1.01
1.08
1.36
1.15
1.11
1.99
0.62
1.24
0.20
0.24
0.22
0.22
2.17
3.13
1.86
2.39
1.22
1.33
1.58
1.38
29.21
41.19
49.02
39.81
2.00
1.97
3.09
2.35
HgpOut
O-H
0.05
0.02
0.06
0.04
0.01
0.01
0.01
0.01
HgpOut
O-H
0.42
0.34
0.05
0.27
0.03
0.07
0.08
0.06
Hg Out
O-H
0.42
0.17
0.14
0.24
0.25
0.46
0.41
0.37
HGT Out
O-H
0.88
0.53
0.25
0.55
0.29
0.54
0.50
0.44
%R HT
Wet FGD
59.42
83.13
86.76
76.43
76.15
59.67
68.68
68.16
%R HgT
PM+FGD
96.78
98.66
98.95
98.13
98.11
96.80
97.52
97.48
Average 0.03 1.12 0.73 1.88 21.08 0.03 0.16 0.31 0.50 72.30 97.80
Minimum
Maximum
STDEV



0.01
0.08
0.03
1.00
1.36
0.13
0.20
1.99
0.71
1.22
3.13
0.70
1.97
49.02
21.47
0.01
0.06
0.02
0.03
0.42
0.17
0.14
0.46
0.14
0.25
0.88
0.23
59.42
86.76
11.66
96.78
98.95
0.92
                                     C-12

-------
Table C-8. Cyclone-fired boilers.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In Hg** In Hg In
No. O-H O-H O-H
ND Lignite, Cyclone Boiler with CS-ESP
Leland Olds
Leland Olds
Leland Olds
Average
1
2
3

0.56
0.26
2.85
0.41
0.23
0.46
0.81
0.34
3.30
8.80
4.77
6.05
HgTln
O-H
4.09
9.51
8.43
6.80
Sub-bituminous/Pet. Coke, Cyclone Boiler with HS-ESI
Nelson Dewey
Nelson Dewey
Nelson Dewey
Average
1
2
3

0.01
0.01
0.01
0.01
0.49
0.24
0.12
0.28
3.20
2.19
2.06
2.48
3.69
2.43
2.18
2.77
Lignite, Cyclone Boiler with Mechanical Collector
Bay Front
Bay Front
Bay Front
Average
1
2
3

0.76
1.08
0.09
0.64
0.78
0.67
0.77
0.74
Lignite, Cyclone Boiler with SDA/FF
Coyote
Coyote
Coyote
Average
1
2
3

0.69
1.18
1.69
1.19
1.62
2.98
3.07
2.34
2.17
1.94
1.74
1.95
13.68
13.90
14.91
14.29
3.70
3.69
2.60
3.33
15.99
18.06
19.66
17.82
HgTln
Coal
5.63
10.18
7.94
7.90
6.62
6.47
6.09
6.39
3.58
3.01
3.36
3.32
10.51
18.55
11.39
10.95
Bituminous, Cyclone Boiler with PS and Wet FGD Scrubbers
Lacygne
Lacygne
Lacygne
Average
1
2
3

6.70
6.52
5.98
6.40
3.99
3.34
0.59
2.64
1.30
0.60
0.61
0.84
12.00
10.46
7.18
9.88
no inlet flow
no inlet flow
no inlet flow
no inlet flow
HgpOut
O-H
0.00
0.00
0.00
0.00
0.10
0.04
0.04
0.06
1.19
0.86
0.48
0.84
0.08
0.14
0.08
0.08
0.04
0.05
0.09
0.06
Bituminous, Cyclone Boiler with CS-ESP and wet FGD Scrubber
Bailly
Bailly
Bailly
1
2
3
0.04
0.04
0.09
3.18
2.37
3.01
2.57
2.95
2.58
5.79
5.36
5.68
4.41
5.20
4.08
0.00
0.00
0.00
Hg^Out
O-H
0.82
1.09
1.60
0.95
0.26
0.16
0.25
0.22
0.60
2.75
3.57
2.30
0.04
0.24
0.44
0.24
0.44
0.43
0.41
0.43
0.36
0.31
0.39
Hg Out
O-H
4.04
5.26
LS
4.65
3.33
2.40
2.44
2.72
1.91
1.80
1.78
1.83
13.97
LS
18.06
16.02
8.74
7.41
5.10
7.08
2.85
2.62
2.78
HGT Out
O-H
4.86
6.35
NA
5.60
3.69
2.60
2.73
3.00
3.69
5.40
5.84
4.98
14.10
NA
18.58
16.34
9.22
7.89
5.59
7.57
3.22
2.93
3.17
%R HT
O-H
-18.68
33.26
NA
7.29
0.13
-6.90
-24.95

0.34
-46.54

-57.07
11.81
NA
5.48
8.64
23.18
24.53
22.17
23.29
54.24
54.95
54.11
%R HgT
Coal
13.66
37.64
NA
25.65
44.27
59.83
55.22
53.11
-2.95
-79.21

-51.99
-34.23
NA
-63.12

no inlet flow
no inlet flow
no inlet flow
no inlet flow
27.09
43.53
22.31
                                      C-13

-------
Table C-9. Fluidized-bed combustion.
Hg Speciation at Inlet and Outlet (|jg/dscm@ 3%O2) : % Reduction for O-H Train and Coal Data
Plant ID Run Hgp In
No. O-H
Lignite, FBC with CS-ESP
R.M. Heskett
R.M. Heskett
R.M. Heskett
Average
1
2
3

4.73
2.93
7.43
5.03
Hg^ln
O-H
5.39
0.96
0.44
2.26
Anthracite Waste, FBC with FF
Kline Township
Kline Township
Kline Township
Average
1
2
3

44.54
43.12
44.97
44.21
0.12
0.06
0.06
0.08
Bituminous Waste, FBC with FF
Scrubgrass
Scrubgrass
Scrubgrass
Average
1
2
3

184.04
124.11
76.68
128.28
0.68
0.42
0.22
0.44
Hgln
O-H
3.83
2.61
3.08
3.17
0.45
0.40
0.34
0.40
0.19
0.09
0.07
0.12
Bituminous/Pet. Coke, FBC with SNCR and FF
Stockton Cogen
Stockton Cogen
Stockton Cogen
Average
1
2
3

2.71
1.56
2.08
2.12
0.06
0.07
0.06
0.07
0.06
0.06
0.06
0.06
Subbituminous, FBC with SCR and FF
AES Hawaii
AES Hawaii
AES Hawaii
Average
1
2
3

0.26
0.35
0.36
0.32
Lignite, FBC with CS-FF
TNP
TNP
TNP
1
2
3
21.65
10.65
28.12
0.04
0.17
0.11
0.10
8.68
4.51
13.78
1.29
1.38
1.18
1.28
7.42
6.09
7.04
HgTln
O-H
13.95
6.50
10.94
10.46
45.11
43.58
45.37
44.69
184.91
124.62
76.97
128.83
2.83
1.69
2.20
2.24
1.59
1.90
1.64
1.71
37.74
21.25
48.94
HgTln
Coal
13.54
12.68
11.11
12.44
148.68
212.95
153.77
171.80
100.09
101.35
100.25
100.56
1.68
1.44
1.66
1.59
3.77
3.72
2.51
3.33
63.81
44.22
95.04
HgpOut
O-H
1.06
0.07
0.05
0.39
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.02
0.03
0.03
0.03
0.00
0.00
0.00
0.00
0.04
0.03
0.04
Hg^Out
O-H
1.44
0.41
0.18
0.68
0.06
0.06
0.06
0.06
0.07
0.05
0.04
0.05
0.04
0.05
0.05
0.05
0.02
0.02
0.02
0.02
12.13
6.78
13.54
Hg Out
O-H
4.57
5.31
4.74
4.87
0.06
0.06
0.06
0.06
0.08
0.07
0.07
0.07
0.05
0.05
0.05
0.05
0.68
0.90
0.55
0.71
4.74
2.94
5.07
HGT Out
O-H
7.07
5.78
4.98
5.95
0.12
0.12
0.12
0.12
0.15
0.12
0.11
0.13
0.11
0.13
0.12
0.12
0.70
0.92
0.58
0.73
16.91
9.76
18.66
%R HT
O-H
49.29
11.09
54.49
38.29
99.74
99.73
99.74
99.74
99.92
99.91
99.85
99.89
96.09
92.16
94.48
94.25
55.84
51.35
64.91
57.37
55.20
54.07
61.88
%R HgT
Coal
47.76
54.40
55.19
52.45
99.92
99.95
99.92
99.93
99.85
99.89
99.89
99.88
93.39
90.80
92.67
92.29
81.39
75.16
77.06
77.87
73.50
77.93
80.37
                                   C-14

-------
            Appendix D
Assessment of Mercury Control Options
                for
       Coal-fired Power Plants
                D-l

-------
                               .[  '  f,.,/M '.<   .Kr.-..fc.v-|
&'
                                                             fr' '  .A--V
                                                                  ^, r  ^
                United States Environmental Protection^gency  r
                National R^skManagement Res<
                   r Pollution Prey^ition and Control Division
      li
                                Prepared b
                                3S Departnll
                 /*  National Kiergy Technology Laboratory
r- jP

i*. '
                rt^'i
   -
                                 August, 2000
       r
                                    D-2

-------
              List of Attachments
Attachment 1.  Description of National Energy Technology Laboratory
              Control Performance and Cost Model 	  D-4

Attachment 2.  Description of Mercury Control Performance Algorithms
              Used in the National Energy Technology Laboratory
              Mercury Control Performance and Cost Model	  D-33

Attachment 3.  Summary of Mercury Control Cases Analyzed with
              National Energy Technology Laboratory Mercury
              Control Performance and Cost Model	  D-42

Attachment 4.  Results of all Model Runs	  D-48
                        D-3

-------
            ATTACHMENT 1

              Description of
   National Energy Technology Laboratory
Mercury Control Performance and Cost Model
                   D-4

-------
               Description of the National Energy Technology Laboratory
                      Mercury Control Performance and Cost Model

1.       Model Description

The National Energy Technology Laboratory (NETL) Mercury Control Performance and Cost Model (MCPCM) is
an Excel spreadsheet model that can be used to assess the performance and cost of mercury (Hg) control systems
that utilize activated carbon injection (ACI) and other methods.   The primary goal of the model is to calculate key
performance parameters that are then used to calculate detailed capital costs and O&M costs of a mercury control
technology for either of two types of power plant applications - either a bituminous or subbituminous coal-fired
power plant configuration.  This overview describes the model layout, its specific capabilities,  and its current
limitations.

1.1     MCPCM Layout

The spreadsheet model is currently divided into seven (7) different, functional sheets that are integrated together to
perform the costing goal defined above. The five sheets are identified below:

Hg Control Scenario Definition Sheet  	Characterization of Control Technology Retrofit
                                                  Configurations
Hg Control Performance Models Sheet	Performance Algorithms for Different Control
                                                  Technology Retrofit Configurations
Application Input Sensitivity Sheet       	User Data Input

Technical Model Results Sheet          	PowerPlant and Mercury Control Technical Performance
                                                  Results
Combustion Calculations Sheet          	Power Plant Performance Calculations

Capital Cost Model Sheet               	Capital Cost Calculations

System Economic Model Sheet          	O&M Cost Calculations and Levelized Cost Calculations

1.1.1    Hg Control Scenario Definition Sheet

This sheet documents the different mercury control technology retrofit scenarios that can potentially be evaluated by
the model. The information provided for each scenario is presented below:

Scenario Number: Number, no units
This number uniquely identifies each control scenario within the program.

Configuration Designation: Abbreviation that identifies  a  control technology  scenario,  e.g., ESP-1 (activated
carbon injection upstream of ESP with no spray cooling)

Configuration Definition: Brief description that uniquely defines a control technology scenario, e.g., ESP-1, "ACI
upstream of existing ESP"

Comment: Further descriptive information that qualifies the functionality of the control technology scenario within
the model.

Data Sources: Identifies and documents specific data and information sources used to establish the performance of
a control technology.
                                                D-5

-------
1.1.2     Hg Control Performance Models Sheet

This sheet contains performance models for each control technology configuration identified in the Hg Control
Scenario Definition Sheet (as currently available).  These models currently calculate total sorbent feed (Ib/MMacf
of flue gas) based on a specified total mercury control efficiency (e.g., 50%) and flue gas temperature.

Currently, each  control configuration (e.g., ESP-1) makes use of separate algorithms  for mid- to  high-sulfur
bituminous coals (e.g., Pittsburgh (Pgh) #8) and Western subbituminous coals (e.g., Wyoming PRB). Therefore, the
coal type that is specified in the Application Input Sensitivity Sheet will determine the specific algorithm that is
used for a specified control configuration and application case. These algorithms were developed based on curve-
fitting available pilot- and full-scale test data. The data sources are also documented in the sheet.

1.1.3     Application Input Sensitivity Sheet

This sheet is intended as the primary user interface to run up to 20 different mercury control cases simultaneously.
For each control case, the sheet defines the desired mercury control requirements, a specific power plant variant of
either of two  power plant application types, and some key economic parameters used for costing purposes.
Additional parametric changes can be made to the model to add greater evaluation flexibility, but these need to be
made within the other functional sheets of the model.  Model parameters included here are:

1.1.3.1   Mercury Control Parameters

Case Number: Numerical value, Fixed (e.g., 1,2, etc.) -
Sequentially identifies each application case. Up to 20 cases can be specified.

Hg Removal Configuration Type: Numerical entry by user (e.g., 1,2, etc.) 
Specifies the mercury control configuration as documented in the Hg Control Scenario Definition Sheet.  The user
must select a configuration type.

Hg Control Flue Gas Temperature Specification: Number, units = F (e.g., 250)
Specifies the temperature at which mercury control is to take place.  The program calculates this temperature as
follows:

         Control Temperature = Flue Gas Acid Dew Point Temperature 
-------
Total SI (sorbent injection) Mercury Removal Efficiency - % of Hg Removed: Number as a percentage, no
units (e.g., 50)
The percentage of total mercury that is removed by the control technology.

SI Mercury Removal Efficiency - % Hg Removed: Number as a percentage, no units (e.g., 50)
The percentage of elemental Hg that is removed by the control technology.  This value is currently assumed to be
calculated by the performance model, but this capability is currently not available.

ACI Mercury Removal Efficiency - % HgCl2 Removed: Number as a percentage, no units (e.g., 50)
The percentage of HgQ2 that is removed by the control technology, but this capability is currently not available.

Sorbent Injection Ratio: Number, units = Ib/MMacf (e.g., 3)
The primary feed of sorbent (based on Ib  sorbent per MMacf of flue gas) that corresponds to the specified Hg
removal efficiency. The configuration performance model calculates this value.

Sorbent Recycle Split: Percent, no units (e.g.,  10)
The ratio of recycled spent sorbent to total sorbent feed into the flue  gas (expressed as a percentage). This value is
currently assumed to be calculated by the performance model, but this capability is currently not available.

FGD Mercury Removal Efficiency - % Hg Removed: Number as a percentage,  no units (e.g., 50)
The percentage of Hg  (in the flue gas) that is removed by an existing wet FGD system.  Mercury removal via an
FGD system can be incorporated manually if desired.   This is permitted as an option in case the user wants to
combine FGD with other mercury removal technologies such as ACI. Set values equal to zero if no wet FGD exists
or mercury control scenarios 14-18 are being used.  If utilized, total mercury  control is calculated within the
"Technical Model Results Sheet."

FGD Mercury Removal Efficiency - % HgCl2 Removed: Number as a percentage, no units (e.g., 50)
The percentage of HgQ2 (in the flue gas) that is removed by the wet FGD system. Mercury removal via an FGD
system can be incorporated manually if desired.  This is permitted as an option in case the user wants to combine
FGD with other mercury removal technologies such as ACI.  Set value  equal to zero if no wet FGD exists or
mercury  control  scenarios   14 -  18 are being used.   If utilized,  total mercury control  is calculated within the
"Technical Model Results Sheet."

If Hg speciation  is unknown set value equal to "% Hg  removal efficiency." For example, if total Hg removal via
FGD is to be set at 80% and speciation is unknown, then set both Hg and HgQ2 removal efficiencies equal to the
80% value.

Fabric Filter Pressure Drop: Number, units = inches H2O (e.g., 6)
Differential pressure across  baghouse tubesheet. Input a value if a pulse-jet FF will be added to the plant after the
primary particulate collector to collect Hg sorbent.  This applies to retrofit scenarios 3, 6, 7, 8, and 11.

Fabric Filter Air/cloth Ratio:  Number, units =ft/min (e.g., 12)
Ratio of volumetric gas flow into baghouse (ft /min)  to total bag surface area (ft ). Input a value if a pulse-jet FF
will be added to the plant  after the primary particulate collector to collect Hg sorbent.   This applies to retrofit
scenarios 3, 6, 7, 8, and 11.

Fabric Filter Particle Collection Efficiency: Number as a percentage, no units (e.g., 99.98)
Mass flow of particulate into baghouse/mass flow of particulate emitted from baghouse. Input a value if a pulse-jet
FF will be added to the plant after the primary particulate collector to collect Hg  sorbent. This applies to retrofit
scenarios 3,6,7, 8, and 11.

1.1.3.2   Power  Plant Design Parameters
                                                 D-7

-------
Gross Power Plant Size: Number, units =MWe (e.g., 500)
Gross power plant electricity output (excludes plant auxiliary power).

Reference Power Plant Type: Alphanumeric entry (HS or LS)
HS refers to the high sulfur reference power plant and LS refers to the low sulfur reference plant.

Plant Capacity Factor: Number as a percentage, no units (e.g., 65)
Ratio of the energy generated during some time period to the total energy that could have been generated had the
plant run at its full rating over the entire time period.

Power Plant Coal Type: Alphanumeric entry (e.g., Pgh #8)
Select a coal type from a menu list of six coals. The coal types are Illinois #6, Wyoming PRB, Texas Lignite, Utah
Bituminous (LS),  Appalachian (HS), Pittsburgh #8, and bituminous process derived fuel (LS). The Combustion
Calculations Sheet contains detailed analysis data for each coal (cell range AE61 to AP141).

1.1.3.3  Economic Assessment Parameters

Levelized Carrying Charge Rate: Number, no units (e.g., 0.133)
The levelized amount of revenue per dollar of investment in the mercury control system that must be collected in
order to pay the carrying charges on the investment.

Sorbent Unit Cost: Number,  units = $/lb (e.g., 0.5)
Unit cost of the activated carbon or other sorbent, including the cost of the material and shipping.

Waste Disposal Removal Service?: Yes or No
This logical question is asked to identify the need to treat the  mercury laden  AC as  hazardous waste.  Yes =
hazardous, in which case the AC is removed  and processed to remove the mercury;  a  processing cost  can be
specified by the user. No = non-hazardous, in which case the AC is disposed of with the fly ash; a disposal cost
can be specified by the user.

Normal Waste Disposal Cost:  Number, units =$/ton (e.g., 30)
Unit cost of disposing non-hazardous power plant waste, such as fly ash.  Use  of a negative number indicates a
waste byproduct credit.

Hazardous Waste Disposal Cost: Number, units =$/ton (e.g., 1,750)
Unit cost of disposing hazardous power plant  waste materials.   AC is removed  and  processed to remove the
Power Cost: Number, units =$/MW-Hr  (e.g., 25)
Unit cost of power the plant charges for running auxiliary equipment.

Mercury By-Product Cost: Number, units =$/ton (e.g., 50)
Unit cost of recovered mercury that could be sold in the marketplace.

1.1.4     Technical Model Results Sheet

The purpose of this sheet is to document the case study input data for the power plant and the mercury control
technology, as well as the results of the performance calculations from the Power Plant Combustion Model Sheet.
Up to twenty case studies can be simultaneously created and documented in the sheet.  Two reference power plants
are documented; additional changes to the reference plant design and operating conditions can be made in this sheet
in order to modify plant performance.
                                                 D-8

-------
For the power plant definition, the sheet uses the input data from the Application Input Sensitivity Sheet and
combines it with the reference plant data to create a specific plant for each case study. For example, if the high
sulfur power plant gross power rating is input as 900 MWe in the input sensitivity sheet, then this overrides the
reference plant rating of 541.9 MWe; plant auxiliary power is scaled accordingly.  Sheet data for the reference
plants should not be changed unless modifications are needed to simulate a different plant.   For example, if a
different net heat rate were desired, then the reference plant value would need to be replaced with a new value.

The mercury control  data inputs are  taken directly from both the Input Sensitivity Sheet and the Hg Control
Performance Models Sheet and are listed for the sake of documentation and use by other parts of the spreadsheet.

Calculated  power plant performance results from the Power Plant Combustion Model Sheet are returned to this
sheet for documentation and  use by the sheet  to calculate specific performance results for the mercury  control
technology.

This sheet  also uses results returned from the Combustion Model Sheet to calculate the flue gas acid dew point.
This is  an  important design parameter given the significant influence  of temperature on sorbent-based mercury
control. While test data indicates that lower temperatures enhance mercury capture from the flue gas, reducing the
gas temperature (via water injection) must be limited to a specified temperature approach to the acid dew point.
Maintaining the  gas temperature at  such  an increment will  help prevent  corrosion  within the ductwork and
particulate control devices. The sulfuric acid dew point calculation is based on the following algorithms:

         1000/Tdp = 2.276-0.0294*ln(PH2O)-.0858*ln(PH2SO4) + 0.0062*ln(PH2O)*ln(PH2SO4)

Where,
         Tdp = Acid Dew Point
         PH2O = Partial pressure of water in the flue gas
         PH2SO4 = Partial pressure  of sulfuric acid in the flue gas

The dew point is in degree K and partial pressures in mm Hg.

For example, if a flue gas contains 12 % volume of water  vapor and  0.02 % volume SO2 and say 2 %  of SO2
converts to SOS, compute the sulfuric acid dew point.

Gas pressure = 10 in wg or (10/407) =0.02457 atmg or 1.02457 atma.
PH2O=0.12* 1.02457* 760 = 93.44 mm Hg.
ln(PH2O) = 4.537
PSO3=PH2SO4.
PSO3 = 0.02 * .0002 * (64/80) * 760 * 1.02457 = 0.0024917
ln(SO3)= -6
Note that 2 % conversion is on weight basis and hence we multiplied and divided by the molecular weights of SO2
and SO3 in  the above calculation.

1000/Tdp = 2.276 - 0.0294 * 4.537 + 0.0858 * 6-0.0062 * 4.537 * 6 = 2.489
Or Tdp=402 K or 129 C or 264 F

The percent conversion of SO2 to SO3 is calculated in the Combustion Model  Sheet based on an algorithm that
accounts for the sulfur in the coal and specific ash constituents (Fe2O3).

1.1.5     Combustion Calculations Sheet

The purpose of this sheet is to take the power plant case study design and operating data from the Technical Model
Results  Sheet and calculate the power plant technical performance parameters, such as combustion gas constituent
flows and total gas flows at strategic boiler locations (e.g., after the air heater).  Also, if flue gas cooling via water
                                                  D-9

-------
injection is specified by the user (by specifying a flue gas temperature lower than the reference plant's post air
heater temperature), then this sheet will calculate the quantity of water required via energy balance calculations.
This sheet also contains the coal analysis data described previously. The coal types are Illinois #6, Wyoming PRB,
Texas Lignite, Utah Bituminous (LS), Appalachian (HS), Pittsburgh #8, and bituminous process derived fuel (LS).
Detailed analysis data for each coal  can be found within cell range AE61 to API41.
1.1.6     Capital Cost Model Sheet

The Capital Cost Model Sheet makes use of the power plant and mercury control performance data to calculate the
capital costs associated with a case study's mercury control design configuration. The costing currently covers the
following equipment sections:

Spray Cooling Water System
Sorbent Injection System
Sorbent Recycle System
Pulse-Jet Baghouse and Accessories
Ash/Spent Sorbent Handling System
Continuous Emissions Monitoring System (CEMS) Upgrade

A total mercury control system cost is calculated from the following cost components: 1) equipment, 2) related
materials, 3) field and indirect labor, 4) sales tax, 5) base erected cost, 6) Engineering design, 7) process and project
contingencies, and 8) general facilities.  Maintenance costs are also calculated as a percentage of the bare erected
cost (e.g., 2%). This sheet also contains an "Economic Master  Table" in which a number of key costing parameters,
such as contingency factors, can be changed and incorporated  into the case study calculations. Twenty case studies
can be handled simultaneously in this sheet.

1.1.7     System Economic Model Sheet

This sheet calculates mercury control system O&M, total system investment, total system capital requirement, and
then levelizes the capital and O&M to establish single value for $/lb of mercury removed. Twenty case studies can
be handled simultaneously in this sheet.

1.1.7.1   Mercury Control System O&M Costs

Operating Labor:  cost of system operating  and  administrative personnel;  unit labor rates  and manpower
requirements/shift are specified in the case study O&M table and  can  be  specified independently for each case
study.

Maintenance Labor: 40% of the total maintenance cost calculated in the Capital Cost Model Sheet

Maintenance Material: 60% of the total maintenance cost calculated in the Capital Cost Model Sheet

Administrative and Support Labor: Calculated as a percentage (labor overhead charge rate) of the sum of the
operating and maintenance labor cost; overhead charge specified in the case study O&M table and can be specified
independently for each case study.

Consumable values are taken from the Technical Model Results Sheet.  The O&M cost consists of the following
consumable components:

Water (for flue gas humidiflcation), gallons/Hr: unit cost =  $/1000 gallons (e.g., 0.80), specified in case study
table
                                                 D-10

-------
Sorbent (e.g., Activated Carbon), tons /Hr: unit cost = $/ton (e.g., 1,100), specified in case study table

Incremental Power, kW-Hr: unit cost = $MW-Hr (e.g., 30), specified in case study table
Fan power accounts for the added pressure drop across the mercury control equipment, such as the fabric filter.
Sorbent injection system power is required to transport the sorbent to the flue gas duct.  Humidification system
power is required to pump water to an injection grid in the ductwork.

Waste Disposal, tons/Hr: unit cost = $/ton (e.g., 30), specified in case study table
Waste AC is generated by the  mercury control system.  This mercury-laden AC must be disposed of or processed
for mercury removal and recovery. This material can be disposed of with the rest of the plant's fly ash or it can be
processed separately if deemed a hazardous material.

Mercury Byproduct, Ib/Hr: unit cost = $/ton (e.g., 30), specified in case study table
If the spent AC is processed for recovery of mercury, then a by-product credit can be applied. This will be applied
only if the user has designated the spent AC as hazardous waste material.

1.1.7.2   Total Mercury Control System Capital Requirement

The total calculated investment in the mercury control system includes the total capital investment calculated in the
Capital Cost Model Sheet, interest during construction (AFUDC), and the following additional cost components:

Royalty allowance - possible technology royalty charges may apply

Preproduction Costs  covers the cost of operator training, equipment checkout, major modifications to
equipment, extra maintenance, and inefficient use of consumables. Calculated as 1 month of fixed operating costs
(O&M labor, admin and support labor, and  maintenance  materials);  1  month of variable operating costs (all
consumables) at full capacity; and 2 % of the total system investment.

Inventory Capital  value of initial inventory of activated carbon that is capitalized. This accounts for an initial
storage supply of AC (e.g., 30 day supply).

Initial Catalysts and Chemicals Charge  the initial cost of any catalysts or chemicals contained in the process
equipment, but not in storage. Does not apply to the mercury control systems.

1.1.7.3   Levelized Cost of Mercury Control

The total cost of mercury control must account for the total capital requirement (expressed as $/kW)  and the total
operating  and maintenance  expenses  (expressed as mills/kWh).  In order to  calculate an annualized cost that
accounts for both of these, the capital requirement is annuitized via use  of the "Levelized Carrying Charge Rate."
The Levelized Carrying Charge Rate assumes a 30-year operating period and accounts for return on debt, return on
equity, income  taxes, book depreciation, property tax, and insurance payments.  The Levelized Carrying Charge
Rate is specified in this sheet in the section called "Financial Data-Factors." It is multiplied times the total system
capital requirement to derive  the annualized  value and converted to  units of mills/kWh  based on  the annual
operating hours of the plant (capacity factor x 8,760 h/yr).

The first-year O&M costs that are calculated in this sheet are also levelized in order to  account for both apparent
and real escalation rates of labor, materials, and consumables over the expected operating time period  (e.g., 30
years).  A levelization factor is specified in this  sheet in the section called  "Financial  Data-Factors."  It is
multiplied by the total O&M cost.

The levelized carrying charge and the levelized O&M are summed to yield a total annualized cost which is divided
by the annual mercury removed in order to derive a unique cost of removal with units of $/ton mercury removed.
                                                  D-ll

-------
2.       DOCUMENTATION OF REFERENCE POWER PLANTS USED IN MERCURY CONTROL
         MODEL.

Two (2) representative power plant applications are being employed in this study to investigate the mercury control
costs  for the  baseline designs as well as parametric variations of these baseline plants.  These plants  are
characterized as:

1)       Plant firing high-sulfur, bituminous coal (Pgh #8) with low-NOx burners for NOX control, cold-
         side ESP for particulate control, and a wet FGD system for SO2 control.

2)       Plant firing low-sulfur coal with low-NOx burners for NOX control and a cold-side ESP for
         particulate control.

This section develops  the  basic  specifications for  each power  plant variant:  plant size, coal analysis, boiler
performance  parameters,  flue  gas  mass/volumetric  flow  rates, gas   constituents  (including  HC1),   gas
temperature/pressure profiles, total mercury concentration, and mercury speciation.

2.1      High Sulfur Power Plant Reference Case

The high sulfur reference coal  (PC)  plant design is  comprised of a balanced draft, natural circulation steam
generator, providing steam for a turbine generator set, condensing at 2.5  inches Hg absolute back pressure at the
design point. The plant design and performance reflect current commercial practice in the U.S. utility industry.  The
turbine-generator is a tandem compound machine, with high pressure (HP), intermediate pressure (IP), and low
pressure  (LP)  sections.   The LP turbine is comprised of two double flow sections exhausting downward into the
condenser sections.  The plant uses a 2400 psig/1000 F/1000 F single reheat steam power cycle. The boiler and the
turbine are designed for a main steam flow of 3,621,006 pounds of steam per hour at 2520 pounds per square inch,
gauge (psig) and 1000 F at the superheater outlet, throttled to 2415 pounds per square inch, absolute (psia) at the
inlet to the high pressure turbine. The cold reheat flow is 3,233,808 Ib/hr of steam at 590 psia and 637 F, which is
reheated  to  1000 F before  entering the intermediate pressure turbine section.   The net plant output power, after
plant auxiliary power requirements are deducted, is nominally 508 MWe. The overall net plant higher heating value
(HHV) efficiency is nominally 36.8 percent. Refer to Table 2-1 for the plant performance summary information.

Applicable Federal, State, and Local  environmental standards relating to air, water, solid waste and noise have been
designed into the high sulfur reference plant. Projected plant air emissions are identified in Table 2-2.   The  wet,
limestone FGD system ensures an SO2 emission rate of less than 0.371 Ib/MMBtu (92% reduction).  The use of low
NOX burner technology, combined with over fire air, results in NOX emissions of less than 0.30 Ib/MMBtu.  The
control or reduction of N2O has not been addressed  in this design because N2O levels are presently unregulated.
The flue  gas scrubber is a wet limestone type  system, with scrubbing and demisting occurring in the same vessel.
An organic acid is added to  the circulating reagent to enhance the scrubbing performance.  Air is blown into the
scrubber  module sump to promote forced oxidation of the sulfite to sulfate.  The gypsum byproduct is  dried to a
cake-like consistency in a train of centrifuges, and is ready to landfill.
                                                 D-12

-------
                              TABLE 2-1
HIGH SULFUR PLANT PERFORMANCE SUMMARY -100 PERCENT LOAD
STEAM CYCLE
Throttle Pressure, psig
Throttle Temperature, F
Reheat Outlet Temperature F
POWER SUMMARY
3600 rpm Generator
GROSS POWER, kWe (Generator terminals)
AUXILIARY LOAD SUMMARY, kWe
Pulverizers
Primary Air Fans
Forced Draft Fans
Induced Draft
Seal Air Blowers
Main Feed Pump (Note 1)
Steam Turbine Auxiliaries
Condensate Pumps
Circulating Water Pumps
Cooling Tower Fans
Coal Handling
Limestone Handling & Reagent Prep.
FGD Pumps and Agitators
Ash Handling
Dewatering Centrifuges (FGD byproduct)
Precipitators
Soot Blowers (Note 2)
Miscellaneous Balance of Plant (Note 3)
Transformer Loss
TOTAL AUXILIARIES, kWe
Net Power, kWe
Net Efficiency, % HHV
Net Heat Rate, Btu/kWh (HHV)
CONDENSER COOLING DUTY, 106 Btu/hr
CONSUMABLES
As-Received Coal Feed, Ib/hr
Sorbent, Ib/hr

2,400
1,000
1,000


541,900

2,000
1,840
1,350
7,500
60
2,400
900
1,100
5,070
2,400
230
1,160
3,000
2,000
1,100
1,100
neg.
2,000
1,300
34,100
507,800
36.8
9,279
2,350

378,550
46,790
  Note 1 - Driven by auxiliary steam turbine, electric equivalent not included in total.
  Note 2 - Soot blowing medium is boiler steam. Electric power consumption is negligible.
  Note 3 - Includes plant control systems, lighting, HVAC, etc.
                                 D-13

-------
                                             TABLE 2-2
                  HS PLANT AIR EMISSIONS - 65% CAPACITY FACTOR

sox
NOx
Particulate
Ib/MMBtu
0.46
0.30
0.01
Tons/Year
5,880
3,840
125
The following indicates the state point conditions at various flue gas stream locations;
Flue gas mass flow rate, boiler exit
Flue gas temperature at boiler exit
Flue gas temperature at ESP exit
Flue gas temperature at FGD exit
4,966,633 Ib/hr (1,561,834 acfm)
290 F
290 F
128 F
The flue gas composition and temperatures at various points of the gas stream are detailed in Table 2-3.  This data
was calculated using the ACI Mercury Control Cost Model described in Section 3.

2.1.1    Plant Site Ambient Conditions

The plant  site is assumed to be in the Ohio River Valley of western Pennsylvania/eastern Ohio/ northern West
Virginia.   The site consists  of approximately 300 usable acres, not including ash disposal, within 15 miles of a
medium sized metropolitan area, with a  well  established  infrastructure  capable  of supporting  the required
construction workforce.

The site is within Seismic Zone  1, as defined by the Uniform Building Code, and the ambient design conditions will
be:

Pressure                           14.4psia
Dry bulb temperature               60 F
Dry bulb temperature range  (-)  10 to (+) 110 F
Wet bulb temperature               52 F

2.1.2    Fuel and Sorbent Composition

The plant performance will be based on the Pittsburgh #8 Coal and Greer limestone compositions and data listed in
Tables 2-4 and 2-5.  Unit start-up will use No.2 fuel oil.
                                                D-14

-------
                                TABLE 2-3
HIGH SULFUR (HS) REFERENCE POWER PLANT GAS STREAM DATA
       Boiler & Plant Data Summary                         PGH #8 (HS)

       Btu Input Rate (MMBtu/h)                             4,711.88
       Coal Input Rate (tons/h)                                188.337
       Coal Mass-Energy Ratio (Ib/MMBtu)                      79.94
       MMBtu/hr Out of Boiler                               4,162.49

       Flue Gas @ Economizer Outlet (SCFM/MMBtu)             12,503
       Flue Gas @ Economizer Outlet (SCFM)                  981,877
       Wet Flue Gas @ Economizer Outlet (Ib/h)               4,598,641
       Wet Air Based on Economizer Outlet (SCFM)             934,407
       Wet Air Based on Economizer Outlet (Ib/h)              4,260,670
       Total Air, % (Economizer Outlet)                          120.0
       Excess Air, % (Economizer Outlet)                         20.0

       Flue Gas @ Air Heater Outlet (SCFM/MMBtu)              13,753
       Flue Gas @ Air Heater Outlet (SCFM)                  1,080,065
       Wet Flue Gas @ Air Heater Outlet (Ib/h)                 5,042,924
       Wet Air Based on Air Heater Outlet (SCFM)             1,031,843
       Wet Air Based on Air Heater Outlet (Ib/h)               4,704,952
       Wet Air Leakage @ Air Heater (SCFM)                    97,436
       Wet Air Leakage @ Air Heater (Ib/h)                    444,283
       Equivalent Total Air, % (Air Heater Outlet)                 132.5
       Equivalent Excess Air, % (Air Heater Outlet)                 32.5

       Wet Flue Gas @ Air Heater Outlet (ACFM)              1,689,826
       Air Heater Outlet Flue Gas Temperature, F                   290
       Air Heater Outlet Pressure, inches Hg.                      29.31

       Fly Ash (ton/h)                                         15.48
       Bottom Ash (ton/h)                                       3.87
       Total Ash (ton/h)                                       19.35
       NOTE:  Standard Conditions = 60 F, 29.92 inches Hg
                                   D-15

-------
                                    TABLE 2-4
                         GREEK LIMESTONE ANALYSIS
Calcium Carbonate, CaCO3

Magnesium Carbonate MgCO3

Silica, SiO2

Aluminum Oxide, A12O3

Iron Oxide, Fe2O3

Sodium Oxide, Na2O

Potassium Oxide, K2O

Balance
Dry Basis. %

    80.4

    3.5

    10.32

    3.16

    1.24

    0.23

     0.72

     0.43
                                       D-16

-------
           TABLE 2-5
PITTSBURGH NO. 8 COAL ANALYSIS
Constituent
Carbon
Hydrogen
Nitrogen
Sulfur
Ash
Oxygen
Total

Moisture
Ash
Volatile Matter
Fixed Carbon
Total
Cl, ppm in Coal
Hg, ppb in Coal
Sulfur
Btu Content
Moisture and Ash Free (MAP), Btu

Silica, SiO2
Aluminum Oxide, A1O3
Iron Oxide, Fe2O3
Titanium Dioxide, TiO2
Calcium Oxide, CaO
Magnesium Oxide, MgO
Sodium Oxide, Na2O
Potassium Oxide, K2O
Sulfur Trioxide, SO3
Phosphorous Pentoxide, P2O5
Total



Initial Deformation
Spherical
Hemispherical
Fluid
Air Dry. %
71.88
4.97
1.26
2.99
10.30
8.60
100.00
Dry Basis. %
	
10.57
38.20
51.23
100.00
650
78 24
3.07
13,244
14,810
Ash Analysis. %
48.1
22.3
24.2
1.3
1.3
0.6
0.3
1.5
0.8
0.1
100.5
Ash Fusion Temperature. F
Reducing
Atmosphere
2015
2135
2225
2450
Dry. % As Received. %
73.79 69.36
4.81 5.18
1.29 1.22
3.07 2.89
10.57 9.94
6.47 11.41
100.00 100.00
As Received. %
6.00
9.94
35.91
48.15
100.00


2.89
12,450














Oxidizing
Atmosphere
2570
2614
2628
2685
             D-17

-------
2.1.3     Air Quality Standards

The plant pollution  emission requirements for the High Sulfur Reference  Case reflect current  environmental
emissions standards for a plant sited in a non-attainment area with respect to ambient air standards for ozone. Table
2-2 presents emissions for the plant without site sensitive NOX reduction enhancements.

2.1.4     Flue Gas Acid Dew Point. 1.4

Based on a flue gas SO3 concentration of 22 ppm and a water concentration of 7.65% at the exit of the air heater, the
acid dew point is estimated to be approximately 280 F.

2.1.5     Mercury Emission Assumptions
Table 2-6, presented below, provides basic data for baseline and variant mercury emissions.
                                              TABLE 2-6
                                  MERCURY EMISSIONS DATA
Total Mercury
Concentration
Oig/Nm3)
10
(Baseline)
3
30
Total Mercury
Emissions Rate
(g/h)
18.35
5.5
55.05
Total Annual Mercury
Release1
(Kg/yr)
104.4
31.3
313.2
                  1. Based on a 65% plant capacity factor.
2.2
Low Sulfur Power Plant Reference Case
The  low sulfur reference  pulverized coal (PC) plant design utilizes a balanced draft, natural circulation type,
pulverized  coal subcritical fired boiler, providing steam  for a  turbine generator  set.   The  boiler design and
performance reflect current commercial practice in the U.S. utility industry.  The turbine-generator is a tandem
compound machine, with high pressure (HP), intermediate pressure (IP), and low pressure (LP) sections. The LP
turbine is comprised of two double flow sections exhausting downward into the condenser sections. The low sulfur
reference plant uses a 2400 psig/1000 F/1000 F single reheat steam power cycle.  The boiler and the turbine are
designed for a main steam flow of 2,734,000 pounds of steam per hour at 2520 psig and 1000 F at the superheater
outlet, throttled to 2415 psia at the inlet to the high pressure turbine.  The cold reheat flow is 2,425,653 Ib/hr of
steam at 604 psia and 635 F, which is reheated to 1000 F before entering the intermediate pressure turbine section.
The  net plant  output power, after plant auxiliary power requirements are deducted, is nominally 404 MWe. The
overall net plant higher heating value (HHV) efficiency is nominally 39.1 percent. Refer to Table 2-7 for the plant
performance summary information.

The  plant also is designed to meet applicable Federal, State, and Local environmental  standards relating to air,
water, solid waste and noise.  The plant has baseline SO2 emissions of about 0.70 Ib/MMBtu, and it is designed to
utilize in-duct  spray drying of lime to provide a removal efficiency of 50%. Lime slurry is  sprayed into the flue gas
duct, where it  dries and captures SO2 as a particle.  The particulate is then collected in the electrostatic precipitator
                                                 D-18

-------
for disposal. If placed in service, the FGD system results in a SO2 emission rate of less than 0.35 Ib/MMBtu; the
current investigation assumes that the FGD system will not be in operation. The use of low NOX burner technology,
combined with over-fire air, results in NOX emissions of less than 0.30 Ib/MMBtu.  The control or reduction of N2O
has not been addressed in this design because N2O levels are presently unregulated.

The low Sulfur Case Reference  plant achieves a net plant efficiency of 39.1%, which is an increase of over 6%
above that achieved by the High Sulfur Reference plant (36.8%). This increase in efficiency is achieved although
the units are virtually identical in terms of size, boiler selection, steam cycle configuration, heat sink, and  other
considerations. The net efficiency improvements are due to the following:

        The low sulfur reference auxiliary plant load is  lower than the high sulfur plant's auxiliary load.  This
         reduction occurs because of the reduced ID fan power requirements, which result from the elimination of
         the pressure losses due to a scrubber.   In addition,  elimination of the reagent and byproduct handling
         systems associated with  a  wet scrubber reduce  the auxiliary load but is offset by  the atomizing air
         compressors for the duct injection system.

        The low sulfur plant gas temperature leaving the boiler is set at 270 F versus 290 F for the high sulfur
         plant. This reduction in exhaust temperature, which improves the boiler efficiency, is feasible because of
         the low SO2 concentrations in the gas leaving the  boiler, and the gas passes  directly into the duct injection
         system that is upstream of the ESP.

2.2.1     Plant Site Ambient Conditions

The plant site is  assumed to  be  in the  Ohio River  Valley of western Pennsylvania/eastern  Ohio/ northern  West
Virginia.  The site consists of approximately 300 usable  acres; not  including ash disposal, within 15 miles of a
medium sized metropolitan  area, with a well  established infrastructure  capable of supporting the required
construction workforce.   The  site is  within  Seismic  Zone  1, as defined by  the Uniform Building Code, and the
ambient design conditions will be:
         Pressure                                     14.4psia
         Dry bulb temperature                         60 F
         Dry bulb temperature range          -10 to+110 F
         Wet bulb temperature                         52 F
                                                  D-19

-------
                        TABLE 2-7
LOW SULFUR POWER PLANT PERFORMANCE SUMMARY -
                   100 PERCENT LOAD
STEAM CYCLE
Throttle Pressure, psig
Throttle Temperature, F
Reheat Outlet Temperature
POWER SUMMARY
3600 rpm Generator
GROSS POWER, kWe (Generator terminals)
AUXILIARY LOAD SUMMARY, kWe
Pulverizers
Primary Air Fans
Forced Draft Fans
Induced Draft
Seal Air Blowers
Main Feed Pump (Note 1)
Steam Turbine Auxiliaries
Condensate Pumps
Circulating Water Pumps
Cooling Tower Fans
Coal Handling
Limestone Handling & Reagent Prep.
Ash Handling
Atomizing Air compressors (Duct Injection)
Precipitators
Soot Blowers (Note 2)
Miscellaneous Balance OF Plant (Note 3)
Transformer Loss
TOTAL AUXILIARIES, kWe
Net Power, kWe
Net Efficiency, % HHV
Net Heat Rate, Btu/kWh (HHV)
CONDENSER COOLING DUTY, 106 Btu/h
CONSUMABLES
As-Received Coal Feed, Ib/h
Sorbent, Ib/h

2,400
1,000
1,000


427,060

1,600
1,410
1,020
3,230
50
8,660
800
800
3,400
1,800
180
230
1,600
2,700
900
neg.
2,000
1,020
22,740
404,320
39.1
8,726
1,722

299,204
4,277
 Note 1 - Driven by auxiliary steam turbine, electric equivalent shown.
 Note 2 - Soot blowing medium is boiler steam. Electric power consumption is negligible.
 Note 3 - Includes plant control systems, lighting HVAC, etc.
                           D-20

-------
2.2.2     Fuel and Sorbent Composition

The plant performance will be based on  a, sub-bituminous,  western Powder River Basin (PRB) coal that has
undergone a moisture reduction and stabilization process. The coal analysis and data are listed in Tables 2-8 and 2-
9. Unit start-up will use No.2 fuel oil.

2.2.3     Air Quality Standards

The plant pollution emission requirements for the Low Sulfur Reference Case adhered to Federal and State control
emission regulations.   Although, some environmental  regulations may apply on a  site specific  basis (National
Environmental Policy Act, Endangered Species Act, National Historic Preservation Act, etc.) will not be considered
in this project. The following ranges will generally cover most cases:

         SOX:                      92 to 95% reduction
         NOX:                      0.2to0.451bperMMBtu
         Particulate:                0.015 to 0.03 Ib per MMBtu
         Opacity:                   10 to 20 percent

2.2.4     Flue Gas Acid Dew Point.

Based on a flue gas SO3 concentration of 0.5 ppm and a water concentration of 6.5% at the exit of the air heater, the
acid dew point is estimated to be approximately 175 F.

2.2.5     Mercury Emission Assumptions

Table 2-10 provides basic data for baseline and variant mercury emissions.

2.2.6     LS Reference Power Plant Gas Stream Data

The following indicates the state point conditions at various flue gas stream locations.

         Flue gas mass flow rate, boiler exit            3,821,126 Ib/hr (1,309,735 cfm)
         Flue gas temperature at AH exit               270 F
         Flue gas temperature at ESP exit              270 F

The flue gas composition and temperatures at various points of the gas stream are detailed in Table 2-11. This data
was calculated using the  MCPCM described in Section 1 and coal properties for  a low sulfur processed derived
PRB coal.
                                                 D-21

-------
             TABLE 2-8
TYPICAL PROCESSED PRB COAL ANALYSIS
Constituent
Carbon
Hydrogen
Nitrogen
Sulfur
Ash
Oxygen
Total

Moisture
Ash
Volatile Matter
Fixed Carbon
Total
Sulfur
Btu Content
Moisture and Ash Free (MAP), Btu Content

Silica, SiO2
Aluminum Oxide, Al
Iron Oxide,
Titanium Dioxide, TiO2
Calcium Oxide, MgO
Magnesium Oxide, MgO
Sodium Oxide, Na2O
Potassium Oxide, K2O
Sulfur Trioxide, SO3
Phosphorous Pentoxide, P2O5
Strontium Oxide, SrO
Barium Oxide, BaO
Manganese Oxide, Mri4
Total
Ash Fusion Temperature. "F


Initial Deformation
Spherical
Hemispherical

Drv. % As Received. %
75.25 70.98
3.46 4.00
1.13 1.07
0.56 0.53
8.19 7.72
11.41 15.70
100.00 100.00
Drv Basis, % As Received. %
4.83
8.19 7.72
27.00 25.72
64.81 61.73
100.00 100.00
0.56 0.53
12,389 11,791
13,494
Ash Analysis, %
22.5
13.8
7.4
0.8
26.6
5.9
1.8
0.2
19.3
0.6
0.4
0.6
0.1
100.00

Reducing Oxidizing
Atmosphere Atmosphere
2295 2395
2300 2405
2305 2415
2310 2425
               D-22

-------
                      TABLE 2-9
COMPARISON OF FEED COAL AND MODIFIED COAL BASIS

Heating Value (Btu/lb)
Carbon (%)
Hydrogen (%)
Nitrogen (%)
Volatiles (%)
Feed Coal
12,740
73.4
5.5
1.1
47.0
Process Derived Fuel
13,840
84.0
3.6
1.3
32.0
                     TABLE 2-10
             MERCURY EMISSIONS DATA
Total Mercury
Concentration
Oig/Nm3)
10
(Baseline)
3
30
Total Mercury
Emissions Rate
(g/Hr)
13.8
3.5
138
Total Annual Mercury
Release1
(Kg/Yr)
78.5
23.5
235.5
  1. Based on a 65% plant capacity factor.
                        D-23

-------
                           TABLE 2-11
     LS REFERENCE POWER PLANT GAS STREAM DATA

Boiler & Plant Data Summary                            LS PDF (LS)

Btu Input Rate (MMBtu/h)                                      3,528.10
Coal Input Rate (tons/h)                                         150.739
Coal Mass-Energy Ratio (Ib/MMBtu)                                85.45
MMBtu/hr Out of Boiler                                        3,119.59

Flue Gas @ Economizer Outlet (SCFM/MMBtu)                     12,656
Flue Gas @ Economizer Outlet (SCFM)                           744,192
Wet Flue Gas @ Economizer Outlet (Ib/h)                        3,518,066
Wet Air Based on Economizer Outlet (SCFM)                      710,817
Wet Air Based on Economizer Outlet (Ib/h)                       3,241,152
Total Air, % (Economizer Outlet)                                   120.0
Excess Air, % (Economizer Outlet)                                  20.0

Flue Gas @ Air Heater Outlet (SCFM/MMBtu)                      13,795
Flue Gas @ Air Heater Outlet (SCFM)                             811,169
Wet Flue Gas @ Air Heater Outlet (Ib/h)                         3,821,126
Wet Air Based on Air Heater Outlet (SCFM)                       777,281
Wet Air Based on Air Heater Outlet (Ib/h)                        3,544,213
Wet Air Leakage @ Air Heater (SCFM)                            66,464
Wet Air Leakage @ Air Heater (Ib/h)                              303,061
Equivalent Total Air, % (Air Heater Outlet)                           131.2
Equivalent Excess Air, % (Air Heater Outlet)                          31.2

Wet Flue Gas @ Air Heater Outlet (ACFM)                      1,309,735
Air Heater Outlet Flue Gas Temperature, F                            314
Air Heater Outlet Pressure, inches Hg.                               29.31

Fly Ash (ton/h)                                                   9.83
Bottom Ash (ton/h)                                                2.46
Total Ash (ton/h)                                                 12.28
NOTE: Standard Conditions = 60 F, 29.92 inches Hg
                               D-24

-------
3.
Bituminous and Subbituminous Coals Used in Mercury Control Model Runs
EPA's mercury control technology application matrix cites three different coals for the model runs.  The
purpose of this section is to identify and document these three coals.

The two bituminous coals are from West Virginia - ultimate analysis and ash analysis data were obtained
from the USGS Coal Quality Database. The high sulfur coal has a 3% sulfur content (by weight) and a
HHV of 12,721 Btu/Lb, while the other has a 0.6% S content (by weight) and a HHV of 14,224 Btu/Lb.

The subbituminous coal is from Wyoming's Powder River Basin (PRB) and was already contained in the
model's coal library. Data originally came from EPRI's FGD Cost Program. This coal has  a coal sulfur
content of 0.5% (by weight) and a HHV of 8,335 Btu/Lb.
                                       TABLE 3-1
        Bituminous and Subbituminous Coals Used in Mercury Control Model Runs
                               WYOMING PRB (LS)   E. Bituminous (HS)
COAL ULTIMATE ANALYSIS
(ASTM, as received, weight percent)
                                                                    E. Bituminous (LS)
     Moisture
     Carbon
     Hydrogen
     Nitrogen
     Chlorine
     Sulfur
     Ash
     Oxygen
    TOTAL
                       30.40%
                       47.85%
                       3.40%
                       0.62%
                       0.03%
                       0.48%
                       6.40%
                       10.82%
                       100.00%
3.10%
69.82%
5.00%
1.26%
0.12%
3.00%
9.00%
8.70%
100.00%
                                    2.20%
                                    78.48%
                                    5.50%
                                    1.30%
                                    0.12%
                                    0.60%
                                    3.80%
                                    8.00%
                                    100.00%
 Mott Spooner HHV (Btu/lb)
Acid Dew Point, F
8,335
 224
                                                      12,721
                                                       215
                              14,224
                               292
COAL ASH ANALYSIS
                        WYOMING PRB (LS)  E. Bituminous (HS)
                      E. Bituminous (LS)
     SiO2                      31.60%
     A12O3                    15.30%
     TiO2                      1.10%
     Fe2O3                    4.60%
     CaO                      22.80%
     MgO                      4.70%
     Na2O                     1.30%
     K2O                      0.40%
     P2O5                     0.80%
     SOS                      16.60%
     Other Unaccounted for       0.80%
    TOTAL                    100.00%
                                             29.00%
                                             17.00%
                                             0.74%
                                             36.00%
                                             6.50%
                                             0.83%
                                             0.20%
                                             1.20%
                                             0.22%
                                             7.30%
                                             1.01%
                                             100.00%
                      51.00%
                      30.00%
                      1.50%
                      5.60%
                      4.20%
                      0.76%
                      1.40%
                      0.40%
                      1.80%
                      2.60%
                      0.74%
                      100.00%
                                          D-25

-------
4.      COST ESTIMATION BASIS

4.1     Introduction

This section defines the methodology used to estimate capital and O&M costs within the NETL Mercury
Control Performance and Cost Model.   Two different spreadsheets within the model provide for cost
estimation. The Capital Cost Model Sheet makes use of the power plant and mercury control performance
data to calculate the capital costs associated with a case study's mercury control design configuration.  The
costing covers the following equipment sections:

       Spray Cooling Water System  (Equipment: water storage tank,  pumps, transport piping, and
        injection grid with nozzles, and control system)

       Solid Sorbent Storage and Injection System (Equipment: silo pneumatic loading system,
        storage silos, hoppers, blowers, transport piping, control system)

       Sorbent Recycle System  (Equipment: hoppers, blowers, transport piping, control system)

       Pulse-Jet Fabric Filter and Accessories (Equipment: pulse-jet FF, filter bags, ductwork, dampers,
        andMCCs and instrumentation andPLC controls for baghouse operation. Excludes Ash Removal
        System, power distribution and power supply, and distributed control system)

       Sorbent Disposal System (Equipment: hoppers, blowers, transport piping, control system)

       CEMS Upgrade

       Flue Gas Desulfurization (FGD) System (cost algorithms are currently not provided)

       FGD System Enhancements  (cost algorithms are currently not provided)

       Circulating Fluidized Bed Absorber System (cost algorithms are currently not provided)

A total mercury control system cost is calculated from the following cost components:  1) equipment, 2)
related materials, 3) field and indirect labor, 4) sales tax, 5) engineering and home office fees, 6) process
and project contingencies,  7) retrofit factors, and  9) general facilities.   Maintenance  costs are also
calculated as a percentage of the bare erected cost (e.g., 2%).  Section 4.2 fully describes the methodology
used to estimate the total installed retrofit capital cost.   Section  4.3  defines specific algorithms  used
calculate specific equipment and installation labor costs.

The System Economic Model Sheet  calculates mercury control system O&M, total system investment,
total system capital requirement, and then levelizes the capital and O&M to establish single value for $/lb of
mercury removed.  Twenty case studies are handled simultaneously in this sheet.  Sections 4.4 and 4.5
fully describe the methodology used to estimate these costs.

4.2     Capital Cost Estimation Basis

The cost of each equipment  section, as identified in Section 4.1,  is estimated according to the following
procedure:

        Bare Installed Retrofit Cost = (Process Equipment + Related Field Materials + Field Labor +
                                        Indirect Field Costs + Sales Tax) x Retrofit Factor

                Equipment, field materials and field labor are specified via algorithms.
                Indirect field costs are calculated as percentage of field labor (7% currently specified).
                                              D-26

-------
                Sales tax is calculated as a percentage of the sum of the four cost elements (0% currently
                  specified)
                Retrofit Factor (accounts for retrofit difficulty) = 1.15 (Specified by EPA on 4/4/00)

        Engineering & Home Office Overhead/Fees = Bare Installed Retrofit Cost x  E&HO Percentage

                EH&O Percentage = 10% (Specified by EPA on 4/4/00)

        Process Contingency = Bare Installed Retrofit Cost x Process Contingency Percentage

                Process Contingency Percentage = 5% (Specified by EPA on 4/4/00)

        Project  Contingency =  (Bare  Installed Retrofit  Cost  +  Engineering   &  Home  Office
                Overhead/Fees + Process Contingency) x Project Contingency Percentage

                Project Contingency Percentage = 15% (Specified by EPA on 4/4/00)

        Total Cost of Each Equipment Section = Bare Installed Retrofit Cost + Engineering & Home
                Office Overhead/Fees + Process Contingency + Project Contingency

The total capital cost of the mercury control  system is calculated to include the  sum of all equipment
sections and the total cost of general facilities as follows:

        General Facilities Cost = (Bare Installed Retrofit Cost x General Facilities Percentage

                General Facilities Percentage = 5% (Specified by EPA on 4/4/00)
                Project Contingency Percentage defined above
        Total Control Capital Cost = Sum of Equipment Section Total Costs + General Facilities Cost
4.3     Equipment and Installation Labor Cost Estimation

This section of the memo defines the equipment cost algorithms used in the Capital Cost Model Sheet.
Each equipment section is defined separately below. Costs are updated from their baseline year values via
use of the Chemical Engineering Annual Plant Index (CEI). The costing algorithms relate to a December
1998 baseline for which the CI value equals 389.5.  The ratio of the current index value (e.g., December
1999) and the baseline value therefore yields a cost inflator that adjusts control costs to the specified year.

4.3.1   Spray Cooling Water System

        Process Equipment (x $1000). $E

        $E = 1900 x (GPM/215)065 x CEI/389.5

                GPM = Water flow in gallons/minute

        Field Materials (x $1000). $FM

        $FM = 1700 x (GPM/215)065 x CEI/389.5
                                             D-27

-------
        Field Labor (x $1000). $FL

        $FL = 1500 x (GPM/215)065 x CEI/389.5

        Indirect Field Costs (x $1000). $IF

        $IF = $FL x 0.07

        Bare Installed Cost (x $1000) = $E + $FM + $FL + $IF

Example:       GPM = 27.4 gpm (Cools flue gas from 290 F to 270 F, 472 MWe,net)
               CEI = 399.7 (November 1999)

        Bare Installed Cost (x $1000) = 511 + 457+ 404 + 28 = 1,611 or $3.41/kW


4.3.2    Solid Sorbent Injection System

        Process Equipment (x $1000). $E

        $E = 400 x ((SF x 1000/454)75486)65 x CEI/389.5

               SF = Sorbent Feed, Kg/Hr

        Field Materials (x $1000). $FM

        $FM = 900 x ((SF x  1000/454)/5486)065 x CEI/389.5

        Field Labor (x $1000). $FL

        $FL = 2600 x ((SF x 1000/454)/5486)065 x CEI/389.5

        Indirect Field Costs (x $1000). $IF

        $IF = $FL x 0.07

        Bare Installed Cost (x $1000) = $E + $FM + $FL + $IF
Example:       SF = 157 Kg/Hr (Based on 3.73 Ib/MMacf, 472 MWe,net)
               CEI = 399.7 (November 1999)

        Bare Installed Cost (x $1000) = 68+ 153 + 442 + 31 = 799 or $1.69/kW
4.3.3    Sorbent Recycle System

        Process Equipment (x $1000). $E

        $E = 1200 x (RR/13) x CEI/389.5

               RR = Recycle (sorbent and ash), Tons/Hr




                                           D-28

-------
        Field Materials (x $1000). $FM

        $FM = $E

        Field Labor (x $1000). $FL

        $FL = $E

        Indirect Field Costs (x $1000). $IF

        $IF = $FL x 0.07

        Bare Installed Cost (x $1000) = $E + $FM + $FL + $IF



4.3.4    Pulse-Jet Fabric Filter and Accessories

        Process Equipment (x $1000). $E

        $E = 4800 x ((GFR/ACR)/84,326))080 x CEI/389.5

               GFR = Flue Gas Flow Rate, acfm
               ACR = PJFF air/cloth ratio, ft3/min/ft2

        Field Materials (x $1000). $FM

        $FM = 500 x ((GFR/ACR)/84,326))080  x CEI/389.5

        Field Labor (x $1000). $FL

        $FL = 2700 x ((GFR/ACR)/84,326))080 x CEI/389.5

        Indirect Field Costs (x $1000). $IF

        $IF = $FL x 0.07

        Bare Installed Cost (x $1000) = $E + $FM + $FL + $IF
Example:       GFR = 1,547,360 acfm (Based on 472 MWe,net)
               ACR = 10 ftVmin/ft2
               CEI = 399.7 (November 1999)

        Bare Installed Cost (x $1000) = 8,006+ 834 + 4,503 + 315 = 15,707 or $33/kW
4.3.5    Sorbent Disposal System

        Process Equipment (x $1000). $E

        $E = 100 x (DS/6) x CEI/389.5

               DS = Disposal Solids (spent sorbent and ash), Tons/Hr




                                           D-29

-------
        Field Materials (x $1000). $FM

        $FM = 2 x $E

        Field Labor (x $1000). $FL

        $FL = 6 x $E

        Indirect Field Costs (x $1000). $IF

        $IF = $FL x 0.07

        Bare Installed Cost (x $1000) = $E + $FM + $FL + $IF


4.3.6    CEMS Upgrade

        Process Equipment (x $1000). $E

        $E = 10 x (MW/290.4)075 x CEI/389.5

               MW = Power Plant Application Net Capacity, MWe,net

        Field Materials (x $1000). $FM

        $FM = 0

        Field Labor (x $1000). $FL

        $FL= 1.2 x$E

        Indirect Field Costs (x $1000). $IF

        $IF = $FL x 0.07

        Bare Installed Cost (x $1000) = $E + $FM + $FL + $IF


4.4     Mercury Control System O&M Cost Estimation

4.4.1    O&M Cost Parameters

The O&M cost consists of the following labor and maintenance components:

        Operating  Labor: cost of system operating  and administrative personnel; unit labor rates and
        manpower requirements/shift are specified in the case study O&M table and can be specified
        independently for each case study.

        Maintenance Labor: 40% of the total maintenance cost calculated in the Capital Cost Model
        Sheet

        Maintenance Material: 60% of the  total maintenance cost calculated in the Capital Cost Model
        Sheet
                                            D-30

-------
        Administrative and Support Labor:  Calculated as a percentage (labor overhead charge rate) of
        the sum of the operating and maintenance labor cost; overhead charge specified in the case study
        O&M table and can be specified independently for each case study.

Consumable values are taken from the Technical Model  Results Sheet.  The O&M cost consists of the
following consumable components:

         Water (for flue  gas  humidification),  gallons/Hr: unit cost =  S/1000 gallons (e.g.,  0.80),
        specified in case study table

        Water quantity is taken from Combustion Calculations Sheet.

        Sorbent (e.g., Activated Carbon), tons /Hr: unit cost = S/ton (e.g., 1,100), specified in case study
        table

        Sorbent unit cost is an input in the Application Input Sensitivity Sheet

        Incremental Power, kW-Hr: unit cost = $/MW-Hr (e.g., 30), specified in case study table

        Fan power accounts for the added pressure drop across the mercury control equipment, such as the
        fabric  filter. AC  injection system power required  to transport  sorbent  to  flue  gas duct.
        Humidification system power required to pump water to an injection grid in the ductwork.

        Waste Disposal, tons/Hr: unit cost = S/ton (e.g.,  30), specified in case study table

Waste sorbent is generated by the mercury control system.  This mercury-laden sorbent must be disposed of
or processed for mercury removal and recovery.  For some design configurations, spent sorbent is captured
with the fly ash and must  be disposed  of  with the ash  at the  cost of ash disposal.  For some  design
configurations, spent sorbent is collected with residual ash  from the ESP. This material can be disposed of
with the rest of the plant's fly ash or it can be processed separately if deemed a hazardous material.

        The user specifies  sorbent to be hazardous or non-hazardous  in the Application Input Sensitivity
        Sheet.  Unit costs for both conventional and hazardous waste disposal are also specified there.

        Mercury Byproduct, Ib/Hr: unit cost  = S/ton (e.g., 30), specified in  case study table

        If the spent sorbent is processed for recovery of mercury, then a by-product credit can be applied.
        This will be applied only if the user has designated the spent AC as hazardous waste material.

4.4.2   Key O&M Cost Parameter Values

        Labor:
        Operating Labor Rate (base) - $25/Hr (specified by EPA 4/4/00)
        Total Operating Jobs - 0.833 OJ/Shift

        Consumables:
        Water - 0.42 Mills/gallon (specified by EPA 4/4/00)
        Activated Carbon - $l/Kg  (specified by EPA 4/4/00)
        Sorbent Storage Capacity - 30 days (specified by EPA 4/4/00)
        Electricity - 25 Mills/kW-Hr (specified by EPA 4/4/00)
        Waste Disposal (ash, AC, mercury)  - $30/Ton
        Waste Disposal (Hazardous waste designation) ~ $1700/Ton
                                              D-31

-------
        Plant Capacity Factor - 65% (specified by EPA 4/4/00)
4.5     Total Mercury Control System Capital Requirement

The  total calculated investment in the mercury control  system includes the total capital investment
calculated in the Capital Cost Model Sheet, interest  during construction (AFUDC), and the following
additional cost components:

        Royalty allowance ~ possible technology royalty charges may apply

        Preproduction  Costs ~  covers the  cost of operator training,  equipment  checkout,  major
        modifications to equipment, extra maintenance, and inefficient use of consumables. Calculated as
        1 month of fixed operating costs  (O&M labor,  admin and support labor,  and maintenance
        materials);  1 month of variable operating costs (all consumables) at full capacity;  and 2 % of the
        total system investment.

        Inventory  Capital  - value of initial inventory of activated carbon that is capitalized.  This
        accounts for an initial storage supply of AC (e.g., 30 day supply).

        Initial Catalysts and Chemicals Charge - the initial cost of any catalysts or chemicals contained
        in the process equipment, but not in storage. Does not apply to the mercury control systems.

4.5.1    Levelized Cost of Mercury Control

The total cost of mercury control must account for the total capital requirement (expressed as $/kW) and
the total operating  and maintenance expenses (expressed  as  mills/kW-Hr).  In order to calculate an
annualized cost that accounts for  both of these, the capital  requirement  is annuitized via use of  the
"Levelized Carrying Charge  Rate." The Levelized Carrying Charge Rate assumes  a 30 year operating
period and accounts for return on debt, return on equity, income  taxes, book depreciation, property tax, and
insurance  payments. The Levelized  Carrying Charge  Rate is specified in this  sheet  in the section called
"Financial Data-Factors."   It is multiplied  times the total system capital requirement  to derive  the
annualized value and converted to units  of mills/kW-Hr based  on the annual operating hours of the plant
(capacity factor x 8,760 hrs/yr).

The first-year O&M costs that are calculated in this sheet are also levelized in  order to account for  both
apparent and real escalation rates of labor, materials, and consumables over the expected operating time
period (e.g., 30 years).   A levelization  factor is specified in this sheet in the section  called "Financial
Data-Factors."  It is  multiplied by the total O&M cost.

The levelized carrying charge and the levelized O&M are summed to  yield a total annualized cost which is
divided by the  annual mercury removed in order to derive  a unique cost of removal with units of $/ton
mercury removed.

4.5.2    Financial Parameter Values

        Apparent General Escalation Rate - 2.9%/Year
        Royalty Allowance ~ $0
        Levelized Carrying Charge Rate - 0.133
        Federal Income Tax Rate - 34%
        Weighted Cost of Capital (after tax) - 9.4%
        Design and Construction - 1  year
        Book Life - 30 Years
                                              D-32

-------
               ATTACHMENT 2

                 Description of
Mercury Control Performance Algorithms Used in the
      National Energy Technology Laboratory
   Mercury Control Performance and Cost Model
                      D-33

-------
            Description of Mercury Control Performance Algorithms Used in the
                    NETL Mercury Control Performance and Cost Model


The purpose of this report is to document the mercury control performance models that are currently incorporated
into the NETL Mercury Control Cost Model. CMU staff based on available pilot- and full-scale data has developed
these models,  in the form of basic  curve-fitting algorithms.  The algorithms calculate the activated carbon feed
(Lb/MMacf basis)  required to  achieve specified mercury removal efficiency for a particular control method (e.g.,
activated carbon injection upstream of an existing ESP). The performance prediction is based solely on control
method, flue gas temperature and coal type (bituminous vs. subbituminous).

1.       Mercury Control Retrofit Configurations: ESP-1, ESP-4, SD/ESP-1, and SD/ESP-2

            Description: Activated carbon injected upstream of existing ESP. ESP-1 - no flue gas temperature
                        control, ESP-4 - flue gas temperature control via water injection.

1.1     Bituminous Coal

      Coal Source: Eastern bituminous coal, West Virginia (< 1%S and 0.1% chlorine)

      Data Sources:
            PROJECT: ADA Technologies/Public Service Electric and Gas Company/EPRI - Mercury Control
                        in Utility ESPs and Baghouses through Dry Carbon-Based Sorbent Injection Pilot-Scale
                        Demonstration

            Pilot-Scale Tests:  160 acfm  slipstream from  the 620 MWe Hudson Generating Station, Unit 2,
            opposed-fired furnace, Eastern bituminous coal  from West Virginia (< 1%S and 0.1% chlorine), ESP
            SCA = 287 ft2/Kacfm

            Literature Source: Waugh, E., B. Jensen, L. Lapatnick, F. Gibbons, S. Sjostrom, J. Ruhl, R. Slye,
            and R. Chang, "Mercury Control in Utility ESPs and Baghouses through Dry Carbon-Based Sorbent
            Injection Pilot-Scale  Demonstration,"  EPRI-DOE-EPA Combined  Utility  Air  Pollutant  Control
            Symposium, August 1997.
            ICR Data: Baseline  removal due to ash alone is from preliminary  ICR report data compiled by
            Dennis Smith, DOE/NETL (5/1/00)

                       Low sulfur bit coal ash removes constant 57% mercury
                       Total removal is 57% from ash and  0 ==> 43% from ACI
                       Baseline mercury removal (ash alone) is an average of 2 plants with 250F inlet and 4
                       plants with >300F inlet, both approximately 57%
                       Correlation for 225F not included;  no baseline removal available and temperature below
                       H2SO4 dewpoint.
               ICR Results and Pilot-Scale Test Results Combined:
                       Mercury removal due to ash for both temperatures is 57% and is not a function of ACI
                       Mercury removal due to ACI is not a function of ash but is a function of temperature
                       The two removals can be combined: 57% from ash and between 0% and 43% from ACI
                                               D-34

-------
       Algorithm 1: Incorporates ICR Results

               Hg Removal = 100 - [a/(ACI+b)c

               Coefficient
                              Flue Gas Temperature, "F

a
b
c
RA2 (error)
ACI Range
(Lb/MMacf)
225
55.536
1.4351
1
0.99
0-5
250
159.27
3.6838
1
0.85
0-5
275
494.64
11.554
1
0.82
0-5
      Algorithm 2: Excludes ICR Results
               Hg Removal = 100 - [a/(ACI+b)c
               Coefficient
                              Flue Gas Temperature, "F
      Where,

a
b
c
RA2 (error)
ACI Range
(Lb/MMacf)
225
128.69
1.4284
1
0.99
0-5
250
370.98
3.6937
1
0.85
0-5
275
1218.6
12.14
1
0.89
0-5
a,b,c = numerical coefficients
ACI = Activated Carbon Injection Feed rate, Lb/MMacf
Hg Removal = % total mercury removed, inlet to outlet
1.2     Subbituminous Coal

      Coal Source: PRB Subbituminous coal
      Data Source:
               PROJECT:
               Pilot-Scale Carbon Injection for Mercury Control at Comanche Station
               Pilot-Scale Tests: 600 acfm slipstream from the 350 MWe Comanche Station, Unit 2 PSCo,
               opposed-fired furnace, PRB coal from Belle Ayr mine, Pulse-Jet with A/C ratio = 12 ft/min, most
               fly ash removed upstream, Flue gas contained little HC1, 275 to 325 ppm SO2 (@ 3% O2 dry), 180
               to 250 ppm NOX (@ 3% O2 dry)

               Literature Source: AWMA 99-524, S.M. Haythornthwaite,  J. Smith, G. Anderson, T. Hunt, M.
               Fox, R. Chang, T. Brown, 1999
                                              D-35

-------
      Algorithm:

               Hg Removal = 100 - [a/(ACI+b)c

               Coefficient	Flue Gas Temperature, "F

a
b
c
RA2 (error)
ACI Range
(Lb/MMacf)
230
1373.11
32.1071
1
-
0-1
280
247.772
3.3867
1
0.89
0-5
300
296.714
4.2911
1
0.84
0-5
345
319.587
3.6636
1
0.83
0-5
       Where, a,b,c = numerical coefficients
               ACI = Activated Carbon Injection Feed rate, Lb/MMacf
               Hg Removal = % total mercury removed, inlet to outlet

2.       Mercury Control Retrofit Configurations: ESP-3, ESP-6

      Description: Pulse-Jet FF (PJFF) retrofitted after existing ESP.  Activated carbon injected upstream of PJFF.
                 ESP-3  - no flue  gas temperature control, ESP-6 ~ flue  gas temperature control via water
                 injection.

2.1    Bituminous Coal

      Coal Source: Eastern bituminous coal, West Virginia (< 1%S and 0.1% chlorine)

      Data Source:
               PROJECT: ADA Technologies/Public  Service Electric and Gas  Company/EPRI - Mercury
                           Control in Utility ESPs and Baghouses through Dry Carbon-Based Sorbent Injection
                           Pilot-Scale Demonstration

               Pilot-Scale Tests: 4,000 acfm slipstream from the 620 MWe Hudson Generating Station, Unit 2,
               opposed-fired furnace, Eastern bituminous coal from West Virginia  (< 1%S and 0.1% chlorine),
               Pulse-jet FF installed downstream of cold ESP. A/C ratio = 12 ft/min, tests conducted with AC
               and fly ash

               Literature Source: Waugh, E., B. Jensen, L. Lapatnick, F. Gibbons, S. Sjostrom, J. Ruhl, R. Slye,
               and R. Chang,  "Mercury Control  in Utility ESPs  and  Baghouses  through  Dry Carbon-Based
               Sorbent Injection Pilot-Scale Demonstration," EPRI-DOE-EPA Combined Utility  Air Pollutant
               Control Symposium, August 1997.

               ICR Results and Pilot-Scale Test Results Combined:
                       Mercury removal due to ash for both temperatures is 57% and is not a function of ACI
                       Mercury removal due to ACI is not a function of ash but is a  function of temperature
                       The two removals can be combined: 57% from ash and between 0% and 43% from ACI
                                               D-36

-------
      Algorithm 1: Incorporates ICR Results

               Hg Removal = 100 - [a/(ACI+b)c

               Coefficient
                              Flue Gas Temperature, "F

a
b
c
RA2 (error)
ACI Range
(Lb/MMacf)
240
51.038
1.3194
1
0.73
0-5
285
159.4
3.5606
1
0.66
0-5
285
159.4
3.5606
1
0.66
0-5
      Where,
a,b,c = numerical coefficients
ACI = Activated Carbon Injection Feed rate, Lb/MMacf
Hg Removal = % total mercury removed, inlet to outlet
        Algorithm 2: Excludes ICR Results

               Hg Removal = 100 - [a/(ACI+b)c
               Coefficient
                              Flue Gas Temperature, "F

a
b
c
RA2 (error)
ACI Range
(Lb/MMacf)
240
118.69
1.3194
1
0.73
0-5
285
370.69
3.5606
1
0.66
0-5
      Where,
a,b,c = numerical coefficients
ACI = Activated Carbon Injection Feed rate, Lb/MMacf
Hg Removal = % total mercury removed, inlet to outlet
2.2     Subbituminous Coal

      Coal Source: PRB Subbituminous coal

      Data Source:
               PROJECT: ADA Technologies/PS Colorado/EPRI/NETL - Pilot-Scale Demonstration of Dry
                           Carbon-Based Sorbent Injection for Hg Control in Utility ESPs and FFs - Phase I

               Pilot-Scale Tests: 600 acfm slipstream from the  350 MWe Comanche Station, Unit 2  PSCo,
               opposed-fired furnace, PRB coal from Belle Ayr mine, Pulse-Jet with A/C ratio = 12 ft/min, most
               fly ash removed upstream, Flue gas contained little HC1, 275 to 325 ppm SO2 (@ 3% O2 dry), 180
               to 250 ppm NOX (@ 3% O2 dry)

               Literature Source: Ebner, T., J. Ruhl, R. Slye, J.  Smith, T. Hunt, R. Chang, and T. Brown, "
               Demonstration of Dry Carbon-Based Sorbent Injection for Mercury Control in Utility ESPs and
               Baghouses," EPRI-DOE-EPA Combined Utility Air Pollutant Control Symposium, August 1997.
                                               D-37

-------
      Algorithm:
               Hg Removal = 100 - [a/(ACI+b)c
               Coefficient
Flue Gas Temperature, "F

a
b
c
RA2 (error)
ACI Range
(Lb/MMacf)
250
4.2774
0.04793
1
0.99
0-3
280
27.5595
0.31345
1
0.8
0-3
300
148.0419
0.92051
1
0.96
0-3
      Where,   a,b,c = numerical coefficients
               ACI = Activated Carbon Injection Feed rate, Lb/MMacf
               Hg Removal = % total mercury removed, inlet to outlet
3.       Mercury Control Retrofit Configurations: FF-1, FF-2, SD/FF-1, and SD/FF-2

      Description: Activated carbon injected upstream of existing reverse-gas baghouse.  FF-1 - no flue  gas
                  temperature control, FF-2 - flue gas temperature control via water injection.

3.1     Bituminous Coal

No data for bituminous coal applications.

3.2     Subbituminous Coal

      Coal Source: PRB coal from Belle Ayr mine

      Data Source:
               PROJECT: ADA Technologies/PS Colorado/EPRI/NETL  - Pilot-Scale Demonstration of Dry
                           Carbon-Based Sorbent Injection for Hg Control in Utility ESPs and FFs

               Pilot-Scale Tests: 600 acfm slipstream from the 350 MWe Comanche Station, Unit 2 PSCo,
               opposed-fired furnace, PRB coal from Belle Ayr mine, Flue gas contained little HC1, 275 to 325
               ppm SO2 (@ 3% O2 dry), 180 to 250 ppm NOX (@ 3% O2 dry)

               Literature Source: AWMA 99-524, S.M. Haythornthwaite, J. Smith, G. Anderson, T. Hunt, M.
               Fox, R. Chang, T. Brown, 1999
                                               D-38

-------
      Algorithm:
               Hg Removal = 100 - [a/(ACI+b)c
               Coefficient
                                       Flue Gas Temperature, "F

a
b
c
RA2 (error)
ACI Range
(Lb/MMacf)
230
266.119
11.1359
1
0.03
0-1
275
23.20196
0.43006
1
0.88
0-5
330
27.9742
0.31913
1
0.87
0-5
      Where,    a,b,c = numerical coefficients
               ACI = Activated Carbon Injection Feed rate, Lb/MMacf
               Hg Removal = % total mercury removed, inlet to outlet
4.      Mercury Control Retrofit Configurations: WS-1

      Description: Existing ESP for paniculate control and wet FGD for SO2 control.

4.1     Bituminous Coal

      Coal Source: Bituminous coal
      Data Source:
               Mercury Speciation and Wet FGD removal: Memo from D. Smith, DOE/NETL, 4/29/2000

      Methodology:

                     Hg speciation for Bituminous coal: 70% oxidized, 30% elemental
                     Existing ESP assumed to remove Hg at rate predicted by ESP-1, ESP-4 algorithm for
                      bituminous coal (Section 1.1)
                     Wet FGD removes 100% of oxidized Hg and 0% elemental Hg

4.2     Subbituminous Coal

      Coal Source: Subbituminous coal
      Data Source:
               Mercury Speciation and Wet FGD removal: Memo from D. Smith, DOE/NETL, 4/29/2000
      Methodology:
                     Hg speciation for Subbituminous coal: 25% oxidized, 75% elemental
                     Existing ESP assumed to remove Hg at rate predicted by ESP-1, ESP-4 algorithm for
                      Subbituminous coal (Section 1.2)
                     Wet FGD removes 100% of oxidized Hg and 0% elemental Hg
5.1
  Mercury Control Retrofit Configurations: WS-2

Description:  Existing SNCR for NOx control, existing ESP for paniculate control and wet FGD for SO2
            control.

  Bituminous Coal
                                              D-39

-------
      Coal Source: Bituminous coal
      Data Source:
               Mercury Speciation and Wet FGD removal: Memo from D. Smith, DOE/NETL, 4/29/2000
               Impact of SNCR: Not specified

      Methodology:

                     Hg speciation for Bituminous coal: 70% oxidized, 30% elemental
                     Existing ESP assumed to remove Hg at rate predicted by ESP-1, ESP-4 algorithm
                      (Section 1.1)
                     SNCR installation increases oxidized Hg by 0% (e.g., 70% total)
                     Wet FGD removes 100% of oxidized Hg and 0% elemental Hg

5.2     Subbituminous Coal

      Coal Source: Subbituminous coal
      Data Source:
               Mercury Speciation and Wet FGD removal: Memo from D. Smith, DOE/NETL, 4/29/2000
               Impact of SNCR: Not specified
      Methodology:

                     Hg speciation for Subbituminous coal: 25% oxidized, 75% elemental
                     Existing ESP assumed to remove Hg at rate predicted by ESP-1, ESP-4 algorithm for
                      Subbituminous coal (Section 1.2)
                     SNCR installation increases oxidized Hg by 0% (e.g., 25% total)
                     Wet FGD removes 100% of oxidized Hg and 0% elemental Hg

6.      Mercury Control Retrofit Configurations: WS-3

      Description: Existing SCR for NOX control, existing ESP for  paniculate control and wet FGD for SO2
                 control.

6.1     Bituminous Coal

      Coal Source: Bituminous coal
      Data Source:
               Mercury Speciation and Wet FGD removal: Memo from D. Smith, DOE/NETL, 4/29/2000
               Impact of SCR: Mercury control phone meeting 4/28/2000
      Methodology:

                     Hg speciation for Bituminous coal: 70% oxidized, 30% elemental
                     Existing ESP assumed to remove Hg at rate predicted by ESP-1, ESP-4 algorithm
                      (Section 1.1)
                     SCR installation increases oxidized Hg by 35% (e.g., 94.5% total)
                     Wet FGD removes 100% of oxidized Hg and 0% elemental Hg
                                              D-40

-------
6.2    Subbituminous Coal

      Coal Source: Subbituminous coal
      Data Source:
               Mercury Speciation and Wet FGD removal: Memo from D. Smith, DOE/NETL, 4/29/2000
               Impact of SCR: Mercury control phone meeting 4/28/2000
      Methodology:

                      Hg speciation for Subbituminous coal: 25% oxidized, 75% elemental
                      Existing ESP assumed to remove Hg at rate predicted by ESP-1, ESP-4 algorithm for
                       Subbituminous coal (Section 1.2)
                      SCR installation increases oxidized Hg by 35% (e.g., 33.75% total)
                      Wet FGD removes 100% of oxidized Hg and 0% elemental Hg
7.     Mercury Control Retrofit Configurations: ESP-7, Combined AC + Lime Sorbent

      Description: Pulse-Jet FF (PJFF) retrofitted after existing ESP.  Combined activated carbon/Lime sorbent
                  injected upstream of PJFF.

7.1    Bituminous Coal

      Coal Source: Bituminous coal

      Data Source:
               Butz,  J.R., R. Chang, E.G.  Waugh, "Use of Sorbents for Air Toxics  Control in a Pilot-Scale
               COHPAC Baghouse," paper presented  at the Air and  Waste Management Association's 92nd
               annual meeting and exhibition, June 20-24, 1999, St. Louis, Mo.


      Methodology:

                     Assumes AC:Lime ratio = 2:19
                     Assume 90%+ Hg Removal based on ADA Technologies tests at PSE&G
                     1-4 Ib/MMacf Sorbent Concentration yields 90-95% Hg Removal
                                              D-41

-------
                  ATTACHMENT 3

Summary of Mercury Control Cases Analyzed with National
            Energy Technology Laboratory
      Mercury Control Performance and Cost Model
                        D-42

-------
    Summary of Mercury Control Cases Analyzed with NETL's Mercury Control
                              Performance and Cost Model
The purpose of this document is to summarize the mercury control cases evaluated with NETL's Mercury
Control Performance and Cost Model.

1.       Original Cases Designated by EPA

Table 1 identifies  the original matrix of cases that were designated by EPA for evaluation.  Of those
designated in the table, the following model plant types were actually evaluated: 1, 4, 7, 8, 10, 13, 16, and
17. The others were not assessed due to lack of control performance data or similarity to another model
plant type. Table 2 describes the mercury control retrofit scenario configurations used in Table 1.
2.       EPA Evaluation Requirements

For each combination of model plant and pertinent mercury control technology (see Tables 1 and 2), EPA
requested  estimates  of capital  cost ($/kW), fixed O&M  cost (mills/kWh),  and variable O&M  cost
(mills/kWh) using the EPRI TAG methodology. These cost estimates were in 1999 constant dollars. EPA
also designated the following analysis assumptions:

(1)  Mercury removal of 50%, 60%, 70%, 80%, and 90% for each of the model plants;

(2)  Flue gas temperature at activated carbon injection location of 150 C for cases without spray cooling
    (SC) and an approach to saturation of 10 degrees Celsius (18 degrees F) for cases with SC;

(3) plant capacity factor of 65%;

(4) activated carbon cost of $1.0/kg;

(5) water cost of 0.42 mills/gallon;

(6) energy cost 25 mills/kWh;

(7) 30 days of sorbent storage;

(8) labor cost of $25/h; and

(6) other economic assumptions
        (i)      general facilities - 5% of direct process capital (DPC)
        (ii)      engineering and home office expense - 10% of DPC
        (iii)     process contingency - 5% of DPC
        (iv)     project contingency - 15% of DPC + (i) + (ii) +  (iii)
        (v)      pre-production cost - 2% of total plant investment (TPI)
        (vi)     retrofit factor -1.15
        (vii)    fixed O&M - 1.5% of TPI
                                            D-43

-------
                                TABLE 1
        ORIGINAL MERCURY CONTROL CASES DESIGNATED BY EPA
MODEL
PLANT
#

1
2
3
Same as 1

4
5
6
Same as 4

7
8
9


10
11
12

13
14
15
Same as 13

16
17
18
Same as 16
POWER
PLANT
SIZE
(MW)

975
975
975

975
975
975

975
975
975


100
100
100

100
100
100

100
100
100
COAL
Type3
Bit
Bit
Bit

Bit
Bit
Bit

Subbit
Subbit
Subbit


Bit
Bit
Bit

Bit
Bit
Bit

Subbit
Subbit
Subbit
S%
3
o
J
3

0.6
0.6
0.6

0.5
0.5
0.5


3
o
J
3

0.6
0.6
0.6

0.5
0.5
0.5
EXISTING
PLANT
EMISSION
CONTROLS

ESP + FGD
FF + FGD
HESP + FGD

ESP
FF
HESP

ESP
FF
HESP


SD + ESP
SD + FF
HESP + FGD

ESP
FF
HESP

ESP
FF
HESP
MERCURY
CONTROL(S)

ESP-1, ESP-3
FF-1
HESP-1

ESP-4, ESP-6
FF-2
HESP-1

ESP-4, ESP-6
FF-2
HESP-1


SD/ESP-1
SD/FF-1
HESP-1

ESP-4, ESP-6
FF-2
HESP-1

ESP-4, ESP-6
FF-2
HESP-1
CO-BENEFIT
CASE(S)
with

SCR
SCR
SCR

SCR
SCR
SCR

SCR
SCR
SCR













a.
Bit = bituminous coal; Subbit = subbituminous coal.
                                  D-44

-------
                  Table 2
Mercury Control Technology Retrofit Scenarios
CASE
ESP-1
ESP-3
ESP-4
ESP-6
ESP-7
ESP-8
ESP-9

HESP-1

FF-1
FF-2
FF-3

SD/FF-1

SD/ESP-1

WS-1
WS-2
WS-3
SCR-SD-1
EXISTING EQUIPMENT
Cold-side ESP (ESP)







Hot-side ESP (HESP)

Fabric filter (FF)



Spray dryer (SD) + FF

SD + ESP

ESP + wet scrubber (WS)
SNCR + ESP + WS
SCR + ESP + WS
SCR + SD + FF
RETROFIT SCENARIO
ACI
ACI + PFF
SC + ACI
SC + ACI + PFF
SC + AC + lime + PFF
SC + ACI + CFBA
SC + AC + lime + CFBA

SC + ACI + PFF

ACI
SC + ACI
SC + AC + lime + PFF

ACI

ACI





                   D-45

-------
3.      Sensitivity Cases

In addition to the original cases  described above, EPA also requested  five sensitivity cases that  are
described below.

3.1     Power Plant Size

The purpose of this sensitivity analysis was to add 500 MWe cases for Model Plant Applications 1, 4, 7, 8.

This work was originally completed on 6/6/2000.  The sensitivity runs were updated on 6-14-2000 to
correct a programming error.  The results are presented via table and graphs  in Excel file "Mercury
Control Results 6-15-OO.xls."

3.2     Mercury Control Operating Temperature

The purpose of this sensitivity analysis was to change the mercury control operating temperature to:  Acid Dew
Point (ADP) + 40 F for the following cases:

        Plant 4, 500 MW
        Plant 7, 500 MW
        Plant 8, 500 MW
        Plant 13,  100 MW
        Plant 16,  100 MW
        Plant 17,  100 MW

High sulfur cases are not impacted by the change.  975 MW cases were not run.

This work was originally completed on 6/6/2000. The sensitivity runs were updated on 6-14-2000 to
correct a programming error.  The results are presented via table and graphs in Excel file "Mercury
Control Results 6-15-OO.xls."

3.3     COHPAC with Recycle

The  purpose of this sensitivity analysis was to add 20% recycle of AC to the  COHPAC-type mercury
control scenarios (ESP-3 and ESP-6).  This sensitivity applies only to retrofit scenarios ESP-3 and ESP-6.
Mercury control temperature was set at ADP+18 F  (ADP+40 F cases were not run) for the following model
plant applications:

        Plant 1, 500 MW (ESP-3)
        Plant 4, 500 MW (ESP-6)
        Plant 7, 500 MW (ESP-6)
        Plant 13,  100 MW (ESP-6)
        Plant 16,  100 MW (ESP-6)

This work was originally completed on 6/6/2000.  The sensitivity runs were updated on 6-14-2000 to
correct a programming error.  The results are presented via table and graphs  in Excel file "Mercury
Control Results 6-15-OO.xls."
                                             D-46

-------
3.4     Addition of Ductwork to Increase Flue Gas Residence Time

The purpose of this sensitivity analysis was to add the capital cost of additional ductwork to the cost of
mercury control for a specified model plant application. The model plant application that was selected was
Plant 4, ESP-4, 500 MW. The following assumptions were used to complete this effort:

        Application: Plant 4, ESP-4, Ductwork added upstream of ESP
        Results presented with and without added ductwork
        Plant sizes: 975, 500 and 100 MWe
        Type of ductwork: carbon steel, polymer-lined, insulated (reflects a conservative
          selection of material)
        Cost of ductwork: $134/sq ft
        Installation labor: 0.8 hrs/sq ft
        Number of ducts: 2
        Duct gas velocity: 2800 ft/min
        Retrofit factor: 1.3
        Gas residence time in new duct: 1 second

This work was completed on 6/14/2000. The capital costing results indicate the following:

        975 MW application: $2.51/kW for 2 ducts @ 47 feet long (22.3 ft x 22.3 ft)
        500 MW application: $3.50/kW for 2 ducts @ 47 feet long (16 ft x 16 ft)
        100 MW application: $5.54/kW for 2 ducts @ 47 feet long (10 ft x 10 ft)

The complete cost results are presented via table and  graphs in Excel file "Mercury Control Results 6-15-
OO.xls" (Plant 4, W&WO Added Ductwork).
3.5     Use of a Combined AC/Lime Sorbent

The purpose of this sensitivity analysis was to assess the potential economic impact of  using a sorbent
consisting of AC and lime. The assumptions were:

        Application: Model Plant 4, ESP-6, AC sorbent (50-90% Removal); Model Plant 4, ESP-7,
                     AC-Lime Sorbent (90%+ removal)
        Plant size: 500 MWe
        AC sorbent Cost = $908/Ton
        AC+Lime Sorbent Cost = $149/Ton, Assumes ACLime ratio = 2:19
        ESP-7 Sensitivity Cases Assume 90%+ Hg Removal based on ADA Technologies tests at PSE&G
        ESP-7 Sensitivity Cases run for 1, 2, 3, and 4 Ib/MMacf Sorbent Concentration
        ESP-6 Comparison Cases Run for 50, 60, 70, 80, 90% Hg Removal
The complete cost results are presented via table and graphs in Excel file "Mercury Control Results 6-15-
OO.xls" (Plant 4,500 MW, Lime-AC Sorbent).
                                            D-47

-------
   ATTACHMENT 4




Results of all Model Runs
          D-48

-------
               Table 1.  Mercury Control Technology Retrofit Configurations
  Mercury Control       Existing Equipment (a,b)            Retrofit Technology (a)
       ESP-1                      ESP                               ACI
       ESP-3                                                      ACI + PFF
       ESP-4                                                       SC + ACI
       ESP-6                                                   SC + ACI + PFF
       ESP-7                                                 SC + AC + lime + PFF
       ESP-8                                                   SC + ACI + CFBA
       ESP-9                                                SC + AC + lime + CFBA

      HESP-1                     HESP
                             SC + ACI + PFF

       FF-1                        FF                                ACI
       FF-2                                                        SC + ACI
       FF-3                                                  SC + AC + lime + PFF

      SD/FF-1                    SD + FF                             ACI
     SD/ESP-1                  SD + ESP                             ACI
a. ESP = cold-side electrostatic precipitator; HESP = hot-side electrostatic precipitator; FF= fabric filter; SD = spray dryer;
       ACI = activated carbon injection; PFF = polishing fabric filter.
b. Existing equipment may include wet scrubber and NOx controls such as selective catalytic reduction (SCR).
                                            D-49

-------
Table 2. Mercury Control Technology Applications and Cobenefits Definition
Power Coal
Model
Plant #
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Plant
Size,
MWe
975
975
975
975
975
975
975
975
975
100
100
100
100
100
100
100
100
100
Coal Sulfur
Type Content,
%S
Bit
Bit
Bit
Bit
Bit
Bit
Subbit
Subbit
Subbit
Bit
Bit
Bit
Bit
Bit
Bit
Subbit
Subbit
Subbit
3
3
3
0.6
0.6
0.6
0.5
0.5
0.5
3
3
3
0.6
0.6
0.6
0.5
0.5
0.5
Existing
Controls
ESP + FGD
FF + FGD
HESP + FGD
ESP
FF
HESP
ESP
FF
HESP
SD + ESP
SD + FF
HESP + FGD
ESP
FF
HESP
ESP
FF
HESP
Mercury
Controls
ESP-1 , ESP-3
FF-1
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
SD/ESP-1
SD/FF-1
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
ESP-4, ESP-6
FF-2
HESP-1
CoBenefit
Case(s) with
SCR
SCR
SCR






SCR
SCR
SCR






a.      Bit = bituminous coal; Subbit = subbituminous coal.
b.      Mercury controls are shown in Table 1.
                                        D-50

-------
RESULTS FOR MODEL PLANTS 1 AND 4            DATE: 5/22/00
(Accounts for ICR Data Modification)
Comments:
1) Model Plant 1, ESP-1: Minimum Hg removal = 87% for ESP and FGD Combination with Eastern Bituminous Coals
2) Model Plant 1, ESP-1: Minimum Hg removal = 97.6% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals
3) Model Plant 1, ESP-3: Minimum Hg removal = 86.6% for ESP and FGD Combination with Eastern Bituminous Coals
4) Model Plant 1, ESP-3: Minimum Hg removal = 97.6% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals
5) Model Plant 4, ESP-4: Minimum Hg removal = 58% for ESP with Eastern Bituminous Coals
7) Model Plant 4, ESP-6: Minimum Hg removal = 61.3% for ESP with Eastern Bituminous Coals

See plot of results below table
Model
Plant #
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
4
4
4
Plant
Size,
MWe
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
Total
Mercury
Removed,
%
87.16%
90.00%
95.00%
97.64%
86.58%
90.00%
95.00%
97.54%
58.00%
60.00%
70.00%
80.00%
90.00%
61 .30%
70.00%
80.00%
90.00%
95.00%
Coal
Type
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Coal
Sulfur
Content,
% by Wt
3
3
3
3
3
3
3
3
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Existing
Pollutant
Controls
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
Hg Control
Configuration
ESP-1
ESP-1
ESP-1
ESP-1
ESP-3
ESP-3
ESP-3
ESP-3
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
Added
Equipment
forHg
Control
None
SI System
SI System
SI System
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
Co-Benefit
Cases with
N/A
N/A
N/A
SCR
N/A
N/A
N/A
SCR
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
3.29
18.12
0.00
0.00
1.22
6.00
0.00
0.00
0.06
0.79
2.24
6.61
0.00
0.38
1.23
3.78
8.89
Capital
Cost,
$/kW
0.11
2.48
8.54
0.11
43.45
44.59
47.10
43.45
5.88
6.01
6.67
7.54
9.53
46.02
46.50
47.11
48.46
50.59
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.002
0.058
0.200
0.002
1.015
1.041
1.100
1.015
0.137
0.140
0.156
0.176
0.223
1.075
1.086
1.100
1.132
1.182
Fixed
O&M Cost,
Mills/kW-Hr
0.000
0.040
0.048
0.000
0.116
0.118
0.121
0.116
0.047
0.047
0.048
0.050
0.052
0.121
0.122
0.122
0.124
0.127
Variable
O&M Cost,
Mills/kW-Hr
0.000
0.022
0.026
0.000
0.063
0.064
0.065
0.063
0.025
0.025
0.026
0.027
0.028
0.065
0.065
0.066
0.067
0.068
Consuma
bles,
Mills/kW-
Hr
0.003
0.307
1.683
0.003
0.080
0.193
0.637
0.080
0.022
0.027
0.088
0.211
0.580
0.092
0.124
0.196
0.412
0.845
Total
Annual
Cost,
Mills/kW-
Hr
0.006
0.427
1.956
0.006
1.274
1.416
1.924
1.274
0.232
0.240
0.319
0.464
0.883
1.353
1.397
1.485
1.735
2.222

-------
    ECONOMIC RESULTS - GRAPHICAL FORMAT



    MODEL PLANT #1
MERCURY CONTROL COST - MODEL PLANT 1 (Highest Value of Hg
Removal Includes SCR), 975 MWe, Bituminous Coal, 3% Sulfur
Is 2.000 -
i
K 1.500-
8

^
c
c
3
1.956
^r^S^^*X_
1 274 1 -416 ^^-""/"^l 924\\

 	 / \ ^"1.274

/ \
n nnfi ^^^0.427 \
0.006^-^ \ 0 OQ6

-^ESP-1
--ESP-3




86% 88% 90% 92% 94% 96% 98% 100%
Total Hg Removed, %
    MODEL PLANT #4
to
MERCURY CONTROL COST -- MODEL PLANT 4, 975 MWe,
Bituminous Coal, 0.6% Sulfur
Total Annual Cost, Mills/kW-H

1 .500 -


2.222^
	 ^1.735
^ ^ ^5
^^--^0883
*^240 0319 '464
55% 65% 75% 85% 95%
Total Hg Removed, %

-^ESP-6
--ESP-4


-------
RESULTS FOR MODEL PLANTS 1 AND 4 (ADP+40)                              DATE: 6/6/00
(Utilizes Original Performance Algorithms - Excludes ICR Data Modification)
Comments:
1) Model Plant 1, ESP-1: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
2) Model Plant 1, ESP-1: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals
3) Model Plant 1, ESP-3: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
4) Model Plant 1, ESP-3: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals

See plot of results below table
Model
Plant #
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
4
4
4
Plant Size,
MWe
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
Total
Mercury
Removed,
%
70.00%
80.00%
90.00%
94.50%
70.00%
80.00%
90.00%
94.50%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Coal
Sulfur
Content,
% by Wt
3
3
3
3
3
3
3
3
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Existing
Pollutant
Controls
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
Hg Control
Configuration
ESP-1
ESP-1
ESP-1
ESP-1
ESP-3
ESP-3
ESP-3
ESP-3
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
Added
Equipment
forHg
Control
None
SI System
SI System
SI System
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
Co-Benefit
Cases with
N/A
N/A
N/A
SCR
N/A
N/A
N/A
SCR
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
6.14
24.42
0.00
0.00
2.00
7.56
0.00
5.31
7.96
12.37
21.19
47.64
1.98
2.98
4.65
8.00
18.05
Capital
Cost,
$/kW
0.12
4.61
12.54
0.12
49.65
51.64
54.83
49.65
7.95
9.22
11.13
14.55
23.42
51.91
52.53
53.46
55.10
59.25
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.003
0.108
0.293
0.003
1.160
1.206
1.281
1.160
0.186
0.215
0.260
0.340
0.547
1.212
1.227
1.249
1.287
1.384
Fixed
O&M Cost,
Mills/kW-Hr
0.000
0.078
0.089
0.000
0.163
0.166
0.170
0.163
0.084
0.086
0.088
0.093
0.103
0.166
0.167
0.168
0.170
0.176
Variable
O&M Cost,
Mills/kW-Hr
0.000
0.042
0.048
0.000
0.088
0.089
0.091
0.088
0.045
0.046
0.048
0.050
0.056
0.089
0.090
0.090
0.092
0.095
Consuma
bles,
Mills/kW-
Hr
0.003
0.572
2.266
0.003
0.080
0.266
0.782
0.080
0.459
0.683
1.055
1.800
4.035
0.248
0.333
0.474
0.758
1.609
Total
Annual
Cost,
Mills/kW-
Hr
0.006
0.800
2.695
0.006
1.489
1.727
2.324
1.489
0.774
1.030
1.451
2.282
4.741
1.715
1.816
1.982
2.307
3.263

-------
ECONOMIC RESULTS S GRAPHICAL FORMAT




MODEL PLANT#1
MODEL PL

MERCURY CONTROL COST -MODEL PLANT 1 (Highest Value
Removal Includes SCR), 500 MWe, Bituminous Coal, 3% Sulf


to
0 1.500 -
 1 nnn
c

2^ I
1 797 ^/^^\.
1.489__ -^y \\ 1.489
/ \
^XO.800 \
0.006^^--^ V 0.006
of Hg
ur
  ESP-1
--ESP-3

65% 70% 75% 80% 85% 90% 95% 1 00%
Total Hg Removed, %
fl
NT #4

MERCURY CONTROL COST -- MODEL PLANT 4,500 MWe,
Bituminous Coal, 0.6% Sulfur



0 I J'bUU

3 W 2-5
o = 2.000 -
o> s 1 ^nn
5 1.000-


4.741

/ ^ 3.263
2.307X^
1.715 1.816 1_982__ -^"
 	  	 _^-^2.282
	   ""1*451
" 1 030
UY/4 '---

- ESP-6
--ESP-4

40% 50% 60% 70% 80% 90% 100%
Total Mercury Removed, %

-------
RESULTS FOR MODEL PLANTS 1 AND 4 (w Recycle for ESP-3 and ESP-6)                         06/06/2000
(Utilizes Original Performance Algorithms - Excludes ICR Data Modification)
Comments:
1) Model Plant 1, ESP-1: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
2) Model Plant 1, ESP-1: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals
3) Model Plant 1, ESP-3: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
4) Model Plant 1, ESP-3: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals

See plot of results below table
Model
Plant #
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
4
4
4
Plant Size,
MWe
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
Total
Mercury
Removed,
%
70.00%
80.00%
90.00%
94.50%
70.00%
80.00%
90.00%
94.50%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Coal
Sulfur
Content,
% by Wt
3
3
3
3
3
3
3
3
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Existing
Pollutant
Controls
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
Hg Control
Configuration
ESP-1
ESP-1
ESP-1
ESP-1
ESP-3
ESP-3
ESP-3
ESP-3
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
Added
Equipment
forHg
Control
None
SI System
SI System
SI System
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
Co-Benefit
Cases with
N/A
N/A
N/A
SCR
N/A
N/A
N/A
SCR
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
6.14
24.42
0.00
0.00
2.00
7.56
0.00
1.94
2.95
4.64
8.03
18.17
1.07
1.66
2.65
4.63
10.58
Capital
Cost,
$/kW
0.12
4.61
12.54
0.12
49.67
51.39
54.15
49.67
9.23
9.86
10.79
12.44
16.62
54.33
54.71
55.26
56.23
58.64
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.003
0.108
0.293
0.003
1.160
1.200
1.265
1.160
0.216
0.230
0.252
0.291
0.388
1.269
1.278
1.291
1.313
1.370
Fixed
O&M Cost,
Mills/kW-Hr
0.000
0.078
0.089
0.000
0.163
0.165
0.169
0.163
0.087
0.088
0.090
0.092
0.097
0.170
0.171
0.172
0.173
0.177
Variable
O&M Cost,
Mills/kW-Hr
0.000
0.042
0.048
0.000
0.088
0.089
0.091
0.088
0.047
0.047
0.048
0.049
0.052
0.092
0.092
0.093
0.093
0.095
Consuma
bles,
Mills/kW-
Hr
0.003
0.572
2.266
0.003
0.081
0.231
0.648
0.081
0.185
0.271
0.414
0.700
1.557
0.166
0.206
0.274
0.409
0.816
Total
Annual
Cost,
Mills/kW-
Hr
0.006
0.800
2.695
0.006
1.491
1.686
2.173
1.491
0.535
0.637
0.804
1.132
2.095
1.697
1.747
1.829
1.989
2.457

-------
    ECONOMIC RESULTS S GRAPHICAL FORMAT

    MODEL PLANT#1
                     MERCURY CONTROL COST -MODEL PLANT 1 (Highest Value of Hg
                        Removal Includes SCR), 500 MWe, Bituminous Coal, 3% Sulfur
                         65%    70%   75%   80%    85%   90%
                                         Total Hg Removed, %
                                                                 95%   100%
ON  MODEL PLANT #4
MERCURY CONTROL COST - MODEL PLANT 4,500 MWe,

1

i
^j-

^

C


i-
Bituminous Coal,
3.000 -

9 nnn










0.6% Sulfur


2.457
_^>^~
-i 	
* r 1.829
/ 2 095
1.989 /
1.697 1.747 ^/
_______--
 	  	 0 . 804
1.132

0.535 0.637
40% 50% 60% 70%

CQ p R
--ESP-4







80% 90% 100%
Total Mercury Removed, %

-------
RESULTS FOR MODEL PLANTS 1 AND 4                      06/05/2000
(Utilizes Original Performance Algorithms  Excludes ICR Data Modification)
Comments:
1) Model Plant 1, ESP-1: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
2) Model Plant 1, ESP-1: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals
3) Model Plant 1, ESP-3: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
4) Model Plant 1, ESP-3: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals

See plot of results below table
Model
Plant #
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
4
4
4
Plant Size,
MWe
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
Total
Mercury
Removed,
%
70.00%
80.00%
90.00%
94.50%
70.00%
80.00%
90.00%
94.50%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Coal
Sulfur
Content,
% by Wt
3
3
3
3
3
3
3
3
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Existing
Pollutant
Controls
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
Hg Control
Configuration
ESP-1
ESP-1
ESP-1
ESP-1
ESP-3
ESP-3
ESP-3
ESP-3
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
Added
Equipment
forHg
Control
None
SI System
SI System
SI System
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
Co-Benefit
Cases with
N/A
N/A
N/A
SCR
N/A
N/A
N/A
SCR
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
6.14
24.42
0.00
0.00
2.00
7.56
0.00
1.94
2.95
4.64
8.03
18.17
1.07
1.66
2.65
4.63
10.58
Capital
Cost,
$/kW
0.12
4.61
12.54
0.12
49.65
51.64
54.83
49.65
9.23
9.86
10.79
12.44
16.62
54.48
54.91
55.55
56.67
59.46
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.003
0.108
0.293
0.003
1.160
1.206
1.281
1.160
0.216
0.230
0.252
0.291
0.388
1.273
1.283
1.298
1.324
1.389
Fixed
O&M Cost,
Mills/kW-Hr
0.000
0.078
0.089
0.000
0.163
0.166
0.170
0.163
0.087
0.088
0.090
0.092
0.097
0.171
0.171
0.172
0.174
0.178
Variable
O&M Cost,
Mills/kW-Hr
0.000
0.042
0.048
0.000
0.088
0.089
0.091
0.088
0.047
0.047
0.048
0.049
0.052
0.092
0.092
0.093
0.094
0.096
Consuma
bles,
Mills/kW-
Hr
0.003
0.572
2.266
0.003
0.080
0.266
0.782
0.080
0.185
0.271
0.414
0.700
1.557
0.182
0.232
0.316
0.484
0.987
Total
Annual
Cost,
Mills/kW-
Hr
0.006
0.800
2.695
0.006
1.489
1.727
2.324
1.489
0.535
0.637
0.804
1.132
2.095
1.717
1.779
1.879
2.075
2.650

-------
    ECONOMIC RESULTS S GRAPHICAL FORMAT

    MODEL PLANT#1
                     MERCURY CONTROL COST -MODEL PLANT 1 (Highest Value of Hg
                        Removal Includes SCR), 500 MWe, Bituminous Coal, 3% Sulfur
                         65%    70%   75%    80%   85%   90%
                                         Total Hg Removed, %
                                                                 95%   100%
oo
    MODEL PLANT #4
MERCURY CONTROL COST - MODEL PLANT 4,500 MWe,
Bituminous Coal, 0.6% Sulfur
Total Mercury Removed,
Mills/kW-Hr






2.650>
^^
 	 - 	 "* 2 075 / 2'95
.  . 1 879 /
1.717 TTTTS ^^X
 	 ?,^ 0 04
0.535 U'J7

--ESP-6
--ESP-4

40% 50% 60% 70% 80% 90% 1 00%
Total Mercury Removed, %

-------
RESULTS FOR MODEL PLANTS 1 AND 4                      05/22/2000
(Utilizes Original Performance Algorithms  Excludes ICR Data Modification)
Comments:
1) Model Plant 1, ESP-1: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
2) Model Plant 1, ESP-1: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals
3) Model Plant 1, ESP-3: Minimum Hg removal = 70% for ESP and FGD Combination with Eastern Bituminous Coals
4) Model Plant 1, ESP-3: Minimum Hg removal = 94.5% for ESP, FGD, and SCR Combination with Eastern Bituminous Coals

See plot of results below table
Model
Plant #
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
4
4
4
Plant Size,
MWe
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
Total
Mercury
Removed,
%
70.00%
80.00%
90.00%
94.50%
70.00%
80.00%
90.00%
94.50%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Bit
Coal
Sulfur
Content,
% by Wt
3
3
3
3
3
3
3
3
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Existing
Pollutant
Controls
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP, FGD
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
Hg Control
Configuration
ESP-1
ESP-1
ESP-1
ESP-1
ESP-3
ESP-3
ESP-3
ESP-3
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
Added
Equipment
forHg
Control
None
SI System
SI System
SI System
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
PJFF
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
PJFF, SI
System, Wl
System
Co-Benefit
Cases with
N/A
N/A
N/A
SCR
N/A
N/A
N/A
SCR
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
No FG Cooling
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
6.14
24.42
0.00
0.00
2.00
7.56
0.00
1.94
2.95
4.64
8.03
18.17
1.07
1.66
2.65
4.63
10.58
Capital
Cost,
$/kW
0.11
3.85
10.73
0.11
43.43
45.08
47.78
43.43
7.37
7.90
8.70
10.10
13.72
47.00
47.37
47.90
48.84
51.23
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.002
0.090
0.251
0.002
1.014
1.053
1.116
1.014
0.172
0.185
0.203
0.236
0.320
1.098
1.106
1.119
1.141
1.197
Fixed
O&M Cost,
Mills/kW-Hr
0.000
0.042
0.050
0.000
0.116
0.119
0.122
0.116
0.049
0.050
0.051
0.053
0.057
0.122
0.123
0.124
0.125
0.128
Variable
O&M Cost,
Mills/kW-Hr
0.000
0.023
0.027
0.000
0.063
0.064
0.066
0.063
0.027
0.027
0.028
0.029
0.031
0.066
0.066
0.067
0.067
0.069
Consum
ables,
Mills/kW-
Hr
0.003
0.572
2.266
0.003
0.080
0.266
0.782
0.080
0.185
0.271
0.414
0.700
1.557
0.182
0.232
0.316
0.484
0.988
Total
Annual
Cost,
Mills/kW-
Hr
0.006
0.727
2.594
0.006
1.273
1.501
2.086
1.273
0.434
0.533
0.696
1.017
1.966
1.468
1.528
1.625
1.817
2.381

-------
ECONOMIC RESULTS S GRAPHICAL FORMAT




MODEL PLANT#1
MERCURY CONTROL COST -- MODEL PLANT 1 (Highest Value of Hg
Removal Includes SCR), 975 MWe, Bituminous Coal, 3% Sulfur
I


"w

c
"ro
2J94
Z3
1.273 1-5^^ 2.086 \
  	 / \  1 .273
x/ \
jf 0 797 \
^^ ' \


--ESP-1
--ESP-3

65% 70% 75% 80% 85% 90% 95% 100%
Total Hg Removed, %
MODEL PLANT#4


i

1
^
to

"ro
c
<
o
1-

MERCURY CONTROL COST --MODEL PLANT 4,975 MWe
Bituminous Coal, 0.6% Sulfur


2.500 -





0.500 -

2 3g i ^
^^__
. 	  /1. 966

1.468 1.528 1-625 '^^
	  	
	 	  	 "~ I .0 I /
n c-3-3 0.696
0.434 U'bJJ
45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100
Total Hg Removed, %





  ESP-6
--ESP-4








-------
RESULTS FOR MODEL PLANTS 7 AND 8 (w Recycle for ESP-6)
DATE:
6/6/00
Comments:
1) Model Plant 7, ESP-4: Minimum Hg removal = 56.2% for ESP with Western Subbituminous Coals
2) Model Plant 8, FF-2: Minimum Hg removal = 50% for Reverse-Gas FF with Western Subbituminous Coals

See plot of results below table
Model
Plant #
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
Plant Size,
MWe
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
Total Mercury
Removed, %
56.20%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Coal
Sulfur
Content,
% by Wt
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Existing
Pollutant
Controls
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
FF
FF
FF
FF
FF
Hg Control
Configuration
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
FF-2
FF-2
FF-2
FF-2
FF-2
Added
Equipment
forHg
Control
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
Co-Benefit
Cases with
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
2.36
4.87
9.00
21.39
0.04
0.06
0.09
0.17
0.38
0.01
0.04
0.09
0.73
1.89
Capital
Cost,
$/kW
6.59
8.79
10.27
12.34
17.55
55.15
55.19
55.25
55.34
55.57
6.65
6.74
6.84
7.56
8.47
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.154
0.205
0.240
0.288
0.410
1.288
1.289
1.290
1.293
1.298
0.155
0.157
0.160
0.177
0.198
Fixed
O&M Cost,
Mills/kW-Hr
0.083
0.086
0.088
0.091
0.098
0.172
0.172
0.172
0.172
0.173
0.083
0.083
0.083
0.085
0.086
Variable
O&M Cost,
Mills/kW-Hr
0.045
0.046
0.048
0.049
0.053
0.093
0.092
0.093
0.093
0.093
0.045
0.045
0.045
0.046
0.046
Consuma
bles,
Mills/kW-
Hr
0.019
0.233
0.459
0.833
1.953
0.097
0.099
0.101
0.107
0.122
0.020
0.023
0.027
0.085
0.190
Total
Annual
Cost,
Mills/kW-
Hr
0.300
0.571
0.835
1.261
2.513
1.650
1.652
1.657
1.664
1.686
0.303
0.308
0.315
0.392
0.520

-------
    ECONOMIC RESULTS S GRAPHICAL FORMAT
    MODEL PLANT#7
                                MERCURY CONTROL COST-MODEL PLANT 7, 500 MWe,
                                            Subbituminous Coal, 0.5% Sulfur
                  3.00
                g  2.50
                f-  2-
                8  1-50
                |  1.00
                <  0.50
                o  0.00
2.513
1.650 1.652 1.657 1.664 /I 686
_ _ _ - /

0.835 ___^^.261
0.300 ^* 	 ' *
^
                      45%  50%  55%  60%   65%  70%  75%   80%   85%  90%  95% 100%

                                               Total Hg Removed, %
O  MODEL PLANT #8
i
to



I

^

to
o
O
ro
c

^















0


MERCURY CONTROL COST -- MODEL PLANT 8, 500 MWe,
Subbituminous Coal, 0.5% Sulfur











25 -

y 0.520
~7
/
/
^T
s* 0.392
	
0.303 0.308 0.315^^^
_i 	 ^ 	





--FF-2







40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100%
Total Hg Removed, %

-------
RESULTS FOR MODEL PLANTS 7 AND 8
                                                       DATE: 06/06/2000
Comments:
1) Model Plant 7, ESP-4: Minimum Hg removal = 56.2% for ESP with Western Subbituminous Coals
2) Model Plant 8, FF-2: Minimum Hg removal = 50% for Reverse-Gas FF with Western Subbituminous Coals

See plot of results below table
Model
Plant #
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
Plant Size,
MWe
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
Total Mercury
Removed, %
56.20%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Coal
Sulfur
Content,
% by Wt
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Existing
Pollutant
Controls
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
FF
FF
FF
FF
FF
Hg Control
Configuration
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
FF-2
FF-2
FF-2
FF-2
FF-2
Added
Equipment
forHg
Control
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System,
PJFF
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
SI System,
Wl System
Co-Benefit
Cases with
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
2.36
4.87
9.00
21.39
0.04
0.06
0.09
0.17
0.38
0.01
0.04
0.09
0.73
1.89
Capital
Cost,
$/kW
6.59
8.79
10.27
12.34
17.55
55.15
55.20
55.26
55.37
55.64
6.65
6.74
6.84
7.56
8.47
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.154
0.205
0.240
0.288
0.410
1.288
1.289
1.291
1.293
1.300
0.155
0.157
0.160
0.177
0.198
Fixed
O&M Cost,
Mills/kW-Hr
0.083
0.086
0.088
0.091
0.098
0.172
0.172
0.172
0.172
0.173
0.083
0.083
0.083
0.085
0.086
Variable
O&M Cost,
Mills/kW-Hr
0.045
0.046
0.048
0.049
0.053
0.093
0.092
0.093
0.093
0.093
0.045
0.045
0.045
0.046
0.046
Consuma
bles,
Mills/kW-
Hr
0.019
0.233
0.459
0.833
1.953
0.097
0.099
0.102
0.108
0.128
0.020
0.023
0.027
0.085
0.190
Total
Annual
Cost,
Mills/kW-
Hr
0.300
0.571
0.835
1.261
2.513
1.650
1.652
1.658
1.667
1.693
0.303
0.308
0.315
0.392
0.520

-------
ECONOMIC RESULTS S GRAPHICAL FORMAT




MODEL PLANT#7

J-
g
1
5
'H
3
03
c
c
"ro

K



2
i

1







MERCURY CONTROL COST -MODEL PLANT 7, 500 MWe,
Subbituminous Coal, 0.5% Sulfur
50 -
nn

50 -





2.513
1.650 1.652 1.658 1.667 /
/ _ 1.693


0.835 ___^n.261
0.571 _____- 	
0.300 ^ 	 
*^~

- ESP-4
--ESP-6






45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100%
Total Hg Removed, %
MODEL PLANT#8
Total Annual Cost, Mills/kW-H
0
0
0
0
0
0
0
MERCURY CONTROL COST -MODEL PLANT 8, 500 MWe,
Subbituminous Coal, 0.5% Sulfur





25 -
40
J 0.520
/
/
-x^O.392
0.303 0.308 0.315^^-^


  FF-2

% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100%
Total Hg Removed, %

-------
RESULTS FOR MODEL PLANTS 7 AND 8 (ADP+40)
                                                                                 DATE: 6/6/00
Comments:
1) Model Plant 7, ESP-4: Minimum Hg removal = 56.2% for ESP with Western Subbituminous Coals
2) Model Plant 8, FF-2: Minimum Hg removal = 50% for Reverse-Gas FF with Western Subbituminous Coals

See plot of results below table
Model
Plant #
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
Plant Size,
MWe
500
500
500
500
500
500
500
500
500
500
500
500
500
500
500
Total
Mercury
Removed,
%
56.20%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Coal
Sulfur
Content,
% by Wt
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Existing
Pollutant
Controls
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
FF
FF
FF
FF
FF
Hg Control
Configuration
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
FF-2
FF-2
FF-2
FF-2
FF-2
Added
Equipment for
Hg Control
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
Co-Benefit
Cases with
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
FG Cooling,
ADP+40F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
2.36
4.87
9.00
21.39
0.04
0.06
0.09
0.17
0.38
0.01
0.04
0.09
0.73
1.89
Capital
Cost,
$/kW
3.92
4.83
6.15
8.22
13.42
51.17
51.28
51.44
51.70
52.36
2.57
2.75
2.95
3.44
4.35
Levelized
Carrying
Charges,
Mills/kW-
Hr
0.092
0.113
0.144
0.192
0.313
1.195
1.198
1.201
1.208
1.223
0.060
0.064
0.069
0.080
0.102
Fixed
O&M Cost,
Mills/kW-Hr
0.078
0.079
0.081
0.084
0.090
0.165
0.164
0.165
0.166
0.167
0.076
0.076
0.076
0.077
0.078
Variable
O&M Cost,
Mills/kW-Hr
0.042
0.043
0.044
0.045
0.049
0.089
0.089
0.089
0.089
0.090
0.041
0.041
0.041
0.041
0.042
Consuma
bles,
Mills/kW-
Hr
0.125
0.243
0.446
0.820
1.939
0.093
0.100
0.111
0.134
0.204
0.009
0.017
0.030
0.072
0.177
Total
Annual
Cost,
Mills/kW-
Hr
0.336
0.478
0.714
1.140
2.392
1.541
1.550
1.567
1.597
1.683
0.185
0.197
0.216
0.271
0.399

-------
ECONOMIC RESULTS S GRAPHICAL FORMAT




MODEL PLANT#7


I

1


o
O
"ro
c
c
|
H

MERCURY CONTROL COST --MODEL PLANT 7, 500 MWe,
Subbituminous Coal, 0.5% Sulfur

1 JC.


1 75 -

1.25 -
Oyc

y 2.392
/f
1541 1.550 1.567 1 .597^/_^83
          y^


0.478 0.714^^-^1.140
0.336^_ 	 	
- ESP-4
--ESP-6




45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100%
Total Hg Removed, %
MODEL PLANT#8
Total Annual Cost, Mills/kW-H
MERCURY CONTROL COST -- MODEL PLANT 8, 500 MWe,
Subbituminous Coal, 0.5% Sulfur






0.399
/*
/
>^271
0.185 0.197^____2JjJ^-^'


- FF-2

40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100%
Total Hg Removed, %

-------
RESULTS FOR MODEL PLANTS 7 AND 8
                                                       DATE: 05/22/2000
Comments:
1) Model Plant 7, ESP-4: Minimum Hg removal = 56.2% for ESP with Western Subbituminous Coals
2) Model Plant 8, FF-2: Minimum Hg removal = 50% for Reverse-Gas FF with Western Subbituminous Coals

See plot of results below table
Model
Plant #
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
Plant Size,
MWe
975
975
975
975
975
975
975
975
975
975
975
975
975
975
975
Total Mercury
Removed, %
56.20%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Coal Type
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Subbit
Coal
Sulfur
Content,
% by Wt
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Existing
Pollutant
Controls
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
FF
FF
FF
FF
FF
Hg Control
Configuration
ESP-4
ESP-4
ESP-4
ESP-4
ESP-4
ESP-6
ESP-6
ESP-6
ESP-6
ESP-6
FF-2
FF-2
FF-2
FF-2
FF-2
Added
Equipment for
Hg Control
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System, PJFF
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
SI System, Wl
System
Co-Benefit
Cases with
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Comments
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
FG Cooling,
ADP+18F
MERCURY
SORBENT
INJECTION
RATIO
Ib/MMacf
0.00
2.36
4.87
9.00
21.39
0.04
0.06
0.09
0.17
0.38
0.01
0.04
0.09
0.73
1.89
Capital
Cost,
$/kW
5.25
7.07
8.32
10.09
14.61
47.74
47.78
47.83
47.92
48.14
5.