Combustion Source Evaluation: Student Manual

United States Air Pollution Training Institute (APTI) June 2003
Environmental Protection Environmental Research Center, MD 17

Agency Research Triangle Park, NC 27711
Air
Combustion Source
Evaluation
Student Manual
APTI Course 427
Third Edition
Author
Brian W. Doyle, PhD, PE
Doyle Engineering Inc
Developed by
ICES Ltd.
EPA Contract No. 68D99022
Acknowledgments
The author gratefully acknowledges the contributions of Dr. Sumana Keener, Ph.D. PE,
who conducted the peer review of this document.
ii
Notice
This is not an official policy and standards document. The opinions
and selections are those of the author and not necessarily those of the
Environmental Protection Agency. Every attempt has been made to
represent the present state of environmental knowledge as well as subject
areas still under evaluation. Any mention of products or organizations does
not constitute endorsement or recommendation by the United States
Environmental Protection Agency.
This project has been funded wholly or in part by the United States
Environmental Protection Agency under Contract No. 68D99022 to ICES Ltd.
Availability
This document is issued by the Air Pollution Training Institute, Education
and Outreach Group, Office of Air Quality Planning and Standards, USEPA.
This workbook was developed for use in training courses presented by the
U.S. EPA Air Pollution Training Institute and others receiving contractual or
grant support from the Institute. Other organizations are welcome to use
the document.
This publication is available, free of charge, to schools or governmental

air pollution control agencies intending to conduct a training course on the
subject matter. Submit a written request to the Air Pollution Training
Institute, USEPA, MD 17, Research Triangle Park, NC 27711.
Others may obtain copies, for a fee, from the National Technical
Information Service (NTIS), 5825 Port Royal Road, Springfield, VA 22161.
Sets of slides and films designed for use in the training course of which
this publication is a part may be borrowed from the Air Pollution Training
Institute upon written request. The slides may be freely copied. Some films
may be copied; others must be purchased from the commercial distributor.
This consent does not extend to copying or transmission by any means

– graphic or electronic – for any other purpose, such as for advertising or
promotional purposes, for creating a new collective work, for resale, or for
information storage and retrieval systems.
State and local air pollution control agencies, U.S. EPA offices, and
federal offices designated by U.S. EPA are authorized to make copies of
this document in connection with telecourses.
Printed on recycled paper in the United States of America.
iii
Course Description
APTI 427: Combustion Source Evaluation is a four-day, resident instructional
course designed to introduce combustion-related pollution problems such as
estimating the actual and potential air pollution emissions from combustion sources;
reviewing applications for permits to construct combustion facilities; and developing
recommendations to improve the performance of malfunctioning combustion
equipment. The course is intended primarily for those with an engineering and/or
scientific degree, or completion of course RE100 and OL 2000, or six months of
equivalent work experience. Major topics discussed include:
• Combustion sources burning • Combustion principles

liquid and solid wastes
• Combustion sources burning • Design and operational

fossil fuel parameters
• Combustion engines (diesel • Selected fundamental

engines, gas turbines, etc.) calculations
• Pollution control devices • Fuels
How to Use This Workbook

This workbook is to be used during classroom instruction, telecourse
sessions, and individual study. The workbook contains instructional
objectives and materials for each of the seven chapters.
Each chapter provides a lesson goal, instructional objectives, subject

narrative, and reference materials that may guide your study. Each chapter
also contains a reproduction of selected lecture slides intended to guide
your notetaking. The slides are presented to generally follow the course
outline; however, the instructor may on occasion vary the order of presen-
tation or present material not included in the workbook. Each student,
therefore, should take thorough notes of the lecture content throughout the
course, but not rely solely upon graphic reproductions for the course
content.
iv
DISCLAMER
This document is for education and training only and does not constitute
U.S. Environmental Protection Agency policy. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
v
TABLE OF CONTENTS
TABLE OF CONTENTS............................................................................................................... vi
LIST OF FIGURES .....................................................................................................................xiv
LIST OF TABLES........................................................................................................................xvi
ACRONYMS ..............................................................................................................................xvii
USEFUL DATA…………………………………………………………………………………………..xx
FORWARD .................................................................................................................................xxi
Chapter 1
1.0 Overview of Combustion Sources 1-1
1.1 Significance of Combustion Sources................................................................ 1-2

1.2 Carbon and Energy .......................................................................................... 1-3
1.3 Fuels................................................................................................................. 1-4
1.4 Air Pollution ...................................................................................................... 1-6
1.4.1 Emission Rates ............................................................................ 1-7
1.4.2 Types of Pollutants ...................................................................... 1-8
1.5 Overview of Sources ...................................................................................... 1-10
1.5.1 Engines ...................................................................................... 1-10
1.5.2 Boiler Systems ........................................................................... 1-11
1.5.3 Other Combustion Sources........................................................ 1-12
Review Exercises 1-14

Review Answers 1-16
References 1-17
Chapter 2
2.0 General Types and Characteristics of Combustion Sources 2-1
2.1 Combustion Systems .......................................................................................... 2-3

2.2 Types of Combustion Systems............................................................................ 2-4
2.2.1 Engines and Turbines........................................................................... 2-5
2.2.1.1 Reciprocating Engines: Lean/Rich, Diesel/Gas ................. 2-5
2.2.1.2 Combustion Turbines .......................................................... 2-7
vi
2.2.1.3 Combined Cycles and Cogeneration .................................. 2-9
2.2.2 Boilers...... .......................................................................................... 2-10

2.2.2.1 Suspension Versus Grate Firing ....................................... 2-12
2.2.2.2 The Basic Burner Design .................................................. 2-13
2.2.2.3 Suspension Firing – Burner Arrangements ....................... 2-16
2.2.2.4 Stoker Fired – Coal, Wood, and Solid Waste.................... 2-18
2.2.2.5 Fluidized Bed Combustors ................................................ 2-24
2.2.3 Thermal Oxidizers and Other Combustion Systems........................... 2-27
2.2.3.1 High Temperature Oxidizers ............................................. 2-27
2.2.3.2 Catalytic Oxidizers ............................................................ 2-28
2.2.3.3 Flares ................................................................................ 2-29
2.2.3.4 Kilns and Other Devices ................................................... 2-30
2.3 Fuel Storage, Handling and Processing............................................................ 2-33
2.3.1 Natural Gas and Oil ............................................................................ 2-33
2.3.2 Solid Fuels – Storage, Feeding .......................................................... 2-34
2.3.2.1 Coal................................................................................... 2-34
2.3.2.2 Wood and Waste .............................................................. 2-35
2.4 Combustion Air Pollution Controls..................................................................... 2-35
2.4.1 Load Variations – Base, Swing, Peaking, Emergency........................ 2-35
2.4.2 Control Systems ................................................................................. 2-36
2.4.2.1 Fuel and Overall Control ................................................... 2-37
2.4.2.2 Air Flow and Controlling Air-Fuel Ratio ............................. 2-37
2.4.3 Air Moving Components – Fans and Draft Levels .............................. 2-38
2.5 Steam System Components.............................................................................. 2-39
2.5.1 Steam Generation............................................................................... 2-40
2.5.2 Steam Turbines .................................................................................. 2-42
2.6 Ash Handling ..................................................................................................... 2-42
2.6.1 Boiler Surface Deposits ...................................................................... 2-42
2.6.2 Bottom and Fly Ash Management ...................................................... 2-44
2.6.3 LOI and Introduction to Ash Chemistry............................................... 2-44
2.7 Conclusion......................................................................................................... 2-44

Review Answers 2-47
References 2-48
Chapter 3
vii
3.0 Fuel Characteristics 3-1
3.1 Introduction.......................................................................................................... 3-3
3.2 Fuel Characteristics and Usage Trends .............................................................. 3-3
3.2.1 General Types ...................................................................................... 3-4

3.3 Fuel Analyses...................................................................................................... 3-7
3.3.1 Basis of Fuel Analysis........................................................................... 3-7
3.3.2 Market Characteristics ......................................................................... 3-9
3.3.2.1 Heating Value ..................................................................... 3-9
3.3.2.2 Sulfur Content ................................................................... 3-11
3.3.2.3 Proximate Analysis Data ................................................... 3-11
3.3.2.4 Moisture Content............................................................... 3-12
3.3.2.5 Volatile Matter ................................................................... 3-12
3.3.2.6 Fixed Carbon .................................................................... 3-12
3.3.2.7 Ash or Mineral Content ..................................................... 3-12
3.3.2.8 Fuel Oil Market Properties ................................................ 3-13
3.3.3 Ultimate Analysis ............................................................................... 3-13
3.3.4 Ash Composition Data ........................................................................ 3-14
3.3.5 Size Distributions of Solid Fuels ......................................................... 3-16
3.3.5.1 Measuring fuel size ........................................................... 3-16
3.3.5.2 Stoker Fuel Sizes .............................................................. 3-17
3.3.5.3 Pulverized Coal Size......................................................... 3-19
3.4 Fuel Properties .................................................................................................. 3-20
3.4.1 Gas Fuels ........................................................................................... 3-20
3.4.1.1 Natural Gas ....................................................................... 3-20
3.4.1.2 Propane and Liquefied Petroleum Gas (LPG) .................. 3-22
3.4.1.3 Waste Biological Gases .................................................... 3-22
3.4.2 Fuel Oils.............................................................................................. 3-23
3.4.2.1 Classification and Characteristics ..................................... 3-23
3.4.2.2 Distillate Oil ....................................................................... 3-24
3.4.2.3 Residual Oil....................................................................... 3-24
3.4.2.4 Oil Emulsions .................................................................... 3-26
3.4.2.5 Coal-Water Emulsions ...................................................... 3-26
3.4.3 Coal .................................................................................................... 3-27
3.4.3.1 Classifications and Characteristics ................................... 3-27
3.4.3.2 Coal Properties Related to Combustion............................ 3-31
3.4.3.3 Contaminants, Air Pollution, Cleaning .............................. 3-33
viii
3.4.3.4 Coal Cleaning and Sulfur Levels....................................... 3-35
3.5 Other Solid Fuels............................................................................................... 3-36
3.5.1 Fossil Solid Fuels................................................................................ 3-36
3.5.2 Wood .................................................................................................. 3-37
3.5.3 Biomass Fuels .................................................................................... 3-40
3.5.4 Solid Wastes....................................................................................... 3-41
3.6 Conclusion......................................................................................................... 3-44

Review Answers 3-47
References 3-48
Chapter 4
4.0 Fundamentals of Combustion 4-1
4.1 Introduction.......................................................................................................... 4-3

4.1.1 Chemical Balance................................................................................. 4-3
4.1.2 Heat Input Rate – Conservation of Energy ........................................... 4-4
4.1.3 Fuel and Air Flows – Conservation of Mass ......................................... 4-6
4.1.4 Excess Air............................................................................................. 4-6
4.1.5 Water Vapor in Exhaust Gas ................................................................ 4-9
4.2 Combustion Mass Flows and Stoichiometry ..................................................... 4-10
4.2.1 Stoichiometric Air Flow ....................................................................... 4-11
4.2.2 Excess air ........................................................................................... 4-15
4.2.2.1 Excess Air Measurement.................................................. 4-15
4.2.3 Water Vapor........................................................................................ 4-19
4.2.3.1 Water from Fuel Hydrogen ............................................... 4-20
4.2.3.2 Water from Fuel Moisture ................................................. 4-20
4.2.4 Calculating Gas Volume and Density ................................................. 4-21
4.3 Enthalpy of Combustion and Heating Values.................................................... 4-24
4.3.1 Flame Structure – Primary and Dilution Zones................................... 4-25
4.3.2 Combustion Temperature Levels........................................................ 4-27
4.3.2.1 Stoichiometric Flame Temperature................................... 4-27
4.3.2.2 Adiabatic Temperature with Excess Gas .......................... 4-28
4.3.2.3 Local versus Average Temperature.................................. 4-31
4.3.3 Energy Content of Flue Gas ............................................................... 4-32
4.3.3.1 Latent Heat ....................................................................... 4-32
4.3.3.2 Sensible Energy................................................................ 4-34
4.3.3.3 Specific Heat..................................................................... 4-35
ix
4.4 Combustor Size and Residence Time............................................................... 4-37
4.4.1 Combustor Size .................................................................................. 4-37
4.4.2 Residence Time.................................................................................. 4-38
4.5 Conclusion......................................................................................................... 4-38

Review Answers 4-43
References 4-44
Chapter 5
5.0 Air Pollution Formation 5-1

5.1 Introduction.......................................................................................................... 5-2
5.1.1 Actual Emission Rates.......................................................................... 5-3
5.1.2 Potential Emission Rates...................................................................... 5-3
5.2 Acid Gases: Sulfur Oxides and HCl .................................................................... 5-5
5.2.1 Sulfur Oxides ........................................................................................ 5-5
5.2.1.1 SO2 ..................................................................................... 5-5
5.2.1.2 SO3 or Sulfuric Acid ............................................................ 5-6
5.2.1.3 Ash Retention of Sulfur Oxides......................................... 5-10
5.2.2 HCl...................................................................................................... 5-10
5.3 Particulate Matter .............................................................................................. 5-10
5.3.1 Particle Formation – Large Particles................................................... 5-11
5.3.2 Particle Formation – Fine Particles..................................................... 5-12
5.3.2.1 Pulverized Coal Particulate............................................... 5-13
5.3.2.2 Particulate from No. 6 Oil.................................................. 5-13
5.3.2.3 Black Smoke..................................................................... 5-14
5.3.2.4 Sulfuric Acid and Sulfate Particulate................................. 5-15
5.3.3 Enrichment of Some Chemicals in Fine Particles............................... 5-16
5.3.4 Mass Emission Transients from Soot Blowing.................................... 5-17
5.4 Metals................................................................................................................ 5-17
5.4.1 Volatility of Metals and Compounds ................................................... 5-18
5.4.2 Mercury............................................................................................... 5-19
5.5 Nitrogen Oxides................................................................................................. 5-20
5.5.1 Overview of NOx................................................................................. 5-20
5.5.1.1 Control Approaches .......................................................... 5-21
5.5.1.2 Formation Mechanisms .................................................... 5-22
5.5.1.3 Typical Emission Rates .................................................... 5-22
5.5.2 Thermal NOx Formation...................................................................... 5-23
x
5.5.3 NOx Formation from Fuel Nitrogen..................................................... 5-25
5.5.4 Premixed and Diffusion Combustion .................................................. 5-27
5.5.5 NOx from Typical Combustion Systems ............................................. 5-28
5.5.5.1 NOx from Boilers and Furnaces ....................................... 5-29
5.5.5.2 Fuel Effects on Boiler Emissions ...................................... 5-30
5.5.5.3 NOx from Reciprocating Engines ..................................... 5-31
5.5.5.4 Combustion Turbines........................................................ 5-33
5.6 Smoke, Carbon Monoxide and Organic Compounds ........................................ 5-33
5.6.1 Complete Combustion and Fuel-Air Mixing ........................................ 5-34
5.6.2 Burner Geometry ................................................................................ 5-35
5.6.3 Excess Air........................................................................................... 5-35
5.6.4 Incinerator Temperatures ................................................................... 5-36
5.6.5 Dioxin-Furan Formation ...................................................................... 5-37
5.7 Opacity .............................................................................................................. 5-38
5.8 Conclusion......................................................................................................... 5-40

Review Answers 5-45
References 5-46
Chapter 6
6.0 Air Pollution Control Systems 6-1

6.1 Introduction.......................................................................................................... 6-3
6.2 Particulate Matter and Metal Emissions Control ................................................. 6-4
6.2.1 Particle Collectors................................................................................. 6-5
6.2.1.1 Inertial Collectors ................................................................ 6-6
6.2.1.2 Particulate Scrubbers ......................................................... 6-7
6.2.1.3 Fabric filters ...................................................................... 6-10
6.2.1.4 Electrostatic Precipitators ................................................. 6-13
6.2.1.5 Collector Combinations..................................................... 6-16
6.2.2 Dust Collector Fires ............................................................................ 6-17
6.2.3 Oil Fired Particulate ............................................................................ 6-18
6.3 Sulfur Oxides and Hydrogen Chloride Controls ................................................ 6-19
6.3.1 Approaches to SOx control................................................................. 6-19
6.3.1.1 Fuel switching ................................................................... 6-19
6.3.1.2 Flue Gas Desulfurization .................................................. 6-20
6.3.2 SO3 Control......................................................................................... 6-22
6.3.3 HCl Control ......................................................................................... 6-23
xi
6.4 Nitrogen Oxides Control .................................................................................... 6-23
6.4.1 Approaches to NOx control................................................................. 6-23
6.4.2 Combustion Modifications................................................................... 6-23
6.4.2.1 Excess Air Control ............................................................ 6-24
6.4.2.2 Temperature Reduction .................................................... 6-25
6.4.2.3 Off-Stoichiometric Furnace Combustion ........................... 6-29
6.4.2.4 Low NOX Burners.............................................................. 6-32
6.4.3 Premixed vs. Diffusion Flames ........................................................... 6-33
6.4.3.1 Theoretical Differences..................................................... 6-34
6.4.3.2 Combustion Turbine Burner.............................................. 6-35
6.4.3.3 Catalytic Turbine Combustion........................................... 6-37
6.4.3.4 Reciprocating Engines...................................................... 6-39
6.4.4 Add-on or Back End NOx Control Systems ........................................ 6-41
6.4.4.1 Selective Non Catalytic Reduction - SNCR ...................... 6-42
6.4.4.2 Selective Catalytic Reduction - SCR ................................ 6-43
6.4.4.3 NSCR with Rich Burning................................................... 6-45
6.4.5 Combined NOx Controls..................................................................... 6-45
6.5 Carbon Monoxide and Organic Emissions ........................................................ 6-47
6.5.1 Trade-Offs with NOx ........................................................................... 6-47
6.5.2 Catalytic Control Systems................................................................... 6-47
6.5.3 Hydrocarbon Capture ......................................................................... 6-48
6.6 Conclusion......................................................................................................... 6-49

Review Answers 6-53
References 6-54
Chapter 7
7.0 Combustion Source Emissions 7-1

7.1 Introduction.......................................................................................................... 7-2
7.2 Energy Use and CO2 Emissions.......................................................................... 7-2
7.2.1 Efficiency of Various Systems .............................................................. 7-2
7.2.2 Combustion Efficiency .......................................................................... 7-3
7.2.3 Thermal Efficiency ................................................................................ 7-5
7.2.4 Power Plant Efficiency .......................................................................... 7-9
7.2.5 CO2 Emissions.................................................................................... 7-10
7.3 Emissions Monitoring ........................................................................................ 7-11
7.3.1 Emissions Variability........................................................................... 7-11
xii
7.3.2 Measurement methods ....................................................................... 7-13
7.3.2.1 CEM Systems ................................................................... 7-13
7.3.2.2 Federal Reference Method Sampling Trains .................... 7-15
7.3.2.3 Indirect or Parametric Emission Monitoring ...................... 7-16
7.3.3 Calculating Emissions......................................................................... 7-16
7.3.3.1 Correcting for dilution........................................................ 7-17
7.3.3.2 Emissions in lb/mmBTU.................................................... 7-18
7.3.3.3 Emissions in lb/hr.............................................................. 7-18
7.4 Reciprocating and Turbine Engines .................................................................. 7-19
7.4.1 Characteristics of Engines .................................................................. 7-19
7.4.2 Emissions & Control ........................................................................... 7-20
7.5 Natural Gas, No. 2, and No. 6 Oil Fired Boilers ................................................ 7-21
7.5.1 Characteristics of boilers .................................................................... 7-21
7.5.2 Nitrogen oxides control ....................................................................... 7-22
7.5.3 Sulfur Oxides Control.......................................................................... 7-23
7.5.4 Particulate Emissions ......................................................................... 7-24
7.6 Coal Fired Boilers.............................................................................................. 7-24
7.6.1 Characteristics of Pulverized Coal Boilers.......................................... 7-24
7.6.2 Nitrogen Oxides Control ..................................................................... 7-25
7.6.3 Sulfur Oxides and Particulate Matter .................................................. 7-26
7.7 Wood Firing & Stoker Furnaces ........................................................................ 7-26
7.7.1 Characteristics of Stoker Furnaces..................................................... 7-27
7.7.2 Particulate Matter Emissions .............................................................. 7-27
7.7.3 Nitrogen oxides control ....................................................................... 7-27
7.7.4 PICs and Dioxin-furans....................................................................... 7-28

Review Answers 7-31
References 7-32
xiii
LIST OF FIGURES
Figure 1-1 Global Carbon Cycle 1-3

Figure 1-2 Annual U.S. Energy Production and Consumption 1-4
Figure 1-3 Boiler Schematic 1-11
Figure 2-1 Combustion Source System Components 2-3
Figure 2-2 Operating Principal of a Reciprocating Engine 2-6
Figure 2-3 Components of a Combustion Turbine 2-8
Figure 2-4 A Combined Cycle System 2-9
Figure 2-5 A Basic Boiler 2-10
Figure 2-6 Comparative Sizes of Boilers 2-11
Figure 2-7 Suspension and Grate Firing 2-13
Figure 2-8 Basic Burner Design 2-14
Figure 2-9 Typical Large Burner 2-15
Figure 2-10 Large Boiler Configurations 2-17
Figure 2-11 Cyclone Burner 2-18
Figure 2-12 Stoker Firing Concept 2-19
Figure 2-13 Example Configurations of Industrial Stokers 2-21
Figure 2-14 Stoker in a Large Boiler 2-24
Figure 2-15 Fluidized Bed Combustor 2-25
Figure 2-16 Thermal Oxidizer (Gas Incinerator) 2-28
Figure 2-17 Flare Concept 2-29
Figure 2-18 Rotating Cement Kiln 2-30
Figure 2-19 Multiple Hearth Sludge Burner 2-32
Figure 2-20 Elements of Solid Fuel Handling 2-34
Figure 2-21 Example of Load Changes 2-36
Figure 2-22 Example of a Computer-based Control System 2-37
Figure 2-23 Air Flow and Draft Controls 2-39
Figure 2-24 Boiler Steam and Water Components 2-40
Figure 3-1 U.S. Energy Production Trends 3-5
Figure 3-2 ASTM Sieves, Separated and Stacked for Shaking 3-16
Figure 3-3 Example of Coal Size Specification for a Spreader Stoker 3-17
Figure 3-4 ASTM Oil Viscosity-Temperature Relationship 3-25
Figure 3-5 Geographic Distribution of Coal Reserves 3-30
Figure 3-6 Wood Chips 3-38
Figure 4-1 NOx and CO Emissions vs. Excess Air 4-7
Figure 4-2 Excess Air vs. O2 and CO2 4-17
Figure 4-3 Gas Turbine Combustion 4-25
Figure 4-4 Boiler Combustion 4-26
Figure 4-5 Adiabatic Temperature vs. Excess Air 4-29
Figure 5-1 Fuel Sulfur Conversion to Pollutants 5-6
Figure 5-2 Conversion of Fuel Sulfur to SO3 5-8
Figure 5-3 Sulfuric Acid Dew Point 5-9
Figure 5-4 Formation of Coke and Ash Particulate 5-12
Figure 5-5 Coke Cenospheres from No. 6 Oil Combustion 5-14
Figure 5-6 Theoretical NO vs. Temperature 5-24
Figure 5-7 Fuel Nitrogen Contribution to NOx Emissions 5-26
Figure 5-8 Residual Oil Nitrogen Content vs. Time 5-27
Figure 5-9 Typical NOx and CO vs. Excess Air 5-29
Figure 5-10 Emissions vs. Excess Air for Gas Fired Engines 5-32
Figure 5-11 Organic Material Destruction vs. Temperature and Time 5-36
Figure 5-12 Dust Opacity vs. Particle Size 5-39
Figure 6-1 Particle Motion vs. Gas Streamlines 6-5
Figure 6-2 Basic Inertial Collector 6-6
xiv
Figure 6-3 Elements of a Scrubber 6-9
Figure 6-4 Fabric Filter Components 6-10
Figure 6-5 Elements of an Electrostatic Precipitator 6-15
Figure 6-6 Effect of Water Injection on Combustion Turbine Emissions 6-26
Figure 6-7 Flue Gas Recirculation 6-27
Figure 6-8 Effect of Flue Gas Recirculation on Emissions 6-28
Figure 6-9 BOOS and OFA Ports 6-30
Figure 6-10 Reburning on a Utility Boiler 6-31
Figure 6-11 Low NOx Burner Concept 6-32
Figure 6-12 Flame Temperature and Emissions vs. Air-Fuel Ratio 6-34
Figure 6-13 Dry Low NOx Combustion Concept – Gas Firing 6-36
Figure 6-14 Dry Low NOx CT Emissions – Gas Fired 6-37
Figure 6-15 Xonon Catalytic Turbine Combustor 6-38
Figure 6-16 Low Emission Combustion Chamber Concept 6-40
Figure 6-17 SNCR General Configuration 6-43
Figure 6-18 SCR General Configuration 6-44
Figure 7-1 Combustion Energy Flows 7-6
Figure 7-2 Elements of a Continuous Monitoring System 7-14
Figure 7-3 EPA Method 5 Sampling Train 7-16
Figure 7-4 SCR Installation on a Cogen System 7-21
xv
LIST OF TABLES
Table 1-1 Examples of sources and fuels 1-2

Table 1-2 Fuel characteristics 1-5
Table 1-3 Carbon emissions (106 tons) 1-6
Table 1-4 Typical emission rates (lb/mmBTU) 1-9
Table 2-1 Types of reciprocating engines 2-6
Table 2-2 Examples of boiler size 2-12
Table 3-1 Fuels 3-4
Table 3-2 U.S. Energy Consumption, 1993-1997 3-6
Table 3-3 Fuel Analysis – Common Elements 3-8
Table 3-4 Fuel Heating Values - BTU/lb 3-10
Table 3-5 Typical Ultimate Analysis for Coal and No. 6 Oil (% by wt.) 3-14
Table 3-6 Example of Coal Ash Composition (%) 3-15
Table 3-7 Standard Sieve Sizes 3-16
Table 3-8 Coal Size Range 3-18
Table 3-9 Natural Gas Composition and Characteristics 3-21
Table 3-10 ASTM Fuel Oil Specifications 3-23
Table 3-11 Fuel Oil Properties 3-24
Table 3-12 ASTM (D388) Specifications for Coal 3-27
Table 3-13 Properties of Example Coals 3-28
Table 3-14 Common Use Size Distribution Terms for Bituminous Coal 3-33
Table 3-15 Mercury Content in Coal 3-34
Table 3-16 Wood Properties 3-39
Table 3-17 Bagasse Properties 3-41
Table 3-18 Classifications and Properties of Municipal and Medical Wastes 3-43
Table 3-19 Characteristics of Medical Wastes 3-43
Table 4-1 Typical Minimum Excess Air Levels 4-8
Table 4-2 Water Saturation Temperature vs. Concentration 4-9
Table 4-3 Typical Molecular Weights 4-11
Table 4-4 Composition of Air 4-12
Table 4-5 "F-Factors" for Various Fuels (scf/mmBTU) 4-13
Table 4-6 Approximate Exhaust Latent Heat Energy 4-34
Table 4-7 Specific Heat [BTU/(lb-°F) or kcal/(kg-°C)] 4-36
Table 5-1 Emissions Originating in the Fuel 5-3
Table 5-2 Some Elements Enriched in Fine Coal Fly Ash 5-17
Table 5-3 Vaporization of Metal Compounds 5-18
Table 5-4 AP-42 data for Uncontrolled NOx Emissions 5-23
Table 5-5 Typical Fuel Nitrogen Content 5-26
Table 6-1 Reciprocating Engine NOx Emissions - lb/mmBTU 6-41
Table 6-2 Combinations of NOx Control Technologies 6-46
Table 7-1 Combustion Energy Flows 7-2
Table 7-2 CO2 Typical Emission Rates 7-10
Table 7-3 Typical Variation in Emission Rates 7-12
xvi
LIST OF TERMS AND ACRONYMS
TERM DEFINITION
ACFM Actual cubic feet per minute
BACT Best Available Control Technology
BIF Boilers and Industrial Furnaces - refers to federal waste fuel

use regulations
CAAA Clean Air Act Amendments
CEM Continuous Emission Monitors
CFC Chlorofluorocarbon
CT Combustion Turbine
DAS Data Acquisition System
DP Differential Pressure or Delta P
DSCFM Dry standard cubic feet per minute
engine Either a combustion turbine or a reciprocating engine
EPA Environmental Protection Agency
ESP Electrostatic Precipitator
Fc-factor Standard cubic feet of CO2 generated when 106 BTU of energy
is released by combustion. Value varies from 1000 to 2000
depending on the fuel.
Fd-factor Standard cubic feet of dry flue gas with zero excess air generate
when 106 BTU of energy is released by combustion. Value is
between 8700 and 9900 depending
on the fuel.
Fw-factor Similar to Fd except water vapor is included. Value is about

10,500 for most fuels; somewhat higher for high moisture fuel.
FGC Flue gas conditioning – adding something such as SO3 to the

flue gas to improve performance of a control device such as a
dust collector.
FGD Flue gas desulfurization
FGR Flue gas recirculation - technology used to reduce the

formation of NOx emissions by recirculating flue gas into the
combustion.
FID Flame ionization detector
xvii
FRM Federal Reference Method – all the test methods in 40CFR60
Appendix A.
HAP Hazardous Air Pollutants
Heat input rate Usually the same as fuel firing rate – normal units: mmBTU/hr
HHV Higher Heating Value (of a fuel) - a number obtained by a lab

analysis that includes the energy from condensing all moisture
in the flue gas; usually reported as BTU per pound. This
number is always higher than the LHV.
incineration Combustion for purposes of destruction/disposal; usually

implies there is no energy recovery. A politically incorrect term
in some circles.
LEL Lower Explosive Level – the volume concentration of a gas in

air at which it can be ignited (e.g. explode) - in air this typically
requires an energy content of about 50 BTU/ft3 of mixture.
LHV Lower Heating Value (of a fuel) – a number in BTU/lb obtained

by a lab analysis that excludes the energy from condensing
moisture in the flue gas; usually reported as BTU per pound.
This number is always lower than HHV.
LNB Low NOx Burner – federal and local definitions of what this
means are absent or vague.
MACT Maximum Achievable Control Technology
MW Molecular weight
NAAQS National Ambient Air Quality Standards
NDIR Non-dispensive infrared
NESHAP National Emission Standards for Hazardous Air Pollutants
NSCR Non Selective Catalytic Reduction – NOx destruction by

reaction with CO and hydrocarbons – typically used in
automobile exhaust systems.
NSPS New Source Performance Standards
OFA Over-fire air
PC Pulverized coal
PIC Products of Incomplete Combustion – e.g. carbon, CO and

organic compounds
ppm Parts per million. When referring to gas concentration ppm is

on a volume basis. Otherwise it is usually on a weight basis.
xviii
ppmd Parts per million by volume, dry basis – all water removed from
the gas.
ppmw Parts per million by volume, wet basis – all water vapor
remains in the gas during analysis.
psi Pounds per square inch
RM Reference Method (test) – see 40CFR60 Appendix A.
SCA Specific collection area
scf Standard cubic feet – gas volume at 68°F and 29.92 inches
mercury atmospheric pressure.
scfm Standard cubic feet per minute
SCR Selective Catalytic Reduction – NOx destruction (control)

method similar to SNCR, but at much lower temperature by
using a catalyst.
SIP State Implementation Plan
SMD Sauter Mean Diameter – a definition of droplet diameter that

yields a number proportional to the surface area of the drops in
a liquid fuel spray.
SNCR Selective Noncalytic Reduction System
Standard pressure Atmospheric pressure at sea level = 29.92 inches = 760 mm

Hg = 14.7 psia
Standard temperature 68°F = 528°R = 20°C = 293°K
Stoichiometric ratio The air/fuel (or fuel/air) ratio compared to the ratio at which the
amount of oxygen is exactly sufficient to completely burn all the
fuel (1.0 means perfectly matched).
STP Standard Temperature and Pressure
Thermal loading Heat release per unit furnace volume, usually BTU/hr per ft3
TRS Total Reduced Sulfur – gaseous sulfur compounds not

including sulfur oxides.
UEL Upper Explosive Limit
VOC Volatile Organic Compound
W.C. or w.g. Water column or water gage – a measurement (usually in

inches) of gas pressure using a manometer.
WtE Waste to Energy – usually refers to municipal garbage fueled

boilers that generate electric power.
xix
USEFUL DATA
ENERGY DIMENSIONS
1 BTU = Energy to heat 1 lb H2O x 1°F

(106 BTU = 1 mmBTU)
1000 Joules = 0.948 BTU (a Joule is a Newton-Meter)
(Approximately 1000 BTU is required to evaporate 1 lb water into steam)
ENERGY RATE vs POWER
1 BTU/hour = 0.293 WATTS

3410 BTU = 1 Kilowatt-hour
1000 BTU/hr = 0.393 Horsepower
(106 BTU/hr . 30 “Boiler HP”; 100 Boiler HP = 3450 lb/hr Sat. Steam)
1 hp = 0.746 kw
CONVERSION FACTORS
Atmospheric Conditions at sea level (standard conditions)
Pressure 14.7 psia

1013 millibars
29.92 inches Hg
760 mm Hg
Air Density 0.0764 lb/ft3 @ 60F

0.075lb/ft3 @ 68F (or 385 ft3 per lb-mole) (EPA Standard)
Flue Gas Density: usually similar to air - for high moisture levels use a weighted average -
0.046xB + 0.075x(1 - B)
where B is the weight fraction of water.
Absolute Temp (ERankine) = 460 + EFahrenheit

Absolute Temp (EKelvin) = 273 + ECentigrade
°R = 1.8 x °K
Standard Temp = 68°F = 528°R = 293°K
Molecular Weight (MW): Air 28.97

Typ. combustion flue gas 28.5 - 30
AIR EMISSION CONVERSIONS
Multiply by To obtain
grains/ft3 2.29 gm/m3

ppm 41.4 x mol. wt. µg/m3
lb/mmbtu 387(106) X 20.9 - y ppm @ y%O2
Fd x MW 20.9
(where Fd = dry F factor)
lb 7000 grains
lb 453 grams
xx
FORWARD
Course Purpose
This course is directed toward the community of individuals responsible for
regulating sources of air pollution – inspectors, permit writers, enforcement agents
and those who write the rules. The overall intent is to provide an understanding of
how various combustion systems work so that regulation of the resulting air pollution
is viable, but no more burdensome than necessary. APTI 427 is an advanced course
that requires students to have some experience or exposure to air emissions
measurements and at least passing familiarity with stationary combustion sources
such as boilers, power plants or stationary engines. That being said, each section of
this book attempts to provide an introduction for those unfamiliar with stationary
source air pollution. This text will serve as the primary document for students taking
this course, and is intended to serve as a useful reference once the course is complete.
Organization of this text follows the intended course sequence. This required
some compromises between the best course sequence and the organization that would
be optimum for a reference book. For example, the discussion of fuels in Chapter 4
overlaps the information on combustion systems presented in Chapter 2, and it is
unnatural to separate NOx formation in Chapter 5 from NOx control in Chapter 6. In
particular, Chapter 7 is a review, which includes material that may fit well in several
preceding chapters.
References are included at the end of each chapter. References in each list are
intended as sources of additional information rather than to identify the origin of any
material in the body of each chapter.
Relationship to Other APTI Courses

The course content of several other APTI overlaps the material in this course.
These courses include:
1. APTI Course 413, "Control of Particulate Matter Emissions"

2. APTI Course 415, "Control of Gaseous Emissions"
3. APTI Course 418, "Control of Nitrogen Oxide Emissions"
4. APTI Courses 474, "Continuous Emission Monitoring"
5. APTI Course OL2000, "Basic Concepts in Environmental Science"
Most of our largest combustion sources are fired with coal and they all have dust
collectors to control particulate emissions. This course provides an abbreviated
xxi
coverage of dust collectors commonly used on combustion sources in Chapter 6.
Similarly many of our larger combustion sources emit SO2 and they increasingly are
using scrubbers to control emissions. Most stationary sources have negligible VOC
emissions, but combustion devices are one way to control VOC emissions from other
types of sources. This course provides a substantial discussion of thermal oxidizers
and incinerators. Since one of the primary methods of controlling NOx emissions is
by combustion system design, NOx emissions are discussed here in considerable
detail. Monitoring methods are not covered in this course, but there is considerable
discussion of how to interpret and use monitoring data.
xxii
Chapter 1: Overview
LESSON GOAL
Demonstrate a general understanding of the significance and types of combustion
sources and the variety of pollutants they emit through successful completion of the
chapter review exercises.
STUDENT OBJECTIVES
When you have mastered the material in this chapter, you should be able to:
1. Provide examples of some common types of combustion sources.
2. Describe how carbon and energy are related in the global carbon cycle.
3. Provide a general perspective on the fuel/energy sources and their uses in the U.S.
4. Describe generally the major types of combustion sources and their potential air
pollutants.
5. Describe the most common ways or dimensions of reporting air pollutant emissions.
6. List, in general terms, three principal categories of air pollution from combustion
sources.
7. List at least one significant pollutant associated with each of the major types of fuel
or combustion source.
8. Describe briefly the two types of engines used in mobile and stationary sources of
combustion and the principal air pollutant(s).
9. Describe a steam-electric power system and draw a simple schematic of its

operation.
10. Discuss other combustion sources such as cement kilns, solid waste incinerators,
hazardous waste incinerators, and thermal oxidizers.
COMBUSTION SOURCE EVALUATION 1-1

Chapter 1: Overview
1.1 SIGNIFICANCE OF COMBUSTION SOURCES

Combustion sources are everywhere in the industrial world. Combustion is arguably
more vital to our society than fire was to early man. As one can see from the list of
common combustion sources in Table 1-1, combustible fuel sources touch our lives
almost continuously in one form or another. Electric power can be generated from
non-combustion sources, such as wind, solar and hydroelectric energy, but the vast
majority of the U.S. power comes from the combustion of fossil fuels, such as coal,
oil, and natural gas.
Table 1-1. Examples of Sources and Fuels

Source Type Principal Fuels
Steam-electric power plant Bituminous coal, subbituminous coal, residual

oil, natural gas (fossil fuels)
Waste-to-energy steam power plant Wood, municipal solid waste, misc.
Factory boiler Gas, No. 2 oil, No. 6 oil, site specific waste
Combustion turbine power plant Natural gas or distillate oil
Cement kiln Fossil fuels, tires, used oil, hazardous waste
Residential or commercial furnace Natural gas or No. 2 oil
Trucks, cars & railroad engines Gasoline or diesel fuel
Incinerators Natural gas fumes, organic liquid and solids
This chapter provides a general overview of the significance and variety of each of
these fuels used during the combustion process. The major sources of combustion,
such as engines, boiler systems, and others, will be discussed in later sections of this
chapter. However, the emphasis in this brief overview is to discuss the sources of
combustion that are most relevant to air pollution control and emission-oriented
projects.

1.2 CARBON AND ENERGY
All the common fuels are hydrocarbons. Most include a large number of compounds
that may include oxygen and various contaminants. But their energy, and therefore
their fuel value, derives almost entirely from the oxidation of carbon and hydrogen.
Since all fuels are, or were, formed with the help of solar energy, fuels represent
stored solar energy, which we release at our convenience in combustion devices.
Figure 1-1 illustrates the cycle in which carbon dioxide is converted to hydrocarbons
by photosynthesis with solar energy, followed by oxidation back to CO2. Although
the actual pathways are complex, the total energy released during oxidation is the
same as the energy absorbed during photosynthesis. Hydrocarbon fuels are an
efficient and practical way to store a large amount of energy.
Solar Energy
CO2 H2O CNHM

(Atmosphere) (Organic Matter)
Fossil
Heat Energy Fuel
Figure 1-1. Energy and Carbon Cycle
In Figure 1-1, hydrocarbons are divided into two groups according to the timescale in
which they participate in the cycle. The “organic matter” hydrocarbons are all living
matter, wood and other biologic material on the surface of the planet. Most of this
material will oxidize and return to the atmosphere within a few months or years either
by combustion, natural decay or conversion in the bodies of living animals. Fossil
fuels, on the other hand, were created originally from green hydrocarbons. However,
they have been sequestered for hundreds of millions of years, so the cycle time for
fossil carbon is on a scale that is hard to comprehend.
The combustion required to power our society takes carbon, which was sequestered
long ago, and returns it to the atmosphere. Atmospheric carbon dioxide levels have,
over the past century, increased from about 270 ppm to about 360 ppm indicating that
the emissions rate to the atmosphere exceeds the rate that natural systems can
sequester or convert carbon dioxide to green hydrocarbons. Figure 1-2 represents

data compiled by the Department of Energy showing recent annual energy production
and consumption in the U.S. (2000). Note that fossil fuels account for about 83% of
the total energy used in the U.S.
Energy Flow, 2000

(Quadrillion Btu)
Coal Exports
2.57 4.10
Coal Other
22.66 2.57
Natural Gas Coal

19.74 22.41
Fossil Fossil
Fuels Fuels
57.39 83.86
Domestic Natural Gas Commercial
Production 23.33 and Industrial
71.90 51.47
Supply Consumption
102.60 98.50
Petroleum
37.96
Imports Transportation
28.52 Nuclear 8.01 26.64
Renewables 6.82
Adjustments
2.18
Other
4.74
Figure 1-2. Annual U.S. Energy Production and Consumption

U.S. Department of Energy, Energy Information Administration, Annual Energy
Review 2000, available at www.eia.doe.gov/oiaf/aeo/index.html.
Thus far, there has been much less public concern with the part of the combustion cycle
where solar energy is stored in living matter than with the use of that stored energy.
Although we have developed ever more effective ways to use hydrocarbon fuels,
combustion systems and their resulting air pollution continue to be a major source of
concern.
1.3 FUELS
Combustion systems and emissions controls are developed around specific fuels. For
example, an automobile engine and its exhaust catalyst will only function properly
using gasoline with specific characteristics; the engine will not run if the tank is filled
with No.2 oil. Some sources are built to run on several different fuels. Boilers are a
good example; most coal-fired boilers can also fire either oil or gas. In addition, any
oil-fired boiler can be retrofitted to fire gas. The fuel limitations of a combustion
device go beyond the fuel handling hardware for a number of reasons – usually
related to the amount of ash in the fuel or the corrosive character of the resulting flue

gas. Table 1-2 summarizes a number of fuels and the dominant characteristics that
relate to air pollution or combustion system design.
Table 1-2. Fuel Characteristics

Fuel Chief Characteristic
Natural Gas Mostly CH4 (methane) - usually burns clean
LPG Mostly C3H8 (propane) - higher density than CH4
Light Oil (No. 2, Diesel) Few contaminants - a mixture averaging CH2
Heavy Oil (No. 4, No. 6, residual) Black (close kin to tar), not readily pumped or
burned when cold, sulfur and other contaminants
Coal (bituminous, anthracite, lignite) Contaminated with sulfur, ash, rock & water.
Composition - C to CH
Wood Moderate ash, water content is high & variable.

Typical composition - CH2O
Waste Oil Distillate oil with some contaminants
Refinery gas, producer gas, Gas mixtures with inerts plus some contaminants
coke oven gas
Landfill gas, sewage treatment gases Mostly CH4 & CO2 with smelly contaminants
Solid wastes Contaminated with almost anything
The owner of a combustion device will choose the least expensive fuel that is practical
and permitted. Their decision is constrained by existing firing equipment (a gas-fired
boiler cannot fire coal without major changes) and environmental restrictions (usually
related to fuel sulfur content). Typically, the cleanest and simplest fuels, such as natural
gas and light oil, are more expensive than the dirty, troublesome fuels, such as coal and
residual oil. So the decision many stationary source operators must face is whether to
pay rising fuel prices or incur the expense of installing, operating and maintaining
expensive fuel handling and air pollution control equipment.
Several of the listings in Table 1-2 deserve comment.
• Distillate oil differs radically from residual oil. When a refinery processes crude oil,
it begins by removing all the lighter constituents that are used to make gasoline, No. 2
oil, lubricating oil and all the “clean” petroleum products. What is left over is called
“residual” or No. 6 oil. Most of the crude oil contaminants remain in the residual oil;
therefore, No. 6 oil tends to have high concentrations of ash, nitrogen and sulfur – all
of which contribute to air pollution. So distillate fuels are relatively clean burning,
while residual oils generate much higher emission levels.

• Coal has a very high ash content which means that all coal-fired boilers must have
dust collectors to control particulate emissions. Coal also tends to have high
levels of both sulfur and nitrogen, which generate emissions of SO2 and NOx.
• While wood is viewed as a natural, renewable fuel it is not always clean burning.
When wood is wet it does not burn easily or well, which usually results in the
emission of visible smoke, CO and other organic compounds. Ironically, waste
wood such as demolition debris, railroad ties and telephone poles are viewed as
“contaminated,” even though as fuels they tend to have low emissions of CO and
other organics. This is true because they are relatively dry. While organic
preservatives like creosote burn cleanly, the CCA (copper-chrome-arsenic) found
in pressure treated wood can generate hazardous particulate matter.
• Natural gas combustion, although considered clean, generates NOx the same as any
other combustion process.
Another way to look at fuel consumption is in terms of the total amount of atmospheric
carbon produced in the form of CO2. Table 1-3 lists the amount of carbon (in the form
of carbon dioxide) generated by various sectors of our society. Note that electric power
generators are the largest emitters of carbon at 557 (sum of the right column) followed
closely by transportation at 496. These two sectors are also the largest source of
various air pollutants.
Table 1-3. Carbon Emissions (106 tons)

Sector Total Carbon Electric*
Transportation 496 1
Industrial 481 180
Residential 290 193
Commercial 243 183

*Electric is part of the total.
U.S. Department of Energy, Energy Information Administration, Annual Energy Review 2000.
1.4 AIR POLLUTION

Ambient air pollution results directly and indirectly from terrestrial contaminants
emitted into the air. Combustion sources commonly emit visible smoke (carbon) as
well as nitrogen oxides (NOx), sulfur dioxide (SO2), and other invisible gases. Some
of these substances react chemically in the atmosphere to create smog, ozone and the
ingredients of acid rain.

Although stack emissions are diluted by at least 1000 to 1 after leaving the stack, the
sum total of emissions in some areas is sufficient to compromise ambient standards,
either locally or regionally. Atmospheric dilution is not an acceptable approach to air
pollution control locally and regionally it has little or no effect. So our approach to
air pollution problems is to reduce emissions rather that manipulate the stack design.
1.4.1 Emission Rates

The amount of emissions produced by combustion sources is proportional to (1) the
amount of fuel burned and (2) the type of contaminants in the fuel. Emissions also
depend on the combustion process. EPA’s Compilation of Air Pollution Emission
Factors (AP-42) (ref. 6) is a useful reference which allows you to relate emission
rates to fuel use. For example, in order to calculate combustion source emissions,
simply identify the specific type of fuel source in AP-42 and then multiply the
appropriate factor by fuel use to determine the amount of typical emissions. The
reliability of these factors will vary depending on the size of the database, and is
classified by EPA on a scale of A (high) thru E (low). However, regardless of the
reliability of these results, EPA takes no responsibility for any of this data.
Combustion source operators must use it at their own risk. For example, source
operators have been penalized after discovering that their actual emissions were
higher than they had assumed based on AP-42 data.
Emission rates can be expressed in several ways. Listed below are some common
measurement units in English dimensions with a brief explanation.
lb/hour or tons/year This form supports a regional air pollution inventory

and/or a cap on total emissions from one or a group of
sources.
ppm or lb/ft3 corrected Nearly all instruments measure pollutant concentration.

Concentration measurements are always expressed in
corrected form (e.g., ppm corrected to 3% O2) to
eliminate the effects of varying amounts of air dilution.
lb/mmBTU Referencing the emissions to the combustion-firing

rate allows regulation of different size sources with a
single limit. The emission in lb/mmBTU can be
calculated from measured stack values of pollutant
concentration and O2 or CO2 concentrations if the fuel
type is known (F-factor method).
lb/ton of product By referencing emissions to production rates (tons,

kw-hr, etc.) a combustion source is rewarded for
efficient production.

The most common unit used to express combustion source emissions is “lb/mmBTU.”
This is generally viewed as the simplest way to express such emissions; therefore, this
convention will be used throughout the remainder of this text.
1.4.2 Types of Pollutants

Emission formation and control is discussed in Chapters 5 and 6. However, for the
purpose of this discussion, emissions from combustion sources can be broken into
three groups:
1. Products of incomplete combustion or PIC (CO, carbon, organic species)

2. Emissions formed from fuel contaminants (SOx, particulates, etc.)
3. Nitrogen oxides (NOx)
As a general rule, PIC emissions are small to negligible from stationary sources, so in
many cases they are not specifically regulated. One of the reasons PIC emissions are
rarely significant is that, for most sources, they are accompanied by black smoke.
Visible smoke was the first air pollutant ever to be regulated and it is still the most
common form of pollution today. Most sources that are not smoking also do not emit
PIC. However, there are exceptions - primarily those sources that burn very clean
fuel. If a source has a low PIC emission rate, it is by definition burning the fuel
completely. Generally, this means the source has 100% combustion efficiency.
Chapter 7 will discuss in greater detail combustion source efficiency and the
difference between combustion and thermal efficiency.
Fuel contaminants, such as inorganic substances, are any chemical in the fuel other
than carbon, hydrogen, or oxygen. Common examples are sulfur, nitrogen and ash
that pass through the combustion zone to form sulfur oxides (SOx), nitrogen oxides
(NOx) and particulate matter (PM). The amount of air pollution formed may depend
on the combustion system as well as how much of the contaminant is in the fuel. For
example, 100% of fuel sulfur will burn to SO2 and will be emitted unless it is
collected or scrubbed out. On the other hand, the amount of fuel nitrogen converted
to NOx is highly dependent on the design and operation of the combustion system.
Fine particulate matter, particularly PM2.5 (particulate smaller than 2.5µm), can be
formed from fuel sulfur as well as from ash, which shows that the emissions may be
interrelated.
Nitrogen oxides are a distinct category of combustion pollution because NOx is emitted
by all combustion sources regardless of the amount of fuel nitrogen present. Atmospheric
N2 and O2 are normally very stable, but they can react with each other in a combustion
zone. The amount formed is strongly dependent on combustion temperature, so some
sources generate a lot, while others generate very little. Thus a major element of NOx
control efforts has been to modify combustion system design to develop “low NOx”
burners. One of the ironies of combustion NOx control is that any adjustment that lowers
NOx emissions tends to increase PIC emissions; with a few notable exceptions. The
reason for this is further described in Chapter 5.

Emission control before 1970 generally referred to visible smoke that could usually be
limited by combustion system adjustment or maintenance. Emission control efforts
have broadened and intensified since. Emissions formed in the combustion zone
(uncontrolled emissions) depend on fuel composition, but combustion system hardware
and emissions control devices determine how much ultimately goes up the stack
(controlled emissions). Table 1-4 presents the emission rates for a variety of fuels and
sources both with and without emission controls.
Table 1-4. Typical Emission Rates (lb/mmBTU)

Source Pollutant Uncontrolled Controlled
Natural gas furnace NOx 0.1 to 0.5 .02 to .02
No. 6 oil fired boiler NOx 0.45 0.25

Particulates 0.10± 0.05±
SO2 2.50 <1.00
Pulverized coal boiler NOx 0.80 0.40

Particulates 10.00 <0.05
SO2 4.50 <1.50
Wood fired boiler Particulates 2.50 0.1 to 0.6
Diesel Engine NOx 3.2 1.9

Particulate – PM10
CO 0.05 – 1.0
0.85
Combustion Turbine NOx 0.50 0.035
Any combustion source can emit CO, particularly during an upset, but many emit
very little. Most boiler sources emit very little carbon monoxide or hydrocarbons
except possibly during a cold start up. Reciprocating engines emit CO and VOC at
predictable low levels. In many cases any operating event that generates significant
CO will flag the problem with a cloud of dark smoke, which prompts corrective
action. Much of the current effort in emissions control is focused on combustion
tuning to reduce NOx emissions. While this may increase CO, it should not lead to
excess CO emissions.

1.5 OVERVIEW OF COMBUSTION SOURCES
This course deals with stationary sources, specifically those large enough to be
individually regulated. Table 1-3 shows that stationary sources generate roughly one third
of the carbon dioxide emissions to the atmosphere. Regulating large sources of air
pollution has no effect on smaller sources such as automobiles, homes and businesses that
may contribute half or more of the pollution in any given area. This is particularly true in
urban areas, many of which are identified as non-attainment for ozone.
Stationary sources come in all sizes from home furnaces to utility boilers. By way
of comparison, the home furnace flue can be as small as two inches in diameter while
an electric utility can have a stack with a 30-foot diameter flue. Emissions from most
small sources are controlled by certification. For example, all units manufactured to
a certain standard are presumed to meet certain emission criteria. This course is only
concerned with those sources large enough to be individually regulated; meaning that
each source has an operating permit and must certify that its emissions meet the
criteria specified in the permit. Certification can be a one-time requirement for some
smaller sources. An annual source test and/or continuous monitoring are generally
required for larger sources.
1.5.1 Engines
The term “engine” in this course includes two types of power source: (1) reciprocating
engines and (2) combustion turbines (CT), also called gas turbines. Both are internal
combustion (IC) devices in which the exhaust gases directly drive the power production
hardware. The two types share a number of characteristics and are used interchange-
ably in some applications. But they differ considerably in their emission
characteristics.
Stationary engines serve a number of common purposes including: emergency (stand-
by) power generation, pipeline pumping and, increasingly, electric utility generators.
Nearly all stationary engines burn clean fuels such as natural gas, or distillate oil, such
as diesel or No. 2 oil. Therefore, the emission species of primary concern is NOx. PIC
including VOC and particulate emissions are also a potential issue, primarily because of
the potential health impacts. Although reciprocating engines can operate on heavy oil,
which will generate emissions in addition to NOx, most of these are marine diesels
rather land based stationary sources. Combustion turbines cannot operate on dirty fuel
for long without the need for significant maintenance or repair. So most turbines and
reciprocating engines are installed with a connection to natural gas, while No. 2 oil is
used as a backup – if it is used at all.
The basic configuration of an engine is that it has a fuel supply line on one end and a
power output shaft on the other. A single manufacturer ships a package that converts
fuel to mechanical power. This is in contrast to steam power plants that are assembled
on site with components from multiple manufacturers. The fact that turbines and

engine sources are “factory built” means the emission rates are very predictable based
on load and ambient conditions.
There are two types of engines: reciprocating and gas turbine. Stationary reciprocating
engines are simply larger versions of ordinary automobile or truck engines. They can
be as large as several thousand horsepower (several megawatts) and the physical
dimensions approach the limits of rail or highway transport. NOx emissions from
reciprocating engines with no emission controls are among the highest of all sources.
But emissions controls have been evolving for about 30 years and reciprocating engines
can now be installed with relatively low emission rates.
Combustion turbines, (as opposed to steam turbines), are aircraft engines adapted for
ground operation, where high altitude operation is not a concern. Until recently they
were not built smaller than about 500 horsepower (0.4 megawatt), while the upper end
ranges to about 200 megawatts (270,000 HP). Given these power levels, the physical
dimensions are surprisingly small and well within U.S. highway transport limits.
Nearly all of the large electric utility facilities built in the U.S. since the mid 1980s used
a combustion turbine as the primary component. NOx emissions have evolved from
moderately high levels in the late 1970s to extremely low levels today.
1.5.2 Boiler Systems

A boiler is nothing more than an extensively developed tea kettle. Fire heats water
until it boils and the steam is piped off for various uses. Boilers are built as part of a
system, so the steam is piped into a factory process that requires heat, or it is piped to
a turbine that drives a pump or an electric power generator. Boilers are the most
common stationary combustion source. They are present in most factories and are at
the heart of the nation’s electric supply system. Figure 1-3 is a schematic showing a
boiler that feeds steam to a turbine connected to an electric generator. The boiler is
an “external” combustion device because the exhaust gases do not contact the turbine
that produces mechanical power.
Exhaust
Steam
Turbine Generator
Hot Electric
Air Steam Power
Combustion
Fuel
Cold
Steam
Condenser
Water
Figure 1-3. Schematic – Boiler, Turbine and Generator

Boilers are the oldest mechanical devices to use the heat from fire indirectly or to
produce mechanical power. Unlike engines, boilers can and do burn any fuel that is
combustible. Therefore, they emit a wide range of potential pollutants, depending on
the fuel. Boilers come in all sizes from home furnaces to utility boilers where the
firebox can be “big as a barn” and 200 feet high. Medium size industrial boilers with
steam generating capacities less than about 200,000 lb/hr are usually factory built,
prefabricated “package boilers.” Above that size boilers are erected on site resulting
in a custom-made, “one-of-a-kind” boiler, although it may have similar siblings.
Steam conditions in such large boilers can have steam pressures up to 3600 psi at
temperatures of 1000°F and higher, figures far removed from a teakettle. Boiler
systems have a lot of supplemental components such as a water supply system, fuel
supply system, various heat exchangers, and fans to move combustion air or flue gas.
Overall, these are relatively complex systems and each one is unique.
1.5.3 Other Combustion Sources

In addition to engines and boilers there are a number of other types of combustors
that, though less common, can still be significant local sources of air pollution.
Cement Kilns. These are large combustors used to manufacture Portland cement – the
key ingredient of concrete. There are less than 200 active kilns in the U.S. They are
well adapted to burn a wide range of fuels and generally burn the cheapest fuel
available. Because they must operate at very high temperatures in order to make their
product, a number of them are used to incinerate hazardous waste. There are also
lime kilns that are similar to cement kilns, but they are usually smaller and operate at
lower temperatures. Cement and lime kilns process limestone (CaCO3) to generate
lime (CaO), so they emit large amounts of CO2 from the process in addition to the
CO2 from combustion.
Solid Waste Incinerators. Most municipal waste in this country goes to a landfill, but
some of it is burned. In most cases, the combustor is a boiler where waste is the fuel
and the facility is called a municipal waste-to-energy plant. This avoids the stigma
associated with the word incinerator. The term incinerator is generally reserved for
facilities where there is no energy recovery. Waste water treatment plants generate
solid sludge, which, in many cases, is fed to a sludge incinerator. Recently designed
units operate cleanly, but the older ones can be a significant source of air pollution.
Hazardous Waste Incinerators. Combustible waste that is legally defined as
hazardous is required to go to a facility that is licensed to burn it. In addition to
certain cement kilns, there are several facilities in this country specifically built and
permitted to burn certain hazardous wastes. All facilities that burn waste must
demonstrate their ability to safely manage the waste. There are extensive monitoring
requirements designed to ensure compliance. However, the legal classification of a
waste may have no relation to its combustion characteristics. Some hazardous
substances, such as benzene, burn quite easily and cleanly, while other non-hazardous
fuels, such as wet wood chips, can be a source of substantial air pollution and,
sometimes, hazardous substances. During normal operation, all modern incinerators
achieve nearly 100% destruction (complete combustion) of organic waste.

Thermal Oxidizers. This term is commonly used for combustors that consume
organic vapors such as solvents, phenols, etc. emitted by industrial facilities. Thermal
oxidizers are pollution control devices that consume fuel without making any steam or
useful energy. When designed and operated properly, they consume essentially 100%
of the target pollutants and their only significant emissions may be NOx.

Review Exercises
1. Bituminous and subbituminous coal are fossil fuels that can be used in which of the
following combustion sources?
a. Combustion turbines and reciprocating engines
b. Steam-electric power plant
c. Residential furnace
d. Incinerators
2. Fossil fuels account for approximately what percentage of the total energy used in the U.S.?
a. 100%
b. 80%
c. 60%
d. 40%
3. The chief characteristic of natural gas is that it ________?

a. has few contaminants and is composed of a mixture averaging CH2.
b. is contaminated with sulfur, ash, rock and water.
c. is a gas mixture with considerable inert species plus some contaminants.
d. usually burns clean and is composed mostly of CH4 (methane).
4. ________ is an example of a common clean fuel for firing industrial equipment.

a. Waste oil
b. Wood chips
c. Light oil
d. Refinery gas
5. Typically stack emissions are diluted more than ________ after entering the atmosphere.
a. 1:1
b. 10:1
c. 100:1
d. 1000:1
6. Which category of combustion source emits the greatest amount of carbon dioxide?
a. Natural gas furnaces
b. Cement kilns
c. Electric power generators
d. Factory boilers

7. The most common, and arguably most useful, way to express combustion source emissions
is ________.
a. lb/mmBTU
b. lb/ton of product
c. lb/hour or tons/year
d. ppm or lb/ft3 corrected
8. Which of the following is an example of an emission from a combustion source?

a. PIC
b. Emissions formed from fuel contaminants
c. Nitrogen oxides
d. All of the above
9. Which of the following, when designed and operated properly, consume essentially 100% of
the organic pollutants?
a. Solid waste incinerator
b. Hazardous waste incinerator
c. Thermal oxidizer
d. All of the above
10. Which of the following pollutants is usually formed primarily from inorganic fuel
contaminants?
a. Particulates
b. Carbon dioxide
c. PIC
d. Organic species

Review Answers
1. b. Steam-electric power plants
2. b. 80%
3. d. usually burns clean and is composed of CH4 (methane).
4. c. Light oil
5. d. 1000:1
6. b. Cement kilns emit the highest concentration of CO2 (lb/mmBTU)

c. Electric-power generators as a group emit the greatest total tonnage of CO2.
7. a. lb/mmBTU
8. d. All of the above
10. a. Particulates

References
1
Data on global carbon and CO2 is available on numerous web sites including
www.epa.gov www.noaa.gov http://cdiac.esd.ornl.gov/ http://globalchange.gov/.
2
3
Handbook of Energy Systems Engineering, Leslie C. Wilbur, Editor, (John Wiley &
Sons, New York, 1985). This is a hefty handbook that covers many areas including
useful sections on producing and combusting gas, oil and coal.
4
Seinfield, John H. and Pandis, Spyros N. Atmospheric Chemistry and Physics:
From Air Pollution to Climate Change, (John Wiley & Sons, Inc., 1997). This
book could be an academic text as well as a reference. The wealth of detail about
the atmospheric behavior of carbon, sulfur and nitrogen compounds as well as
aerosols (PM2.5) makes it relevant to the study of combustion air emissions.

THIS PAGE INTENTIONALLY LEFT BLANK

Chapter 2: General Types and
Characteristics of Combustion
LESSON GOAL
Demonstrate a general understanding of the various types and characteristics of
combustion systems, their methods of operation and their impact upon air pollution
control through successful completion of the chapter review exercises.
STUDENT OBJECTIVES
When you have mastered the material in this chapter, the student should be able to:
1. Describe how fuel and air flows are controlled for combustion systems.
2. Describe generally how reciprocating engines and combustion (gas) turbines convert
fuel into mechanical energy.
3. Define the terms combined cycle and cogeneration as they relate to the use of waste
heat from engine power sources.
4. Describe the typical size range of engines, turbines and boilers.
5. Describe the difference between suspension and grate firing.
6. Explain the general operating concept of burners used in face-fired boilers.
7. Describe the most common burner configurations used in large boilers.
8. Explain how solid fuels are burned in stoker and fluidized bed combustors.
9. Describe the use of oxidizers and flare combustion to control air emissions.
10. Discuss the fuel storage, handling and processing requirements of natural gas, oil,
coal, wood, and municipal waste.
11. Explain generally how air flow controls work.

12. Detail the overall operation of a steam electric power plant and be familiar with the
primary steam system components.
13. Describe how fuel ash impacts the operation of a combustion facility.

Chapter 2: General Types and
Characteristics of Combustion
2.1 COMBUSTION SYSTEMS

There are several types of combustion systems and a number of variations within
each type. Each type of source has its own air emission characteristics and in some
cases its own set of applicable regulations. Thus anyone working with air pollution
controls from various sources needs to understand the source characteristics and
know where that source fits in the world of air pollution regulations.
Fire is a familiar phenomenon, but the fire in a combustion system differs from an
open campfire or the fire in a fireplace in several respects: (1) the fire is completely
enclosed in a combustion chamber, (2) the flow of both fuel and air into the chamber
is carefully controlled, and (3) the burner is designed to mix the fuel and air in a way
that creates a stable flame and promotes clean combustion. Combustion systems are
made up of a number of components, some of which are illustrated in Figure 2-1.
Diagram Key:
5 1. Burner - Combustion Device
3 2. Fuel Supply
3. Air Supply (Fan)
4. Control System
1 5. Combustion or Emissions Monitor
Combustion (Optional)
Device
Figure 2-1. Combustion Source System Components

Major systems include:
• Burner device(s) and combustion chamber (controls combustion performance)

• Fuel storage and supply system
• Air supply system
• Heat transfer components to make use of the heat released by combustion
• Air pollution control systems
• Controls and data management to allow automatic operation
There may also be a number of auxiliary systems or components such as a water

treatment system. Many of these components can affect combustion performance and
air emission levels either directly or indirectly. For example, worn components in a
steam turbine reduce its efficiency, which means the combustion system will fire
more fuel in order to achieve normal full load. The steam turbine is remote from the
combustion system, but its deterioration increases the fuel flow and that, for a variety
of reasons, can substantially increase air emissions.
Whether or not any part of a facility has an apparent connection to potential air
emissions, a person working with the facility should be aware of the various
components in a combustion facility and the way that each interrelates with other
parts of the system as a whole.
Therefore, this chapter provides a basic description of the most common types of
combustion systems and their related components. The chapter begins with a general
overview of the various types and characteristics of combustion systems, their methods
of operation and their impact upon air pollution control. Specifically, this chapter will
discuss major sources of combustion, such as engines, boiler systems, and others. The
chapter concludes with a review of fuel storage and handling procedures, steam system
components, and ash handling. The purpose of this chapter is to introduce the various
types and characteristics of basic combustion systems and their operation.
2.2 TYPES OF COMBUSTION SYSTEMS

Engines of all sizes are becoming a much more common source of power and, from
an air pollution perspective, they are the simplest group of combustion sources to
manage. They only burn clean fuels like gas and distillate oil, so there is usually no
concern over emissions of SO2 and particulates. They are relatively simple in the
sense that an engine is a package from one manufacturer, unlike the conglomeration
of components from various manufacturers that is typical of most other types of
sources. The fact that a turbine or engine is “factory built,” means the emissions
should be quite predictable from one engine to the next; which is not necessarily true
of sources built on-site.

2.2.1 Engines and Turbines
Internal combustion engines are used extensively in emergency power generation and
pipeline pumping applications. By the end of the 20th century market forces such as the
movement toward deregulated electric power and plentiful supplies of natural gas, have
resulted in engines displacing steam turbines as the prime mover of choice for new
electric power generators. Engine based plants can be erected quickly and the air
emissions from gas fired facilities can be controlled to very low levels. Combustion
turbines can only burn clean fuels – preferably natural gas with No. 2 distillate oil as a
backup. As a result, a growing fraction of the nation’s domestic electric power capacity
cannot utilize solid fuel. But since few of the steam plants have been retired, coal is
still the primary fuel used for electric power generation. While some reciprocating
engines, notably marine diesels, have been built to fire heavy oil, most stationary
engines burn clean fuels. In many cases, stationary engines and turbines are built to fire
natural gas backed up with No. 2 oil fuel capability if the gas supply is interrupted.
Engines come in two varieties: (1) reciprocating engines similar to automobile and
truck engines, and (2) combustion/gas turbines similar to “jet” aircraft engines.
Reciprocating engines can be tiny or as large as about 10 MW (13,000 HP). Gas
turbine sizes range from “microturbines” smaller than 75KW (100HP) to more than
200 MW (270,000 HP. Virtually all new power plants and most new combined cycle
or cogeneration units use a gas turbine as the primary component.
2.2.1.1 Reciprocating Engines: Lean/Rich, Diesel/Gas

Reciprocating engines are the oldest type of internal combustion engine, with some
obvious similarities to their predecessor – the reciprocating steam engine. An
example of the early steam engine is the steam railroad locomotive. Reciprocating
engines operate by igniting a fuel-air mixture in a closed chamber then using the force
of the expanding gas to push a piston that is connected to a crankshaft. Figure 2-2
shows the sequence of events, common to most reciprocating engines.
• The piston moves down drawing a fuel-air mixture into the chamber through
the intake valve.
• The piston then moves up and compresses the fuel-air mixture.
• The mixture is ignited and the piston is driven downward – the power stroke.
• The crank continues to rotate and the exhaust valve opens, discharging the
gaseous products of combustion.
An engine will have a number of cylinders, all lined up along a crankshaft. The
cylinders vary in diameter from a few inches (automobiles) to more than a foot
(industrial engines). Operating speeds depend on size: small engines can run at
several thousand RPM while big stationary engines spin at less than 1000 RPM.

Intake Valve Ignitor or Fuel Injector Exhaust Valve
Figure 2-2. Operating Principle of a Reciprocating Engine
There is also a two-stroke version of the cycle in which the intake and exhaust
functions occur during a short period when the piston is near the bottom of its stroke,
the two-cycle engine. Environmentally, a two-cycle engine is somewhat dirtier than a
four-stroke engine because some of the incoming fuel-air mixture invariably escapes
with the exhaust gases, resulting in unburned fuel (VOC) in the exhaust.
Reciprocating engines can be further divided into several categories as summarized in

Table 2-1. First there is a division according to how the fuel is ignited. When the fuel
is natural gas (or gasoline) an ignitor (spark plug) is required to ignite the fuel. The
spark is timed to occur just before the piston reaches top dead center, to achieve the
most power. If the fuel is diesel oil, it can be injected (sprayed) into the chamber and
the temperature of the compressed gases is sufficient to ignite the fuel. So a diesel
injector replaces the spark plug. Diesel engines also have a higher compression ratio
that gives higher compression temperatures than a gasoline engine. Diesel engines are
somewhat more fuel efficient than spark ignition engines. Table 2-1 lists the major
types of reciprocating engines and their characteristics.
Table 2-1. Types of Reciprocating Engines

Type Characteristics
Natural gas fired Gas mixed with the intake air, spark ignited
Diesel fired Diesel oil auto-ignites and burns as it is injected, no spark required.
Dual fuel Essentially a gas fired diesel engine. A small amount of diesel fuel
is injected to ignite the gas with no spark plug.
Lean burn Operates with more than 5% excess air
Rich burn Operates with less than 5% excess air

The dual fuel engine is basically a diesel engine running on natural gas. Since there
is no spark plug, a very small amount of diesel oil is injected to ignite the gas.
Lean and rich burn refer to the operating fuel-air ratio, a concept discussed further in
Chapter 4. Rich burn means the engine operates with barely enough air for complete
combustion. This gives maximum power, but also results in some degree of
incomplete combustion causing increased PIC emissions. While increased PIC
emissions are normally undesirable, these emissions are a necessary element for one
type of catalytic air pollution control as discussed in Chapter 6. Modern automobile
engines operate in a rich burn mode in order for the air pollution control catalyst to
work. In general, rich burning is associated with more power and increased
emissions while lean burning implies less power and lower emissions. These
implications are discussed further in subsequent chapters.
Combustion in reciprocating engines is intermittent; whereas, it is continuous in all

other combustion sources. Combustion is a transitory process with ignition and
combustion being completed in roughly 10 milliseconds; combustion occurs many
times each second in each cylinder. However, there is a fundamental difference in the
combustion process between diesel oil and natural gas (or gasoline which evaporates
before ignition occurs). Natural gas is mixed with the air before combustion starts.
Diesel fuel is a liquid that mixes with the air and evaporates as it is burned. The
difference between premixed and diffusion combustion is a key factor in the ability to
control NOx emissions because premixed combustion allows engine designs that limit
peak flame temperatures. Chapters 5 and 6 will discuss how NOx control efforts
have been much more successful on gas-fired engines than on diesel fueled engines.
2.2.1.2 Combustion Turbines

Combustion turbines, also called gas turbines, were developed initially as aircraft
engines in the middle of the 20th century. Compared with reciprocating engines, a
combustion turbine can produce 10 to 100 times as much power from an engine of
equivalent physical size. A combustion turbine has only a few (three at most) primary
moving parts. The basic process is shown in Figure 2-3. Air is compressed by an axial
or radial compressor (a turbine in reverse) to a pressure of anywhere from 8 – 20
atmospheres. Fuel is burned in a combustor raising the temperature and expanding the
gas that is then expanded through a turbine. The turbine drives the compressor and
there is power left over to turn a propeller, pump or electric generator. Design
variations include two or three concentric shafts: one shaft may be used to drive the
compressor while a second shaft may be used to turn the power turbine.

Air Fuel Exhaust
Combustor
Load
Shaft Coupling
Compressor
Turbine
Figure 2-3. Components of a Combustion Turbine
Combustion turbines can only burn clean fuel, such as natural gas or distillate oil.
Any contaminants ingested with the air or originating in the combustion zone can
corrode, erode or deposit on the turbine blades, resulting in reduced performance.
Without such damage, combustion turbines can operate thousands of hours without
repair.
Combustion turbine efficiency is directly related to the temperature of the gas leaving
the combustor and entering the turbine, usually called T5. That temperature (and hence
engine thermal efficiency) is limited by the design of the turbine. If it is too high, the
turbine blades are damaged. The current maximum temperature using internally cooled
turbine blades is about 2400°F. Uncooled blades might be limited to 1200°F. Both of
these temperatures are much lower than normal flame temperatures that are typically
above 3000°F. So the combustor has a hot primary zone where the flame is stabilized
and most of the combustion takes place. The hot gas from the primary zone is then
diluted with additional air to achieve an acceptable turbine inlet temperature. Of course
all the air comes through the compressor; the combustor simply divides it, supplying
the right amount in the right place to give a stable primary zone and a uniform
temperature around the turbine inlet nozzle. Since a lot of air is required for dilution,
the total air flow through the engine is three to four times as much as is necessary to
burn the fuel.
Recent developments in gas turbine designs are primarily related to NOx emissions
control that is discussed in Chapters 5 and 6.

2.2.1.3 Combined Cycles and Cogeneration
Engines typically have thermal efficiencies of 30% to 40%; that is, the amount of fuel
energy that is actually converted into shaft power. More than half of the energy is lost.
Some of this loss is inevitable, as dictated by the second law of thermodynamics. But a
major part of the loss is waste heat. Combustion turbines dump heat into the exhaust
stack, while reciprocating engines lose heat to cooling water and up the stack. Steam
turbines waste heat in the condenser downstream of the turbine. Chapter 7 provides a
more detailed discussion of typical system efficiencies. If some of the waste heat can
be utilized, the overall efficiency of the system is higher than that of the engine alone.
Combined cycle is a term used to describe a power system in which the waste heat
from the primary engine is used to drive another power source. Figure 2-4 shows such an
arrangement. Waste energy in the exhaust gas from an engine is used to heat a boiler that
drives a steam turbine. The efficiency is increased from perhaps 35% for the engine
alone to about 50% for the combined cycle system. Most of the large electric power
plants built in the U.S. since the mid-1980s are combustion turbine combined cycle
systems.
Steam
Steam Waste Heat

Load Boiler
Turbine
Fuel for
Optional Duct Burner
Fuel
o o Hot
1400 -2400 F
Exhaust
Gas
Combustor
Air
Compressor Turbine Generator
Figure 2-4. A Combined Cycle System
A closely related concept uses waste energy to heat a building or some manufacturing
process instead of operating a steam turbine. When the waste heat from a turbine or
engine is used directly for heating, the system is called cogeneration. Cogeneration
was fairly common in the first half of the 20th century, but gradually disappeared
with the evolution of huge electric generating plants. However, cogeneration has
seen a resurgence in recent years at college campuses, factory complexes and some
urban centers where there is a significant demand for heat.

2.2.2 Boilers
A boiler consists of a steel container with water on the inside and fire on the outside.
Fire heats water until it boils and the steam is piped off for various uses. Steam rises
to the top and is piped off to drive a steam turbine, heat a building or support a
manufacturing process. Figure 2-5 provides a schematic of a basic boiler. Boilers are
ubiquitous, in that they are present in most factories and are at the heart of the nation’s
electric generating system. Steam heat from a boiler is the most common way to heat a
building.
Exhaust
Air Steam
Fuel
Water
Figure 2-5. A Basic Boiler
Boilers were originally developed in the early 1800s to make steam for the first fuel
powered engines, so they are found in some of the oldest mechanical devices used to
produce mechanical power. Unlike engines, boilers can burn any fuel that is
combustible. The original fuel was cordwood (logs). Coal, when it became available,
was usually a more practical fuel and it displaced wood. In practice, boilers can be fired
on virtually anything that burns; it comes down to building a combustor that will burn
the fuel and sticking it in (or on) a boiler. This means that, unlike engines, boilers can
be the source of a wide range of air pollutants derived from contaminants in the fuel.
The mechanical construction of boilers divides into two categories: (1) fire-tube, and
(2) water-tube. The first boiler was probably a tank of water with fire under it. Then to
increase the heat transfer surface, they put tubes through the tank and ducted the hot
combustion gases through the tubes. This is the fire-tube configuration that was used
for railroad locomotives and is still available for small to medium size commercial
boilers. The physical size of the tank limits the size of fire-tube boilers to a maximum
of about 50,000 lb/hr steam generation (65 mmBTU/hr).
Building the boiler with water inside the tubes and fire on the outside (a water-tube
boiler) removes the size limitation of a single cylindrical pressure vessel. There is
almost no limit on the number of tubes in a water-tube boiler, and thus no limit on the
boiler size. [Some electric utility boiler firing rates exceed 10,000 mmBTU/hr.] The

tubes terminate in outlet and inlet manifolds – a steam drum at the top and a mud drum
at the bottom. In between the tubes are welded to each other - side by side in sheets
that form the walls of the firebox.
There is no apparent limit to the size of a boiler. Boilers come in all sizes from small
home furnaces to utility boilers with a firebox that can be 50' x 60' x 200' high.
Medium size industrial boilers with steam generating capacities less than about
200,000 lb/hr are usually factory-built, “package boilers” that fit on a highway truck
or rail car. Above that size boilers are erected on site with the result that each is
“one-of-a-kind”, although it may have similar siblings. Figure 2-6 is a graphic
illustration of the size range of boilers.
Utility Boiler
Industrial
Boiler
Home
Furnace
Figure 2-6. Comparative Sizes of Boilers
Boiler thermal capacities are referenced in different terms depending on their size and
use. Table 2-2 lists some typical uses and measures of thermal capacity.

Table 2-2. Examples of Boiler Size
Use Generic Size mmBTU/hr
Residential heat 50,000 BTU/hr 0.05
Commercial building heat 100 Horsepower 3.3
Factory - medium size 30,000 lb/hr steam flow 40
Manufacturing - large 200,000 lb/hr steam flow 250
Electric Utility 500 MW (electric) 5,000
2.2.2.1 Suspension Versus Grate Firing

Combustion in boilers or furnaces can be divided into two categories: suspension
fired or grate fired depending on the physical form of the fuel. If the fuel is a gas, a
liquid mist or a powdered solid, it can be burned in suspension, meaning that it takes
less than about one second for combustion to be completed after the fuel is injected
into the combustion zone. When the fuel is solid particles larger than about 1 mm
(1/32 inch) it can only be burned over a period of time longer than several seconds.
In this case the fuel is spread on a metal or refractory grate and air is blown up
through the fuel bed. The process is called grate firing and is the principle behind all
stoker type furnaces.
Suspension firing allows much better control of the combustion process than is possible
with stoker firing. Suspension type burners use jets of fuel and of air that can be
tailored for a specific fuel-air mixing pattern, thus controlling the flame. On a stoker,
with air being blown up through the fuel bed, there is almost no control of fuel-air
mixing, which severely limits control of combustion performance and the resulting
emissions. However, some type of stoker (or its off-spring, the fluidized bed) is the
only way to burn a solid fuel that cannot be pulverized to a fine powder.
Solid fuels contain significant amounts of ash and a stoker tends to keep much of this
ash on the grate. By contrast, suspension firing of pulverized coal puts a large
amount of ash in suspension. Some of it melts partially and the ash accumulates in
various forms on all the boiler surfaces. As a result, a major aspect of operating a
pulverized coal boiler is managing the ash and its impact on heat transfer.
Figure 2-7 shows a furnace with both grate and suspension fires. Within the two
categories of grate and suspension fired combustors there are a number of variations
in both the general configuration and the specific hardware design.

Hot Gas
Solid
Fuel
Bin
Liquid or Gas
Fuel
Burner Air
Grate Air
Figure 2-7. Suspension and Grate Firing
2.2.2.2 The Basic Burner Design

Suspension firing requires a burner that injects the fuel and air into the combustion
zone. The burner design must achieve two critical goals:
(1) A stable flame that does not fluctuate wildly or blow out, and
(2) Complete mixing of fuel and air to minimize incomplete combustion

and smoke.
The basic burner design may have secondary objectives such as limiting flame length
or reducing NOx formation, and it does all these things by carefully tailoring the flow
patterns of fuel and air as they enter the combustion zone.
Most suspension fired burners share a number of basic characteristics. Stable

combustion is the most critical objective of a burner design. If the flame in a furnace
should blow out and then be reignited a few seconds later, the accumulated fuel-air
mixture in the furnace will ignite and the resulting explosion could be catastrophic.
Stability is usually achieved by establishing a flow pattern that causes part of the
flame to recirculate back on itself. Figure 2-8 illustrates a section of a circular burner
that shows the elements common to most burners.

Air enters the burner through an adjustable register (dampers) and then flows through
the throat into the furnace. The air register vanes are angled which causes the air to
rotate around the burner centerline. When the swirling air leaves the burner,
centrifugal force causes it to spread away from the centerline leaving a low pressure
region in the middle. This causes some of the flow in the center of the flame to come
back toward the burner – a recirculation pattern that folds the flame back on itself.
This recirculation assures that the fresh fuel and air are ignited as they enter the
furnace. Thus the flow recirculation generated by air swirling as it leaves the burner
is a key contributor to flame stability.
Combustion Zone
Wind Box
Furnace wall tube
Baffle, diffuser, or swirler

Air Flow
Flow recirculates
to stablize flame
Oil Nozzle or Coal Pipe
Gas Injection Area
Furnace wall tube

Register with
adjustable vanes
swirl the air
Figure 2-8. Basic Burner Design
A second component that contributes to flame stability is the baffle or diffuser

positioned near the throat in the center of the burner. When there is no air swirl the
baffle acts like a bluff body creating a flow eddy with some back flow. When the
flow is swirling it tends to make the whole flow pattern more stable. In some burners
this baffle alone is sufficient to serve as a flame stabilizer. In newer burner designs it
takes the form of an axial swirler and becomes a primary element for flame
stabilization.
Burners similar those in Figures 2-8 and 2-9 range in size from about 10 mmBTU/hr
to more than 300 mmBTU/hr. The same basic design features appear in burners as
small as those in a residential furnace.

Air Register
Door
Coal
Nozzle
Lighter
Impeller Burner Throat
Figure 2-9. Typical Large Burner

Adapted from a drawing in Steam – 40th Edition by Babcock & Wilcox, 1992
In addition to stabilizing the flame, the burner design must assure complete mixing of
the fuel with the air. Incomplete mixing means unburned fuel, which constitutes air
pollution. Adjusting both the air flow direction and the fuel injection jets relative to
the air flow controls mixing. A detailed discussion of this topic is beyond the scope
of this text; however, a very general description of the key adjustments is described in
the following paragraphs.

• Air flow is adjusted primarily by moving the air register that changes the amount
of swirl (angular momentum); this changes the length of the flame and the
strength of the recirculation zone.
• Gaseous fuel is usually injected through a large number of jets (several dozen)
around the periphery of the throat. The number and orientation of these jets
determines the distribution of the gas in the air stream, which is the beginning of
the mixing process. A fairly uniform distribution is necessary for complete
mixing and complete combustion.
• Liquid fuel is sprayed from a single nozzle in the center of the baffle with a
conical spray or a series of steam atomized jets in a conical arrangement. The
potential track of an oil spray is shown in Figure 2-8. The oil spray can penetrate
as much as two or three feet before it loses momentum and the oil drops are
carried by the air flow. The axial position of the oil nozzle can be adjusted,
usually with the burner in operation. Both the conical spray angle and the
number/size of the jets can be adjusted in the design of the atomizing nozzle.
These variables determine how the oil is distributed in the air stream and, thus the
uniformity of the fuel-air mixture in the flame.
• Pulverized (powdered) coal is blown into the flame through a pipe and a typical
burner is shown in Figure 2-9. The end of the pipe can be fitted with a set of
conical spreader vanes or a swirler to spread/direct the coal flow. The velocity of
this coal/air flow is lower than the air velocity at the burner throat. The fuel-air
distribution in the flame is determined by a combination of coal pipe exit
conditions and the recirculation zone flow pattern.
2.2.2.3 Suspension Firing – Burner Arrangements

Furnaces and boilers, smaller than 100 mmBTU/hr, are usually fired with a single
burner similar to that described in Figure 2-8 and depicted in Figure 2-9. Larger
industrial and utility boilers use multiple burner arrangements. The most common
arrangement, called wall-fired or face-fired, has rows of burners on one of the four
walls of the firebox. A large utility boiler could have four or five rows of burners,
each with as many as eight burners. A second, somewhat less common arrangement
is opposed wall-fired in which matching rows of burners are mounted on opposite
walls of the firebox. The top two sketches in Figure 2-10 are elevation (side) views
of one row of wall-fired and opposed wall-fired configurations. About half of the
large utility boilers in the U.S. are either wall-fired or opposed wall-fired.
Another burner arrangement common in utility boilers, called tangential firing, uses
burners mounted in the corners of the firebox rather than in the walls. These burners
differ from face fired burners in that they are rectangular rather than round and the
only swirl is generated by a swirler in place of a baffle at the center of the burner.
However, there is swirl for the furnace as a whole as shown in the lower right sketch

of Figure 2-10 showing a plan view (looking from the top) of the furnace. Note that
the burners are aimed not at the center of the furnace, but rather their aim is tangent to
a theoretical circle in the center of the furnace, hence the term tangential firing. This
arrangement does not create any recirculation patterns, but note that each flame
impinges on its neighbor. The flame in the center of the furnace is stable because the
burners light one another even though the individual burners might not be ignited.
Tangentially-fired furnaces were built exclusively by the Combustion Engineering
Company (CE was subsequently purchased by ABB) and account for roughly 40% of
the large utility boilers in the U.S.
Wall-fired, Horizontally Opposed,

Elevation View Wall-fired, Elevation View
Turbo-fired,
Elevation View
Arch-fired, Tangential Firing,

Elevation View Plan View
Figure 2-10. Large Boiler Configurations

Nitrogen Oxide Control for Stationary Combustion Sources, USEPA ORD, July 1986, EPA/625/5-86/020
The two remaining sketches in Figure 2-10 are less common burner arrangements
used in utility boilers. Each sketch shows one (or a pair of) burner in a line of burners
across the furnace. These burners make little or no use of swirl and instead use the
flame’s buoyancy to recirculate the flow for flame stabilization. The population of
these furnaces is small and there is little prospect that any more will be built.
Another burner with a limited population is the cyclone burner, which is designed to
burn crushed, not pulverized, coal. Figure 2-11 is a cut-away view; the internal
diameter of the burner is 6-8 feet. These burners, built primarily by B&W, eliminated

the need to pulverize coal in order to burn it in a large boiler. Coal is crushed to less
than 3/4" (like gravel), fed through the end of the barrel, and burned on the inside
surface of the burner barrel, which is built of water tubes. Air is blown in through a
tangential slot along the top of the barrel at hurricane velocity, about 300 ft/sec. This is
high enough to distribute the coal around the circumference and burn it very intensely.
Hot combustion gases and fine particulate depart through the burner throat into the
main boiler firebox. Most of the coal ash melts to form a liquid slag that drains through
a notch in the furnace wall. While this design overcomes the cost and complexity of a
pulverized coal system, it suffers from a critical limitation: NOx emissions are high and
difficult to control.
Boiler Fire Box
Tertiary Air
Crushed Coal Hot

Combustion
Gases
Primary Air
300 fps
Liquid Slag
Figure 2-11. Cyclone Burner Concept
2.2.2.4 Stoker Fired – Coal, Wood, and Solid Waste

The stoker is the most primitive of current combustor designs, but it is capable of
burning almost any solid fuel. Unless they are powdered, solid fuels can only be
burned in a stoker or a fluidized bed combustor. Fuels fired in stokers cover the
range from crushed coal to municipal solid waste. Stoker furnaces are used primarily
in coal or wood-fired industrial boilers and to burn municipal solid waste. They
range in size up to about 1000 mmBTU/hr. The physical dimensions of the grate set
this upper limit; the maximum firing rate is about 1.0 mmBTU/ft2 and dimensions
larger than about 30 ft. are unmanageable. Stokers were used in utility boilers for the
first half of the 20th century, but as the need developed for larger boilers, pulverized
coal systems took over the market, where combustion uses the volume of the furnace
not just the floor area. About 25 years after the transition from stoker to pulverized

coal furnaces there began another transition, from coal to oil, then to gas. So the
primary use of stoker type furnaces during the last half-century has been for wood
chips and waste fuels – mostly in the forest products industry and for municipal solid
waste. Coal stokers are still in use, but many date to the first half of the last century.
Burning fuel on a stoker is little different from a campfire or the fire in a fireplace.
The evolution of the stoker from the common hearth or campfire has been automation
and adaptation for the larger physical size of a stoker furnace. Fuel is introduced with
some type of mechanical feeder – the actual stoker. Ash is removed either by hand in
smaller furnaces or by some variety of automated mechanical grate in larger systems.
Airflow is no longer natural draft, but is supplied by a forced draft fan. Air flow is
divided, some supplied under the grate and some supplied to the fire above the grate
as overfire air. Figure 2-12 shows the key features of a stoker. The traveling grate
shown here is one of many grate varieties.
Hot Exhaust Gas

Solid Fuel Bin
Overfired Air Ports
Gas phase combustion

above fuel bed
Warm Air Suply Traveling Grate Stoker
Ash Hopper
Figure 2-12. Stoker Firing Concept
Fuel Burning on a Grate
The length of time it takes for fuel to burn on a stoker grate depends on how large the
pieces are because solid fuel burns primarily on the surface. A large piece of fuel burns
faster if it is broken up into smaller pieces. So there has to be a correlation between the
size of fuel particles fed to the furnace, the expected firing rate, and the transport rate
on an automatic grate. For example, one can put pieces of railroad ties into the furnace,

but they will probably take two or three hours to burn completely. So the speed of the
grate must be adjusted accordingly and the furnace will probably need a supplemental
fuel of wood chips or crushed coal in order to sustain its normal load. Typically, coal
or wood is crushed or chipped into pieces smaller than 3/4".
The other aspect of stoker fuel particle size is the behavior of small particles. A
stoker can be operated with fairly small fuel particles, but the result is that the fuel
burns up quickly and the firing rate can actually be somewhat unstable – burning
furiously then dying down. So one normally does not want a lot of fine particles.
Another more serious problem with particles less than about 1/8" (3 mm) is that the
air velocity is sufficient to blow them off the grate and out of the furnace. Clearly
there is a need to limit the minimum size and the amount of small particles. Chapter
3 provides a description of standard methods for stoker fuel sizing.
Most stokers use some kind of automatic fuel feeder to spread fuel onto the grate.
This device must have an adjustable feed rate to control the average furnace-firing
rate and may have other adjustable features. One essential objective of the feeder is
to generate a uniform distribution of fuel on the grate. If the fuel layer on the grate is
a series of mounds and valleys, it affects the distribution of air coming up through the
grate. The air tends to flow up through thin spots in the fuel layer and is blocked
from flowing through the thicker mounds. Good combustion dictates more air where
there is more fuel – not less. Thus fuel mounds burn slowly and the air that should be
used to burn them carries unused oxygen out of the furnace.
Solid fuels take considerable time to burn, depending on the moisture content and on
the size of the pieces of fuel. This means that a stoker boiler cannot respond rapidly to
load changes. The boiler can take 10 or 15 minutes to burn down fuel that is already
in the furnace in response to decreased load. There is a similar restriction on building
up a supply of fuel in the furnace to respond to load increases. The operator can use
air flow controls to increase or decrease the load rapidly. However, dropping the air
flow too much will cause the boiler to smoke and raising it will rapidly decrease the
amount of fuel in the furnace, which limits the use of air flow for load control.
Fuels like green wood chips and municipal waste can contain 50% to 60% water
which requires hot air to dry the fuel before it will burn. So these stokers require hot
air under the grate (400°F – 500°F). On most of these systems the fuel is moved
progressively from one end of the grate to the other. As the fuel moves, it dries and
then begins to burn so that eventually only the ash remains. The traveling grate
shown in Figure 2-12 is one way to achieve this. A stoker boiler can also burn wet
fuel in a thick stationary bed. Once the fire is started, fresh fuel is continuously
spread on top where it dries, burns down and leaves only ash at the bottom against the
grate. Eventually, the fire is allowed to burn down in order to remove the ashes.

Fuel Feeders
A great many feeder devices have been designed in response to various fuels. They
all have various features intended to improve their reliability and support the objective
of a uniform fuel distribution across the grate. These details are beyond the scope of
this text, but a basic description will be provided of two broad groups of feeders:
(1) spreaders, and (2) bottom feeders.
Spreader stokers throw fuel onto the grate as indicated in sketch (c) of Figure 2-13. The
spreader is located two to four feet above the grate. Coal spreaders use a small paddle
wheel device spinning fast enough to throw larger coal particles across the furnace.
There may be several of these across the width of the furnace. Wood waste spreaders
commonly use a pneumatic puffing device in place of the paddle wheel.
Fuel Fuel
Residue
Air Air
(a) Underfeed
Fuel
Residue
Air Ash Pit
(b) Crossfeed Spreader
Fuel
Air Ash Pit
(b) Overfeed Spreader
Figure 2-13. Example Configurations of Industrial Stokers

Nitrogen Oxide Control for Stationary Combustion Sources, USEPA ORD, July 1986, EPA/625/5-86/020

Bottom feeders, such as those shown in sketches (a) and (b) of Figure 2-13, are slow
moving. The traveling grate shown in Figure 2-12 and sketch (b) of Figure 2-13 can take
up to an hour to traverse the length of the furnace, corresponding to the time it takes for
the fuel to burn completely. The height of the bed in sketch (b) of Figure 2-13 or speed
of the ram in sketch (a) of Figure 2-13 is used to control the furnace’s average firing rate.
Air Flow
Early stokers had no fans. They had tall stacks that generated enough draft to draw air
into the combustion zone and through the boiler. A single damper near the stack
controlled air flow. Essentially all furnaces now use forced draft (FD) fans to supply
air to the furnace. Larger furnaces also use an induced draft (ID) fan to maintain a
negative pressure in the furnace and move the exhaust gas. Some smaller boilers may
still use stack draft to move the flue gases downstream of the combustion zone.
Overfired Air
Nearly all stokers distribute some of the air above the grate as overfire air. As solid
fuel is heated and burned it releases various hydrocarbons – volatile matter which can
burn in the gas phase above the grate. Volatile matter can account for more than 50%
of the heat release, which implies that a major portion of the combustion air needs to
be supplied as overfired air to burn some fuels. This poses a dilemma for older stoker
systems. Unless the grate is water cooled, it needs about 2/3 of the total air flow to
keep it cool and prevent long term thermal damage, which leaves insufficient
overfired air for typical fuels. As a consequence, stoker furnaces generally use much
more air than is theoretically necessary for combustion. Excess air levels are usually
at least 35%, but are frequently more than 50% in order to prevent smoking.
Overfired air is necessary to burn all the volatile gases that are released from the fuel
bed as it burns. If the amount and/or distribution of overfired air is inadequate the
furnace will smoke and emit CO. Overfired air does not have to be preheated, but it
does have to be an adequate amount to mix well with the gases coming off the fuel
bed. Usually about 50% or more of the combustion air needs to be supplied by a high
pressure fan and the ports into the furnace need to be the right size and spacing.
Grate Concepts
The grate supports the burning fuel and has openings that supply air to the bed. Air
will be evenly distributed around the grate if there is some pressure drop across the
grate, typically 0.5" w.g. or more. The lower the pressure drop, the more the air will
tend to follow the path of least resistance. Air will not go where the fuel is most
deeply piled on the grate. The operating problems associated with inadequate pressure
drop begin to appear when the grate holes become enlarged or the seals around the
edge of the grate become worn. Thus deterioration of the grate system will have
significant impact on operation before there is obvious need for repair or maintenance.

There are many types of grate design that serve a variety of purposes with different
fuels. General descriptions of a few types follow.
• Stationary Grates. Stationary grates are the oldest as well as some of the newest
designs. They are common in wood chip and wood waste fired facilities. A
stationary grate can be water cooled, which allows most of the combustion air to
be overfired. The problem with a stationary grate is that the ash must be removed
by some other mechanism. Small to medium grates can be raked by hand. Larger
stationary grates require a mechanical rake.
• Traveling Grate. Traveling grates, such as the one shown in Figure 2-12 provide
automatic transport of the fuel bed and removal of the ash, but they require a lot
of air for cooling, which limits the amount of overfired air and the ability to
control smoking. Traveling grates are common in coal-fired systems.
• Oscillating Grates. There are a wide variety of oscillating grate designs where
portions of the grate move periodically and gradually push the fuel downhill
toward an ash hopper. This concept is favored for municipal waste combustors
because the grate action tends to stir the bed and help to expose and burn wet or
non-uniform fuels. Some of these designs also tend to be self-cleaning to remove
molten glass or other sticky ashes.
Stoker Summary
Most of the major components of a stoker-fired boiler are shown in Figure 2-14.
Starting from the left side there is a fuel supply system that controls the rate of fuel
feed and spreads it onto the grate. The air supply is split into under-grate and over-
fire flows. Hot combustion gases pass out the top of the boiler past convectively
heated tubes that superheat the steam. Although the walls of the furnace are normally
water tubes, this boiler also has a convective steam generating section as shown by
the tubes connecting the mud drum to the steam drum.
The exhaust gases from a stoker tend to carry a lot of course particulate, and virtually
all stoker boilers employ a pre-treatment system such as the multiple cyclones shown
here. Some of the particulate will still be burning, particularly if the fuel is waste
wood or bio-fuel, and a cyclone is not damaged by flaming embers. The air heater
provides the hot air necessary for firing wet fuel as well as saving energy by cooling
the exhaust gas. Gas flow through the system is provided by a pair of fans, forced
draft (FD) to supply air and induced draft (ID) to maintain a negative draft in the
combustion chamber.

Steam
drum
Mud
drum
Refractory
Cinder
reinjection
line
RDF Boiler tubes

feed chute Overfire
air headers
Nozzles
RDF
distribution
air Ash and fuel bed Traveling Grate
Hatch
Undergrate
air plenums
Ash pit
Figure 2-14. Stoker in a Large Boiler
2.2.2.5 Fluidized Bed Combustors

The fluidized bed combustor is the most recent type of solid fuel combustor. In fact,
this combustor has come into common use only since about 1975. Fluidized bed
combustors can fire a wide variety of solid, liquid or gaseous fuels and they overcome
most of the problems associated with fuel and air misdistribution across a stationary

grate. Figure 2-15 shows a version of a fluidized bed. The basic device appears
similar to a stoker except that a bed of ordinary sand, about a foot deep, replaces the
grate. When air is blown up through the bed nothing happens until the air exceeds a
critical flow rate, at which point the bed begins to lift. Air flow causes the bed to
expand to a depth of more than three feet, and the air bubbles through it like steam in a
pot of boiling water. The temperature of the sand is raised to about 1500°F by adding
gas to the air and igniting it. Once the bed is up to its normal operating temperature,
small pieces of virtually any solid fuel can be added and they will burn.
Waste Feed Exhaust Gases
Entrained Material
Feed Spray Dispersion

Cyclone Separator
Reaction Vessel
Dilute Phase
Fluidized Bed
Dense Phase
Fluidized Bed Dust Return
Orifice Plate
Fluidizing Air Supply
Solid Product
Figure 2-15. Fluidized Bed Combustor
The stirring action of the bubbling air causes the fuel to mix throughout the bed and
constantly come in contact with fresh air as it rises through the bed. Fuel can be added
at any location and it fairly rapidly disburses through the bed, so the problem of trying
to evenly distribute the fuel disappears. Fuel particles either float randomly on the

surface or mix down into the bed. In either case they burn rapidly. Combustion is
very stable because the bed temperature cannot be changed rapidly, so unlike a flame,
it cannot be blown out. In fact, the bed can be shut down by shutting off the air. An
inactive fluid bed takes several hours to cool and during that time it can continue
operation simply by restarting the air and fuel flow.
A fluidized bed requires a system to continuously make up sand that is lost. Some of
the sand as well as ash from the fuel are continuously carried out of the bed by the
exhaust gas. A cyclone or inertial collector is used to capture the coarser particles
that are returned to the bed. Finer dust particles pass through the cyclone and must be
captured with a dust collector.
A fluidized bed is ideally suited to firing solid waste fuels or very low-grade fuels
such as those with high moisture contents or low heating values. This type of
combustor can be set up to fire fuels in almost any combination. A side benefit of
operating at a very low temperature is that it produces fairly low levels of NOx. In
some cases, adding lime or limestone to the bed has reduced SOx emissions.
Given the advantages of the fluidized bed combustor, why are there are so few of
them? The primary answer is that as long as traditional fuels like natural gas or oil
are cheap, there is little economic incentive to use other fuels. Fluidized bed
combustors have some significant costs that are not incurred by systems fired on
standard fuels.
• Electrical Energy Costs. Lifting the bed of sand requires 30" or 40" w.g. pressure
that is easily four times as much as the forced draft (FD) pressure of a standard
boiler. So the fluidized bed has a substantially higher electrical energy cost to run
the combustion air fan.
• Cost of Bed Cooling. In order to keep the bed temperature low enough that the
sand does not melt it has to be cooled. One approach is to put water tubes in the
bed, which makes a very effective steam generator. However, the erosion rate
from sand scrubbing the tubes is very high, so the tubes have to be replaced quite
frequently. Another approach is to operate with about three times the normal
excess air level. This increases the size of the bed by a factor of three and triples
the electric bill for running the air fan. A third alternative is a system called a
circulating fluidized bed in which the bed sand is continuously circulated through
a heat exchanger to cool it off before it is returned to the combustor bed. This
incurs a significant amount of additional hardware, but avoids most of the
problems with the first two alternatives.
• Ash Management Costs. A fluidized bed combustor, even if it is burning clean

fuel, requires a dust collector to control particulate emissions with associated ash
management systems. Of course it also requires systems to make up sand that is
gradually lost as the system operates.

2.2.3 Thermal Oxidizers and Other Combustion Systems
Although combustion systems can generate air emissions, they are used effectively as
pollution control systems in a number of applications. The word incinerator has
acquired a negative political connotation leading to the use of other terms, such as
thermal oxidizer.
A variety of thermal oxidizers are used to control the emission of organic gases from
other (non-combustion) types of sources. Various industrial processes generate
solvent vapors or other organic gases and combustion is an effective way of
controlling the emission of these pollutants. When properly operated, thermal
oxidizers achieve nearly complete destruction of virtually all organic compounds,
converting them to carbon dioxide (CO2) and water vapor (H2O).
2.2.3.1 High Temperature Oxidizers

Figure 2-16 shows the elements of the basic thermal oxidizer. Air from a process
such as a coating operation is blown into a refractory lined chamber. A gas-fired
burner maintains the chamber temperature above the minimum destruction
temperature, typically 1400°F. If the incoming air is well mixed with the hot gas
from the burning flame, there will be non-detectable levels of organic compounds in
the exhaust. However, if the mixing is poor, there will be large spatial temperature
gradients in the exhaust and some of the incoming air will carry its pollutants up the
stack. The quality of mixing is determined primarily by the geometry of the chamber.
The partition wall in the chamber suggests the sort of geometry necessary to assure
mixing. In practice, a thermal oxidizer should be tested when it is commissioned to
be sure it is working properly. Once the design is shown to work well, at or above a
given temperature, its performance can be reasonably assured by simply maintaining
the exit temperature above the demonstrated level.
While a basic thermal oxidizer, such as the one pictured in Figure 2-16 is a very
effective pollution control device, it uses a lot of fuel and the resulting hot air is
dumped into the atmosphere. One way to reduce energy costs is to install a heat
exchanger, which can reduce fuel consumption by 70% or more. The applicability of
a heat exchanger depends on the kinds of organics being controlled and the type of
heat exchanger used. For example, chemicals like phenolic resins may condense in
the cold end of a metal plate or tube heat exchanger and make it inoperative. Other
types of equipment such as a regenerative heat exchanger can be more tolerant of
troublesome chemicals. Any type of heat exchanger can develop leaks or a valve
malfunction resulting in organic material bypassing the oxidizer. Thus the heat
exchanger becomes the primary cause of low destruction efficiency.

Exhaust
Burner
o
1400 F or Higher T
Gas Fuel
Optional
Heat
VOC Laden Air Exchanger
Figure 2-16. Thermal Oxidizer (Gas Incinerator)
2.2.3.2 Catalytic Oxidizers

Catalytic oxidizers are similar to high temperature oxidizers, except that they work at
much lower temperature and are therefore cheaper to operate. By passing the
contaminated air through an oxidation catalyst, organic compounds are oxidized at
temperatures as low as 500°F or 600°F. Oxidation catalysts are precious metals –
usually platinum, but rhodium and palladium have been considered. The configuration
is not very different from Figure 2-16, except that the burner only has to heat the air to
the catalytic reaction temperature and there is a catalyst between the burner and the
stack. The catalyst can be on the surface of a fixed support, like a honeycomb grid, or
on the material in a bed.
The destruction efficiency of a catalytic oxidizer depends on a number of factors

including the type of catalyst and the organics species being oxidized. Oxidation
catalysts work better at higher temperatures, so low temperature regions in the
oxidizer can cause reduced destruction efficiency. Performance also deteriorates if
there is insufficient catalyst surface area or if the air velocity through the catalyst is
too high. These are concerns that can be evaluated by the manufacturer during design
or by demonstration tests after the system is operational.
The performance of a catalytic oxidizer will deteriorate if the catalyst is damaged or

poisoned. Catalysts work by reacting with or adsorbing molecules of the original
species and then, when the reaction is completed, the catalyst is returned to its
original state. If there are species in the gas that react with or absorb on the catalyst
permanently, the catalyst does not return to its original condition and it gradually

becomes deactivated. The vulnerability of the catalyst to any particular type of
poisoning or contamination depends on the exact composition of the catalyst, its
operating temperature and other factors. Catalyst manufacturers invest heavily in the
development of proprietary formulations suited to specific applications. There is no
“silver bullet” and all catalysts are vulnerable under certain conditions. Thus there
needs to be some method to periodically monitor or measure the performance of a
catalytic oxidizer to be sure that it is still operating as intended.
2.2.3.3 Flares
Historically, flares were used by chemical processors and refineries to destroy waste
gases. Many of these served primarily as safety devices. Combustible gases that are
released accidentally can accumulate and pose an explosion hazard unless a continuous
ignition source at the point of release prevents inadvertent accumulation. Thus the
effectiveness of using a flare for pollution control is a relatively recent issue. Flares are
extensively used today at landfills, where the odorous gases generated in the landfill are
sucked out of the fill and fed to a flare for destruction. In this case, the effective
destruction of the gases is the primary purpose of the flare. At larger landfills the gases
are used beneficially as fuel to operate engines that generate electric power.
A flare has two basic purposes: (1) assure stable continuous combustion, and (2)
consume the organic gases efficiently with minimum emissions. The design usually
must accommodate variable rate gas flows and sometimes a significant variation in
heating value. Figure 2-17 is one design concept that achieves these objectives.
Stack
3ft. Diameter X 10 ft. High
Steel Shell
Refractory
Ignitor Pilot Gas
Pilot Gas
Commercial
Venturi
Burners Waste Gas
Figure 2-17. Flare Concept

Combustion occurs inside the refractory tube, which makes the flare relatively
insensitive to wind disturbance, and makes it feasible to measure the emission
performance of the flare. A number of venturi burners at the base give stable
fires. These burners use jets of gas (fuel) to draw in the right amount of air to
give a flammable mixture. The amount of air entrained by the venturi is directly
proportional to the amount of gas passing through the jet, so the fuel-air ratio is
held fairly constant and the flame does not smoke or blow out. The pilot gas at
the base provides continuous ignition. This is one of numerous flare designs.
2.2.3.4 Kilns and Other Devices
Cement Kilns
Cement kilns are large combustion devices used in the manufacture of Portland
cement. Their basic function is to heat a pulverized mixture of lime, silica and other
constituents to a temperature above 2300°F to create clinker, the principal ingredient
of Portland cement. Any lower temperature does not form the crystal structure of
clinker. Figure 2-18 shows the key elements of a cement kiln. The kiln is a long
sloping tube that rotates slowly. Meal, a careful blend of limestone, silica and other
minerals in powder form is fed into the upper end. The carbon is liberated from the
limestone as carbon dioxide and the remaining constituents partially melt and fuse
into the substance called clinker. After it leaves the kiln, the clinker is cooled and is
ground down to a powder that is Portland cement.
Exhaust gases @ 400° - 600°F

go to a high performance dust
collector
Pulverized rock blend meal

Burner - usually Kiln rotates slowly at ambient temperature
pulverized coal
Clinker
(Hot)
Product is heated
o
to about 2500 F
Figure 2-18. Rotating Cement Kiln
Kilns can burn any reasonably dry fuel; usually the most inexpensive is medium to
high sulfur coal. The coal is pulverized and blown into the lower end of the kiln
through a single burner. Air comes in parallel to the burner and is preheated by the
hot clinker that has left the kiln. The hot combustion gases flow up the kiln,
transferring heat to the meal and the meal flows in the opposite direction. The kiln is
a counterflow heat exchanger. In the process, much of the coal ash and any other

incidental minerals are combined into the clinker. The process also absorbs much of
the coal sulfur into the clinker as part of the crystal structure, so SO2 emissions are
reduced. In short, the process operates at a very high temperature and most of the
minerals introduced to the kiln are absorbed into the clinker. Cement kilns exhaust to
a large and very efficient dust collector.
Kilns can be significant sources of NOx, although progress in kiln burner design since
about 1980 has reduced these emissions. The one thing that kilns cannot easily control is
organic emissions originating in the minerals fed to the kiln. As the mineral feed moves
down the kiln, it is heated and any volatile constituents in the meal are released and
carried out by the exhaust gases. Only those that volatilize at very high temperature and
are carried far down the kiln will encounter gases hot enough to destroy them. Some
minerals contain traces of fossil organics (oil shale) or compounds that break down to
form ammonia. The normal cement kiln cannot control these resulting emissions.
The fact that kilns are large stable combustors, which operate at a very high temperature
and have very efficient dust collectors, makes them attractive for hazardous waste
incineration. In fact, a number of commercial cement kilns have permits to manage and
destroy hazardous wastes. They compete with facilities built specifically for hazardous
waste incineration. Organic wastes are fuel for the kiln and their use reduces the
consumption of purchased fuel, sometimes by 50% or more. For some older less efficient
kilns, this has made the difference between economic operation and having to shut down.
In addition to liquid wastes, some kilns can fire solid wastes. Long kilns as shown in
Figure 2-18 can be fitted with a mechanism that inserts a five-gallon pail into the
middle of the kiln once each revolution. This location is hot enough that the waste
burns completely. With other kiln systems, such as a preheater design, a short kiln is
used that includes only the hottest part of the process. The cooler part of the process
is handled in a separate preheater tower so the kiln feed is already at a temperature
above 1500°F. These kilns can accept whole tires as well as containers of solid waste
into the upper end of the rotating kiln. The steel in the tire cords is incorporated into
the clinker and the rubber provides energy at useful point in the process.
Sludge Burners
Wastewater treatment plants generate a semi-solid sludge, which tends to be odorous
and can be a disposal problem. This sludge typically contains as much as 80% water.
At one time most sewage sludge was land applied; that is once again becoming an
attractive and viable alternative. However, the contents of sludge can be a problem
for land application and the sludge needs to be dewatered in order to make
transportation viable from urban treatment plants to remote rural sites. So treatment
facilities that dewater, pelletize, and track the composition of the resulting product are
becoming more common. Disposal by incineration continues to be an option
particularly with dry pellets that have significant fuel value. Sludge pellets can be
fired in a stoker or fluidized bed boiler that is part of a waste to energy facility.

Historically, sludge incinerators were primitive devices intended primarily to get rid
of the sludge, with air emissions being a secondary consideration. The traditional
device is a multiple tray sludge burner as shown in Figure 2-19.
The multiple hearth furnace is a counterflow heat exchanger. Sludge is fed in the top
and drops by gravity gradually from one hearth to the next. As the sludge moves
downward, it is heated, dried, and eventually incinerated near the bottom of the
furnace. The heat produced by the sludge incineration near the bottom of the furnace
provides some of the heat necessary for drying sludge in the upper furnace. Residual
ash drops out the bottom. Drying removes not only water, but also any other volatile
constituents including much or most of the organic material in the original sludge.
These gases with much of the resulting organic vapor are emitted with the exhaust.
Cooling air discharge
Floating damper
Sludge inlet
Flue gases out

Rabble arm
at each hearth
Drying zone
Combustion
air return
Combustion zone
Cooling zone
Rabble arm
Ash discharge drive
Cooling/combustion
air fan/blower
Figure 2-19. Multiple Hearth Sludge Burner
The multiple hearth furnace is also used in some industrial processes. If the feed
material is free of volatile contaminants, the only pollutants will be from the
combustion process as well as any particulate generated from the product.

2.3 FUEL STORAGE, HANDLING AND PROCESSING
2.3.1 Natural Gas and Oil
Natural gas is the cleanest and easiest fuel to handle. However, because it is delivered
by pipeline, large end users must install branch piping from the gas company’s line to
their facility. In addition, smaller facilities use gas at pressures of only a few psi,
while larger facilities take gas at more than 100 psi. However, long distance pipelines
operate at pressures on the order of 1000 psi. Therefore, both large and small end
users must install a pressure reduction station.
In many locations, the price of gas is reduced if the user is willing to accept an
interruptible supply. During times of high demand (typically winter) a gas company
may run short of supply unless it has customers willing to have their gas shut off.
Thus it pays some companies to install hardware to fire a different fuel (usually oil)
and therefore be able to accept a gas curtailment. This may apply both to engines that
can burn No.2 oil and to large boiler facilities that typically burn No. 6 oil.
Oil may be delivered by pipeline in some locations, but typically it is delivered by

truck, barge or rail. Small to medium size facilities typically take delivery by truck.
Oil-fired electric utilities use too much fuel to make truck delivery practical and
typically take delivery by barge or rail. In some cases, they run a pipeline a few miles
from the nearest barge delivery point to the facility. Virtually all oil-fired facilities
have oil storage capacity, usually at least enough to operate one full day. Some oil-
fired utilities have large tanks that allow the plant to operate all winter when barge
deliveries are not possible.
No. 2 oil is usually stored at ambient temperatures and it can be pumped even when
the temperature is less than 0°F. By contrast, No. 6 oil has a pour point temperature
below which it is too viscous to be pumped. Pour points are commonly in the range
of 60°F to 100°F, but it can be higher. Therefore, No. 6 oil storage tanks must be
insulated. If the oil is not used rapidly, supplemental heaters are used to keep the oil
above its pour point.
Another concern with No.6 oil is stratification in the tank. If there is no mixing
device, oil added to the tank can form a layer above or below the existing oil. When
this transition layer drops past the oil pick-up line, it can cause a minor upset in the
source operation or it can alter the emissions. Water is also a concern because No. 6
oil deliveries occasionally contain significant amounts of water. Normally the water
settles to the bottom of the tank where, if there is enough of it, it will begin to
interrupt combustion as slugs of water are fed to the burner. Occasionally, No. 6 oil
is dense enough to cause water to float on the surface of the oil.

2.3.2 Solid Fuels – Storage, Feeding
Solid fuels, primarily coal and wood chips, are transported by truck or rail and
occasionally by barge or ship. Figure 2-20 shows the primary elements of solid fuel
handling. Perhaps the most significant element of solid fuel handling is the exposure
to weather. Rain during transport or on-site storage can substantially increase the
water content of the fuel with significant impacts on combustion performance. Thus
a source can have seasonal emission characteristics that are governed by the
accumulation of rainwater in the fuel. When the fuel is fed to the plant it may pass
through a crusher or shredder, past one or more screens to achieve a target size, and
under a magnet to remove “tramp iron.”
Rain
Truck or Train
Car or Truck
Dumper
Rain
Stockpile
Crusher or
Shredder To Burner
Hopper or
Feeder
Figure 2-20. Elements of Solid Fuel Handling
2.3.2.1 Coal
Coal is hauled from mines to electric utility plants in unit trains, where the entire train
carries coal to one plant. At one time many plants were located at or close to the
primary supply mine. However, with the advent of SO2 emission limits, a number of
plants switched to low sulfur coal. In a number of cases, coal is hauled from
Wyoming and Montana to the eastern or midwestern U.S. to meet the increased
demand for low sulfur coal. Natural moisture content for bituminous coal is less than
4%; however, surface moisture from rain can raise that figure to 10%. Wet coal
burns quite well, especially if it is pulverized, but it tends to stick to conveyors and
hopper walls which can make a lot of work for operators to keep all the coal feed
systems operating in a large plant.

Coal is stored in very large piles at the plant site. A car dumper empties coal rail cars
one at a time, and the coal is moved to an initial pile by a conveyor. Then it may be
pushed to another part of the pile by a large bulldozer. A key part of the bulldozer’s
operation is to compact the coal and shape the pile to promote drainage of any rainfall.
If the pile is porous and it absorbs a lot of water, spontaneous combustion can ignite a
section of the pile. The concept of water promoting combustion seems contradictory,
but the water enables chemical oxidation, which begins to warm the coal ultimately
leading to combustion. If properly managed, coal can be stored in a pile almost
indefinitely.
2.3.2.2 Wood and Waste

Wood chips and wood waste are usually consumed fairly close to the place they are
generated. Truck hauls of more than 100 to 150 miles are typically not economical.
As a general rule, good quality wood chips are used to make paper or engineered
lumber. Most of the fuel wood comes from the residue that remains after logging, or
is generated at the mill from bark or trimmings that cannot be used for product.
Construction and demolition activities also generate a significant amount of wood
scrap that can be burned as fuel if there is a facility nearby.
Combustion performance and air emissions are strongly influenced by the moisture
content of the wood being fired. Green or wet wood can contain 55% moisture or
more. Dry wood is usually less than 15% moisture. Wood from demolition activity
or waste trimmings at a wood products factory is usually fairly dry if it has not been
stored outdoors.
Wood stored in a pile deteriorates with time and, like coal, can catch fire. Unless
wood is kept dry, it promptly begins to rot and can lose 1% to 2% of its value every
month, either as fuel or for wood products. Hence a wood chip pile is a perishable
commodity that should be used as promptly as possible.
2.4 COMBUSTION AIR POLLUTION CONTROLS
2.4.1 Load Variations – Base, Swing, Peaking, Emergency

Most combustion systems are a part of a facility in which the load varies with time.
Electric utility generators can be broken into three categories:
• Base load. Base load units are the first to be brought to full load and they operate
at steady full load most of the time.
• Swing load. Swing load or topping units are intended to follow load swings and
make up the difference between base load and current demand.

• Emergency. Emergency units are normally off line (shut down) and are used only
when there is extraordinary demand or when there is a failure of the normal
generating and supply system.
2.4.2 Control Systems

With any of these types of systems, a control system is required to drive the fuel feed
system, match the air flow to the fuel flow, and operate other elements of the system.
A simple example of the type of load variation seen by a typical system is shown in
Figure 2-21 - a hypothetical load versus time for a swing load electric utility boiler.
The boiler is shut down when the trace starts about midnight. About 0300 it is started
and runs at minimum load till 0700 at which time the demand for electric power
increases and the boiler load rises to near maximum. The boiler then operates at
varying high load throughout the day, dropping to a minimum load about 2100 and is
finally shut down shortly after midnight.
CO
Load
NOx
0600 1200 1800 0000

Time of Day
Figure 2-21. Example of Load Change Effects
Until about 1980, control systems relied entirely on mechanical, pneumatic and
electro-mechanical devices. Then the programmable logic controller (PLC) appeared
and control systems became much more sophisticated. Thus the population of
systems in service today is a mix of pre-1980 systems, those that have been at least
partially converted with modern controls and modern electronic systems. There has
been a rapid evolution of control systems since 1980 and some work much better than
others.

2.4.2.1 Fuel and Overall Control
Fuel control is relatively simple. The basic operation of the control is to add more
fuel when load demand rises and decrease fuel flow when demand drops. The control
needs to be stabilized, so it does not fluctuate. The control may also set limits on the
minimum load, maximum load and rate of load change. A PLC controller can also
compare flow control position to measured fuel flow or to actual load in order to
detect abnormal operating conditions.
2.4.2.2 Air Flow and Controlling Fuel-Air Ratio

Air flow should vary with the fuel flow in order to keep a fixed ratio of air flow to
fuel flow. A good controller should be able to keep this ratio steady within 1% to 2%
of an operating set point. This may be possible based on measured fuel flow and
measured air flow, but if the system uses measured exhaust O2 concentration, then air
flow can be controlled much more precisely. Exhaust O2 is a direct indicator of the
actual air flow relative to the theoretically required amount. Measured O2 is about
one order of magnitude more sensitive an indicator than using the comparison of
measured fuel and measured air flow. Figure 2-22 shows one form of a current PLC
control system. The measured fuel and air flow signals are the basis for setting an
approximate air damper position. Measured O2 is then used to trim the damper to a
more precise position.
Load Signal O2
PLC
Flow
Signal
Fuel Valve
Air
Air Damper
Figure 2-22. Example of a Computer-based Control System

2.4.3 Air Moving Components – Fans and Draft Levels
Air flow through a boiler is driven and controlled by components illustrated in Figure
2-23. Flow through the system shown here is driven by a pair of fans, forced draft
(FD) and induced draft (ID). This is a balance draft system that is common for large
boilers. The term draft refers to the difference in pressure between ambient
atmosphere and the gas pressure inside the boiler. Some large boilers and most small
package boilers use only a single FD fan. Either system is capable of pushing air and
flue gas through the system. The difference between the FD fan and balanced draft
systems is the gas pressure at various places in the boiler. A balanced draft system
keeps the combustion zone and exhaust gases below atmospheric pressure. Any leaks
in the boiler walls or ducts will cause air to leak into the system rather than exhaust to
leak out of the system. If the boiler is built inside a building (typical of northern
climates) a balanced draft system prevents accumulation of flue gas in the building.
Another advantage of the balanced draft system is that it prevents very hot gases from
leaking out of the firebox with resulting damage to adjacent hardware.
In most systems, the flow through the FD fan is governed by the boiler air flow
requirement. The ID fan is linked to the firebox draft, so the ID fan normally follows
the movement of the FD fan. Although for all practical purposes, the two fans move
in unison. Flow through a fan is usually controlled by a damper as indicated in
Figure 2-23. Newer systems might have a variable speed drive in addition to the fan
damper.
In a typical arrangement, the FD fan draws air at ambient temperature; although,

some systems may include a preliminary air heater to keep the fan above freezing
during winter weather. The air heater raises the combustion air temperature with heat
drawn from the flue gas. Most utilities use rotating regenerative air heaters built by
Ljungstrom. The heated combustion air is supplied to the burners and/or under the
grate of a stoker. In the firebox, the flame heats the water walls by radiation to
generate steam. In the top and back of the boiler are convectively heated tubes, most
of which superheat the steam. Then the gas passes through the other side of the air
heater to the ID fan and up the stack. There is a pressure drop across the burners and
across each of the heat transfer elements. So the draft steadily decreases in value
from the exit of the FD fan to the inlet of the ID fan. Typical values in inches of
water are indicated. In a balanced draft system, the firebox is normally maintained
at -0.3" w.g. to -0.6" w.g.

Balanced Draft Utility Boiler
Typical Pressures Shown in Inches W.G. Relative
to Atmospheric Pressure
- 0.25
Superheat and Preheat Sections
2000°F
Economizer
600°F
Combustion Zone
Face Fired Boiler
with Three Rows
of Burners
550°F
+5 -4
Air Heater
350°F
-0.5
Fuel Feed +8 -9 -0.3
Forced Draft
(FD) Fan
Control
System
Induced Draft
(FD) Fan
Figure 2-23. Air Flow and Draft Controls
2.5 STEAM SYSTEM COMPONENTS

The gas side and water side of a boiler are physically separate, but they are
operationally linked. While the focus here is on the gas side, one invariably needs to
understand the basic elements of the water side and how they relate to the gas side.
Figure 2-24 provides a somewhat detailed cross section of a typical large utility
boiler. A smaller boiler will have many of the same components.

Attemperator
Steam
Drum
Reheat
Superheater
Platens
Primary
Superheater
Secondary
Superheater Economizer
Gas
Proportioning
Dampers
Furnace
Downcomers
Burners
Burners
Secondary
Air Duct Secondary
Coal Air Heater
Bunker
Primary
Air Heater
Feeder
Primary Gas Outlets
Air Duct
Tempering
Air Duct
Pulverizers
Primary Steam Coil

Air Fan Air Heater
Figure 2-24. Boiler Steam and Water Components

Adapted from a drawing in Steam – 40th Edition by Babcock & Wilcox, 1992
2.5.1 Steam Generation

Perhaps the simplest way to go through the boiler is to follow the water as it passes
through the system. The amount of water flowing into the boiler is at least as large as
the amount of steam flowing out of the boiler. In larger boilers water is preheated,
sometimes in several steps beginning with a steam heater. One or more boiler feed

pumps operate at high enough pressure to supply water to the boiler. The first water
heater in the boiler is called an economizer. The economizer may be the last heat
exchanger in the gas pass or located just ahead of the air heater.
When water leaves the economizer it enters the steam generating part of the system
and is piped directly to the steam drum. The steam generator consists of a steam
drum at the top, water tubes which transfer heat to the water, and a manifold or mud
drum at the bottom. In a circulating boiler (most of them), there are also downcomers
that pipe water from the steam drum to the mud drum. Water boils in the tubes and
the steam-water mixture enters the steam drum that separates the steam from the
water. Water flows from the steam drum, through the downcomers to the mud drum,
and then back into the tubes to generate more steam. The amount of water circulating
in this manner can be six to eight times as much as the steam flow; that is equivalent
to saying that water passes up through the tubes an average of six times before it is
converted to steam. Some boilers rely on natural circulation while others have
pumps for forced circulation. Still other boilers have no water recirculation at all.
In a once-through design, all the water changes to steam before it reaches the drum,
which simply serves as a manifold to collect and distribute the steam.
Saturated steam leaving the drum can be used directly or it can be superheated to a
temperature higher than the boiling point. Small commercial boilers generate steam
for heating at pressures less than 15 psi, with corresponding saturated steam
temperatures below about 250°F. However, steam used to drive a power turbine is
superheated. Steam conditions in electric utility boilers can have steam pressures up
to 3600 psi at superheat temperatures of 1000°F and higher. In Figure
2-24 steam goes from the drum at a temperature of about 690°F to a primary
superheater that might raise the temperature to 850°F before it goes to the secondary
superheater, which raises the temperature to 1000°F. The final steam temperature is
controlled primarily by an attemperator between the primary and secondary
superheaters, which can spray water into the steam to cool it. Superheated steam is
then used to drive a turbine – usually connected to an electric generator.
When water is converted to steam any mineral contaminants stay in the water or are
deposited on tube and drum surfaces. These deposits interfere with heat transfer and
can lead to corrosion and tube rupture. If the steam drives a turbine, the turbine
blades will be adversely affected by contaminants in the steam. As a general rule, the
higher the steam pressure, the cleaner the water must be. Thus any facility with a
medium to high pressure boiler must install equipment to purify the water used in the
boiler. Purified water fed to the boiler is called makeup. The facility will also have
instrumentation and at least some simple chemical analysis ability in order to track
water quality. Steam used for heating or to operate turbines is usually condensed and
returned to the boiler. The method of removing contaminants from this loop is to
dump some of the water, a process called blowdown. Blowdown valves in the mud
drum are opened periodically (as necessary) and the blowdown rate can be 1% or 2%
of the steam flow rate.

2.5.2 Steam Turbines
Most steam-electric power plants have two turbines: a high pressure turbine and a
low pressure turbine. Steam leaving the high pressure turbine is sent back to the
reheat superheater in the boiler where the steam temperature is again raised to
1000°F. The steam then drives the low pressure turbine. Steam conditions at the
exhaust of the last turbine are typically less than 100°F and pressures measured in
inches of mercury absolute – a substantial vacuum. A condenser maintains these
conditions, which is either water or air-cooled. A water-cooled condenser’s
performance is usually influenced by seasonal water temperature variations and by
any tidal variations in water level. Any influence that raises condenser temperature
reduces turbine efficiency and this telegraphs back to the boiler, which has to fire
harder to make more steam to make up for the lost efficiency.
There are two other factors that commonly reduce steam turbine efficiency. One is
mechanical deterioration of the turbine, which normally occurs over a period of many
years. This can reduce turbine efficiency, and thus peak power output, by 5% or
more. The second is turbine inlet steam temperature. Normal steam temperatures on
newer units (since about 1955) are 1000°F. If steam temperature drops by even 15°F,
there will be a measurable increase in boiler fuel flow to make up the difference. A
discussion of the factors affecting superheated steam temperature is beyond the scope
of this text, but some of the significant factors are listed below.
• Size of the superheater

• Deposits on the inside or outside of the superheater tubes
• Type of fuel being burned – gas, oil or coal
• Excess air level
• Boiler load
• Burners in service
Note that conditions on the gas side of the boiler are as important as conditions on the
water/steam side in determining steam temperatures.
2.6 ASH HANDLING
2.6.1 Boiler Surface Deposits

Gas and distillate oil fuels carry little or no ash into the combustion zone. Residual
oil carries a small amount of ash, which can, over a period of weeks or months, generate
a fairly thick coating on tubes throughout the boiler. Coal is about 10% ash (or higher),
which means that a ton of coal carries 200 pounds of incombustible rock into the

combustion zone. In a stoker furnace most of the ash stays on the grate, but in a
pulverized coal furnace all the ash is carried through the flame zone in suspension.
There are two major consequences: (1) an efficient dust collector is necessary on all
pulverized coal-fired boilers in order to control particulate emissions, and
(2) accumulation of dust and slag on boiler surfaces is a major factor in the design and
operation of pulverized coal boilers.
The dust generated by firing pulverized coal is a dry powder at room temperature.
The composition is silica, alumina, iron oxide and other mineral oxides – essentially
pulverized rock. In the combustion zone, much of this dust is semi-molten and some
constituents actually vaporize. The dust cools very rapidly after it leaves the flame
zone, but if it is semi-molten when it hits a surface it sticks and accumulates. Thus
furnace temperatures and the melting temperature of the ash are key variables in
determining how much dust will stick to the walls and where it will do so. Any
deposit on the boiler walls and tubes reduces the heat transfer rate. Deposits that
grow uncontrollably also present a physical hazard inside the furnace. They block the
normal gas passes and if large deposits break free, they can damage the bottom of the
boiler. In general, the operators attempt to keep the deposits relatively dry and avoid
situations where liquid slag is running down the walls of the firebox.
Boiler operators use several methods to control gas side deposits inside the boiler:
• Purchase coal with an appropriate ash melting or softening temperature.
• Use soot blowers to remove deposits.
• Limit the load, excess air or other parameters to minimize problems.
Utility power plants are selective about which mines supply their coal. Anytime a
plant proposes to fire coal from an untried mine, it will conduct a test burn to
determine if there are any unmanageable problems.
Soot blowers are devices that can blast the walls or tubes with jets of steam to remove
loose deposits. They are permanently mounted on the boiler walls and positioned on
long pipes that penetrate deep into all the convective heating tube banks in the boiler.
The soot blowers are controlled by the boiler operator with the objective of
maintaining normal steam temperatures. The operator selectively cleans certain
surfaces to restore heat transfer rates that control the temperatures.
Increasing load or decreasing excess air tends to raise boiler temperatures and high
temperatures lead to liquid or sticky slag. Load, excess air, and the choice of which
burners to use are parameters that the operators use from experience to control ash
accumulation problems.

2.6.2 Bottom and Fly Ash Management
In a stoker boiler most of the fuel ash stays on the grate and is dumped into a hopper
for disposal. However, in a suspension-fired boiler, less than 20% of the fuel ash drops
out in the boiler hoppers; the remainder, fly ash, must be captured in the dust collector.
Bottom ash is usually dropped into water to cool it. The recovered material has the
consistency of gravel. Coal bottom ash has been used for fill in most of our coastal
cities and can also be used on icy streets in the winter. Fly ash recovered by a dust
collector is a fine powder that is frequently mixed with water to prevent fugitive
emissions. Almost all fly ash is land filled somewhere close to the power plant.
2.6.3 LOI and Introduction to Ash Chemistry

Bottom ash typically has a furnace residence time of an hour or more. This is long
enough to burn out any residual carbon from the fuel so bottom ash usually has very
little carbon. However, biomass-fired stokers are an exception because the ash is
frequently removed before the fuel is completely burned. Fly ash carbon contents can
be significant, particularly from biomass-fired stokers. On pulverized coal-fired
boilers, fly ash carbon contents should normally be less than 4% and at this level
represent a heat loss of less than 0.3%. Ash carbon content is usually characterized
by a Loss on Ignition (LOI) laboratory test. The sample is heated to about 1400°F
and the change in weight is attributed to carbon loss.
Ash chemistry reflects the composition of ash in the fuel. However, the proportions
of various elements in the fly ash will be skewed from the original as discussed in
Chapter 5. This change is caused by ash constituents that are volatile or semi-volatile
at flame temperatures. Any substance that vaporizes in the combustion zone will
condense as the gas cools and deposit on all available surfaces. Most of this
condensation will be on ash particles and, since fine particulate matter has more
surface area than course particulate, these volatile species will appear preferentially
on the finest fly ash particulate. This can result in the fly ash having a very different
chemical signature from the bottom ash.
2.7 CONCLUSION
This chapter has attempted to provide a brief description of the types of combustion
sources that someone in the regulatory community might reasonably expect to
encounter. The focus is on an overview and general description of combustion sources.
A more detailed description of the specific types of sources can be found in one or
more of the references indicated in this section. Subsequent chapters of this text will
present more information relevant to air emissions from these sources.

Review Exercises
1. Reciprocating engines are commonly found in a(n) ________.
a. jet aircraft
b. gas turbine power plant
c. automobile
d. steam electric generating plant
2. Combustion turbine power plants differ from steam power plants because they ________.
a. typically have 10 to 100 times less power
b. have more operating parts that require greater maintenance
c. have less thermal efficiency
d. can only burn clean fuel, such as natural gas or distillate oil
3. A chief characteristic of a rich burn reciprocating engine is that it ________.

a. has less power, but lower emissions than a lean burn engine
b. has more power, but higher emissions
c. is spark ignited with a lean gas mixture in the intake air
d. is exclusively used with diesel oil
4. Which of the following boilers has the largest thermal capacity?

a. Electric utility boilers
b. Large factory built boilers
c. Residential furnace
e. Commercial building HVAC boilers
5. Which of the following is not true of suspension type burners?

a. Allows better control of the combustion process than stoker firing.
b. Operates by injecting fuel into the combustion zone.
c. Capable of burning almost any solid fuel.
d. Permits greater control of fuel-air mixing.
6. Which of the following correctly characterizes the evolution of electric power generating
plants in the U.S.?
a. Pulverized coal furnaces / stoker combustors / oil-fired boilers / gas turbines
b. Stoker combustors / pulverized coal furnaces / oil-fired boilers / gas turbines
c. Oil-fired boilers / stoker combustors / pulverized coal furnaces / gas turbines
d. Pulverized coal furnaces / oil-fired boilers / stoker combustors / gas turbines

7. Which combustion system correctly identifies the following illustration?
Boiler Fire Box
Tertiary Air
Crushed Coal Hot

Combustion
Gases
Primary Air
300 fps
Liquid Slag
a. Large industrial boiler

b. Cyclone burner
c. Combustion gas turbine
d. Fluidized bed combustor
8. Which of the following stoker grates would be preferred for burning municipal waste?
a. Stationary grate
b. Traveling grate
c. Oscillating grate
d. A and C
9. Which of the following oxidizers or combustion systems operate at a relatively low

temperature?
a. High temperature oxidizer
b. Catalytic oxidizer
c. Utility boiler
d. Cement kiln
10. In a suspension-fired boiler, approximately ________ of the fuel ash drops into the boiler
hopper for disposal.
a. 20% or less
b. 50%
c. 70%
d. 100%
11. Air flow through a combustion turbine is ___________.

a. about the same as a boiler that fires a comparable amount of fuel
b. kept low to minimize exhaust noise
c. very high in order to keep the turbine at an acceptable temperature
e. very high in order to maximize fuel efficiency.

Review Answers
1. c. automobile
2. d. can only burn clean fuel, such as natural gas or distillate oil.
3. b. has more power, but higher emissions.
4. a. Electric utility boilers
5. c. Capable of burning almost any solid fuel.
6. b. Stoker combustors / pulverized coal furnaces / oil-fired boilers / gas turbines
7. b. Cyclone burner
8. c. Oscillating grate
9. b. Catalytic oxidizer
10. a. 20%
11. c. very high in order to keep the turbine at an acceptable temperature

References
Babcock and Wilcox Company. Steam: Its Generation and Use, (Babcock and
Wilcox Company, Barberton, Ohio, 1992). A huge book with in depth coverage of
large boiler furnaces, particularly coal-fired boilers. It also covers fluidized beds and
refuse firing. The book has grown over the years and the material covered has
changed.
Bryers, Richard W. Ash Deposits and Corrosion Due to Impurities in Combustion

Gases, Proceedings of an International Conference in Henniker, NH, (Hemisphere
Publishing Corp in association with McGraw-Hill International Book Co, 1977).
Gill, James H. and Quiel, John M. Incineration of Hazardous, Toxic, and Mixed
Wastes, (North American Mfg. Co., Cleveland, OH, 1993). This book focuses on
waste combustion in rotating kilns, but much of the data is broadly applicable to other
systems and fuels.
Pritchard, R., Guy, J.J., and Conner, N.E. Handbook of Industrial Gas Utilization,
(Van Nostrand Reinhold Co., 1977). The authors are/were at the University of
Salford (GB) and the detail/clarity are typically British. The book covers all aspects
of combustion hardware and furnace design for any system that fires gas or gas in
combination with other fuels.
Reed, Richard J. and Schilster, Robert E. North American Combustion Handbook,

(North American Manufacturing Company, Cleveland, OH, 1993). A good basic
reference for small to medium size gas and oil-fired combustion systems.

Volume II, (North American Manufacturing Company, Cleveland, OH, 1995).
Provides good detail on burner types, control systems, emissions, heat recovery and
other special topics.
Singer, Joseph G. Combustion Fossil Power Systems, (Combustion Engineering,

Inc., Windsor, CT, 1981).
U.S. Environmental Protection Agency. Alternative Control Techniques Document:

NOx Emissions from Stationary Gas Turbines, January 1993, EPA Report 453/R-93-
007, NTIS PB93-156586. This weighty document is a fairly a comprehensive, if
slightly dated, description of combustion turbine technology with a significant
database included.

NOx Emissions from Stationary Reciprocating Internal Combustion Engines, July
1993, EPA Report 453/R-93-032, NTIS PB94-104494. This weighty document is a

fairly a comprehensive, if slightly dated, description of engine NOx technology with a
significant database included.
U.S. Environmental Protection Agency. Introduction to Boiler Operation: Self-

instructional Guidebook, APTI Course SE:428A, 1984. The descriptions of boilers
and related combustion systems in the Student Manual are clearly written and well
illustrated.
Wilber, Leslie C. Handbook of Energy Systems Engineering, (John Wiley & Sons,
New York, 1985). A hefty handbook that covers many areas including useful
sections on gas, oil and coal combustion.


Chapter 3: Fuel Characteristics
LESSON GOAL
Demonstrate an understanding of how various gaseous, liquid, solid, and emulsion
fuels are used in combustion sources through successful completion of the chapter
review exercises.
STUDENT OBJECTIVES
1. Describe past and current trends in the types of fuel used in stationary sources in the U.S.
2. Describe the trends in fuel production in the U.S. over the past 30 years.
3. List the differences between fuel analysis based on “as received” and “mineral and
ash free” basis.
4. Describe the difference between an ultimate fuel analysis, a trace element analysis
and a market driven analysis such as a proximate analysis.
5. Discuss the issues involved in obtaining a representative fuel sample for lab analysis.
6. Describe the difference between the lower and higher heating value of a fuel.
7. List the primary non-fuel contaminants in bituminous coal and No. 6 oil.
8. List the typical ranges in sulfur and ash content in No. 2, No. 4, and No. 6 fuel oils.
9. Describe how the size distribution of solid fuels is obtained in the laboratory.
10. List some of the common minor constituents present in natural gas.
11. Describe the difference in potential air pollutant formation between residual and
distillate oils based on how they are manufactured.
12. Describe the differences in heating value, ash content, and sulfur levels among
anthracite, bituminous, sub-bituminous, and lignite coals.
13. Describe how agglomerating/caking tendencies may affect stoker coal combustion.

14. Describe how ash softening and fusion temperatures can affect pulverized coal
combustion.
15. Describe the range in mercury levels in coals throughout the U.S.
16. List the various types of wood fuels used in combustion processes.
17. List various types of solid waste and describe their general characteristics.
18. Describe the common factors that affect the moisture levels in municipal waste fuels.

Chapter 3: Fuel Characteristics
3.1 INTRODUCTION
Fuel characteristics are the single most important factor that must be considered in the
design and operation of a combustion process. The fuel characteristics significantly
affect almost every aspect of the combustion system from (1) the fuel storage and
handling equipment, (2) the configuration and size of the combustion furnace and
burners, (3) the air pollution control systems, and (4) the ash handling systems. Fuel
characteristics also have a major impact on the quantity and types of air pollutants
generated during combustion. The variability of fuel quality over time introduces
routine variability in the performance of combustion systems.
This chapter begins with a general summary of the wide variety of fuels used during
the combustion process. Information concerning ultimate and proximate analyses is
presented because these data are frequently used to evaluate the fuel properties. In
later sections of this chapter each of the major fuels is discussed. The emphasis in the
fuel specific sections is on the fuel characteristics that are most relevant to air
pollution control and emission oriented projects.
3.2 FUEL CHARACTERISTICS AND USAGE TRENDS

In the 19th century wood was the most common fuel. By the beginning of the 20th
century it was replaced by coal in stationary sources. Coal supplied more energy and
was available in larger quantities. In the early 1960s oil began to compete with coal
in terms of cost and usually generated lower air emissions than coal. By the end of
the 20th century natural gas had become a preferred fuel for new stationary sources,
but coal and oil continued to be used where they were practical and when they were
more economical than natural gas. The Clean Air Act Amendments of 1990 (CAAA
1990) have encouraged the use of clean fuels to reduce emissions that contribute to
acid rain and ozone formation. Concerns over carbon dioxide emissions are tending
to increase the use of biomass and other renewable types of fuels. Decreasing landfill
capacity and concerns over the environmental adequacy of landfills have also
stimulated interest in the increased firing of waste materials. Collectively, these
trends have begun to have an impact on combustion process design and on a facility’s
choice of fuel. In the future we will probably see the processing of coal, waste or bio-
fuels to create clean gas and/or liquid fuels.

3.2.1 General Types
There are five major categories and numerous subcategories of fuels used in
combustion processes as indicated in Table 3-1. However, this list omits most of the
minor fuels. Furthermore, new fuel types – based on wastes or biomass – may be
commercially developed and used in the future.
Table 3-1. Fuels

Fuel Category Examples
Solids Bituminous coal, sub-bituminous coal, lignite, peat, wood
Liquids Kerosene, No. 2 oil, No. 6 oil, coal-water emulsions, oil-emulsion
Gaseous Fuels Natural gas, propane, landfill and biological gases
Wastes Municipal and medical wastes, hazardous wastes, sewage sludge, tires
Biomass Wood and wood waste, bagasse, straw
The dominant fuels for stationary sources since 1990 have been coal, natural gas, oil
and wood, generally in that order. The consumption and production of dominant fuels
in the U.S. is indicated in Figure 3-1. Note the steady increase in the production of
coal and the decrease in the production of oil. The steady increase in coal use is due
primarily to increased demand for electrical power and abundant U.S. coal supplies.
However, total oil consumption has actually increased with the growth in transportation
use; although stationary source use has been partially displaced by natural gas. The
decrease shown in Figure 3-1 is offset by an increase in imported oil.
Table 3-1 collectively addresses all types of coal including anthracite, bituminous,
sub-bituminous, and lignite coals. Due primarily to the cap in SO2 emissions
mandated in the CAAA of 1990, there has been a trend toward sub-bituminous and
lignite coals that have low sulfur content. Most wood fuels are used near the point of
generation, primarily in the forest products industry (paper and saw mills) as well as
in furniture manufacturing plants. Although there has been some shift from the use of
No. 6 to No. 2 oil, driven primarily by the need to minimize sulfur dioxide emissions,
the use of No. 2 oil has remained relatively steady.
A variety of liquid fuels are beginning to be applied for electrical power generation
and industrial process heating. One of these new fuels includes ori-emulsion, an oil-
water emulsion extracted primarily in Venezuela. Large deposits of ori-emulsion
appear to be available for future years. In addition, various energy companies are
experimenting with coal-water emulsion fuels that burn like No. 6 oil, but require
major air pollution control devices. Depending on market prices and other factors
these fuels may compete for fuel oil combustion process applications at some
time in the future.

By Major Source, 1949-2000
25
Natural Gas
20
1
Crude Oil
15
Coal
10
Nuclear Electric
Power
5
Hydroelectric Power3
NGPL4
Wood and Waste2
0
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
1
2
Includes lease condensate
3
Includes ethanol blended into motor gasoline
4
Conventional and pumped-storage hydroelectric power
Natural gas plant liquids
Figure 3-1. U.S. Energy Production Trends

As indicated in Figure 3-1, the use of natural gas increased substantially between
1990 and 2000. However, a marked increase in the consumption of natural gas for
electrical power production has played a large part in driving up total consumption.
Natural gas use peaked about 1970 then significantly decreased from 1973 through
the early 1980s. There was a perception that gas supplies were dwindling and that a
crisis loomed. The oil crisis of 1973 spurred Congress to pass a law that prohibited
new applications of natural gas for electrical power generation. However, the law
was repealed in 1990 after it became apparent that natural gas supplies were more
than adequate for the foreseeable future.
Natural gas use for electrical power production has increased significantly since the
mid-1980s. This increase is due primarily to the availability of natural gas at
competitive prices. As a result, a number of coal- and oil-fired boilers have been
converted to primarily gas-fired units. This change has reduced both sulfur dioxide

and nitrogen oxide emissions. Natural gas use has also increased due to the
construction of new power plants driven by gas turbines.
A small increase in natural gas use is associated with thermal oxidizers used for air
pollution control that was mandated by the promulgation of Maximum Achievable
Control Technology (MACT)-based standards for hazardous air pollutants.
Natural gas will be the dominant fuel used in a wide variety of small combustion
units installed during the next twenty years as the electric generation industry
increases “distributed power” capability. Almost certainly this will result in the
increased use of electrical generation at or near the point of use rather than at large
centralized power stations.
Table 3-2. U.S. Energy Consumption, 1993-1997

Energy Use Expressed in Quadrillion BTU
Energy Source Energy Type 1993 1994 1995 1996 1997
Coal 19.837 20.027 20.090 21.011 21.439
Fossil Fuels Natural Gas 20.827 21.288 22.163 22.560 22.588
Petroleum 33.841 34.735 34.663 35.864 36.314
Total Fossil Fuels 74.522 76.073 76.943 79.434 80.360
Hydroelectric Power 3.147 2.996 3.472 3.914 3.932
Geothermal Energy 0.393 0.395 0.339 0.352 0.322
Renewable Fuels Biomass 2.784 2.838 2.846 2.938 2.723
Solar Energy 0.071 0.072 0.073 0.075 0.074
Wind Energy 0.031 0.036 0.033 0.035 0.035
Total Renewable Fuels 6.426 6.309 6.763 7.315 7.086

[2]
Source: DOE/EIA –0603(98)/1 Energy Information Administration
Table 3-2 shows specific data for fuel use over a recent five-year period. Note that
renewable energy is only about 10% of the total, so the U.S. has a long way to go to
substantially replace fossil fuel with renewable energy sources. Most of the renewable
energy comes from hydroelectric power and it is unlikely that additional dams will be
built in this country to allow a significant increase from this source. In fact, environmental
concerns could actually reduce hydroelectric power generation. An expansion in the use of
bio-fuels implies development of large land areas devoted to fuel crops, which clearly has
limited potential. Wind and solar energy production can be expanded, but these sources
also will have to grow by two orders of magnitude in order to begin to significantly
displace fossil fuel. Although displacement of fossil fuels with renewable energy is a
highly desirable goal, its implementation is challenging.

3.3 FUEL ANALYSES
3.3.1 Basis of Fuel Analysis
A fuel analysis is a laboratory report with information useful to the combustion
source operator, regulatory agency and others. Originally the purpose of a fuel
analysis was to help assure that a fuel was suitable for a particular combustion
system, or to determine the economic value of the fuel. The octane rating of gasoline
is an element of a fuel analysis that is most familiar to the general public. Each
automobile engine has specific requirements and the analysis of the gasoline must
show that the fuel meets or exceeds these requirements. The fuel requirements for
stationary sources are related to the ability of the source to burn the fuel as well as
environmental issues. In this section we will examine those fuel analyses that are
most commonly conducted and relate their significance to stationary combustion
sources.
A fuel analysis is the product of a series of laboratory tests on a small sample of the
fuel. When reviewing any analysis we need to know what was analyzed, or the basis
for the analysis. Examples of the basis for fuel analysis include:
• As-received: The fuel as it was received in the laboratory.
• Air dried: The fuel was dried at room temperature to remove surface moisture.
• Moisture free: The fuel was dried in an oven to remove moisture.
• Mineral free: The ash and inorganic materials were subtracted from the analysis
to report the composition with mineral and ash excluded.
As-received fuel is undried fuel in the same condition as it was received at the
laboratory. Due to the possibility of surface moisture loss from solid fuels and wastes
during sampling and shipment, the reported as-received moisture level can be slightly
lower than the as-fired fuel.
Air-drying is conducted at room temperature, which removes surface moisture that

reflects water accumulated during transportation and storage. This basis is sometimes
used in the analysis of coals that are being fired in pulverized coal boilers. The fuel
analysis expressed on an air-dried basis is a reasonable description of the fuel as it
leaves the pulverizing mill enroute to the combustion zone. ASTM procedure D3302
is used to determine the amount of air-dried moisture that is used to convert the fuel
analysis data to an air-dried basis.
Moisture-free fuel has been oven dried in accordance with ASTM D3302 to remove
the inherent moisture trapped in pores and cracks of solid and waste fuels. Fuel must
be dried very carefully to avoid excessive volatilization of organic vapors from the
fuel. Loss of volatile matter will reduce fuel heating value and alter elemental

composition. Fuel analyses are often expressed on a moisture-free basis to eliminate
variability introduced by the moisture in the fuel.
A mineral-free fuel analysis allows comparison of different fuels without the variable
influences of ash and water. However, the mineral content of many solid fuels can
vary widely.
Analysis can be broken into three types: (1) market driven (includes proximate), (2)
ultimate and (3) trace element. What we have called the market analysis gives a
description that is useful to buyers and sellers in a fuel market. Ultimate analysis,
gives a detail listing of the elemental composition of the fuel. Trace element analyses
are supplemental analysis conducted for whatever elements are requested. Table 3-3
lists common constituents of these analyses.
Table 3-3. Fuel Analysis – Common Elements

Type of Analysis Coal and Solid Oil Gas
Market (Proximate) • Heating value per lb. • Heating value per gal. • Heating value per ft3
• Moisture • Density or API gravity • Specific gravity
• Sulfur • Sulfur
• Volatiles • Viscosity
• Fixed carbon • Pour point
• Ash • Water & sediment
Ultimate (Chemical) • C, H, S, N • C, H, S, N • CH4, C2H6, C3H8, etc.

• Oxygen by difference • Oxygen by difference • CO2, N2, H2O
• Moisture • Moisture • O2
• Ash • Ash
Ash or Trace Element • SiO2, Al2O3, TiO2 • V, Na, • Others by request

• Fe2O3, CaO, MgO • Others by request
• K2O, Na2O, Cl, Hg
• Others by request
A fuel analysis is the result of a series of laboratory tests that begin with a one or two
pound fuel sample. The results are representative of the fuel only if the sample is
representative; however, it is not difficult to obtain a representative oil sample. Most
facilities have a hand valve located on a tee off the main line leading to the burners.
The sample is taken some time during an air emissions test (or other pertinent event).
Some oil is drained from this tee before filling a quart sample bottle that is then sent
to the lab for analysis.
Sampling coal, wood or any other solid is another matter entirely. Solid fuel is not
homogeneous and a small sample taken from one place at one point in time can be
very different from the average. The original procedures used to sample coal
illustrate the difficulty in obtaining a representative sample. Coal is transferred into a
combustion facility on one or more conveyor belts. At the beginning of a test that
requires a fuel sample, a technician is stationed at each feed belt. At five or ten
minute intervals the technician rakes a two or three pound sample off the belt into a

large garbage can. At the end of the test all the garbage cans, perhaps totaling half a
ton, are collected and emptied onto a pile on a clean floor or a tarpaulin. Using
shovels, this pile is mixed to the maximum degree practical. The pile is then divided
in half; one half is discarded and the other is divided again. One of these quarters is
then divided again. This process continues until the sample is reduced to about 10
pounds, which can be sent to the lab. The lab then takes out about one pound and
grinds it up to obtain smaller samples for each of the tests. Mechanical samplers have
largely replaced the need for technicians to rake the belts, but the challenge of getting
a representative sample remains.
While representative sampling of coal and wood chips is feasible, it is all but
impossible to obtain a small, representative sample of municipal waste. Many of the
individual items in the waste stream are larger than the sample required by the lab.
Combustion assessments of solid waste facilities must be conducted without the
benefit of a representative fuel analysis.
3.3.2 Market Characteristics

When purchasing fuel, a buyer wants to know how much energy they will get and
whether the fuel is suited to their use. The essential information about the fuel is
contained in an analysis that goes by different names for different fuels. Proximate
analysis is the term used for solid fuels (e.g., coal, etc.). No single name is used for
this type of analysis of oil or gas. The normal components of this analysis are listed
in Table 3-3.
3.3.2.1 Heating Value

For economic reasons the price of any fuel is expressed as dollars per million BTU
($/mmBTU). The price of fuel can be determined from the cost per pound and the
heating value that is expressed as BTU per pound of solid, per gallon of liquid, or per
cubic foot of gas. Heating values are determined by burning a small sample
(typically one gram) in a calorimeter that measures the amount of energy released.
A British Thermal Unit (BTU) is the amount of energy required to heat one pound of
water one degree Fahrenheit. The energy released when burning a pound of fuel
should be equally unambiguous, but unfortunately there are two different definitions
commonly used. The difference has to do with the water vapor generated by
burning hydrogen in the fuel and whether or not this water vapor condenses. The
higher heating value (HHV) assumes that all the water condenses, which means that
the flue gas is cooled to room temperature. This is what happens in the laboratory
test apparatus that measures heating value. In real combustion systems, the flue gas is
almost never cooled below the water dew point, so the uncondensed water vapor
carries energy with it up the stack. If one excludes the heat of condensation from the
definition, you have the lower heating value (LHV). The difference is 1030 BTU/lb
of water generated by combustion. Table 3-4 lists typical higher and lower heating
values for several fuels.

Table 3-4. Fuel Heating Values - BTU/lb
Fuel HHV LHV
Natural Gas 22,200 20,000
No. 2 Oil 19,000 17,860
No. 6 Oil 18,200 17,300
Bituminous Coal 14,000 13,600
Doug Fir - dry 9,000 8,400
HHV is used for emission calculations and is nearly always the value used in any
regulatory context, but it also is used by people who operate boilers and furnaces.
HHV is generally used by anyone selling fuel because no vendor would voluntarily
use the lower value to characterize his fuel. On the other hand, engine and turbine
manufacturers use LHV. The argument for this approach is: “since the heat of
condensation is never available, we will exclude it from all calculations”. A resulting
benefit is that their reported engine efficiency is higher (more impressive) when they
use the lower heating value. Problems arise when an unsuspecting engineer combines
data from two different sources and does not realize that the BTU values from the two
sources are not comparable. For example, if a permit writer combined engine
efficiency data (BTU per kilowatt hour)from an engine manufacturer with emissions
data (lb/mmBTU)from an EPA Method test, he could underestimate the total
emissions by more than 10%. So it is important to know which definition of BTU is
used and to convert everything to one basis, preferably HHV.
The significance of fuel heating value is illustrated in two example problems.
Example 1: Importance of Fuel Heating Value
A combustion process must operate at a heat release rate of 10 x 106 BTU per
hour (10 mmBTU/hr) in order to maintain the desired boiler-operating load.
Fuel A has a heating value of 11,000 BTU/lb. while Fuel B has a heating
value of 12,750 BTU/lb. What is the difference in the quantities of fuel
needed for alternative Fuel A and Fuel B?
Solution:
HT Release Rate 10 × 106

Fuel A = = = 909 lb
Fuel HHV 11,000 hr
HT Release Rate 10 × 106

Fuel B = = = 784 lb
Fuel HHV 12,750 hr
As illustrated in this example, considerably less of Fuel B is needed.

The fuel storage, feeding and processing equipment necessary for a plant
designed for Fuel B could be smaller than for a plant designed to handle Fuel A.
The heating value also has a direct impact on the quantity of ash that must be
handled. The relationship between heating value and ash quantities is
illustrated in Example Problem 2.
Example 2: Relationship Between Heating Value and Ash Quantities
The plant in Example 1 is considering converting from coal supply B to coal

supply A. Both fuels have an ash content of 10.5%. The boiler operates 7000
hours per year. What change will occur in the total quantity of ash that will
have to be disposed? Assume that none of the ash escapes into the
atmosphere.
Solution:
0.105
(Coal B) Present Ash Quantity = × 10 × 106 × 7,000 = 576,500 lb
12,750 yr
0.105
(Coal A) Future Ash Quantity = × 10 × 10 6 × 7,000 = 668,200 lb
11,000 yr
Where:
0.105 = lb of ash/lb of coal

12,750 = BTU/lb of coal
11,000 = BTU/lb of coal
10x106 = BTU/hr
7,000 = hr/yr
So the amount of ash will increase by almost 92,000 lb/yr. Examples 1 & 2
illustrate that the quantity of both ash and fuel handled in the combustion
system is inversely related to the heating value.
3.3.2.2 Sulfur Content

Although sulfur content is a chemical constituent included in an ultimate analysis,
most market analyses report sulfur because coal and oil markets are based on fuel
sulfur content. Permits that limit fuel sulfur content control SO2 emissions from most
stationary sources. So, for example, a source operator will specify oil with less than
1% sulfur because that limit is written in the air permit for the source.
3.3.2.3 Proximate Analysis Data

Proximate analyses are used to characterize coal and other solid fuels. The data are
measured using ASTM procedure D271. In addition to heating value and sulfur, the

proximate analysis specifically includes fixed carbon, volatile matter and moisture.
The analysis procedure involves heating a sample and measuring the weight loss at
several temperatures. Very gentle heating or placing the sample in a dry atmosphere
will drive off the surface moisture. Further heating drives off inherent moisture that
is more tightly bound. In the case of coal, this is moisture that was present when it
was mined. Raising the temperature further in an oxygen free atmosphere drives off
volatile constituents and essentially all the hydrogen. Finally, by raising the
temperature to about 1400°F with oxygen present, all the carbon is oxidized leaving
just the incombustible ash (mineral matter). Each of these measured values has
significant implications for combustion performance.
3.3.2.4 Moisture Content

The moisture content of the fuel can impact combustion processes in several ways.
First, water increases the amount of fuel that must be fired. A ton of wood with 50%
water is actually only half a ton of dry fuel. Second, the water absorbs energy when it
evaporates during combustion. This increases the amount of fuel that has to be fired
and the water decreases flame temperature. Third, a significant amount of water in
the fuel generates enough steam to cause a measurable increase in the moisture
content in the flue gas. This has implications for emissions measurement and for the
performance of emissions control devices like dust collectors. Water content is
particularly significant in wood and waste fuels where it can constitute more than
50% of the “as received” fuel.
3.3.2.5 Volatile Matter

Volatile matter includes organic compounds that vaporize during the heating of the
fuel entering the combustion chamber. These can include methane, hydrocarbons,
and carbon monoxide. A fuel with a high volatile content will be easier to ignite and
burn in the combustion zone. Most pulverized coal combustion systems require a
certain minimum fuel volatility in order to sustain stable flames and achieve complete
combustion of the fuel.
3.3.2.6 Fixed Carbon

Fixed carbon is the combustible residue remaining after departure of organic
compounds and other volatile matter in the fuel. In a stoker furnace the fixed carbon
levels provide a general indication of the quantity of the fuel that will remain on a
grate and burn as a char. Since solid carbon (char) burns only on the surface of the
fuel particle, combustion of fuel with a high fixed carbon level will be relatively slow.
3.3.2.7 Ash or Mineral Content

A few minerals are volatile at combustion temperatures, but most of the inorganic
fuel constituents remain after combustion in the form of ash. Ash must either be
removed from the bottom of the furnace or collected from the flue gas to control
particulate emissions. Thus the fuel ash content is an indicator of potential air

emissions as well as the amount of solid waste requiring disposal. There are also
implications for the accumulation and management of dust and slag on the interior
surfaces of a furnace, as discussed later in this chapter and in Chapter 7.
3.3.2.8 Fuel Oil Market Properties

In addition to heating value and sulfur content, there are several properties significant
to the fuel oil market.
Oil is purchased by the gallon, and density or specific gravity is needed to convert
heating value from BTU/lb to BTU/gallon. Since the heating value in BTU/lb is
relatively constant for each oil category, the density determines the variation in
BTU/gallon. There is some high density oil on the market that is heavier than water,
meaning water will float on the surface or be suspended in the oil rather than settling
to the bottom of the storage tank. Density is frequently reported as °API gravity,
which is related to specific gravity by the following formula:
141.5
sp.gr@60oF =
API + 131.5
No. 6 oil is a viscous liquid that must be heated to thin it out before it is sprayed into
a combustion chamber. In fact, it frequently is so thick at ambient temperatures that
it has to be heated in order for pumps to transfer it for transportation or to move it to a
burner. The minimum pumping temperature is called the pour point, which is about
the temperature that paraffins begin to condense and form solid crystals in the oil.
Storage tanks must be maintained above this temperature. No.2 oil also has a pour
point, but this is usually well below 0°F.
Oil viscosity is measured and used to determine what minimum temperature is necessary
to fire the oil. The viscosity can be measured in the laboratory at any temperature, but
the oil may have to be heated to make the measurement and a common temperature used
is 122°F. Viscosity is discussed further in Section 3.4.2, Fuel Oils.
Water and sediment can be suspended in small quantities in No.6 oil. Note that the
sample analyzed by a laboratory probably does not include any accumulation of
foreign material at the bottom of the tank. Any significant amount of water and
sediment reported by the lab could be an indicator that there is a significant
accumulation at the bottom of the tank.
3.3.3 Ultimate Analysis

An ultimate analysis is usually performed for the benefit of an air emissions
assessment. The results are used to calculate the F-factor (see Chapter 4), a parameter
essential to reduction of air emissions test data. The ultimate analysis presents the
elemental composition of the fuel as determined by ASTM D3176. The elements
include, but are not necessarily limited to the following:

• Carbon • Nitrogen
• Hydrogen • Sulfur
• Oxygen • Chlorine
In addition to these elements, the ultimate analysis usually includes data on ash and
moisture levels. The moisture levels presented in ultimate analyses usually represent
moisture present on a “moisture free” basis after being gently oven dried. Table 3-5
gives some typical values for the ultimate analysis of coal and oil samples.
Table 3-5. Typical Ultimate Analysis for Coal and No. 6 Oil (% by wt.)
Analyte Coal – As Received Coal – Dry Basis No. 6 Oil
Moisture 3.03 0 nil
Carbon 60.68 62.58 85.78
Hydrogen 4.35 4.49 10.59
Nitrogen 1.12 1.16 0.03
Chlorine 0.12 0.12
Sulfur 4.29 4.42 2.13
Ash 19.18 19.78 0.10
Oxygen (diff.) 7.23 7.45 1.37
Total 100.00 100.00 100.00
Note that only one coal analysis was performed and the second column is derived
from the first by adjusting for zero moisture. The fact that the composition always
totals exactly 100.00% is not fortuitous. The oxygen concentration is determined by
the difference after all other figures, including the total, are entered in the column.
The reported oxygen concentration may be approximately correct but, by definition, it
covers up any errors in the other analysis.
3.3.4 Ash Composition Data

Ash is the solid material remaining after combustion is complete and all the organic
fuel is converted to gas. Coal generates large amounts of ash, but even residual oil
with less than 0.1% ash can generate substantial furnace ash accumulations over a
period of weeks or months. Ash settles in the bottom of the furnace, accumulates on
the walls and heat transfer surfaces, and is carried to the dust collector or stack by the
flue gas. Wherever it goes, it needs to be managed, cleaned off the walls, removed
from hoppers or prevented from causing excess particulate emissions. The amount of
ash to manage is simply the amount introduced with the fuel. For example, 150 tons
of coal with 10% ash content will generate 15 tons of ash in the furnace.

The composition of the ash affects operation of the combustion source as well as
several aspects of ash management. Ash chemistry determines the ash melting
temperature that in turn affects if and where molten sticky ash will accumulate as slag
on the furnace walls. This in turn affects the firing rate and maximum achievable
load. The ash properties directly affect corrosion and erosion rates of boiler surfaces.
The ash composition of the fuel can affect the ability of an electrostatic precipitator or
fabric filter to collect fly ash. For example, the sodium, potassium and sulfur
contents of the ash exert an influence on the dust collecting performance of an
electrostatic precipitator. An additional discussion of these subjects is presented in
Chapters 5, 6, and 7.
Table 3-6. Example of Coal Ash Composition (%)

Mineral Ash No. 1 Ash No. 2
SiO2 38.94 55.93
Al2O3 22.75 25.02
TiO2 1.03 1.15
Fe2O3 30.97 9.73
CaO 1.59 0.81
MgO 0.94 1.07
K2O 1.82 2.36
Na2O 0.46 0.46
SO3 0.31 0.85
P2O5 0.10 0.12
SrO 0.02 0.00
BaO 0.08 0.12
Mn3O4 0.10 0.02
Undetermined 0.89 2.36
Ash compositional data are often presented based on the assumption that all of the
elements are present in an oxide form. A typical composition analysis for two coal
ashes is provided in Table 3-6. This example is analysis of ash prepared in the
laboratory from coal samples. One could also analyze samples of ash taken from the
furnace hoppers or from the dust collector. On a given furnace, the results of these
analyses should be similar in most respects to the results of the coal analysis.
However, the ash concentrations of some of the more volatile constituents, such as
sulfur, chlorine and lead may differ substantially between the fuel ash and the
resulting furnace ash. These differences are discussed in more detail in Chapter 5.
The examples shown here are relevant to the behavior of ash in a pulverized coal
furnace. Analysis for other elements can be requested to serve other purposes. Toxic
elements such as arsenic, beryllium or lead could be requested if there was an interest

in potential air emissions. Note that mercury and its compounds are volatile and are
not likely to be found in ash that is generated by burning away the organic
compounds, either in the laboratory or in an actual furnace. Thus the methods of
sample preparation and analysis may be critical to interpreting the results.
3.3.5 Size Distribution of Solid Fuels

Solid fuel particles burn on the surface, so large particles take longer to burn than small
particles. Hence there is both fundamental and practical significance to the particle size
distribution of solid fuel fed to a combustor. We will briefly discuss the effects of fuel
particle size on typical combustion systems and then describe size analysis.
3.3.5.1 Measuring Fuel Size

ASTM sizing screens can be used by plant personnel to measure fuel size distribution.
These are 8" or 12" pans that have an open wire mesh weave as illustrated in Figure
3-2. The pans are stacked with the largest openings at the top and smaller openings in
each successive screen. The fuel sample is placed in the top pan and the entire stack
is shaken for a standard period of time. After shaking the contents, each screen is
weighed to determine the percentage of the sample in each size range.
Figure 3-2. ASTM Sieves, Separated and Stacked for Shaking
Table 3-7. Standard Sieve Sizes

Mesh Size (mm) Size (inch)
200 0.075 0.003
100 0.15 0.006
50 0.30 0.012
35 0.50 0.020
18 1.00 0.039
10 2.00 0.079
8 2.36 0.093

Standard sieves for testing small particles are made of wire mesh and the mesh number
refers to the number of wires per inch. Thus a No.8 mesh has openings a bit smaller than
1/8 inch. Standard mesh sizes run from No.8 to No.400. Sieves with openings 0.25
inches and larger are designated by the size of the openings and come either with round
holes or wire mesh. A few of the standard sizes are listed in Table 3-7.
3.3.5.2 Stoker Fuel Sizes

In a stoker furnace, air is blown up through a grate that carries a bed of burning fuel.
Combustion of the fuel takes a number of minutes. Maximum coal or wood particle
size is typically on the order of an inch (2 cm), but these furnaces can be designed or
operated with much larger particles. Typically, a facility limits the maximum
allowable particle size either through a fuel purchase agreement or by on-site fuel
processing (crushing, chipping, shredding, etc.). This assures no fuel particles are too
large to burn completely in the available furnace residence time. Stokers may also
limit the amount of fine particulate for two reasons. First, fine particulate burns
rapidly while the stoker is designed and operated for fuel that takes a number of
minutes to burn completely. So too much fine particulate disrupts normal stoker
operation. Second, particles smaller than one or two millimeters (about 1/16 inch),.
95
90
80
70
60
50
40
30
25
20
15
10
8
6
4
3
2
1
200 100 50 40 30 20 16 12 8 .25 .38 .50 .75 50 1 1.50 2
U.S. standard sieve Round hole screen

designation (inches)
Figure 3-3. Example of Coal Size Specification for a Spreader Stoker

but larger than 0.1 mm, tend to get blown out of the furnace before they burn
completely. So fuel with a lot of fine particulate tends to put a lot of unburned carbon
into the fly ash that is captured in the dust collector. This wastes fuel and can make
the dust a fire hazard.
An example specification for coal size distribution in spreader stoker boilers is shown
in Figure 3-3. The percentage of coal in each size range should fall within the shaded
area. In this example, between 14% and 32% of the coal should be nominally 0.25
inch diameter. If more than 32% passes through a 0.25 inch screen, the coal is too
fine. If less than 14% reaches the 0.25 inch screen, the coal is too coarse. The
problem in Example 3 illustrates how the fuel-sizing curves are interpreted.
Example 3: Interpretation of Fuel Sizing Charts
A coal supply to be used for a spreader stoker has a size distribution as

indicated in the following table. Is this material appropriately sized for
use in the boiler if the size curve shown in Figure 3-3 applies?
Coal Size Range, Example 3
Table 3-8. Coal Size Range

Coal Size Range Wt. % of Fuel (passing sieve)
< 3/8 80
< 1/8 25
<20 mesh 8
< 50 mesh 6
Note: Mesh sizes apply to standard ASTM sizing screens
Solution:
80% is less than 3/8 inch, which is outside the required range of 18% to 40%.
More of the coal should be in pieces larger than 3/8 inch.
25% is less than 1/8 inch, which is close to the required range of 8% to 22%
8% is less than 20 mesh, which is in the required range of 2% to 9%.
6% is less than 50 mesh, which is outside the required range of less than 4%.
Too much of the coal is finer than 50 mesh.
Therefore, the example coal has more fines and much less larger sizes than the
desired size distribution. The lack of larger sizes means the coal will burn more
rapidly than intended for this grate, because smaller pieces mean more surface area.
The excess of fines less than 50 mesh means that more partly burned coal will blow
out of the furnace, wasting fuel and increasing particulate emissions.

Having discussed the theory of stoker coal sizing it should be noted that it is common
for coal-fired stokers to burn fuel that lies outside the original size specifications.
The reasons are market based. Properly sized stoker coal frequently sells for
considerably more than readily available supplies that have too much fine material. It
is much cheaper to take what is available than it is to require specially prepared coal.
When dealing with wood or waste-fired stokers, the subject of fuel particle sizing
becomes academic. For the most part these stokers take whatever is supplied. With
wood the key issues involve conveyors, feeders and other hardware that can be
jammed or disrupted by oversize or odd shaped pieces. Once the fuel is in the
furnace, problems with incomplete combustion are addressed by approaches other
than trying to adjust particle size.
Municipal and medical wastes are usually fired in whatever form they arrive at the
facility, in which case size distribution has little meaning. The furnace is fed by a
large grapple that drops huge handfuls of waste directly into the chute that feeds the
grate. The only exceptions are a few municipal waste facilities that prepare and burn
refuse derived fuel (RDF). RDF is made by running municipal waste through a large
shredder that reduces everything to pieces about the size of a credit card. As part of
the process, the facility separates out most of the metal and glass to support recycling.
Conveyors and feeders similar to those managing wood chips handle the resulting
stream of RDF. There are a relatively small number of these facilities, but an
extensive explanation about RDF is beyond the scope of this course.
3.3.5.3 Pulverized Coal Size

Grinding coal to a powder creates pulverized coal. With only a few exceptions it is
blown directly from the grinding mill into the furnace. In order to work well,
pulverized coal furnaces require the powdered fuel to meet specific particle size
criteria. The size distribution of pulverized coal is normally measured with 50, 100,
and 200 mesh screens. As a general rule, 90% to 95% of pulverized coal will pass
through a 200 mesh screen (75µm) although with some types of coal this number can
be as low as 70%. This fine size assures sufficient particle surface area that the fuel
will ignite promptly and burn rapidly.
It is also necessary on most furnaces that the amount of fuel retained on a 50 mesh
screen (0.3 mm) be less than 1%. These particles are too large to burn completely in
the typical pulverized coal furnace. When a pulverized coal furnace produces fly ash
with high carbon levels, coal particles larger than 0.2 to 0.3 mm are usually one of the
primary causes. Operation and maintenance of the coal mills determines the size
distribution.

3.4 FUEL PROPERTIES
3.4.1 Gas Fuels
Gas fuels are easy to burn because, unlike liquid and solid fuels, a gas mixes directly
with air. Another advantage is that, with the exception of sulfur, gas fuels are usually
free of contaminants that form air pollutants. So gas is usually clean burning. The
most common constituents of commercially available gas fuels are methane (CH4)
and propane (C3H8). Small amounts of other gases are usually present in commercial
gas fuels. Acetylene (C2H2) is a gas that is sold in small quantities, usually for use in
cutting or welding applications. In some industrial facilities carbon monoxide (CO)
or hydrogen (H2) are available as a process waste or as surplus fuel; but they are used
on site and not sold commercially. Both landfills and sewage treatment facilities can
generate gas with significant methane content. In a number of cases, this gas is being
collected and piped to a commercial use, such as electric power generation.
In the 19th and early 20th centuries, gas was manufactured locally and distributed
commercially to provide illumination. The manufacturing process heated coal in the
absence of air to create gas, coke and some heavy liquid hydrocarbons. The only
current reminders of these early process plants are hazardous waste sites created when
the liquids were discarded in on-site waste pits. While a few plants may currently be
operating, coal gasification, at the present time, is primarily a research activity.
3.4.1.1 Natural Gas

Natural gas is a naturally occurring fuel that comes from wells and is transported by
pipeline to the end user. Natural gas can also be transported over seas by
compressing and refrigerating it to form liquefied natural gas (LNG). At one time
natural gas was an unwanted byproduct of oil wells and it was flared or incinerated at
the well. This practice disappeared once long distance pipelines made it feasible to
transport it to market. A predicted shortage about 1970 failed to materialize and
today natural gas is used wherever a pipeline makes it available. Most new electric
power plants in the U.S. since about 1985 use natural gas as the primary fuel.
Natural gas, as it comes from the well, is mostly methane, but frequently contains
some impurities that must be scrubbed out. Before the gas is distributed a few ppm of
a mercaptan is added to give the gas an odor. Methane is odorless, but humans can
detect the odor of mercaptans at extremely low concentrations. This use of
mercaptans means that all commercially available natural gas has a few ppm of
sulfur. The natural gas composition data shown in Table 3-9 illustrate the typical
composition of natural gas from several sources. It usually consists of more than
80% methane with a wide variety of other combustible and noncombustible
compounds. The specific constituents present in natural gas depend on the source
and, to a lesser degree, upon the processing to remove hydrogen sulfide and other
reduced sulfur compounds.

The heating value of natural gas is usually assumed to be approximately 1,000 BTU/ft3.
As indicated in Table 3-9, the heating value can vary significantly. When there is a
significant amount of carbon dioxide or nitrogen (N2) present, the heating value is
reduced. If there are substantial amounts of high molecular weight organic compounds
(>C2), the heating value is increased. Accordingly, the heating values (HHVs) for
natural gas can range from less than 850 BTU/ft3 to more than 1500 BTU/ft3.
Table 3-9. Natural Gas Composition and Characteristics

Fuel Composition, Wt. %
Fuel Pennsylvania S. California Ohio Louisiana Oklahoma
CH4 83.40 84.00 93.33 90.00 84.10
H2 – – 1.82 – –
C2H4 – – 0.25 – –
C2H6 15.80 14.80 – 5.00 6.70
CO – – 0.45 – –
CO2 – 0.70 0.22 – 0.80
N2 0.80 0.50 3.40 5.00 8.40
O2 – – 0.35 – –
H2S 0.18
Total 100.00 100.00 100.00 100.00 100.00
S – – 0.34 – –
H2 23.53 23.30 23.20 22.68 20.85
C 75.25 74.72 69.12 69.26 64.84
N2 1.22 0.76 5.76 8.06 12.90
O2 – 1.22 1.58 – 1.41
Total 100.00 100.00 100.00 100.00 100.00
Physical Properties
Sp. Gravity 0.636 0.636 0.567 0.600 0.630
3
BTU/ft 1,129.00 1116.00 964.00 1,002.00 974.00
BTU/lb. fuel 23,170.00 22,904.00 22,077.00 21,824.00 20,160.00
There are four general terms used to describe natural gas fuels:
• Sour gas • Lean gas

• Sweet gas • Wet gas
Sour gas is natural gas that has significant quantities of hydrogen sulfide and other
reduced sulfur compounds. Due to the corrosion problems and health risks associated
with these compounds, the hydrogen sulfide and other reduced sulfur compounds are
removed before the natural gas is commercially distributed. Natural gas that is free of
hydrogen sulfide and other reduced sulfur compounds is referred to as sweet gas.

Lean gas has a very high methane concentration implying that it will have a
moderate-to-low heating value.
Wet gas contains moderate concentrations of high molecular weight organic

compounds. As a result, condensation of organic vapor may result in condensed
organic liquids accumulating in the transport system.
3.4.1.2 Propane and Liquefied Petroleum Gas

Propane (C3H8) gas changes to a liquid when compressed to about 15 psi at typical
ambient temperatures. This makes it a convenient fuel that can be sold by the gallon
and used as a gas. Commercial propane is typically composed of approximately 97%
propane with the remainder being ethane and butane. Commercial propane has a
heating value of 2,540 BTU/ft3 and has inherently low sulfur content due to the
processing steps involved in its preparation. Propane is more expensive than natural
gas, so it is only used in those industrial applications where natural gas is not
available but where a gas fuel is required.
Liquefied Petroleum Gas (LPG) is usually propane, with some other gases that
condense to liquid form at moderate pressures. Examples of gases in commercial
LPG include propane, isobutane, n-butane, pentane, and other paraffinic compounds.
3.4.1.3 Waste Biological Gases

Anaerobic biological processes (those without exposure to air) generate methane and
other combustible gases. Prominent processes include animal digestive tracts, waste
decomposition in closed landfills, septic systems and some sewage treatment
processes. Waste gases from large landfills and from sewage treatment plants have
been collected and used as fuel for engines that drive electric generators. At sites
where the gas is not used it is usually necessary to flare the gas to control off- site
problems such as odors and the uncontrolled accumulation of flammable gasses.
The rate of fuel gas formation in a landfill depends on a number of variables including
the amount of water in the landfill, the landfill cover and degree of oxygen exclusion
from the waste, waste temperature, and the composition of the buried wastes. Typically,
a landfill can generate commercial amounts of fuel gas for about 20 years. Landfill gas is
highly variable and can have methane contents of approximately 50% by volume.
Biogas is a mixture of methane, carbon dioxide, water, and some other organic
compounds formed by the anaerobic digestion of sewage sludge. Biogas usually has a
heating value of about 600 BTU/ft3. Biogas can be used on-site at the sewage treatment
plant, or it can be treated and transported via pipeline to an industrial customer.

3.4.2 Fuel Oils
3.4.2.1 Classification and Characteristics
Fuel oils are classified into several categories based on ASTM standards. A partial
list of the fuel oil specifications used to classify fuel oil is summarized in Table 3-10.
Note that these are limiting values and the market place may be more restrictive.
Although Table 3-10 implies a kind of continuum from light to heavy, in practice
there are two general categories of fuel oil: distillate and residual. Distillate oil
includes No. 1 and No. 2 and it differs radically from the heavier residual oils. The
difference is clear if one understands the refining process.
Table 3-10. ASTM Fuel Oil Specifications

Grade Flash Point, °F (°C) Water and Sulfur Ash Content, Viscosity Specific Gravity
(EC) Sediment, Content, (% by vol.) (centistokes) @60 °F
(% by vol.) (% by vol.)
Min. Max. Max. Max. Min. Max.
No. 1 100 or legal (38) trace 0.5 2.2 0.85
No. 2 100 or legal (38) 0.10 0.5 3.6 0.85
No. 4 130 or legal (55) 0.50 – 0.10 26.4
No. 5 (Light) 130 or legal (55) 1.00 – 0.15 65.0
No. 5 (Hvy) 130 or legal (55) 1.00 – 0.15 194.0
No. 6 140 or legal (60) 2.00 2,200.0
[2]
When a refinery processes crude oil, it begins by distilling off the lighter constituents
that are used to make gasoline, No. 2 oil, lubricating oil, and all the “clean” petroleum
products. What is left is called “residual” or No. 6 oil. Most of contaminants in the
original crude oil stay in the residual oil. Thus the residual oil generally contains
concentrations of ash, nitrogen and sulfur which all contribute to air pollution. So
distillate fuels are relatively clean burning, while residual oil generates much higher
emission levels.
Residual fuel is cheaper than distillate fuel, but it is so viscous that it requires special
handling to use it. No.6 oil must be heated just to pump it, so it is stored in insulated
tanks with some means of heating it. It must be heated to temperatures above about
180°F to thin it sufficiently to atomize. Failure to spray it properly can result in coke
formation (the result of cooking the residual fuel to the point that it is solid carbon) in
the furnace and in the burner hardware. So residual fuel is only used by electric
utilities, industrial, or large commercial facilities where an operator is continuously
present to monitor and maintain the combustion system. Table 3-11 summarizes
some typical properties of common fuel oils

Table 3-11. Typical Properties of Common Fuel Oils
Grade Name Color Sulfur Ash BTU/lgallon Comments
No. 1 Kerosene Light 0.0 – 0.5 Nil 137,000
No. 2 Distillate Amber 0.0 – 0.7 Nil 141,000 Home heating fuel
No. 4 Light Residual Black 0.2 – 2.0 0.0 – 0.1 146,000 Fluid at ambient temperature
No. 6 Residual Black 0.7 – 5.0 0.1 – 0.5 150,000 Heat required to pump or
burn
3.4.2.2 Distillate Oil

No. 1 oil is usually called kerosene. No. 2 is common heating oil. Although it can
have as much as 0.5% sulfur, most No. 2 oil has sulfur levels less than 0.1%. Diesel
engine fuel is made by blending No.2 oil with kerosene. The proportions vary by
season – cold weather fuel has more kerosene to lower the viscosity. The sulfur
content of most diesel fuel has been progressively lowered from 0.5% to 0.05% and at
some point to less than 0.01% (100 ppmw). This reduction was first applied to
highway diesel fuel but will eventually apply to virtually all diesel fuel. The main
reason for this reduction is that sulfur fouls the catalysts that are needed for effective
control of air pollution from diesel engines. Note that reciprocating engines can
operate on residual oil and that most marine diesel engines do just that. Thus diesel
engine emissions in harbor areas may be very different from inland diesel emissions.
3.4.2.3 Residual Oil

Residual oil is what remains after the refinery has removed all the distillate fuel from
the original crude oil. This means that residual contains sulfur, nitrogen and ash
(metals) in significant amounts. Residual oil is much less expensive than distillate
oil. Refiners extract as much distillate product as they can from the crude and still
have a marketable residual. This means they stop removing distillate when sulfur or
viscosity levels of the residual reach a marketable upper limit. There is actually a
small market for petroleum coke. But most residual fuel has to meet a sulfur limit of
1% or less and a viscosity that allows the fuel to be burned at temperatures of 240°F
or less. Most facilities that were once permitted to burn 1.5% or 2.2% sulfur fuel
have had to reduce their SO2 emissions. Since very few oil-fired facilities have SO2
scrubbers, their only choice is to reduce the fuel sulfur content. So residual oil is
considerably lighter now than it was 20 years ago.
The main difference between No. 4 and No. 6 oil is that No. 4 oil does not have to
be kept warm in order to pump it. No.4 oil is basically thinned out No. 6 oil, and a
distributor can produce it by blending No. 2 and No. 6 oils.
No. 5 is a designation that is not commonly used.

Industrial boilers fire No. 6 oil, sometimes termed “Bunker C” oil, in large quantities
for electric power generation and other industrial use. Its density approaches, and can
exceed, the density of water. At ambient temperatures it may resemble bitumen (tar)
used to make highway asphalt. However, it does not evaporate and No. 6 oil spills
are very messy to clean up.
In order to fire heavy oil, the viscosity must be reduced to less than about 200 SSU
(Saybolt Seconds Universal) or 45 centistokes. At or below this level an atomizer can
generate a fine spray that will burn rapidly. Above these levels, the nozzle will generate
a coarse spray with large droplets, which do not completely burn in the combustion zone.
Burner operators use the viscosity from a fuel analysis and a special ASTM chart shown
in Figure 3-4 to determine the minimum oil temperature. Any oil has a viscosity-
0 10 20 30 40 50 60 70 80 90 100 110 120130 140 150 160 170 180 190 200 220 240 260 280 300
10000 100000
5000 50000 A
2000 20000
1000 10000
Maximum Practical Limit for Pumping
500 5000
4000
300 3000
2000
150 1500
1000
75 750
50.2 500
400
32.0 300 Viscosity Range
23.2
for Atomization
200
No. 5 and No. 6 Oil
19.2 150
15.2 100
90
13.7 80
70
12.1 60
55
11.3 50
11.0 45 B
40
Insoluble
5%
35
NO2 NAAQS
Ozone NAAQS 33
Acid Rain 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 220 240 260 280 300
PM2.5 Temperature, Degrees Fahrenheit
Regional Haze
Figure 3-4. ASTM Oil Viscosity-Temperature Relationship

American Society for Testing and Materials (ASTM, West Conshohocken, PA),.

temperature relationship that is a straight line parallel to one of the diagonal lines in the
chart. Viscosity and temperature data from the lab is plotted on the chart and then a
straight line plotted down and to the right until it reaches the desired firing viscosity of
approximately 150 SSU. The firing temperature can then be read on the abscissa. Note
that in order to make the plot a straight line, both the vertical and horizontal scales are
completely non-linear. This is the classic example of creating crooked graph paper in
order to make the data fall on straight lines.
A significant aspect of No. 6 oil is the amount of vanadium and other metals in the
ash. At certain temperatures, V2O5 is active in the catalytic oxidation of sulfur
dioxide to sulfur trioxide. In boilers firing high vanadium content oil, the fraction of
sulfur dioxide oxidized to sulfur trioxide can increase to levels well above those
experienced in boilers firing low vanadium oil. Sulfur trioxide reacts with water at
ambient temperatures to form sulfuric acid and the result is corrosion of the flue gas
system. SO3 concentrations of only about 10 ppm in the flue gas result in the
condensation of sulfuric acid into a highly visible plume downwind of the stack. This
will be discussed further in Chapter 5.
3.4.2.4 Oil Emulsions

Oil emulsions are fuel oils prepared by dispersing fine atomized oil droplets into
water. Oil droplet combination (coalescence) is prevented by using a dispersing
agent. There are two general uses of oil emulsions.
1- A viable liquid can be prepared from oil supplies that are too thick to be readily
usable as fuel. The resulting firing characteristics are similar to a No. 6 oil.
These proposed emulsions could have high sulfur, vanadium, and asphaltene
levels and hence incur pollution control costs that offset their low cost.
2- Emulsions of residual oil have been used to improve combustion and reduce
emissions. Water in the oil reduces flame temperature, which reduces NOx
formation. Water in an oil spray can improve the atomization, which may improve
combustion. But the drawback to this use of emulsions is that there are usually
cheaper and simpler ways to gain the same advantages. Thus oil emulsions to date
have been more of a gimmick than a true advance in combustion technology.
3.4.2.5 Coal-Water Emulsions

Finely ground (highly pulverized) coal can be mixed with water and dispersing agents
to prepare an emulsion suitable for firing in oil burners. This fuel can have properties
similar to No. 6 oil. The storage tanks must be designed to provide mixing to prevent
any settling of the dispersed coal particles. The heating values, sulfur contents, and
ash contents of the coal-water emulsion fuel depend entirely on the coal used to
prepare the emulsion, and that is the difficulty. Although a coal-water emulsion can
be fired through pipes, valves and atomizing nozzles similar to the oil handling
hardware, a number of critical differences are not easily resolved. (1) Coal ash
content is at least 100 times that of oil and boilers designed for oil firing are not

configured to handle such great quantities of ash. The ash will tend to bridge over the
closely spaced tubes. The furnace volume for oil firing is about half that required for
coal, so the maximum load may have to be reduced by a third or more. (2) Most of
the coal ash will be carried by the flue gas, meaning that a good dust collector is
required to meet particulate emissions standards. Very few oil-fired boilers are built
with dust collectors. (3) The coal has to be pulverized to make an emulsion, so why
not blow it directly into the furnace and not bother with the emulsion?
The idea of coal-water emulsions received serious consideration during the oil crisis
in the 1970s when there were fears of ever escalating oil prices and ongoing
shortages. But coal-water emulsions hold little appeal as long as oil and natural gas
are readily available and relatively inexpensive.
3.4.3 Coal
3.4.3.1 Classification and Characteristics
Coal continues to be the dominant fuel used for power generation and industrial
process kilns and furnaces, and is readily available in most areas of the U.S. All coals
carry significant amounts of inorganic contaminants: 5% to 25% ash, 0.5% to 5%
sulfur, and 1% to 2% nitrogen. There are four distinct classes (ranks) of coal having
progressively lower carbon contents and progressively higher moisture levels.
Table 3-12. ASTM (D388) Specifications for Coal (Ash and Moisture Free Basis)
Class Group Fixed Carbon Volatile Matter Heating Value Agglomerating

% by wt. % by wt. BTU/lb Character
Anthracite Meta anthracite ≥98 <2
Anthracite ≥92 >98 ≥2 <8 Nonagglomerating
Semi anthracite ≥86 >92 ≥8 <14
Bituminous Low Volatile Bituminous ≥78 >86 ≥14 <22
Med. Volatile Bituminous ≥69 >78 ≥22 <31
High Volatile A Bituminous >69 ≥31 ≥14000 Commonly
High Volatile B Bituminous ≥13000 <14000 Agglomerating
≥11500 <13000
High Volatile C Bituminous
≥10500 <11500
Subbitumino Subbituminous A ≥10500 <11500

us
Subbituminous B ≥9500 <10500
Subbituminous C ≥8300 <9500 Nonagglomerating
Lignite Lignite A ≥6300 <8300
Lignite B <6300 Nonagglomerating

Table 3-13. Properties of Example Coals
Analyses (dry
Bituminous Subbituminous Lignite
basis), Anthracite
Pittsburgh Seam Spring Creek North Dakota
% by wt.
Proximate Analysis, % by wt
Volatile Matter 6.40 40.20 43.10 43.60
Moisture 7.70 5.20 24.10 33.30
Fixed Carbon 83.10 50.70 52.20 45.30
Ash 10.50 9.10 5.70 11.10
Ultimate Analysis, % by wt
Hydrogen 1.90 5.10 5.00 4.50
Carbon 83.70 74.00 70.30 63.30
Sulfur 0.70 2.30 0.35 1.10
Nitrogen 0.90 1.60 0.96 1.00
Oxygen 2.30 7.90 17.69 19.00
Ash 10.50 9.10 5.70 11.10
Heating Value
BTU/lb, as received
11,890 12,540 9,190 7,090
Ash Analysis, % by wt
SiO2 51.00 50.58 32.61 29.80
Fe2O3 3.50 17.16 7.53 9.00
TiO2 2.40 1.10 1.57 0.40
Al2O3 34.00 24.62 13.38 10.00
CaO 0.60 1.13 15.12 19.00
MgO 0.30 0.62 4.26 5.00
Na2O 0.74 0.39 7.41 5.80
P2O5 – 0.39 0.44 –
K2O 2.65 1.99 0.87 0.49
SO3 1.38 1.11 14.56 20.85
Cl 51.00 50.58 32.61 29.80
Ash Fusibility Temp, °F, Reducing Conditions
Initial Deformation – 2220 2180 2030
Softening – 2470 2320 2170
Fluid – 2750+ 2370 2300
Ash Fusibility Temp, °F, Oxidizing Conditions
Initial Deformation – 2560 2180 2160
Softening – 2650 2320 2220
Fluid – 2750+ 2700 2300

Beginning with anthracite, which is almost pure carbon, they are:
• Anthracite
• Bituminous
• Subbituminous
• Lignite
Coal is categorized in more detail in accordance with the ASTM specification D388
summarized in Table 3-12. This classification system is based fixed carbon levels or
heating value from proximate analyses. Older anthracite and bituminous coals are
categorized based on fixed carbon and while coals with less than 69% fixed carbon
are categorized based on their heating value (HHV on a mineral free basis).
Examples of coal ultimate and proximate analyses for the four categories of coal are
provided in Table 3-13. These data are intended only to provide a general indication
of the properties of each type of coal. Within each category of coal, there is
considerable variability in the properties. Even coal supplied from a specific mine
also can have considerable variability over time due to spatial differences in the coal
deposit. Coal is inherently a non-homogeneous material.
Anthracite is a hard, brittle coal mined primarily in central Pennsylvania and is slow
burning in most combustion processes due to its low volatile content. Meta-
anthracite, the most aged form of anthracite coal, burns so slowly it is sometimes
blended with other coals with higher volatile contents. It can have very low moisture
levels of less than 3% and is generally low in sulfur. The low sulfur content and its
clean burning characteristics make it the premium coal for residential furnaces. The
heating value of anthracite is moderately high ranging from 12,000 to 13,500 BTU/lb.
(as received).
Bituminous coals have high heating values and moderate to high volatile levels. They
burn well when pulverized and are used extensively in utility power stations in the
Midwest, south, and east. Bituminous coals are well suited for the production of
metallurgical coke because the coal becomes sticky and cohesive when heated in the
absence of air. In fact, the ability to form the sticky bituminous material during coke
formation was the basis for naming this type of coal as “bituminous.” The formation
of sticky, partially swollen material during combustion can cause problems when
bituminous coals are burned in some types of stoker furnaces.
Many of the bituminous coals can have moderate-to-high sulfur levels resulting in
high emissions of sulfur dioxide. Some bituminous coals also have moderate-to-high
mercury levels. The emission of mercury from large combustion sources is an issue
of increasing concern.
Unlike some bituminous coals, the subbituminous coals do not swell appreciably
during heating. The subbituminous coals have high moisture levels and can be
extremely vulnerable to spontaneous combustion in coal storage piles. They have
more ash than bituminous and anthracite coals, but generally has low sulfur levels and
reasonable heating values. Accordingly, they have been substituted for bituminous

coals in some sources to reduce SO2 emissions. However, subbituminous coals are
entirely non-coking.
Lignite is the lowest rank coal that sometimes has recognizable plant debris;
therefore, it has a high moisture level, up to 30% by weight, and is vulnerable to
spontaneous combustion in coal storage piles. Even when the fuel appears to be dry,
the moisture levels can be quite high. This type of coal also has a high volatile
content. Due primarily to the low heating value, it may not be economical to ship
lignite fuel a long distance.
Three main regions of the country produce coal: the Appalachian, interior, and
western regions. Figure 3-5 illustrates the geographic distribution of coal reserves
throughout the U.S.
RANK FIELD
Anthracite
Bituminous Coal
Subbituminous Coal
Lignite
Figure 3-5. Geographic Distribution of Coal Reserves

Source: DOE, IEA
As this figure indicates, anthracite is available primarily in eastern Pennsylvania.
Bituminous and subbituminous coals are largely available in numerous coal seams

present in western Pennsylvania, West Virginia, Kentucky, Ohio, and Illinois.
Subbituminous coals are available throughout the Midwest, including Wyoming and
Colorado. Lignite is available in portions of the Dakotas and Montana.
Due to the cost associated with shipping coal, most plants use coal supplies that are
relatively close. Accordingly, most bituminous coal combustion occurs at plants in
the Midwest, East, and South, while western boilers and industrial combustion
processes use primarily subbituminous and lignite coals. There are also some
bituminous coal supplies in the West as indicated in Figure 3-5.
The low sulfur content of subbituminous and lignite coals have become increasingly
attractive to some utility boiler operators in the Midwest and East. In some cases,
these coals are economically attractive because they allow compliance with SO2
emission requirements without the cost of scrubbers or alternative SO2 control
techniques.
3.4.3.2 Coal Properties Related to Combustion

Operators must consider all of the properties of the coal supplies being considered
when considering fuel changes for SO2 compliance or other purposes. There are a
number of physical and chemical properties of coal that can make their use
impossible in some types of boilers. These properties include, but are not limited to
the following:
• Agglomerating/caking tendencies
• Ash softening and fusion temperatures
• Coal size distribution
• Grindability and friability
• Mercury content
These characteristics are described briefly in the following subsections and include
the ASTM testing procedures used to quantify the levels of the coal characteristics.
Agglomerating/Caking Tendencies
The agglomerating/caking tendencies of coal can be especially important with respect
to some types of bituminous coal-fired in stoker boilers.
The agglomerating/swelling characteristics of coal can be categorized by the free

swell index, also known as the “Coke Button”. This value is determined in
accordance with ASTM procedures D720. This test involves burning a one-gram
sample of coal in a covered crucible under a specific time temperature profile. The
free swell scale ranges from a value of 0 (free burning) to 9 (severe swelling). Coals
with free swell indices less than approximately 5 are usually not subject to the various
stoker-firing problems potentially caused by caking coals. These are termed “free
burning” coals. In these types of coal, the coal particles burn separately and do not
fuse together during heating. When the free swell index is greater than 5, there is

some vulnerability to combustion problems due to the fusing of coal particles that
chokes off air flow through the grate.
Ash Softening and Fusion Temperatures

The ash softening and fusion temperatures indicate the tendency of coal ash to melt or
become sticky in the combustion chamber. Most coal-fired boilers are designed to
handle solid (dry) ash that can be easily removed from the boiler. However, some
coal ashes will soften and liquefy at the operating temperature in some regions of the
boiler. When this type of fuel is burned, the boiler cleaning and ash handling system
are challenged and the boiler may have to reduce load or shut down for cleaning. A
boiler designed with a “wet bottom” is intended to handle the coals that form molten
ash.
ASTM Procedure D1857-68 defines three temperatures that can be used to evaluate
ash behavior in the combustion furnace:
• Initial deformation temperature

• Softening temperature
• Fluid temperature
The initial deformation temperature is the value at which a cone of coal ash first
begins to deform due to softening. The softening temperature is defined as the point
at which the cone has deformed to a hemispherical lump. The fluid temperature is
defined as the temperature at which the ash forms a thin molten puddle.
The ash fusion temperature is the result of complex inorganic ash chemistry. For
blended coal supplies, the ash fusion temperature of the blended coal is difficult to
predict from the ash fusion temperatures of the individual coal supplies. In fact, two
coals – each having a low ash fusion temperature – can conceivably yield a blended
coal with a high ash fusion temperature. The converse is also true. Due to the
complexity of the ash chemistry, it is appropriate to test the blended samples rather
than attempting to calculate the values from the separate results for each coal supply.
Stoker Fuel Size Distributions

The coal size distribution can affect the operation of stoker-fired boilers. Excessive
coal fines can accumulate in piles close to the coal distributor. Oversize particles can
be thrown completely across the boiler and accumulate on the rear walls. In either
case, the size distribution problems create a fuel-air maldistribution that contributes to
increased air pollutant emissions.
Coal Size Terminology

A variety of “common use” terms have been applied to coal sizing. These are defined
for each major category of coal. The size terms for bituminous coal are provided in
Table 3-14. (Different terms exist for anthracite, subbituminous, and lignite coals.)

Table 3-14. Common Use Size Distribution Terms for Bituminous Coal
Coal Category Size Description Size Characteristics
Bituminous Run-of-mine No size segregation
Run-of-mine 8 inch Lumps greater than 8 inches broken
Lump Larger than 5 inches
Egg Smaller than 5 inches and larger than 2 inches
Nut Smaller than 2 inches and larger than 1.25 inches
Stoker Coal smaller than 1.25 inches and larger than 0.75 inches
Slack Coal smaller than 0.75 inches
Grindability and Friability

Coal that is hard to grind can reduce the throughput of coal pulverizers and thereby
limit the capacity of a pulverized coal-fired boiler. The ability to grind coal is
measured using the Hardgrove grindability value (ASTM D4090). In this procedure,
a 50-gram sample of 16 x 30 mesh coal is ground for a standard time period. The
amount of coal less than 200 mesh after grinding is used to calculate the grindability.
The 200-mesh weight fraction is then used to calculate the Hardgrove value.
The friability of coal is measured to determine the vulnerability of coal to produce

excessive quantities of fine coal particles during coal mining, processing, shipping,
and firing. The friability is measured using either a “drop-shatter” test or a tumbler
test. Coal that is highly friable is vulnerable to the formation of fines as the fuel is
transported and fed into the boiler. Friability does not matter when it is fed to a
pulverized coal furnace, but it can cause excess fines to be fed to a stoker.
3.4.3.3 Contaminants, Air Pollution, Cleaning
Mercury content
Mercury is a volatile metal that is released into the exhaust gases during combustion.
It forms elemental mercury, or mercuric oxide (HgO), or mercuric chloride (Hg2Cl2)
depending on temperature and the presence of oxygen and chlorine. Unless there is
specific control equipment to capture it, most of the mercury will penetrate air
pollution control systems such as fabric filters, electrostatic precipitators, and flue gas
desulfurization systems. Once released into the environment, mercury accumulates in
fish as methyl mercury, moves up the food chain and becomes a threat to human
health and welfare.
Mercury is neither created nor destroyed during combustion and it is not collected
efficiently by standard pollution control equipment. Accordingly, one way to
minimize mercury emissions is simply to select fuels with low mercury contents.

While data are limited regarding mercury levels, there is some general information
regarding the levels of mercury in U.S. coals. A summary of typical mercury levels is
provided in Table 3-15.
Note there can be significant variability in the coal mercury levels within a given
region and within a given coal mine. The data in Table 3-15 provides only a general
indication of the regional variability in mercury levels. While the amount of mercury
in coal is small, electric utility plants use so much coal that the mercury emissions can
be several pounds daily.
Table 3-15. Mercury Content in Coal

Higher Mean Mercury Median Mercury Mercury Content
State Coal Rank Heating Value, Concentration, Concentration, Variance,
Btu/lb. ppm (w/w) ppm (w/w) ppm (w/w)
Alabama Bituminous 12,575 0.22 0.18 0.028
Alaska Subbituminous 7,756 0.06 0.05 0.002
Arizona Subbituminous 10,834 0.04 0.05 0.000
Arkansas Lignite 6,220 0.33 0.24 0.075
Colorado Bituminous 11,250 0.06 0.03 0.005
Colorado Subbituminous 9,527 0.06 0.05 0.003
Illinois Bituminous 11,605 0.09 0.06 0.002
Indiana Bituminous 11,386 0.10 0.08 0.006
Iowa Bituminous 99,937 0.16 0.14 0.010
Kansas Bituminous 12,014 0.17 0.14 0.019
Kentucky Bituminous 12,595 0.16 0.11 0.024
Louisiana Lignite 7,955 0.12 0.12 0.000
Maryland Bituminous 12,786 0.33 0.18 0.160
Missouri Bituminous 10,839 0.17 0.15 0.016
Montana Lignite 6,701 0.13 0.10 0.017
Montana Subbituminous 8,124 0.08 0.06 0.003
Montana Bituminous 6,790 0.23 0.23 0.003
New Mexico Subbituminous 9,297 0.06 0.04 0.005
North Dakota Lignite 6,265 0.14 0.11 0.017
Ohio Bituminous 11,948 0.20 0.17 0.022
Oklahoma Bituminous 13,135 0.23 0.11 0.084
Pennsylvania Bituminous 12,759 0.27 0.20 0.069
Tennessee Bituminous 13,364 0.15 0.10 0.024
Texas Lignite 7,628 0.17 0.14 0.018
Utah Bituminous 10,061 0.09 0.05 0.011
Virginia 13,580 0.11 0.07 0.009

3.4.3.4 Coal Cleaning and Sulfur Levels
Ash
Mineral matter in coal can be either inherently part of the organic coal, or separate in
the form of layers within or adjacent to the coal seam. To some extent the amount of
mineral matter in coal can be controlled by careful mining procedures that minimize
the amount of overburden that is accidentally (or carelessly) included with the coal
that is shipped to market. Coal cleaning can remove some of this incidental mineral
matter. There are no economical ways to remove ash that is chemically bound with
the organic coal.
Sulfur
Sulfur is present in coal in three different chemical forms: (1) pyritic sulfur, (2)
organic sulfur, and (3) sulfates. The fraction of the sulfur that is present as pyritic
sulfur is important in some boiler applications because this can contribute to
clinkering and slagging in the boiler. The pyritic content of cement kiln feed is also
closely associated with the sulfur dioxide emissions from the kiln. As the kiln feed is
heated, pyritic sulfur emits sulfur dioxide that is then vented into the atmosphere.
Some pyritic sulfur is present in large particles that can be partially removed during
physical coal cleaning because pyrites are denser than coal.
ASTM procedure D 2492 is used to chemically analyze the sulfur-bearing materials

within the coal. This speciation of the sulfur bearing materials is performed only
when coal cleaning is being evaluated or when there are operating problems
potentially associated with the sulfur forms in the coal.
Coal Cleaning
Much of the ash in coal is rock and dirt from overburden or layers within the coal
deposit. When contaminants are not chemically combined with the coal, physical
processes can potentially remove them. Coal cleaning is usually performed at or
close to the mine to reduce the ash and sulfur content of the fuel. This reduces the
fuel transportation costs and decreases the sulfur dioxide emissions resulting from the
combustion of the coal.
Early in the 20th century coal was cleaned by manually picking rock from the coal as
it moved slowly up a conveyor. Today there are two main techniques used to clean
coal: (1) a washing technique that uses the difference in densities of organic fuel
particles and ash or sulfur particles, and (2) a froth flotation technique that uses the
difference in surface characteristics of organic fuel particles and the ash or sulfur-
containing particles. Coal cleaning can usually reduce the quantity of ash in the coal
by 25% to 50% and the quantity of sulfur by 5% to 25%. However, some fuel loss
will occur if coal cleaning is very aggressive.

Coal washing can also increase the moisture content of the fuel. However, drying the
coal using good drainage techniques or centrifuges can minimize moisture content.
3.5 OTHER SOLID FUELS

3.5.1 Fossil Solid Fuels
Peat
Peat is a highly variable fuel with a heating value as low as 3,000 BTU/lb. and a
moisture level as high as 70%. The color of the fuel can also vary from yellow to
brownish black. The sulfur content of peat is usually less than 1% (dry basis), while
the ash content can vary from 2% to 70% by weight (dry basis). Due primarily to its
very low heating value, peat is rarely used for the types of boilers and manufacturing
processes that are within the scope of this course.
Coke
Coke is a solid, carbonaceous residue that remains after bituminous coal and other
similar coal fuels are heated in the low oxygen environment. Coke usually has a very
low volatile content and fixed carbon levels approaching 90% by weight. Coke
heating values are usually above 14,000 BTU/lb.
Most coke is produced and used by the steel industry. However coke can and is used
in combustion processes such as cement kilns, cement kiln precalciners, industrial
heaters, and industrial dryers: Coke falls into three classifications:
• Metallurgical coke
• Petroleum coke
• Coke breeze
Metallurgical coal has ash characteristics compatible with steel composition and it is
used to make metallurgical coke. A coke plant is built and operated as part of the
steel mill that it serves.
Petroleum coke is a solid residue that remains after crude oil is heated in either the
delayed coked or fluid coking processes. Petroleum coke has volatile matter levels of
less than 12% by weight and fixed carbon levels that vary from 70% to 95% by weight.
The sulfur levels in petroleum coke can range from 2.5% to 8% by weight, while ash
contents of petroleum coke are lower, usually below 3%. The heating value is usually
14,100 to 15,600 BTU/lb, slightly above the levels usually associated with coal.
Coke breeze is physically degraded coke particles that form, in part, due to the
relatively high friability characteristics of coke materials. As much as 40% (by
weight) of coke breeze material will measure 1/8 inch diameter, or less. The ash
levels of coke breeze are usually higher than those of standard coke fuels.

Another form of coke is used in a combustion process not covered in this course;
charcoal briquettes. They are typically made from wood and/or coal and are used in
backyard barbeques.
3.5.2 Wood
A variety of wood materials are used for fuels in utility and industrial boilers, and
some manufacturing process combustion systems. The most significant variables
affecting wood firing characteristics are moisture content and the physical form of the
wood. Wet wood does not burn easily and a stoker furnace must dry the wood before
combustion starts. Wood forms fed to furnaces range from shredded or chipped wood
to sawdust to sander dust. The form and moisture content determine how the fuel is
fed and the design of the combustion system.
Wood fuels derive from the following general types of source:
• Green chips and bark from logging or lumber mill waste
• Construction, demolition and mixed waste that is usually shredded in a hog mill
• Dry chipped or shredded wood waste from a manufacturing plant
• Sander dust from a manufacturing plant.
Green Wood
Green wood chips can come from clean up operations at a logging site or they are
generated at a mill from trimmings and bark. Green wood typically has a moisture
content in the range of 50% to 60%, which results in a heating value in the range of
4,000 to 5,000 BTU/lb. The main inorganic contaminant is likely to be sand or dirt
clinging to the bark. Green wood can be either chipped or shredded. A chipper uses
less power and the blades can easily be kept sharp if the wood is free of metal and
rock. As illustrated in Figure 3-6, chipped wood is often accompanied by small wood
sawdust or similar material that is generated simultaneously.

Figure 3-6. Wood Chips
If logs are stored in salt water, the wood can acquire salt contents of 0.5% to 2% by
weight. During combustion, all of the sodium and chlorine are liberated and emitted
as a sodium oxide (Na2O) fume and hydrochloric acid (HCl).
Mixed Wood Waste

Wood is frequently the largest component of construction and demolition waste. If it
can be separated from most of the inorganic material it may be relatively dry and is
thus an excellent fuel. Mixed wood waste is usually broken up in some kind of
hammer mill (wood hog) that is not damaged by nails, door hinges and other light
hardware that would dull the blades of a chipper. Hogged wood is uneven fragments
that can be limited to less than 4" to 6" in length. If mixed wood waste contains small
amounts of pressure treated or painted wood, there will not be a significant effect on
air emissions. Pressure treated lumber used for exterior construction can have
moderate-to-high levels of arsenic and chromium. If substantial amounts of treated
wood are burned there could be a measurable increase in emission of metals from the
additives in the paints and coatings.
Dry Wood Waste

Dry wood from a facility such as a furniture factory has heating values in the range of
8,000 to 10,000 BTU/lb. Waste wood materials from fiberboard operations can have
higher heating values due to the presence of dried phenol-formaldehyde and urea
binders – values can be in the range of 12,000 to 14,000 BTU/lb. Nitrogen levels in
this type of wood fuel are low, with the possible exception of fiberboard scraps that
can have nitrogen levels of 0.5% to 1% by weight due to the presence of small
quantities of nitrogen-containing binders.

Sawdust and Sander Dust
When fine sawdust or sander dust are collected during the production of dimensional
lumber or furniture they can be fired in suspension similar to pulverized coal.
Sawdust and sander dust usually originate from the dimensional lumber that has been
kiln dried. Accordingly, these fuels have a low moisture content of less than
approximately 10% by weight and a high heating value of more than 8,000 BTU/lb.
Some of the general characteristics of wood fuels are summarized in Table 3-16.
There are substantial differences in the characteristics of hardwood and softwood
fuels and, within each of these two major categories, there are substantial differences
from one species to another.
Table 3-16. Wood Properties (Dry Basis)

Analyses (dry
basis), % by wt Jack Pine Birch Maple Western Hemlock
Proximate Analysis, % by wt
Ash 2.1 2.0 4.3 2.5
Volatile Matter 74.3 78.5 76.1 72.0
Fixed Carbon 23.6 19.2 19.6 25.5
Ultimate Analysis, % by wt
Carbon 54.3 57.4 50.4 53.6
Hydrogen 5.9 6.7 5.9 5.8
Sulfur 0 0 0 0
Nitrogen 0.1 0.3 0.5 0.2
Ash 2.0 1.8 4.1 2.5
Oxygen 38.6 33.8 39.1 37.9
Heating Value
Btu/lb. 8930 8870 8190 8885
Ash Analysis, % by wt
Si02 16.0 3.0 9.9 10.0
Al2O3 6.3 0 3.8 2.1
Fe2O3 5.0 2.9 1.7 1.3
CaO 51.6 58.2 55.5 53.6
CaCO3 4.9 13.0 1.4 9.7
MgO 5.5 4.2 19.4 13.1
MnO 1.6 4.6 1.0 1.2
P2O5 2.8 2.9 1.1 2.1
K2O 4.1 6.6 5.8 4.6

Table 3-16. Wood Properties (Dry Basis) cont.
Analyses (dry
basis), % by wt Jack Pine Birch Maple Western Hemlock
Mn2O 3.1 1.3 2.2 1.1
TiO2 0.2 trace trace trace
SO3 2.6 3.2 1.4 1.4
Ash Fusibility Temp, °F
Initial 2450 2710 2650 2760
Softening 2750 2720 2820 2770
Fluid 2760 2730 2830 2780
Weight
3
lb/ft (bone dry) 29 37 to 44 31 to 42 26 to 29
Source: Babcock & Wilcox, Steam 40th Edition, Page 8-18.
It is apparent that all wood fuels have a high volatile content, low ash content, low
sulfur content, and low nitrogen content. The nitrogen content of wood is
considerably lower than the levels usually present in coal. Accordingly, the NOx
emissions from combustion systems burning wood are lower than those firing coal.
3.5.3 Biomass Fuels

Biomass fuel refers to crops specifically grown for fuel purposes as well as to
agricultural wastes used as fuel. At the present time there are a number of wood-fired
facilities that rely on trees cut specifically for fuel as well as on wood waste. There
are also facilities like sugar cane mills that extract the sugar from the cane and then
burn the waste (bagasse) to produce energy to run the plant. Ethanol and diesel fuel
can be produced from corn or from some agricultural wastes.
Bagasse is a biomass fuel that has been used extensively over the last 30 years. This
is fiber and associated plant material remaining after the sucrose and glucose has been
removed from sugar cane. Bagasse fuels can have moisture levels above 50% by
weight. Ash content can be as high as 2.5%, primarily due to the presence of small
soil particles that are on the surface of the sugar cane after harvesting. A proximate
and ultimate analysis of a typical bagasse fuel is provided in Table 3-17.

Table 3-17. Bagasse Properties
Analyses (as-fired), % by wt Value
Proximate
Moisture 52.0
Volatile Matter 40.2
Fixed Carbon 6.1
Ash 1.7
Analyses (as-fired), % by wt Value
Ultimate
Hydrogen, H2 2.8
Carbon, C 23.4
Sulfur, S trace to 0.6
Nitrogen, N2 0.1
Oxygen, O2 20.0
Moisture, H2O 52.0
Ash 1.7
Heating Value
Btu/lb. 4,000
kJ/kg 9,304
Studies have been conducted on the potential to grow and use fuels such as hybrid
poplars, hybrid willows, silver maples, switch grass, rice straw, bagasse, poplar, and
general vegetation. With the exception of those facilities already operating, most of
these concepts must wait for substantially higher fuel market prices before they are
economically attractive.
If one ignores economic considerations for the moment, biomass fuel has certain
attractive features. (1) Biomass fuel is renewable. Use of biomass fuel reduces our
dependence on finite (but substantial) supplies of fossil fuels. (2) Biomass fuels do
not increase the amount of carbon in the global carbon cycle, so their use does not
contribute to the increase in atmospheric CO2 and global warming. Energy crops are
being actively investigated and developed by the U.S. Department of Energy's
Biofuels Feedstock Development Program (BFDP).
3.5.4 Solid Wastes

The vast majority of our municipal wastes are buried in landfills. However a number
of municipalities have constructed waste-to-energy plants where waste is burned as
fuel to generate electric power and/or steam heat. Municipal and medical wastes are
also burned in incinerators where there is no energy recovery, but this is a
diminishing population. Most incinerators built prior to about 1980 have been shut

down by air pollution concerns and new facilities need energy recovery in order to
compete economically with landfills.
Municipal waste can include anything that a person could conceivably place in a trash
can or put out on the curb. So there is no simple way to summarize what is or is not
present. In communities where recycling is successful, municipal waste will include
less glass, paper and metal than is present in the overall waste stream. Municipal and
medical waste incinerators can inadvertently receive and process a variety of wastes
that are undesirable. These inappropriate materials can be classified into two groups:
(1) waste materials that could generate toxic emissions and (2) waste materials that
could damage the combustion equipment. A partial list of these materials includes
the following:
1. Waste that could emit toxic materials:
a. Asbestos-containing insulation
b. Batteries containing mercury
2. Wastes that could damage combustion equipment:
a. Containers with more than two or three pounds of liquid fuel

or solvent
b. Compressed gasses such as Medical oxygen or acetylene

cylinders
c. Ammunition
d. Coils of steel cable
In a well-designed and operated furnace, most hazardous wastes will be burned up

and any potential emissions controlled. Asbestos is not flammable and the
procedures for managing it do not allow for it to be commingled with municipal
waste. Mercury will slip through the air pollution control equipment unless there are
specific controls to capture mercury. Containers of gas or liquid fuel can explode in
an incinerator. If there is enough fuel, the explosion will damage the furnace. Steel
cable can get tangled in either the feeder mechanism or the ash handling hardware.
Table 3-18 lists some general categories of waste and typical properties. Non-
hazardous wastes are regulated at the State level, so precise definitions are not
attempted. While all of these waste streams have components with fuel value, they
also have components which can be recycled or which have no fuel value.

Table 3-18. Classifications and Properties of Municipal and Medical Wastes
Heating Value
Description Example materials
BTU/lb.
Municipal Solid 4,000 to 6,000 Mixtures of highly combustible waste, such as paper, cardboard,
Waste wood, animal and vegetable wastes from restaurants, hospitals,
and markets. Typically contains 20% - 50% moisture and 15%
to 30% non-combustible solids.
Construction varies Any non-regulated waste from construction & demolition work
and Demolition including wood, roofing, siding, sheet rock, plaster, plumbing,
Debris wiring, fixtures, broken pavement, dirt etc.
Industrial Waste varies Items such as rubber, plastics, wood waste, packaging, etc. from
industrial operations, boiler ashes, etc. BTU values must be
determined for the individual materials to be destroyed.
Medical Waste, 7,000 to 10,000 A mixture of general hospital waste containing non-regulated
Black Bag waste. Includes paper, plastics (both chlorinated and non-
chlorinated), and cardboard. Waste is typically from the
hospital's food service, administration, patient rooms, operating
rooms, and laboratories.
Medical Waste, 8,500 Regulated wastes – usually associated with infectious hazards.
Red Bag Includes waste saturated with blood or body fluids. Includes
sharps – needles and glassware. May include wastes generated
by pharmaceutical firms, research and diagnostic laboratories,
and other health care facilities.
The heating values of municipal waste will vary from approximately 1,000 BTU/lb.
to approximately 8,500 BTU/lb. This variation is due primarily to variation in
moisture content. There can be considerable seasonal variability in the waste
characteristics burned in a municipal incinerator and wet weather will increase the
moisture content. In localities that do not recycle yard waste, grass clippings and wet
leaves will have a major impact on combustor operation.
Medical waste composition is somewhat similar to municipal waste except that it

usually has a fairly low moisture content. It also contains more packaging waste,
which generally has a high heating value. As indicated in Table 3-19, medical waste
components can have heating values that range from less than 100 BTU/lb for high
metal or moisture-containing materials to more than 20,000 BTU/lb for wastes
containing mainly plastics.
Table 3-19. Characteristics of Medical Wastes

Category Bulk density Moisture Content Heat value as-fired,
as-fired, lb/ft3 of component, % BTU/lb.
by wt
Human Anatomical 50-75 70-90 800-3,600
Plastics 5-144 0-1 13,900-20,000
Swabs, Absorbants 5-62 0-30 5,600-12,000
Alcohol Disinfectants 48-62 0-0.2 11,000-14,000

Table 3-19. Characteristics of Medical Wastes (cont.)
Category Bulk density Moisture Content Heat value as-fired,
as-fired, lb/ft3 of component BTU/lb.
weight, % by wt
Animal Infected 30-80 60-90 900-6,400
Anatomical
Glass 175-225 0 0
Beddings, shavings, 20-45 10-50 4,000-8,100
paper, fecal matter
Gauze, pads, swabs,
garments, paper, 5-62 0-30 5,600-12,000
cellulose
Plastics, PVC, syringes 5-144 0-1 9,600-20,000
Sharps, needles 45-50 0-1 60
Fluids, residuals 62-63 80-100 0-2,000
Medical wastes can have high levels of chlorine due to the consumption of polyvinyl
chloride (PVC) plastics. The concentrations of metals such as cadmium, arsenic,
mercury, and lead can also be high.
Small hospital incinerators are rapidly being retired and medical waste is going to
large regional incinerators or into the municipal waste stream. So in the near future,
medical waste will only be a distinct fuel at a few facilities.
3.6 CONCLUSION
This chapter is an overview of fuels, how they are characterized, methods of burning
them and some of the implications for air pollution. Additional information is available
in the references. Subsequent chapters of this text will draw on the information
presented here in presenting methods of combustion analysis and air pollution control.

Review Exercises
1. Which fuel is produced in the greatest amount, in terms of energy, in the U.S.?
a. Oil
b. Natural gas
c. Nuclear
d. Coal
2. Ultimate analysis of a fuel would provide which of the following?

a. Data required by an agent purchasing fuel for a specific plant
b. Amount of moisture absorbed by coal during rail shipment
c. Chemical composition
d. Ash characteristics
3. Obtaining a representative sample for lab analysis becomes more challenging in which
sequence?
a. Natural gas, coal, municipal waste
b. Wood waste, oil, coal
c. Natural gas, coal, oil
d. None of the above
4. Use of the lower heating value rather than the higher heating value will ________.
a exclude the heat of condensation from the BTU/lb value.
b. reduce the figure for BTU/lb.
c. have a greater effect for natural gas than for coal.
d. All of the above.
5. The difference between residual oil and distillate oil is ________.

a. distillate oil evaporates completely when heated and residual will not
b. residual oil contains higher levels of inorganic contaminants than distillate
c. residual oil must be heated to fire it while distillate is fired at ambient temperature
d. All of the above
6. Municipal solid wastes ________.

a. are usually fairly dry and thus burn easily
b. may be composed of 30% non combustible material
c. never include containers of fuel or waste classified as hazardous.
d. are uniformly mixed so it is easy to obtain a representative sample for analysis.

7. Medical wastes will ________.
a. sometimes burn intensely because of the amount of plastic and packaging material
b. always burn poorly because of the amount of wet anatomical remains
c. be safe to handle because all sharps are kept separated in special containers
d. never include ordinary non-infectious waste
8. Coke is a fuel that is ________.

a. used primarily in metal manufacturing such as steel making
b. is manufactured by driving all the volatile material out of coal
c. is composed almost entirely of carbon
d. All of the above
9. Inorganic material in coal ________.

a. is composed mostly of hazardous and toxic metals
b. is composed mostly silicon and metal oxides
c. may include high concentrations of mercury
d. cannot be determined from laboratory analysis
10. A combustion source will normally buy fuel that ________.

a. is marginally compatible with the installed hardware
b. is least expensive and allowed in the operating permit
c. has sulfur content higher than that listed in the facility’s operating permit
d. has a high water content to reduce combustion temperatures

Review Answers
1. d. Coal
2. c. Chemical composition
3. a. Natural gas, coal, municipal waste.
6. b. may be composed of 30% non combustible material
7. a. probably burn intensely because of the amount of plastic and packaging material
8. d. All of the above.
9. b. is mostly silica and metal oxides
10. b. is least expensive and allowed in the operating permit

References
1
Review 2000, available at www.eia.doe.gov/oiaf/aeo/index.html. The DOE is an
excellent source of information about the properties, availability and usage of fuels
in this country.
2
Meyers, Robert A., Editor. Handbook of Energy Technology and Economics,
(John Wiley & Sons, Inc., 1983). The book has major sections on all types of fossil
fuels as well as renewable sources like wind, solar and geothermal power.
3
North American Manufacturing. North American Combustion Handbook, (North
American Manufacturing Company, Cleveland, OH, 1952). This is a good basic
reference for gas and oil properties and utilization.
4
Schmidt, Paul F. Fuel Oil Manual, Fourth Edition, (Industrial Press Inc., New
York, 1985). The book provides descriptions of test methods and a broad
discussion of oil properties in relation to oil use, but does not list ASTM test
method numbers.
5
Yen, T. F. The Role of Trace Metals in Petroleum, (Ann Arbor Science, Ann
Arbor, Michigan, 1982). The book provides some basic data on typical metal
levels, as well as chemical detail relevant to refining oil to control trace metal
content.
6
Schmidt, Richard A. Coal in America: An Encyclopedia of Reserves, Production
and Use, (Coal Week, McGraw-Hill Publishing, 1979). All you ever wanted to
know about coal.
7
American Society for Testing and Materials (ASTM, West Conshohocken, PA),
www.astm.org. Essentially all the test procedures used by laboratories to analyze
fuel samples are adopted and published by ASTM.

Chapter 4: Fundamentals of Combustion
LESSON GOAL
Demonstrate an understanding of the fundamentals of combustion; including the role
of excess air and the quantitative relationship between mass flow and heat release
through successful completion of the chapter review exercises.
STUDENT OBJECTIVES
1. Describe basic combustion chemistry and list the products of combustion.

2. Use the Law of Conservation of Energy to determine firing rate from fuel flow or
visa versa.
3. Use the Law of Conservation of Mass and the definition of excess air to relate
exhaust flow rate to the gas flow rate when the air-fuel equivalence ratio is unity.
4. Define the terms stoichiometric air-fuel ratio, F-factor, and excess air.
5. Calculate the combustor-firing rate (mmBTU/hr) for a combustion source using
measured data for exhaust gas flow rate and O2 concentration.
6. Convert gas volume at a known temperature and pressure to the volume at standard
conditions.
7. Calculate the combustor firing rate for a source fired with a fuel for which there is a
known Fw-factor using stack test data.
8. Calculate the excess air correction factor and the percent excess air from measured
O2 data.
9. Calculate combustor residence time from measured furnace volume and the
measured exhaust flow rate.
10. Calculate the approximate maximum average temperature in the combustion
chamber of a thermal oxidizer using an exhaust O2 measurement.

11. Calculate what fraction of fuel energy goes up the stack using temperature and O2
data.
12. List the major and minor elements of energy lost up the stack and discuss the energy
balance for a boiler.

Chapter 4: Fundamentals of Combustion
4.1 INTRODUCTION
Combustion is a complex science. A typical furnace flame starts with turbulent fluid
mechanics that is then combined with heat transfer, mass transfer and chemical
reactions. The design of combustion systems frequently includes a substantial
understanding of all these sciences. However, it is possible to perform a great deal of
useful analysis with relatively simple tools. This chapter will discuss the complex
aspects of combustion in general terms and present several analytical tools that are
useful due to their simplicity. Most of the analytical techniques are based on the
basic laws of conservation of mass and energy.
This chapter will discuss the basic physical and chemical rules that govern
relationships between combustion parameters. These rules will be simplified
wherever possible so that the calculations can be performed on a hand calculator with
minimal need for complex data reduction. The chapter begins with a general
description of a typical furnace flame, the principles of fluid mechanics and heat
transfer, and the chemical reactions that result. This chapter will further discuss the
complex aspects of combustion in general terms and present many analytical tools
used in combustion source evaluation.
4.1.1 Chemical Balance

Conservation of mass requires that the mass of fuel and air equal the mass of exhaust
products. We can illustrate this with the simple chemical equation for the combustion
of methane. There are the same number of atoms on each side of the equation.
CH4 + 2O2 → HEAT + CO2 + 2H2O (4-1)

(fuel) (air) (exhaust gas)
Methane yields an equation with a simple chemical balance, but a similarly balanced
equation can be written for any hydrocarbon fuel where the relative amounts of
carbon and hydrogen are known. The relative amounts of fuel and air define the
stoichiometric ratio and when they are balanced, as in equation 4-1, the stoichiometric
ratio equals one. Equation 4-1 can be used to illustrate several concepts:

• Since the combustion products are CO2 and water, combustion in its cleanest form
is non-polluting. Most stationary sources come very close to 100% conversion of
fuel to CO2 and H2O.
• We can burn many hazardous organic substances (wastes) to non-polluting end

products for the same reason. Contaminants in the fuel (waste in this case) are the
potential pollutants.
• Combustion forms water, so the flue gases will carry a lot of water vapor. The
effects are felt when water condenses in emission monitoring equipment and in
the methods of calculating emission rates. It also generates confusion about the
composition of a visible plume.
• The mass balance of an equation shows that the amount of oxygen, and hence the
amount of air required, is determined by the amount of fuel burned. So in real
systems the air flow rate must be controlled to match fuel flow rate.
Three elements important to real combustion are missing in this simple equation.
• Air is composed of about 79% nitrogen (N2), so there is a lot of N2 that enters the
combustion with the O2 and is released with the flue gas. This has a huge effect
on the basic size of a combustion system, but in practice is only important when
dealing with oxygen enriched systems. However, one should keep in mind that
the mass ratio of air to fuel is usually more than 10:1. Fuel is a very small part of
the mass flow through the system.
• Real systems don’t operate at a stoichiometric ratio = 1.00. A little extra air is
required for all combustion, so there is extra O2 (and N2) on both sides of the
equation. How much extra air, how we control it and how we measure it are
important subjects we will discuss later.
• Most fuel has contaminants (ash, sulfur, nitrogen, etc.) that may or may not
participate in the combustion, but which will appear in the flue gas as air
pollutants.
4.1.2 Heat Input Rate – Conservation of Energy

Energy must be conserved in a combustion device. Chemical energy in the fuel is
released during combustion and appears as heat. Most of the heat is then transferred
from the exhaust gases to the product (steam, warm air, etc.). Fuel is fired into any
combustion source at a rate sufficient to maintain load; whether load is defined as
megawatts to maintain the voltage in a power grid or heat to maintain the temperature
of a building. Fuel flow varies from a minimum up to a maximum design firing rate.
(Note: The terms “firing rate” and “heat input rate” will be used interchangeably.) A
combustion source has a design capacity, which can be expressed in various forms
such as kilowatts for a power plant, pounds of steam per hour for a boiler or

horsepower for an engine. Underlying any of these design capacity terms is a heat
input or firing rate – normally expressed as millions of BTU per hour (mmBTU/hr).
There is a direct relationship between firing rate and horsepower, megawatts or
whatever the published dimension of the source. In this course a combustion source
will be characterized by the size of its designed heat input rate expressed in
mmBTU/hr.
A new source will achieve its design capacity (megawatts, horsepower, steam flow
rate, etc) at or slightly below its design firing rate. With age and deterioration, the
source may have to increase its firing rate in order to achieve the design capacity.
This loss in efficiency can, in severe cases, amount to 10% or more.
The design heat input rate appears on the permit of most significant combustion
sources. This firing rate is important because it determines which regulations apply.
For example, sources smaller than 250 mmBTU/hr have different rules than larger
sources. The design firing rate is always listed in the specifications of a new source,
but it may be harder to locate for an old existing source. There can be several
definitions for a source such as: “normal maximum”, “maximum continuous rated”
(MCR) or “emergency maximum” that refer either to firing rate or to overall capacity.
Not to add confusion, but older sources, boilers in particular, can be uprated or
derated in such a way as to yield a design heat input different from the original. Thus
the design heat input for permit purposes can be a judgment call.
All furnaces are heat transfer devices so the basic design parameter is heat flux –
normally measured (in the U.S.) in BTU/hr or million BTU/hr (mmBTU/hr). The
capacity of the furnace or boiler to transfer heat is not changed if the fuel changes
from liquid to gas or to a solid. So one should think of the size of a combustion
device in terms of its heat input or firing rate in BTU/hr or, more commonly, in
mmBTU/hr. From the firing rate one can derive the fuel use in units such as:
• cubic feet per hour (gas fuels)
• gallons/hour (oil)
• tons/hr (coal, wood or waste)
Equation 4-2 is the relation between the fuel energy (HHV or heating value), fuel
flow rate and firing rate. One can write similar equations for gas or liquid fuel.
Fuel energy (BTU/lb) * Fuel mass flow (lb/hr) = Firing rate (BTU/hr) (4-2)
Example 4-1.
Determine the oil flow rate to a boiler operating at 85 mmBTU/hr. The typical
heating value of residual oil is 150,000 BTU/gallon.

Solution:
85 * 106 BTU/hr
= 567 gal/hr (9.4 gal/min)
150,000 BTU/gal
One advantage of working with firing rate (mmBTU/hr) rather than fuel flow rate is
that it reduces confusion when dealing with a facility that fires multiple fuels. Other
advantages will appear later when we discuss exhaust flow rates and emissions
measurements. Note that the BTU basis – higher or lower heating value must either
be specified or implicitly understood.
4.1.3 Fuel and Air Flow – Conservation of Mass

Combustion requires fuel and air – the exhaust is the sum as indicated in Equation 4-3.
Fuel mass flow + Air mass flow = Exhaust gas mass flow (4-3)
With rare exception, the mass of air required for combustion is 10 to 15 times the fuel
flow. In other words, the exhaust flow is only a little larger than the inlet air flow on
virtually all combustors. The only exceptions are when firing low grade gas fuels
composed primarily of N2 or CO2 and in combustion systems that use pure oxygen
rather than air.
As noted earlier, the amount of fuel fed to a combustor is governed by systems load
demand. On nearly all regulated combustion devices the load, and hence the fuel
flow, varies with time. Fuel flow varies over the range defined by the minimum
practical firing rate and the design (maximum) firing rate. The amount of air required
is directly proportional to the fuel flow, so it also varies with time as it tracks the fuel
flow. In practice it is possible to leave the air at a fixed level at or above that required
for the maximum firing rate; but a properly operated source will have the air flow
track the fuel flow as closely as possible.
4.1.4 Excess Air

The chemical balance, illustrated in Equation 4-1, is a quantitative link between the
amount of air required and the amount of fuel used in a combustor. The amount of air
chemically required to burn the fuel is frequently called the stoichiometric air flow.
Alternatively, when the fuel and air flow are chemically or stoichiometrically
balanced, we say that the equivalence ratio is equal to one. The equivalence ratio is
defined as “the actual air-fuel ratio divided by theoretical or stoichiometric air-fuel
ratio.” (Caution: Different references may use either air-fuel or fuel-air ratio when
presenting the stoichiometric ratio.)
Real combustion systems operate with more air than the stoichiometric air
requirement in order to avoid the emissions, such as CO, that result from incomplete
combustion. Products of incomplete combustion result when some of the fuel does

not mix with enough air. If one could achieve a completely uniform air-fuel mixture
in the combustion zone, very little excess air would be required. But perfect mixing
is impossible and the overall air flow must be high enough to assure complete
combustion in the most poorly mixed region of the flame. Excess air, as expressed in
Equation 4-4, is that fraction of the total air flow that is not chemically required for
combustion.
Total air flow = Stoichiometric air + Excess air (4-4)
Excess air is normally expressed as a percent of the stoichiometric flow and can be
any value from near zero to several hundred percent.
Excess air is arguably the most important operating variable in most combustion
systems. Part of the reason is illustrated in Figure 4-1, which shows the typical
relationship between excess air and emissions. Products of incomplete combustion
(smoke, CO and VOC) are minimal until the air flow drops below some critical level,
then they increase sharply. The “knee” in this curve defines the minimum practical
excess air level. NOx behaves differently; it usually decreases linearly with decreasing
excess air. This different behavior means that emission control efforts always deal with
a trade-off between NOx and PIC. Lower NOx means more PIC and visa versa. Figure
4-1 illustrates this relationship.
CO & Smoke
Best
Operating
Range
Excess Air or O2
Figure 4-1. NOx and CO Emissions vs. Excess Air
The amount of excess air required by a combustion system to avoid PIC emissions
depends on the design, load and other factors. As a general rule, the better the
system, the less excess air is required to prevent smoke and/or incomplete combustion
– the knee in the curve occurs at a lower excess air level. Table 4-1 lists some typical
excess air ranges and flue gas oxygen levels for common combustion systems.

Table 4-1. Typical Minimum Excess Air Levels
Type of Combustion Excess Air % Oxygen
Suspension firing: gas, oil, pulverized coal 5% - 15% 1% - 3%
Cyclone (crushed coal) 10% - 15% 2% - 3%
Stoker grate: coal, wood, solid waste 30% - 75% 5% - 9%
Fluidized bed 5% - 150% 1% -13%
Combustion Turbine 250% 15%
Lean burn reciprocating engine >5% >1%
Oxygen and excess air levels shown here refer to the exit of the combustion zone and
may be lower than those measured at the stack. The first four systems listed in Table
4-1 are operator controlled; that is, the operator can adjust the excess air level.
Systems that can operate at lower levels than those listed here, without generating
excessive smoke or CO, are doing very well. Those operating at substantially higher
levels probably have room for improvement.
Operation at the minimum practical excess air is desirable for several reasons:
• Lower air flow means a lower stack flow rate – less hot gas going up the stack,
which means improved thermal efficiency;
• NOx emissions are lower at lower excess air levels;
• Maximum load on some boilers is limited by the size of the air fans, so reducing
the excess air requirement increases the maximum load.
So two objectives of good combustion performance – and hence low emissions – are to:
• design and maintain a system which is capable of low excess air levels, and
• continuously operate that system very close to the minimum practical air flow.
Typical data for engines and combustion turbines is also shown in Table 4-1, but the
operator usually has no control over excess air in these systems. Excess air level is
built into the engine or turbine fuel injection system and adjustments over a very
limited range are done at the factory or on site by the manufacturer’s technicians.

4.1.5 Water Vapor in Exhaust Gas
Combustion of the hydrogen in hydrocarbon fuels generates water vapor. The
amount of water vapor is quite substantial - when firing oil the amount (mass) of
water vapor produced is greater than the amount (mass) of oil-fired. With natural gas
the mass of water produced is more than double the mass of gas-fired. Firing a wet
fuel such as wood also adds to the water vapor in the exhaust gas. The typical
volumetric amounts of water vapor in the exhaust gas from some typical combustion
sources are:
• Natural gas 20% by vol.

• Oil 10% by vol.
• Coal 6% by vol.
• Green wood 18% by vol.
The presence of all this water affects the combustion system design as well as
combustion diagnostics and air emissions. One effect, discussed in Chapter 3, is that
the higher heating value (HHV) of the fuel depends on how much water vapor is in the
exhaust gas. HHV is used both in combustion system design and in emission rate
calculations.
When the water condenses it produces several affects. As long as the exhaust gas
temperature remains above the water dew point, the water appears in vapor form. But
condensation occurs locally on cool surfaces and when the gas is cooled in emission
sampling systems. Table 4-2 lists several water vapor concentrations and their
corresponding dew point temperatures.
Table 4-2. Water Saturation Temperature vs. Concentration

Temperature °F Volume % Weight %
32 0.6 0.4
60 1.7 1.1
80 3.4 2.1
100 6.5 4.1
120 11.5 7.5
140 19.7 13.2
This shows that at typical ambient temperatures, air (or exhaust gas) can only hold
about 2% moisture by volume; higher concentrations will condense. With a few
exceptions real system stack gas temperatures are always above 250°F and water
vapor in the flue gas does not condense until the gas leaves the stack and begins to
mix with the atmosphere. One of the exceptions is flue gas sampling systems used to

measure air emissions. Most of these systems cool the flue gas sample to room
temperature in order to analyze it, which means these systems include a condenser to
remove liquid water that would disable the instruments. So they analyze a dry sample
that has 10% to 20% less volume than the original wet sample. The concentration of
the air pollutants in the remaining dry sample gas is higher and this difference makes
the resulting data incompatible with measurements of exhaust velocity and exhaust
flow volume. Chapter 7 provides a more detailed discussion of exhaust sampling.
Exhaust gas moisture also affects flame temperature. The specific heat of water
vapor is about 0.50 BTU/lb/°F, which is roughly twice that of air and of dry, exhaust
gas. As a result, water vapor variation can have a subtle but profound effect on
combustion. Briefly stated, it takes a lot more energy to raise the temperature of wet
flue gas than for dry flue gas. Thus flame temperatures are lower when there is a lot
of water vapor present. This is particularly significant when dealing with NOx
emissions, as discussed in Chapters 5 and 6.
4.2 COMBUSTION MASS FLOWS, STOICHIOMETRY

A lot of information can be derived about a combustion source from measurements of
the exhaust gases in the stack (from a stack test). A principal objective of this section
is to demonstrate the following statement:
Using measured data for exhaust gas flow rate and O2 concentration, the combustor
firing rate (mmBTU/hr of heat input) can be accurately calculated for most sources.
Note that only two measured parameters are required and there is no reference to
other source data. In most cases, the kind of fuel a source is firing does not need to
be known. The calculation can be done on a hand calculator with only a few steps.
The exceptions to this statement are:
• If the source is firing a wet fuel like green wood or municipal solid waste, a
measurement of exhaust gas moisture is also required.
• If the source is a process, such as a cement kiln that adds CO2 or H2O to the
exhaust gases, the interpretation of the stack data is more complex.
There are at least two good reasons why a person dealing with air emissions would
want to be able to use stack test data to determine the source-firing rate.
1- The source firing rate may be unknown or in question based on data from source
operations.
2- Comparison of firing rates determined from (a) source data and (b) stack data is a
powerful quality control check of emissions data determined during a stack test.

Conservation of mass is the primary basis for this section, but we will also make use of
some properties of fuels. Previous versions of this course and many academic authors
have used voluminous equations to reach results; however, this course cuts through the
heavy derivations and focuses on simpler approaches that lead to a better understanding
of the results.
4.2.1 Stoichiometric Air Flow

The air flow that must be supplied to a combustion system is governed by two
parameters: (1) the firing rate, or heat input rate, and (2) the amount of excess air
necessary to assure complete combustion and to prevent smoke. First, consider the
stoichiometric air flow, or the amount chemically required for complete combustion.
For example, use a simple fuel, pure methane, to illustrate the derivation process.
CH4 + 2O2 → CO2 + 2H2O
Equation 4-1 shows that one molecule of methane requires two molecules of oxygen
to burn completely and form one molecule of carbon dioxide and two molecules of
water. Counting molecules gives the volume of gases involved – one molecule of
methane produces three molecules of combustion products. In order to relate the
mass or the weight of these quantities we need to know the atomic and molecular
weights of the constituents. These values are listed in Table 4-3.
Table 4-3. Typical Atomic and Molecular Weights

Constituent Weight Constituent Weight
C 12.011 CO2 44.01
H 1.008 H2O 18.016
O 16.00 CH4 16.042
N 14.007 S 32.06
The mass of the constituents relative to the mass of methane derive from the ratio of
molecular weights to methane as follows:
44.01/16.042 = 2.74 lb CO2 produced per lb CH4

2*18.016/16.042 = 2.246 lb H2O produced per lb CH4
64.00/16.042 = 3.99 lb O2 consumed per lb CH4
This gives the relative mass of each combustion product but does not tell us the total
amount of exhaust gas. Actual combustion takes place in air that is mostly nitrogen.
So there is actually a large amount of nitrogen that simply passes through the
combustion process without chemical participation. The approximate composition of
air is listed approximately in Table 4-4.

Table 4-4. Composition of Air
Compound Actual Vol. Approx. Vol. Approx Wt.
N2 78.08 79.1 76.9
O2 20.95 20.9 23.1
Argon 0.93 0 0
Other 1.0 0 0
Here the “other” category is mostly argon with small amounts of CO2 and other gases.
The ratio of nitrogen to oxygen shows us the mass of nitrogen relative to oxygen:
76.9/23.1 = 3.33 lb N2 per lb O2 (mass ratio)

79.1/20.9 = 3.78 ft3 N2 per ft3 O2 (volume or mole ratio)
Hence the amount of nitrogen present during stoichiometric combustion of methane is
3.33 (lb N2 per lb O2) * 3.99 (lb O2 / lb CH4) = 13.28 lb N2 per lb CH4
This gives all the components of the exhaust gas from stoichiometric combustion of
one pound of methane:
2.74 lb CO2
2.246 lb H2O
13.28 lb N2
-----------
18.27 lb exhaust gas (E.G.) per lb CH4
This is the basis of the standard accounting procedure used to relate mass flow of air
to fuel mass flow. This procedure can be performed for a fuel of any composition;
although the calculations will be more involved than for methane.
However, this derivation can be carried two steps further and reach a much more
useful result. The heating value for pure methane is 23,875 BTU/lb. If the mass of
exhaust gas is divided by the heating value,
18.27 (lb E.G./lb CH4) / 23875 (BTU/lb) = 765 (lb E.G./mmBTU)
the result will be the weight of stoichiometric exhaust gas generated when one
mmBTU of methane is burned. This weight in turn can be converted to a volume by
multiplying it by the density of flue gas.
765 (lb E.G./mmBTU) * 385 (std ft3 E.G./lb-mole E.G.) / 29 (lb E.G./lb-mole E.G.) =
10156 ft3/mmBTU
Where: 29 is the typical molecular weight of flue gas (lb/lb-mole)

385 is the number of cubic feet in a pound mole (ft3/lb-mole)

The result is the volume of exhaust gas generated by burning 1.0 mmBTU of methane
with no excess air. This is a useful form because it relates a commonly measured
parameter, flue gas volume, to a key source parameter – firing rate. It is a number
that was derived using the conservation of mass, in the form of a balanced chemical
reaction, and the composition of the fuel. No other parameters were used; it is
independent of source design and operating parameters.
This parameter, standard cubic feet of gas per million BTU (std ft3/mmBTU or
scf/106 BTU), can be derived from any fuel. In fact, it has been calculated for many
samples of different fuels and the results published. The relationship between
stoichiometric stack flow and firing rate is called, in EPA parlance, the F-factor.
F-factors are useful for several reasons.
• The F-factor is the key that links exhaust flow to firing rate.
• F-factors generally are fairly constant for a given fuel. In some cases they are
nearly constant for different fuels.
• Default F-factors published by the EPA may be as good or better than those
derived from individual fuel analysis.
The availability of reliable F-factors means that one can do a calculation to relate
stack gas volumetric flow and firing rate without having to obtain a fuel analysis or
go through lengthy calculations. In short, the F-factor is a key component of a very
useful tool for analyzing combustor performance.
Table 4-5 lists values tabulated in 40 CFR 60 Appendix A, Method 19.
Table 4-5. "F-Factors" for Various Fuels (scf/mmBTU)
Fuel Types Fd - dry Fw - wet Fc - carbon
Anthracite coal 10100 10540 1970
Bituminous coal 9780 10640 1800
Lignite 9860 11950 1910
Wood 9240 ------ 1830
Oil 9190 10320 1420
Natural gas 8710 10610 1040

Source: 40CFR60 Appendix A, Method 19
In the preceding example for methane gas, it was determined that the total exhaust flow,
including water vapor, which corresponds to Fw, is the wet F-factor. The wet F-factor,

Fw, is the total volume of flue gas (standard cubic feet) generated when one million BTU
of heat is released by combustion with no excess air.
A similar derivation can be used to exclude water vapor to determine the amount of
dry exhaust gas. Fd, the dry F-factor is the volume of dry flue gas (standard cubic
feet) generated when one million BTU of heat is released by combustion with no
excess air. Formulas for calculating Fd and Fw from a fuel analysis are as follows.
106
Fd = × [3.64 (%H) + 1.53 (%C) + 0.57 (%S) + 0.14 (%N) − 0.46 (%O)] (4-5)
HHV
106
Fw = × [5.56 (%H) + 1.53 (%C) + 0.57 (%S) + 0.14 (%N) − 0.46 (%0) + 0.21(%H2O ]
HHV
(4-6)
Where: %H, %C, %S, %N, and %O are concentrations (% by weight)

from an ultimate analysis of fuel and HHV is the higher heating
value in BTU/lb.
In order to calculate Fw, the analysis for the fuel “as fired” that includes the water in
the fuel must be used. Fd can be calculated either from an “as fired” analysis or a dry
fuel analysis provided the matching HHV is used.
Table 4-5 shows a third column, Fc, which is the volume of carbon dioxide generated
when a million BTU of energy is released. In order to use this factor in most
calculations, the CO2 concentration as well as the total flue gas flow rate is needed.
Before starting an example of how to use an F-factor, an explanation of the F-factor

concept is in order. One usually thinks of measuring combustion energy by measuring
the amount of fuel consumed. But combustion requires both fuel and O2 from
combustion air. The F-factor looks at the other side of the combustion equation and
measures how much air is consumed to release the energy. Interestingly enough, the
amount of flue gas generated (which is directly related to and only slightly larger than
the amount of air consumed) is remarkably predictable and nearly constant for various
fuels. In short, if how much oxygen was consumed is known, there is a very good basis
from which to determine the amount of heat released. Equation 4-7 expresses the
relationship between the stoichiometric gas volume and firing rate.
⎛ std ft 3 ⎞ ⎛ mmBTU ⎞ ⎛ std ft 3 ⎞

E.G. Stoich. Flow ⎜ ⎟ = Firing Rate ⎜ ⎟ × Fw ⎜ ⎟ (4-7)
⎝ hr ⎠ ⎝ hr ⎠ ⎝ mmBTU ⎠
And: Fw is taken from Table 4-5 or is derived from a fuel analysis.
Wet F-factors show the total flue gas volume, while dry F-factors show the volume of
flue gas after all the water vapor is removed. Note that the wet F-factors are nearly

constant for coal, oil and gas. Wet fuels, such as lignite and wood, generate more
water vapor that increases the amount of exhaust gas. Since neither type of F-factor
varies much, default values such as those in Table 4-5 may be preferable to F-factors
calculated from a single fuel analysis; especially where there is significant uncertainty
in the lab results. Given that Fw is a readily available “constant”, Equation 4-7 shows
the direct relationship between firing rate and exhaust flow.
In order to relate the actual exhaust flow to firing rate, one must account for excess
air – which is discussed in subsequent paragraphs. But assuming the stoichiometric
exhaust flow is known, the amount if there were zero excess air, it is easy to calculate
the firing rate.
Example 4-2.
Given that the stoichiometric exhaust flow is 45,000 standard cubic feet per minute
(scfm) on a source fired with oil and gas, what is the combustor firing rate?
Solution:
If the relative amounts of oil and gas being fired are known, it is safe to choose a
middle value for Fw and the error will be small. The Fw-factors for oil and gas are
10,320 and 10,610 respectively, so, for this example, pick the rounded-up average of
10,500.
45,000 (ft3/min) * 60 (min/hr) / 10,500 (ft3/mmBTU) = 257 (mmBTU/hr)
Obviously we could just as easily have started with the firing rate and determined the
stoichiometric stack flow.
4.2.2 Excess Air

Under normal operation the air flow to most combustion sources provides
stoichiometric air plus some excess air. Any operation with less than stoichiometric
air assures incomplete combustion – meaning that some partly oxidized components
of the fuel such as CO, will be present in the exhaust gas. A wood stove with the
dampers closed to bank the fire for the night is a good example of an operation with
less than stoichiometric air. Adding excess air assures there is enough air for
complete combustion, although PIC can still be generated if the fuel and air do not
mix completely in the combustion zone. In this section we will discuss how excess
air is measured, and how it is used diagnostically.
4.2.2.1 Excess Air Management

In most stationary sources, the air flow to the combustor is controlled, in part, by
measuring the excess air and maintaining a fixed excess air level. In fact, controlling
excess air exactly is now an integral part of the air pollution control systems used in

automobiles as well as many stationary sources. Theoretically, one could determine
the excess air level in a system by measuring the fuel flow, calculating the air required
for combustion and then subtracting this from the measured air flow. Fortunately,
there is a far simpler and more precise method, and that is to measure the oxygen
concentration (% O2) in the flue gas. Air enters the combustion zone with 20.9% O2
concentration. So whatever amount of O2 remains in the exhaust gas, by definition,
represents the excess air. Using instruments developed since about 1970, O2
concentration in exhaust gas can be measured accurately, reliably and inexpensively.
Note that when calculating air emissions it is standard practice to use O2 measured in a
dry gas sample (% by vol., dry basis).
There is a simple relationship between excess air and flue gas oxygen concentration,
given by Equation 4-8.
% O2
% Excess Air = × 100 (4-8)
20.9 - % O2
This is more of a definition of excess air than a rigorously derived formula, but it is
the form that pervades all the current air pollution data reduction methods. % O2 (by
vol., dry basis) is the amount of oxygen remaining after combustion. 20.9 – %O2 is
the initial minus the final amount of air, which, by conservation of mass, must be the
amount of air consumed in the flame. So the excess air formula is simply the ratio of
the amount of oxygen remaining divided by the amount consumed.
This formula is not exactly the form that results if one started with a balanced
chemical equation and went through a rigorous derivation. An assumption implicit in
Equation 4-8 is that the combustion air volume is the same as the flue gas volume –
which is only approximately correct. The exact formula for excess air includes fuel
composition as well as % O2. Equation 4-8 is simple, it does not depend on fuel
composition and it is widely accepted. However, note that 20.9% is the dry volume
concentration of O2 in the atmosphere, so O2 should be measured dry. In most
sampling systems the gas sample is extracted and cooled before it enters the
instrument, yielding a dry measurement. If O2 is measured wet (hot flue gas with no
condensation or using a dilution type sampling system), the O2 value must be
corrected upward by the ratio of wet to dry flue gas volumes before it is used in
standard data reduction formulas.
Equation 4-8, the relationship between excess air and %O2, is represented in graph
form in Figure 4-2. This figure also shows a series of curves relating CO2 to excess
air for three different fuels. While CO2 can be used to determine excess air, one
needs a fuel analysis to determine the specific relationship.
[CO2 was commonly used in the past to determine relative excess air level and it is
still used in some cases today. It was used in preference to O2 because exhaust gas
was analyzed with an Orsat device – a laboratory apparatus that determined

composition by sequentially absorbing CO2, O2 and CO in chemical solutions and
measuring the associated change in gas volume. CO2 was the first gas analyzed, so it
was easier to measure than O2 for routine determination of excess air. The Orsat is a
cumbersome and, except in the hands of a skilled technician, a highly unreliable
device. Modern instruments developed since about 1970 render the Orsat apparatus
archaic, although EPA Method 3b still calls for it.]
If one has a fuel analysis, O2 and CO2 become duplicate measurements. On any
given source, one can determine one from the other. Figure 4-2 provides one way to
do this; although we have provided CO2 curves for only three specific fuels. The
relationship between CO2 and O2 is codified by the EPA in 40 CFR 60, Appendix A,
Method 3B. In this case, it is used as a data quality control check when measured
data for both O2 and CO2 are available for a specific test. This will be covered in
great detail in Chapter 7.
Excess Air versus O2 and CO2
18
18
Coal - CH0.85
% CO2
16
15.9
Oil - CH1.65
% CO2
14 12.2
Gas - CH3.8
% CO2
% O2
12
20.9 - % O2
10
1
0 20 40 60 80 100 120 140
% Excess Air
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4

Equivalence Ratio
Figure 4-2. Excess Air versus O2 and CO2

Equation 4-8 can be manipulated algebraically to define another commonly used term,
the excess air correction factor that is also called the air-fuel equivalence ratio. If 100%
is added to the excess air, the result is the amount of air relative to the stoichiometric
amount. In other words, zero excess air corresponds to 100% of the stoichiometric
requirement; 15% excess air would be 115% of stoichiometric air flow rate.
The excess air correction factor is defined as:
20.9
E.A. Correction Factor = (4-9)
20.9 - % O2
The lower abscissa in Figure 4-2 shows this term.
The total exhaust flow including excess air is simply the stoichiometric flow times the
excess air correction factor that gives Equation 4-10.
⎛ 20.9 ⎞
Exhaust Flow = Stoich. Flow × ⎜ ⎟ (4-10)
⎝ 20.9 - % O2 ⎠
Used in combination with Equation 4-7 this provides the relationship between
exhaust flow, firing rate and exhaust O2 concentration.
⎛ ft 3 ⎞ 20.9 − O2
Exhaust Flow ⎜ ⎟ ×
⎛ mmBTU ⎞
Firing Rate ⎜ = ⎝ hr ⎠ 20.9
(4-11)
⎟
⎝ hr ⎠ ⎛ ft 3 ⎞
Fw ⎜ ⎟
⎝ mmBTU ⎠
Given any two variables, the third can be calculated. Thus, a source-firing rate can be
calculated from data that is normally measured during a stack emissions test – stack
flow rate and O2. Conversely one can determine the stack flow rate (and thus fan
size, dust collector size, etc.) if the firing rate is known together with a reasonable
estimate of the excess air level.
Example 4-3.
Stack test measurements showing 20,000 scfm total exhaust flow and 4.0% O2.
Calculate the source-firing rate during the test.
Solution:
Calculate the stoichiometric flow rate using Equation 4-10 – the numerator of
Equation 4-11.
⎛ std ft 3 ⎞ 20.9 − 4.0 ⎛ std ft 3 ⎞

20,000 ⎜⎜ ⎟⎟ × = 16,172 ⎜ ⎟
⎝ min ⎠ 20.9 ⎝ min ⎠

Then use equation 4-7 or 4-11 to calculate the firing rate.
⎛ std ft 3 ⎞ 60 min
hr( ) ⎛ mmBTU ⎞
16,172 ⎜ ⎟× = 92.4 ⎜
⎝ min ⎠ 10,500 std ft (
3
mmBTU
⎝ )hr
⎟
⎠
Where: The generic value of 10,500 (std ft3/mmBTU) was used for Fw.
Comparing this calculated firing rate to the reported firing rate is an effective tool for
checking the quality of reported emissions data or to verify the source operating load
during an emissions test.
Oxygen measurement is quite simple, but there are some caveats. O2 instruments are
used either as part of the operating system or to support emissions measurements.
Boiler operators put their gas sampling point as close to the combustion zone as
practical, while instruments used in conjunction with emission monitors are further
downstream, usually in the stack. Two factors deserve note. First, O2 (or any other
flue gas constituent) may not be uniformly distributed in a large duct, so a single
probe can give a value different from a multi-probe or multi-point average. This
problem is prevalent on large boilers with multiple burners and large ducts. Multiple
probes or a traverse of the duct may be required to obtain an approximate average.
Second, an O2 instrument does not distinguish between O2 coming out of the combustion
zone and O2 introduced by air leaks. An O2 measurement just shows the net amount at
that location. As a general rule the excess air increases as the flue gases move farther
from the combustion zone because of air leakage. So an O2 measurement tells the
amount of excess air at the measurement location - values at two locations will be
different depending on the amount air leakage and the degree of stratification. In regard
to emission measurements, it is critically important that excess air (O2 or CO2) be
measured at the same location, and preferably with the same probe, used to measure the
emission species. By doing this, the amount of air dilution is determined at the emission
sampling point and not some other unrelated position.
4.2.3 Water Vapor

Water vapor in exhaust gases originates from hydrogen in the fuel, from water in the
fuel, or from water in the combustion air. All fuels, except nearly pure carbon, generate
more water in the flue gas than typical ambient saturation levels, so condensation and
liquid water are always an issue when sampling or handling exhaust gas. Normally the
water vapor content is measured during any emissions sampling program, but the
amount of moisture can also be estimated from the fuel properties. Estimated water
vapor contents can be compared to measured data as part of a quality control program.

4.2.3.1 Water from Fuel Hydrogen
The amount of water vapor generated by burning the hydrogen in the fuel can be
derived starting with a balanced chemical reaction like Equation 4-1. For example,
combustion of a fuel with average composition of CHy gives the following reaction.
CHy + (1+y/4)(O2 + 3.78N2) –> CO2 + y/2 H2O + (1 + y/4)*3.78*N2
The right side shows y/2 molecules of water vapor. Working some algebra yields the
volume percent of water vapor from fuel hydrogen in the stoichiometric exhaust gas:
y
Vw1 = 2 × 100 (4-12)
4.78 + 1.45 × y
Here the fuel composition has been normalized to one atom of carbon. The hydrogen
content ‘y’ can be calculated from the weight percent carbon and hydrogen in a fuel
analysis using the following relationship:
y = 12 * (%H) / (%C)
Where: 12 is the atomic weight of carbon.
Many fuels have y ≈ 2, which results in a stoichiometric water vapor content of about
13% by volume. In real systems there is excess air present, which reduces or dilutes
the water vapor concentration – meaning that water vapor from fuel hydrogen will
always be lower than given by Equation 4-12. Dividing Vw from Equation 4-12 by
the excess air correction factor (Eqn 4-9) gives the concentration of fuel hydrogen
water vapor in the actual flue gas.
⎛ 20.9 − % O2 ⎞
Vw ( actual exhaust concentration ) = Vw1 × ⎜ ⎟
⎝ 20.9 ⎠
4.2.3.2 Water from Fuel Moisture

The other major source of water vapor in the exhaust gas is moisture in the fuel.
Fuels like wood and municipal waste can be more than 50% water as fired, so they
add a lot of water vapor to the flue gas. The concentration of water vapor (from fuel
moisture) in the exhaust gas can be calculated, but it depends on the fuel heating
value as well as the amount of moisture in the fuel. For wood or municipal waste, the
approximate volume of water in the flue gas resulting from water in the fuel is given
by the formula:

⎛ w ⎞
% Flue Gas Water Vapor = 25 × ⎜ ⎟ (4-13)
⎝ 100 - 0.75 w ⎠
Where: w = % moisture in the fuel as fired and the flue gas has zero excess
air.
For example, 50% moisture in wood fuel will generate about 20% moisture in the flue
gas before air dilution. The moisture from hydrogen in the fuel will add about 13%
more bringing the total to roughly 33%. However, any source fired with wet wood
will operate with at least 50% excess air, and that reduces the water vapor
concentration in the exhaust gas to less than 22%.
Drying operations or sources like cement kilns add water to the flue gas from the
process, so the exhaust gas moisture is higher than the amount generated by the fuel.
In order to estimate the flue gas moisture content from these sources, one needs
information about the process as a basis for a mass balance.
This section has presented the basic concepts behind the amount of water vapor in
combustion exhaust gas. A rigorous presentation has not been attempted because that
is seldom useful in practice. The properties of wet fuels fluctuate enough to frustrate
attempts to perform accurate mass balances. In general, the main need for flue gas
water vapor information is to set up the emissions test and to reduce the data after the
test. However, there are cases where one needs to estimate water vapor content
where measured data is not available or is in question.
4.2.4 Calculating Gas Volume and Density

Analysts who deal with emissions, stack flows and combustion frequently encounter
gas volume calculations. Most of the required calculations are based on the Ideal Gas
Law - written as
PV =nRT
Where: P is pressure in lb/ft2

V is volume in ft3, n is number of moles and
V/n = 385 ft3/mole for gas at 68°F (528°R)
R is the universal gas constant = 1545 ft-lb/(lb-mole °R)
T is temperature in degrees Rankin
While this law is not frequently used in this form, from this rule, one can readily
derive the pressure-volume-temperature relationships that are commonly used.
Anyone who deals with exhaust flow data will, from time to time, have to adjust a flow
rate for changes in temperature and/or pressure. This is a simple calculation that is easily
misunderstood. The basic rule is that gas volume is proportional to temperature and

inversely proportional to pressure. Heating the gas makes it expand. Compressing it
(reducing its volume) raises the pressure. Beyond that the only thing to remember is that
both temperature and pressure must be in absolute units – where 0° is -460°F (-273°C)
and zero pressure is a perfect vacuum. So temperature is measured in °R (°F + 460) or
°K (°C + 273) and pressure in inches of mercury or psia (psi absolute). Expressed in a
formula, gas volume changes obey the rule:
V1 = V2 x (T1/T2) x (P2/P1) (4-14)
Although this relationship can be used for any temperatures and pressures, it is most
commonly used to correct an actual gas volume (i.e., at stack temperature and
pressure) to standard conditions. Standard pressure is normally the average
atmospheric pressure at sea level which is 29.92 inches Hg = 760 mm Hg = 14.7 psia.
Standard temperature varies from one discipline to the next. EPA methods and the air
pollution community use 68°F = 528°R = 293°K = 20°C. Note that the F-factors all
refer to gas volumes at standard conditions.
Measurements of temperature and pressure are not the focus of this course, but
deserve brief comment. When measuring gas flow rate, the pressure is the static
pressure of the gas in the stack or duct. Stack pressure, where emission
measurements are normally made, is usually very close to atmospheric (barometric)
pressure outside the stack. But occasionally measurements are made upstream of a
fan or an emissions control device with an associated pressure change. Stack thermal
draft or very high velocity can slightly lower the pressure at the sampling location
Thus the pressure relative to atmosphere (draft) in the duct should be recorded and
used to determine the absolute pressure where ever the flow is measured. Thus the
sample point static pressure is the sum of the atmospheric pressure and the draft:
Ps = Pbar + ∆Pdraft /13.6
Where: Ps = Sample point static pressure in inches Hg

Pbar = Barometric pressure outside the duct in inches Hg
∆Pdraft = Duct or stack draft in inches of water
13.6 is the difference in density between mercury and water.
In most cases correcting Ps for stack draft has a very small affect on the reported
emissions.
Example 4-4. Correcting gas volume.
Stack flow is measured at 8,000 ft3/min at a temperature of 310°F, a barometric

pressure of 28.67 inches Hg, and a stack draft of –0.45 inches w.g. Determine the
flow rate in standard cfm (std ft3/min).

Solution:
First determine the absolute static pressure of the stack flow.
Ps = 28.67 – 0.45/13.6 = 28.64 inches Hg
Second, use Equation 4-14 where the stack flow is condition No. 1 and standard
conditions are condition No. 2.
Vstd = 8000acfm * [528°R/(460 + 310)°R] * [28.64 inHg/29.92 inHg] = 5251 scfm
Flue gas density changes with pressure and temperature in the opposite direction from
pressure – as volume increases, pressure decreases. For the most part, numeric values
for density are only needed at standard conditions. The formula for gas density at
standard conditions is
ρ(lb/std ft3) = MW(lb/lb-mole) / 385 (std ft3/lb-mole) (4-15)
Where: MW is the molecular weight of the gas.
The molecular weight of dry air is 29.0. This can easily be derived from the ratio of
the weight of the constituents using the composition of the atmosphere (Table 4-4)
and the molecular weight of the constituents (Table 4-3).
MW dry air = (78.08*28 + 20.95*32 + 0.93*40)/(78.08 + 20.95 + 0.93) = 28.96
Where: 40 is the molecular weight of argon
The molecular weight of flue gas is generally close to air, 29.0, and this is usually a
safe assumption when the composition is unknown. The weights of the two
combustion constituents, CO2 and H2O are respectively heavier and lighter than air,
so their combination frequently averages to about 29. The exception is flue gas with
high moisture content, in which case the average molecular weight is lower, which
reduces the gas density. EPA Method 3 for molecular weight, allows the use of the
value 30.0 as a default for dry gas molecular weight when the source is burning coal
oil or gas.
The density of exhaust gas can be calculated with Equation 4-15 using the gas
composition and the weighted average of the molecular weights of individual species
in the same way the density of air was calculated.
Example 4-5.
The exhaust gas from a coal-fired combustion source has measured concentrations of
water vapor = 8% by vol. and O2 = 5% by vol. What is the approximate density of
the exhaust at standard temperature and pressure?

Solution:
An approximate composition of gas is needed. The curves in Figure 4-2 can be used
to infer CO2 concentration for a coal-fired source. 5% O2 means 31% excess air,
which means the CO2 must have been about 14%. Most of the balance of the gas was
nitrogen, so:
%N2 = 100 - 8 - 5 - 14 = 73%
Now use the weighted average of the molecular weights to find the average molecular
weight:
MWavg = [73*28 + 14*44 + 5*32 + 8*18]/100 = 29.6
[This is identical to the equation for the molecular weight (MWd ) of dry flue gas
presented in EPA Method 2.]
Use Equation 4-15 to determine density:
ρ = 29.6 (lb/lb-mole) / 385 (std ft3 / lb-mole) = 0.077 (lb/ std ft3)
When flue gas moisture content is known, the relationship between wet and dry flue
gas is given by an equation presented in EPA Method 2,
MWs = (1-Bws) * Md + 18 * Bws
Where: MWs is molecular weight of (stack) exhaust gas

Bws is the proportion by volume of water in the gas
MWd is molecular weight of dry exhaust gas
4.3 ENTHALPY OF COMBUSTION AND HEATING

VALUES
Enthalpy is the engineering term for the sensible energy content of a substance and is
usually expressed as BTU/lb. Fuel combustion releases stored chemical energy and then
transforms it into the enthalpy of the exhaust gases. As these gases transfer heat to their
surroundings, they cool and their enthalpy drops. In a boiler, the heat is transferred to
water or steam thereby increasing the enthalpy of the water or steam. Tracking the
energy flow in a combustion device can be a useful exercise. This chapter will show how
evaluation of energy flows can be useful in combustion source diagnostics.

4.3.1 Flame Structure – Primary and Dilution Zones
Fire is a very simple, but familiar concept. However, the physical and chemical
phenomenon involved is complex and engineering modeling of combustors always
requires major simplifying assumptions. A detailed discussion will not be attempted
here, but some basic concepts are essential to subsequent sections of this chapter.
Combustion can occur at any temperature, as evidenced by spontaneous combustion

which starts at room temperature and progresses to the point of creating flames. The
chemical reactions are slow at low temperatures, but if the heat is not dissipated, the
fuel can get hot enough to burst into flame. As a general rule, flames involve high
temperatures – typically 2500°F and higher. Cooler fires can exist, but they usually
require some special design effort to make them stable. Whenever a flame temperature
gets too low, the flame usually becomes unstable and goes out. Most of the combustion
reactions take place very rapidly in an initial flame zone that is highly visible; the
exception is a natural gas flame that is barely visible. As the gases move beyond the
initial flame zone, they cool rapidly, but some residual combustion takes place until the
temperature drops below 1400°–2000°F.
Below about 1400°F chemical reactions are too slow to support any flame. Stated
more succinctly, flames are usually hot or not at all. Cool combustion is a special
situation not encountered in most combustors. [Catalytic reactors are a different
concept that does not require a traditional flame.]
Engine Casing
Combustor
Turbine
Nozzle
Fuel Primary Zone Dilution Zone
Injector (3000°F)
Air From Compressor
Engine Center Line
Figure 4-3. Gas Turbine Combustor
A review of two typical combustion systems will be conducted to illustrate some

concepts of flame structure. Figures 4-3 and 4-4 show two combustion systems – one
for a combustion turbine and the other a typical furnace. The gas turbine combustor
depicted in Figure 4-3 consists of a stainless steel “can” with a bunch of holes and a
fuel injector at one end. Air from the compressor flows through the holes into the
combustion zone and the hot combustion gases then flow through the turbine. The
turbine inlet temperature, depending on the engine design and the power level, is

anywhere from about 1200°F to 2400°F – which is less than the temperature of a
satisfactory flame. So the combustor is designed with a primary zone next to the fuel
injector that operates well above 2500°F. The primary zone provides a hot stable
flame. Downstream of the primary zone additional “dilution” air is added to cool the
gases to an acceptable turbine inlet temperature. So the combustor has a hot stable
fire with a moderate exhaust temperature. The holes in the combustor are all
carefully tailored to provide just the right amount of air in the right place to make the
combustor work well and generate an appropriate temperature profile at the turbine
inlet. A gas turbine combustor operates nearly adiabatically where most of the heat
radiated from the flame is carried back into the combustor by the dilution air that
flows around the combustor.
Exhaust
(350°F Typical)
Combustion Air
Ambient Temperature
Windbox
Burner 3000°F Typical
Fuel
Figure 4-4. Boiler Combustor
Figure 4-4 is a basic furnace or boiler. It is similar to the turbine combustor in the
sense that the burner generates a hot stable fire and the furnace exit is much cooler –
typically 2000°F. A major difference is that the combustion gases cool off by
radiating heat to the furnace walls, not by dilution with additional cold air. Another
difference is that a furnace operates at atmospheric pressure rather than several
hundred psi and the residence time in the furnace is one to three seconds versus 15
milliseconds in the turbine combustor. Both boiler and turbine combustors rely on a
hot primary fire to achieve complete combustion and maintain stable operation.
A brief conceptual discussion can relate physical and chemical processes to flame
characteristics. In most commercial combustion systems the fuel mixes with the air
in the flame, as opposed to being mixed prior to ignition. So the rate of combustion is
governed by the slower of two mechanisms: mixing and chemical reaction. The fuel
and air have to be mixed down to the molecular level, because all carbon and
hydrogen molecules must come in contact with oxygen molecules before a chemical
reaction can occur. The chemical reaction rate is very fast at typical flame
temperatures. So most combustion rates are limited by mixing – meaning the size
and character of the flame are controlled by how well the fuel is mixed with the air.

And the mixing, in turn, is controlled by the design of the burner that injects the fuel
and air into the flame zone. The chemical differences in fuel hydrocarbons usually do
not make much difference [provided they are not predominantly halogens such as
carbon tetrachloride which absorb energy]. Designing burners for different fuels
means dealing with the physical properties that control the quality of a liquid fuel
spray, the size of pulverized fuel particles, or other mixing related parameters.
In combustors with slow or imperfect mixing, chemical reaction rates can become a
significant operating limitation. As discussed in the next section, temperatures drop
rapidly after the combustion gases leave the flame zone. If mixing is incomplete and
the temperature drops far enough, the chemical reaction rates slow to the point that
PIC are left over. So it is generally true that combustion is either fast or incomplete.
A closely related concept is that flames either burn hot or they go out – they are either
on or off, not half way. This leads to the next section on flame temperatures.
4.3.2 Combustion Temperature Levels

How hot is the flame in a typical combustor? The answer depends on the burner
design, the location in the flame and perhaps on whether it is instantaneous or time
averaged. Set aside “low NOx” combustors for the moment and consider only
traditional burner flames. In these flames the fuel and air are mixing in the flame as
opposed to being premixed and then ignited in the combustion zone. Most combustor
flames reach peak temperatures in the vicinity of 4000°F at some point in the flame.
By following a fuel particle through the flame, one might note the temperature will
rise rapidly to a peak level as the fuel mixes with air and burns. But this peak
temperature level is very short lived for two reasons: (1) the hot gases continue
mixing with excess air in the combustion zone which cools them and (2) radiative
heat transfer is intense and the gases begin cooling the instant the chemical energy is
released. The points where the fuel reaches stoichiometric proportions with air are
the hottest and, if the flame is steady, will be fixed regions in the flame. Regions
downstream of this point will be cooler. While an attempt to map out flame zone
temperatures will not be made at this time, an upper limit for the peak flame
temperature can be established.
4.3.2.1 Stoichiometric Flame Temperature

If combustion occurs in a closed furnace with very well insulated walls (and the
inside refractory can withstand the temperature) there will be very little heat transfer
from the flame to its surroundings. The resulting temperature will approach the
“adiabatic” temperature. This temperature can be determined from a simple energy
balance. First consider the case of stoichiometric air flow. The hottest possible
adiabatic flame temperature is reached when there is no excess air to dilute or cool
the flame. This is the highest temperature that can occur anywhere or anytime during
combustion. A basic energy balance results when the exhaust gases absorb all the
energy released by fuel combustion.

Fuel energy released = Enthalpy of the exhaust gas (E.G.)
⎛ 29 ⎞ ⎛ lb ⎞ ⎛ BTU ⎞
( )
Q (BTU) = V std ft 3 × ⎜ ⎟ ⎜ 3 ⎟
⎝ 385 ⎠ ⎝ std ft ⎠
× Cp ⎜ º ⎟ × ∆T
⎝ lb F ⎠
( F)
º
(4-16)
Where: Q = Fuel energy released

V = E.G. volume at standard conditions
Cp = E.G. specific heat
∆T = E.G. temperature rise
6
Set Q = 10 BTU and solve for ∆T :
106 385
∆T ( ºF ) = × (4-17)
V × Cp 29
This is one equation with only one unknown variable – ∆ T. Cp, the specific heat of
exhaust gas, is a number that is tabulated and can be taken from a table. Q and V are
coupled, as shown earlier, by chemical balance using the fuel analysis. When Q is
106 BTU, the associated volume of flue gas should be equal to the F-factor, Fw. This
is almost correct, except that the F-factors are derived using the higher heating value
(HHV) of the fuel. The sensible energy balance expressed in Equations 4-16 and 4-
17 does not include the heat of vaporization for water vapor. So we need to use an
adjusted F-factor, Fw` that is based on the lower heating value (LHV). So we define
Fw` = Fw * HHV/LHV
Now use Fw from Table 4-5 and heating values from Table 3-4 and set V = Fw`.
Using these substitutions and the typical values of Fw = 10,320, HHV = 19,000, LHV
= 17,800 and Cp = 0.30 in Equation 4-17, the result is ∆ T = 4017°F
In other words, the theoretical maximum possible combustion temperature for a

typical fuel is roughly the initial temperature plus 4000°F. The value for specific
cases varies considerably because the value of Cp is not constant at high temperatures
and it is strongly dependent on the amount of water vapor in the combustion gases.
Knowing an exact value for the adiabatic flame temperature is seldom useful, but it is
valuable to know (1) the form of Equation 4-17 and (2) the fact that the adiabatic
temperature rise for most flames is in the vicinity 4000°F.
4.3.2.2 Adiabatic Temperature with Excess Gas

The adiabatic flame temperature, as derived in Equation 4-17, is for stoichiometric
combustion because V was obtained from a table of F-factors that, by definition, are
stoichiometric volumes. The same derivation could be used for a flame with excess
air or excess gas, in which case V would be larger than the Fw -factor value.
Increasing the amount of combustion gas, without increasing the amount of energy
released, spreads the energy to more gas and lowers the adiabatic flame temperature.

For example, if we increased the amount of combustion gas by 25%, the temperature
rise would be reduced approximately to
∆ T2  ≈ ∆ T1/1.25 = 0.8* ∆ T1 = 3360°F
“Approximately” is used here because changes in the value of specific heat would
modify this result. One could use Equation 4-17 with applicable values of V and Cp
to get a more accurate value, but the basic concept is that temperature rise is inversely
proportional to the amount of gas involved in combustion.
Extra gas in a combustor can be either air or recirculated exhaust gas. Flue gas
recirculation (FGR) is a NOx control technology that will be discussed in Chapter 6.
Here we will consider the effect of extra (excess) air. Nearly all combustion systems
operate with some excess air and its effect on average flame temperature is easily
predicted. We simply multiply the volume, V = Fw`, in Equation 4-17 by the excess
air correction factor from Equation 4-9.
106 20.9 − % O2 385

∆T = × × (4-18)
Fw` × Cp 20.9 29
If typical values of Fw = 10,500 and Cp = 0.30 were inserted into this equation, the
results would appear as plotted in Figure 4-5. The line left of the peak represents
incomplete combustion and is estimated. The actual peak temperature is uncertain
because Cp is not well defined when the gas begins to ionize. Figure 4-5 is not exact
because a fixed value for Cp was assumed; however, that is not correct, especially at
typical flame temperatures.
6000
5000
TPEAK
4000
Effect of Gas Ionization
3000
2000
1000
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Air-Fuel Equivalence Ratio (A-F)
0 5 8 10 11 12 13 14 15 16
Oxygen (% by volume)
Figure 4-5. Adiabatic Temperature vs. Excess Air

In general the adiabatic flame temperature is inversely proportional to the amount of
gas present, provided there is at least the stoichiometric oxygen requirement. Figure
4-5 shows graphically the dependence of adiabatic flame temperature rise on excess
air. This relationship is not just theoretical; it applies to real systems with low heat
losses such as the exit of a thermal oxidizer or the exit of a gas turbine combustor.
This relationship can also be applied on a microscopic scale to give the upper limit of
the temperature in some region of the flame with a known air-fuel ratio. As seen in
Section 4.3.2.3, the temperature in various regions of the flame can be very different
from the average flame temperature and from the combustor exit temperature.
While the calculation of peak flame temperature tends to be an esoteric exercise, there
can be many benefits in determining the temperature of thermal oxidizers and some
other devices.
Example 4-6. Thermal Oxidizer Temperature
Stack tests are performed on a gas-fired thermal oxidizer such as that shown in Figure
2-16. Contaminated air enters the incinerator from an oven at 250°F. The measured
stack oxygen concentration measured during the emission tests is 14.0%. What was
the maximum possible average temperature in the combustion chamber?
Solution: There are two approaches:
1. A temperature of about 1350°F is the approximate temperature rise for an

adiabatic combustion system operating at 14% O2 (see Figure 4-5). Adding the inlet
temperature gives us: 1350°F + 250°F = 1600°F as the maximum combustion
chamber temperature. The actual temperature would have been slightly lower due to
heat loss to the surroundings.
2. Use Equation 4-18 to calculate the temperature rise. A specific value for Cp can
be determined, as is done in the following section, but for now use a general value of
Cp = 0.30. The adjusted F-factor, using values for natural gas from Table 4-5 and
Table 3-4 is:
Fw` = 10610 * 22,200/20,000 = 11,777 (std ft3/mmBTU)
Inserting these numbers in Equation 4-18 gives:
106 20.9 − 14.0 385

∆T = × × = 1240ºF
11,777 × 0.30 20.9 29
Adding the inlet temperature gives the final temperature:
Tmax = 250° + 1240° = 1510°F

The two approaches yield different results because Figure 4-5 is based on oil fuel and
approach No. 2 used data specific to natural gas. If exhaust gas composition were
used to determine a more exact value of Cp, there would be a third answer. However,
quibbling about the exact value belies the object of the example, which is that: the
average exhaust gas O2 concentration and the maximum possible average combustion
zone temperature are linked and you can determine one from the other. This is useful
in checking the quality of emissions test data for some sources. If the maximum
temperature computed from the O2 concentration is significantly lower than the
reported actual temperature, it means there is an error in either the O2 measurement or
in the temperature measurement. Of course a measurement error could seriously
compromise some or all of the test data.
Note that in this example there could have been a heat exchanger in the system and
the calculation is still valid. However, calculating temperature from oxygen
concentration does require an airtight system that does not take in atmospheric air
between the combustion chamber and the measuring point.
4.3.2.3 Local versus Average Temperature

Figures 4-3 and 4-4 illustrate that temperatures are not uniform in most combustors,
even if there is little heat loss from the flame. The average combustor exit temperature is
lower than the peak combustion temperature for two reasons: (1) radiative heat loss from
the flame, and (2) the fact that fuel mixes with air in the flame. To understand the second
reason, consider a small parcel of fuel (a tiny drop of oil or a puff of gas) and follow its
progress in Figure 4-5. The fuel enters the combustor on the left side as pure fuel mixed
with zero air. As it mixes with air it progresses to the right, eventually reaching the
excess air level at the combustor exit conditions – the average exhaust air-fuel
equivalence ratio which is greater than 1.0. The temperature of this fuel parcel rises as it
burns, but the temperature can never exceed the adiabatic temperature line. However,
since its equivalence ratio will have to pass through 1.0 at some time, the temperature at
that time can reach the peak level, about 4000°F above the air inlet temperature. So when
fuel mixes with air during combustion (diffusion flame), all the fuel can potentially pass
through the adiabatic peak temperature as it burns. This concept is explored further when
we discuss NOx formation. However, the purpose of this discussion here is simply to
show that flames can have very high local temperatures even if the exit temperature or
even the overall average flame temperature is relatively low.
If the fuel and all the air are premixed, then ignited in the combustor, none of the fuel
passes through the point when equivalence ratio equals 1.0 (provided there is some excess
air) and none of the flame can reach the corresponding peak temperature. In a premixed
system the flame temperature can never be higher than the point on the temperature curve
that corresponds with the air-fuel equivalence ratio at the combustor exhaust. So a
premixed flame will not have the hot spots that characterize a diffusion flame.
Figure 4-5 shows us the upper limit of the average exit temperature for any type of
combustor. The higher the excess air level, the lower the average exit temperature.
In a system where the fuel and air are completely premixed the maximum flame

temperature and maximum exhaust temperature are one and the same. If the combustor
injects fuel into the combustion zone, so there is mixing in the flame, then local peak
temperatures can go as high as the adiabatic temperature at the air-fuel equivalence ratio
equals 1.0, which is somewhere around 4000°F above the inlet temperature.
4.3.3 Energy Content of the Flue Gas

Energy from combustion is transferred initially to the combustion gases (also called
exhaust gas or flue gas) and then to the walls and surfaces of the furnace. The total
energy released equals the amount transferred to the surroundings plus the amount
carried by the flue gas exiting the stack. The energy in the flue gas is made up of
three components: (1) the sensible energy that is proportional to the temperature of
the gas, (2) the latent heat of vaporization of the water vapor, and (3) energy in
unburned fuel – products of incomplete combustion.
QEG = QLH + QS + QPIC
Where: QEG = total flue gas energy content

QLH = flue gas latent heat
QS = flue gas sensible heat
QPIC = energy in the products of incomplete combustion
QEG and QLH will be discussed here, while QPIC, which is usually small to negligible,
will be discussed in Chapter 5.
4.3.3.1 Latent Heat

Latent heat energy in the exhaust gas is contained in the water vapor from fuel
moisture and hydrogen combustion. Latent heat energy is only released if the vapor is
condensed to liquid water. This means that latent heat is not usable energy in most
systems. If energy input were measured using the lower heating value of the fuel
(LHV), then latent heat energy would be excluded from consideration. However,
since we normally characterize fuels by their HHV rather than by their LHV, we have
no choice but to account for the latent heat in any energy balance. The amount of
latent heat is about 1000 BTU per pound of water vapor. That is the energy that
would be available if we cooled the gas to about ambient temperature and condensed
most of the vapor into liquid water. Written in an equation, the latent heat flow is:

⎛ BTU ⎞ ⎛ lb ⎞ ⎛ % H2 O ⎞ ⎛ 18 ⎞ ⎛ BTU ⎞
QLH ⎜ ⎟ = WEG ⎜ ⎟×⎜ ⎟×⎜ ⎟ × 1000 ⎜ ⎟ (4-19)
⎝ hr ⎠ ⎝ hr ⎠ ⎝ 100 ⎠ ⎝ 29 ⎠ ⎝ lb ⎠
Where: QLH = flue gas latent heat, BTU/hr

WEG = flue gas mass flow rate, lb/hr
%H2O = concentration of water vapor in the E.G., % by volume
18 = molecular weight of water, lb H2O/lb-mole H2O
29 = molecular weight of E.G., lb E.G./lb-mole E.G.
18/29 = factor to convert water fraction from a volume to a
mass basis
Knowing the absolute value of the latent heat flow is useful in some situations, but
knowing the heat loss relative to total firing rate (QTot) is usually more informative.
If Equation 4-19 is divided by the firing rate given by Equation 4-11, the result is:
QLH % H2O 20.9 F 18

= × × w × (4-20)
QTot 100 20.9 − % O2 1000 385
Note that the exhaust flow rate cancels out and the fraction of total energy carried by
uncondensed water vapor depends primarily on two variables – % H2O and % O2.
In fact, if we put in a typical value of 10,500 for Fw (valid for coal, oil or natural gas),
the result is:
QLH % H2O 20.9

= × × 0.49 (4-21)
QTot 100 20.9 − % O2
This simple approximate equation gives values well within typical measurement
accuracy for most fuels. This equation also yields low values for wet fuels such as
green wood, municipal waste or lignite.
Example 4-7. Energy in uncondensed water vapor
What is the energy carried by uncondensed water in the stack of a gas-fired source
where the exhaust flow has 4.5% O2 and 17% water vapor by volume?
Solution:
QLH 17 20.9
= × × 0.49 = 0.106
QTot 100 20.9 − 4.5
10.6% of the fuel energy (based on higher heating value) is in the form of
uncondensed water vapor. Note that this amount of energy will be carried out the
stack unless the gas can be cooled enough to condense the water. While some

residential furnaces and water heaters capture most of this energy with near ambient
stack temperatures, very few industrial or commercial facilities do so.
The equations presented here are useful to a person using stack test data. The latent
heat content could have been formulated from just a fuel analysis. Normally, fixed or
standard values are assigned to the latent heat stack losses for predictable fuels such
as natural gas, oil and most coals. Table 4-6 lists three fuels and the approximate
amount of latent heat energy derived from fuel hydrogen in the exhaust.
Table 4-6. Approximate Exhaust Latent Heat Energy
Fuel C/H Ratio HHV (BTU/lb) QLH/QT
Natural Gas CH3.8 23,000 0.094
Heavy Oil CH1.7 18,500 0.060
Bitumin. Coal CH0.8 13,500 0.042
Fuel moisture also adds water to the exhaust gas and this affect is not included in the
last column of Table 4-6.
4.3.3.2 Sensible Energy

The sensible energy carried by the flue gas is proportional to temperature difference
between flue gas and ambient air, and can be expressed as:
QS (BTU/hr) = WEG (lb/hr) + Cp (BTU/lb/°F)* ∆T (°F) (4-22)
Where: QS = flue gas sensible heat, BTU/hr

WEG = E.G.mass flow rate, lb/hr
Cp = E.G. specific heat, BTU/lb/°F
∆T = temperature difference between flue gas and ambient air, °F
Gases just beyond the combustion zone are hot and carry most of the combustion
energy. By the time they reach the stack they have cooled and carry relatively little
energy. Divide Equation 4-22 by Equation 4-11, and the fraction of the total fired
energy is present as sensible energy.
QS 20.9 C ×F MWFG
= ∆T × × p 6w × (4-23)
QTot 20.9 − % O2 10 385
where: MWFG is the molecular weight of exhaust gas

[MWFG/385 is the gas density in lb/ft3]
Fw is the wet F-factor (Table 4-5)

Here, as in the case for latent heat, there are two significant variables and several
constants. If the following values are used for the constants that are typical of a
coal-, oil- or gas-fired boiler stack: Cp = 0.30; Fw = 10,500; and MWFG = 29, then
Equation 4-23 reduces to:
QS ∆T 20.9
= × (4-24)
QT 4200 20.9 − % O2
Equations 4-23 and 4-24 show that the sensible heat carried by the exhaust gas
increases when either the temperature or the O2 increase. This is logical because
increases in either temperature or flow rate should increase the energy. Equation 4-24
suggests that when the temperature rise reaches 4200°F at O2 = 0, then QS = QT and
all the combustion energy is converted to sensible energy. In other words, the
implied adiabatic flame temperature is 4200°F. The reason this number differs from
the 4000°F we derived earlier is that Equation 4-24 did not reckon with some of the
heat being absorbed by water vapor. Note that Equations 4-23 and 4-24 are not
intended for use at temperatures approaching typical flame temperatures.
Example 4-8. Exhaust sensible heat
Boiler stack measurements show T = 375°F and O2 = 3.3% on a day when the
ambient temperature is 50°F. What fraction of the fuel energy is wasted in the form
of hot gas going out the stack?
Solution:
QS 375 − 50 20.9
= × = 0.092
QTot 4200 20.9 − 3.3
In this example about 9.2% of the fuel energy is in the form of hot gas going out the
stack.
The equations derived here for latent and sensible heat contents show that a
considerable amount of information can be derived from a few simple measurements.
In many combustion sources, the fuel energy is either used or it goes up the stack.
Thus if we know the amount of energy carried up the stack, we know what fraction
was used. Therefore, we can infer the thermal efficiency of many sources just by
measuring stack temperature and O2 concentration. The subject of source efficiency
is explored further in Chapter 7.
4.3.3.3 Specific Heat

Reference has been made to specific heat several times during the preceding pages,
usually designated by Cp, which stands for specific heat at constant pressure. The
scale for specific heat is defined simultaneously with the definition for a BTU or

calorie. Water is defined to have a specific heat of 1.00 BTU/(lb-°F). The specific
heat for other materials is derived by comparing them to water.
The specific heat of exhaust gas depends on both composition and temperature.
Table 4-7 gives values for several gases at several temperatures. These are values at
the stated temperature. One can also find tables that give average specific heat values
over a range of temperatures.
Table 4-7. Specific Heat [BTU/(lb-°F) or kcal/(kg-°C)]
Temperature Air CO2 Water Vapor
68°F 0.242 0.200 0.445
212 0.244 0.218 0.452
500 0.249 0.245 0.470
1100 0.260 0.285 0.526
2200 0.278 0.315 0.622
3000 0.297 0.325 0.673
3800 0.303 0.330 0.709
One of the first observations about the values in Table 4-7 is that the values for water
vapor are approximately double the values for air and CO2. So exhaust gases with
large amounts of water vapor will have significantly higher specific heat than dry gas.
It is frequently necessary to estimate the approximate water content of exhaust gas in
order to make reasonably accurate energy balances.
Many energy related calculations for combustion sources deal with exhaust gases at
or below about 1500°F. A second observation is that in this temperature range Cp for
dry gas is about 0.25 and Cp for water vapor is about 0.5. These approximations can
be used for most calculations without inducing substantial errors.
Example 4-6.
What is the specific heat of a combustion source exhaust gas with 18% O2 and a stack
temperature of 475°F?
Solution:
0.18 * 0.47 + (1 - 0.18) * 0.25 = 0.290 BTU/(lb-°F)
The specific heat of the exhaust gas from most sources fired with fossil fuels falls in
the range of 0.26 to 0.29. However, it can go considerably higher for the exhaust
from sources fired with very wet fuel such as many bio-fuels or municipal waste.

This means that when wet fuel is introduced, the combustion zone temperature will be
lower and the exhaust will carry more energy up the stack than when firing dry fuel.
4.4 COMBUSTOR SIZE AND RESIDENCE TIME

4.4.1 Combustor Size
The size of any combustion device is directly related to the design firing rate, although
different types of combustion devices are not necessarily comparable. The combustor
volume required depends on several factors.
• Combustor volume is roughly proportional to the fuel firing rate or heat input rate
(BTU/hr).
• Combustor volume is inversely proportional to the operating pressure. For

example, engines have smaller chambers than atmospheric pressure furnaces.
• Combustor volume depends on the fuel mixing/combustion rate. For example,

pulverized coal furnaces are bigger than oil-fired furnaces that are usually bigger
than gas-fired furnaces.
Combustion intensity or thermal loading is defined as the heat release per unit volume and is
generally measured in units of BTU/hr per cubic foot of combustor volume (BTU/(hr-ft3).
Boilers and furnaces generally are designed with thermal loadings in the range of 50,000 to
150,000 BTU/(hr-ft3). Combustion turbines, operating at 10 atmospheres or more of
pressure, have design heat release rates more than ten times as high.
Example 4-9. Furnace Thermal Loading
What is the thermal loading of a furnace with firebox dimensions of 7 ft x 8 ft x 25 ft,

fired at design maximum rate of 120 mmBTU/hr?
Solution:
The furnace volume is: 7 * 8 * 25 = 1400 ft3

Thermal loading is 120(106)/1400 = 85,700 BTU/(hr-ft3)
Knowing the thermal loading of a combustor relative to other similar combustors can
give an indication of the flexibility to modify it. A furnace with a low thermal
loading is more amenable to up-rating than a furnace that is highly loaded. A furnace
with high thermal loading will usually be less amenable to low NOx technologies that
stage the combustion and increase the volume of the flame.

4.4.2 Residence Time
Regulations for certain types of incinerators specify a minimum combustion residence
time. This is based on the philosophy that the wastes, if given long enough at a high
enough temperature, will be destroyed. In practice, residence time is not usually a
critical parameter governing performance, but the concept provides regulatory
reassurance.
Combustor residence time is the average time it takes fuel to transit the combustion
zone. If fuel is injected at t = 0, the resulting exhaust products depart the combustion
zone at t = tres. Residence time is defined and calculated as the furnace volume (ft3)
divided by the gas flow, in actual ft3/second.
tres = Furnace volume (ft3)/Actual gas flow (ft3/second) (4-25)
Example 4-10. Residence Time.
Determine the residence time of a thermal oxidizer with an exhaust flow of 2600 std
ft3/min and an operating temperature of 1450°F in a 350ft3 chamber.
Solution:
First, determine the gas flow at actual ft3 per second.
Actual flow = 2600 * {(1450+460)/528}/60 = 157 acf/sec
Now use Equation 4-25
Residence time = 350/157 = 2.2 seconds
Residence time is directly linked to thermal loading because the gas flow rate is
proportional to the firing rate via the F-factor. Specifying a minimum residence time
assures that the firing rate or flow rate through the device does not exceed a specified
(permitted) value for the size of the device. Of course residence time alone is not
sufficient to assure good combustion performance or the complete destruction of
wastes. This issue is covered in more detail in Chapter 6.
4.5 CONCLUSION
Combustion is such a complex process that combustor design continues to rely
heavily on “cut and try” methods rather than modeling using computers. However,
the combustion process obeys the basic laws of conservation of mass and energy.
Using these basic rules, it is possible to develop several simple tools that can be used
for inspection and diagnosis of typical combustion systems.

Combustion diagnostics has always relied heavily on the analysis of the exhaust
gases. In recent years, interest in air emissions has elevated the science of exhaust
gas analysis. Using a few exhaust or stack measurements and some very simple
calculations, we can determine the source firing rate in addition to the emissions. The
amount of energy going up the stack can be determined, and that, in many cases, is
enough to accurately determine the thermal efficiency of the source. In short, stack
data can usually be used to determine the useful load of the source, which provides a
means to check the quality of measured emissions data.
This chapter has discussed the basic physical or chemical rules used to generate
relationships between combustion parameters. These rules are simplified where
possible so that the calculations can be performed on hand calculators with minimal
need for complex data reduction.

Review Exercises
1. Calculate the approximate density of the exhaust gas from an oil-fired boiler at standard
temperature and pressure if the water vapor content is 10% by volume and O2 = 3.3%
(dry measurement).
a. 0.070
b. 0.074
c. 0.078
d. 0.083
2. What is the maximum possible average temperature in the combustion chamber of a gas fired
thermal oxidizer, when the contaminated air temperature = 150ºF, and the exhaust gas O2 =
14.3% dry volume. (Use the calculation method and assume an average specific heat of 0.30
BTU/lb-°F.)
a. 1265°F
b. 1186°F
c. 1481°F
d. 1317°F
3. What fraction of fuel energy for a coal-fired boiler goes up the stack in the form of hot gas?
When: Stack temperature, T = 400ºF

Exhaust gas O2 = 5%
Ambient Temperature = 75ºF
a. 8.7%
b. 10.2%
c. 11.7%
d. 15.1%
4. Determine the oil flow rate to a boiler.
Where: Firing rate = 90 mmBTU/hr.

Heating Value of Oil = 125,000 BTU/gal.
a. 12 gal/min
b. 720 gal/min
c. 8 gal/min
d. 15 gal/min

5. Determine the flow rate from a combustion source stack in standard cfm.
When: Actual (measured) flow rate = 7,000 ft3/min

Temperature = 310ºF
Atmospheric pressure = 28.00 inches Hg
Stack draft at the test location = -0.05 inches w.g.
a. 6100 ft3/min
b. 10,208 ft3/min
c. 4491 ft3/min
d. 4492 ft3/min
6. Which of the following correctly characterizes the relationship between wet and dry flue gases?
a. More energy is required to raise the temperature of wet flue gas than dry flue gas.
b. More energy is required to raise the temperature of dry flue gas than wet flue gas.
c. Flame temperatures are generally lower when there is very little water vapor present.
d. The specific heat of water vapor is roughly half the value of that for dry exhaust gas.
7. Where is the primary mixing zone located in the following illustration?

a. Location A
b. Location B
c. Location C
Engine Casing
d. Turbine A
Combustor
Turbine
Nozzle
Fuel Primary Zone Dilution Zone
Injector (3000°F)
Air From Compressor C.

B.
Engine Center Line
8. The size of any combustion device is usually directly related to its design firing rate.
The design combustor volume will also depend on which of the following?
a. Fuel moisture
b. Stoichiometric flame temperature
c. Operating pressure
d. Stack height and temperature

9. Which of the following definitions best describes enthalpy?
a. The amount of air chemically required to burn fuel.
b. An engineering term for the sensible energy content of a substance.
c. A process in which water vapor is extracted from fuel moisture by hydrogen combustion.
d. The relationship between stoichiometric stack flow and firing rate.
10. Which of the following accurately describes latent heat?

a. Latent heat is energy produced in the exhaust gas used to power industrial processes.
b. Latent heat is approximately 5000 BTU/lb. of water vapor.
c. Latent heat is usually designated Cp, which stands for specific heat at a constant pressure.
d. Latent heat is not usable energy in most combustion systems.

Review Answers
1. Determine the dry gas composition. Use Figure 4-2 and 3.3% O2 to obtain about 13.5%
CO2. Obtain the nitrogen content by difference: 100 – 3.3% - 13.5% = 83.2% N2. The
molecular weight of the dry exhaust gas is 0.832*28 + 0.033*32 + 0.132*40 = 29.75.
Determine the molecular weight of the wet (total) exhaust gas as 29.75*0.9 + 18*0.10 =
28.58. The density of this gas is 28.58/385 = 0.74 lb/ft3
c. 0.074
2. Use equation 4-18 to determine the maximum possible temperature rise from the exhaust
gas O2 content: ∆T = 106/(10,610*0.30) + (20.9-14.3)/20.9 * 385/29 = 1317°F. We used
an Fw value for natural gas of 10,610 from table 4-5. The adiabatic (maximum average)
temperature is the sum of ∆T + Tinlet = 1467°F
c. 1467°F
3. Equation 4-24 gives a simple approximate formula for the fraction of heat input lost in the
form of hot gas leaving the stack: QS/QT = (400-75)/4200 * 20.9/(50.9-5) = 0.102.
b. 10.2%
4. Dividing the firing rate by the energy content of a gallon of fuel gives:
90 (mmBTU/hr)/0.125 (BTU/gal) = 720 gal/hr.
Dividing this by 60 minutes per hour gives 12 gal/min.
a. 12 gal/min
5. Use equation 4-14 to correct the gas volume from stack (2) to standard (1) conditions:
V1 = 7000 * 520/(460+310) * (28.00 – 0.05/13.6)/29.92 = 4491 std ft3/min.
c. 4491 ft3/min
6. a. More energy is required to raise the temperature of wet flue gas than dry flue gas.
7. a. Location A
8. c. Operating pressure
9. b. An engineering term for the sensible energy content of a substance.
10. d. Latent heat is not usable energy in most combustion systems.

References
Buschmann, J. C., and Larsson, A. C. Worldwide Experience with Selective Catalytic

Reduction. Presented to the Institute of Clean Air Companies (ICAC) Forum ’98,
Durham, NC, March 18-20, 1998.
Clean Coal Technology. Control of Nitrogen Oxide Emissions: Selective Catalytic

Reduction (SRC). Topical Report Number 9. July 1997.
Donnelly, J.R., and Brown, B. JOY/Kawasaki Selective Catalytic Reduction De-NOX

Technology. Presented at the 82nd Annual Meeting of the Air & Waste Management
Association, Anaheim, CA, June 25-30, 1989.
Durilla, M., Chen, J. M., and Speronello, B. K. Composite SCR Catalysts for NOx
Reduction. Paper prepared by Engelhard, Inc. 1990.
Kokkinos, A., Cichanowicz, J.E., Eskinaze, D., Stallings, J., and Offen, G. NOx
Controls for Utility Boilers: Highlights of the EPRI July 1992 Workshop. Air Waste
Management Association, November 1992, pp. 1498-1505.
Sharreef, G.S., Stone, D.K., Ferry, K.R., Johnson, K.L., and Locke, K.S.. Selective
Catalytic Reduction NOx Control for Small Natural Gas-Fired Prime Movers. Paper
92-136.06. Presented at the 85th Annual Meeting of the Air and Waste Management
Association, Kansas City, MO. June 21-26, 1992.
Tonn, D. P., and Uysal, T. A. 2200MW SCR Installation on New Coal-Fired Project.
Presented to the Institute of Clean Air Companies (ICAC) Forum, Durham, NC,
March 18-20, 1998.

NOx Emissions from Cement Manufacturing. EPA-453/R-94-004. March 1994.

NOx Emissions from Glass Manufacturing. EPA-453/R-94-037. March 1994.

NOx Emissions from Process Heaters (Revised). EPA-453/R-93-034. March 1994.
U.S. Environmental Protection Agency, Code of Federal Regulations, 40CFR60

Appendix A – Test Methods, available at www.epa.gov/ttn/emc

Chapter 5: Air Pollution Formation
LESSON GOAL
Demonstrate an understanding of how various gaseous, liquid, solid, and emulsion fuels are
used in combustion sources through successful completion of the chapter review exercises.
STUDENT OBJECTIVES
1. Describe the relationship between various fuel contaminants and the related emissions.
2. List the potential air pollutants resulting from incomplete combustion.
3. Calculate the maximum emission rate of most air pollutants from an ultimate fuel analysis.
4. Describe the formation of sulfuric acid and the resulting air pollution impacts.
5. Discuss the effects of SO3 on the emissions from a boiler fired on No. 6 oil.
6. Describe the formation mechanisms for small (<1µm) and large (>2µm) particles in
the combustion process.
7. Describe why fine combustion particulate has a chemical composition different than
that of the fuel ash.
8. Discuss the effects of atomization on emissions from a No. 6 oil-fired boiler.
9. Describe how the grinding and size distribution of pulverized coal particles affects
the amount of carbon in fly ash (carbon carry over).
10. Describe how and why black smoke is formed.
11. Describe the two primary NOx formation mechanisms.
12. Discuss the typical effect of excess air on NOx emissions and the trade off with
products of incomplete combustion (PIC).
13. Describe the key elements necessary for complete combustion.
14. Describe the formation and control of dioxins and furans.
15. Describe how opacity is affected by the size distribution of particles in a plume.

Chapter 5: Air Pollution Formation
5.1 INTRODUCTION
Combustion sources emit a number of different pollutants that can be divided into
three categories:
• Products of Incomplete Combustion (PIC): black smoke/soot, CO, organic

compounds, and in some cases particulate matter.
• Pollutants resulting from inorganic contaminants in the fuel: SOx, particulates,

HCl, etc.
• NOx: nitrogen oxides from atmospheric N2 and O2 breaking down in the

combustion zone and reacting
The emission rates of some of these pollutants are regulated because excess emissions
could threaten ambient air quality. Note that there are air pollutants such as ozone
which are not emitted from combustion sources, but which are formed in the
atmosphere from combustion pollutants. For example, NOx is regulated in order to
control ozone. A less clear cut example is PM2.5 some of which is emitted directly and
some of which is formed from SOx, NOx, or other gas phase species. Future rules
may regulate various combustion emissions in order to address PM2.5 and/or to
preserve visibility in scenic areas.
The first several sections of this chapter address the first two categories of emissions
listed above. The last category, NOx, occupies the largest section of the chapter.
NOx is a pollutant of considerable concern, because of its relationship to ozone
formation, and because the amount of NOx formed during combustion is substantially
influenced by the design and operation of the combustor.
The purpose of this chapter is to explore air pollutant formation in combustion

sources, while air emission control will be reviewed in the next chapter. For most
pollutants this is a logical separation of topics. However, NOx formation and control
are closely linked when combustion controls are used. This is addressed in part by
providing a modest overlap in these two chapters in the area of NOx emissions.

5.1.1 Actual Emission Rates
The basic chemistry of pollutant formation combined with a mass balance provides a
relatively simple basis for predicting emission rates of some pollutants. Unfortunately,
this approach is too imprecise to be useful for predicting emissions of NOx, any
product of incomplete combustion, and several other pollutants. These emission rate
predictions must be based on data from actual sources. The EPA Compilation of
Emission Factors (AP-42) is the most comprehensive compilation of data currently
available. While it is produced and continually updated by the EPA, the Federal
Government does not take responsibility for the accuracy or applicability of the data.
When a facility underestimates emissions from a specific source it cannot attribute the
mistake to the generic data in AP-42.
5.1.2 Potential Emission Rates

Conservation of mass tells us that what goes into a combustor must come out with the
exhaust gases. All the inorganic species in the fuel come out in some form of
pollutant, usually as gases or particulates. In addition, all combustion sources emit
NO and NO2, however the amount will vary. The sum of these two gases is termed
NOx. Table 5-1 lists the contaminants of two common fuels (No. 6 oil and coal) and
the resulting pollutants. Most sources burn fuel almost completely, but any of them
can emit excessive products of incomplete combustion (PIC) if poorly designed or
improperly operated. Boiler sources do not normally emit PIC in significant
quantities; however, NOx control strategies tend to degrade combustion performance,
leading to a trade off between NOx and PIC. Combustion turbines and reciprocating
engines normally emit small, but potentially significant, amounts of PIC, and here
too, the PIC emissions may be affected by NOx controls.
Table 5-1. Emissions Originating in the Fuel

Fuel Fuel Concentration Pollutant Concentration Primary Method of
Constituent Control
No. 6 Oil Coal Species Conversion
Sulfur(1) 0.5 – 2% 1 – 4% SO2 99% Low sulfur fuel
H2SO4 1% Very low excess air
Ash <0.05% 10% Particulate 20 – 98% Dust collector
PM-10, 2.5 20 – 80% Dust collector
Nitrogen(2) <0.5% 1% NOx 10 – 50% Combustion mod
Chloride (low) (low) HCl 100% Fuel specs
CnHm 98% 85% C, CO, HC 0 – small Combustion tuning
Notes: [1] SO4 in the fuel does not convert to SO2 or sulfuric acid.
[2] Only organic nitrogen contributes to NOx formation.

The column labeled “Conversion” refers to the fraction of the contaminant that, based
on conservation of mass, typically leaves the combustion zone as an air pollutant.
Ash, for instance, can settle to the bottom of the furnace where it can be shoveled out
and this portion is not a potential air pollutant. If we assume 100% conversion, the
calculation of maximum potential emissions using fuel analysis and a mass balance is
straight forward. All combustion source emissions can be expressed in units of
pounds per million BTU (lb/mmBTU). Divide the amount of contaminant in the fuel
(expressed as lb contaminant/lb fuel), by the fuel heating value expressed as
(mmBTU/lb fuel), the result is (lb contaminant/mmBTU). One pound of contaminant
can convert to more than a pound of pollutant if the contaminant is oxidized or
combines with other species. So there is a need to specify the amount of pollutant
generated by each lb of contaminant in order to determine the potential emissions in
units of (lb/mmBTU). The general formula for converting fuel analysis to maximum
potential emissions is:
lb % contaminant/100 lb pollutant
= × (5-1)
mmBTU ⎛ BTU ⎞ lb contaminant
HHV ⎜ ⎟ / 106
⎝ lb ⎠
When the pollutant is a gas, the ratio of pollutant to fuel constituent is the ratio of
molecular weights. When particulate is formed from fuel ash, the ratio of pollutant to
fuel ash is unity because the fuel ash measured in the laboratory is generated in
essentially the same manner as particulate is formed in a furnace.
Example 5-1.
Determine the potential SO2 emission rate for 3% sulfur coal with HHV = 12,000
BTU/lb.
Solution:
In the case of sulfur oxidizing to SO2 (S + O2 → SO2) the conversion number is the
ratio of molecular weights of SO2 and S, which is 64/32 = 2 (lb of SO2 / lb of S).
So the potential SO2 emissions from 3% S coal is:
⎛ lb S ⎞
0.03 ⎜ ⎟
⎝ lb coal ⎠ × 64 = 5 ⎛ lb of SO2 ⎞
⎜ mmBTU ⎠⎟
(5-2)
⎛ mmBTU ⎞ 32 ⎝
0.012 ⎜ ⎟
⎝ lb coal ⎠
So 3% sulfur coal will generate, at most, 5 lb/mmBTU of SO2 emissions. Actual

emissions will be lower if any of the SO2 is collected, combined or absorbed with
other species in the combustor.

One can do a similar calculation to determine the potential emissions of any inorganic
contaminant in the fuel. This sets an upper limit on the potential emissions – 100%
conversion and all the contaminant is emitted. The trouble is that the conversion to
air emissions is less than 100% for nearly all the common contaminants except sulfur
and chloride. So while this simple mass balance sets a useful upper limit, knowing
both the conversion rate and the furnace collection/absorption rate aids in determining
a more useful emissions estimate from the fuel analysis.
Before discussing the individual pollutant species, note that fuels like natural gas burn
relatively clean because they contain no inorganic contaminants and hence produce
none of the resulting pollutants. In addition, relatively clean flue gas benefits air
pollution control components such as the catalysts used to control NOx and CO. The
only significant pollutants from natural, gas-fired sources are NOx and CO, and the
formation of these species can usually be controlled to very low levels both in the
combustion process and with catalytic control devices.
This reduced pollution formation potential is true to a lesser extent for No. 2 oil-
fired sources. One exception is that ordinary No. 2 oil can contain up to 0.5% sulfur
resulting in significant SO2 emissions and potential degradation of catalyst-based
control devices. No. 2 oil, being a liquid, is also not as amenable as natural gas to
low NOx combustion concepts that rely on mixing fuel and air prior to combustion.
5.2 ACID GASES: SULFUR OXIDES AND HCl

5.2.1 Sulfur Oxides
Most fuels contain at least a trace of sulfur and many waste fuels contain significant
amounts of chloride. These inorganic contaminants convert mostly to SO2 and HCl
gases in the combustion zone. The potential emissions can be calculated from a fuel
analysis as described in the preceding section.
5.2.1.1 SO2
As a general rule, unless a source has a SO2 collector (usually a scrubber), virtually
all the sulfur in the fuel goes up the stack in the form of SO2 (sulfur dioxide gas). For
example, a 1% sulfur oil will generate about 1.1 lb/mmBTU emissions of SO2. So it
is common practice to control SO2 emissions by placing a limitation on fuel sulfur
content, which is easier to monitor than SO2 emissions. SO2 remains a gas at ambient
conditions after it leaves the stack but it is water soluble. Once it leaves the stack it
dissolves into rain or fog droplets and can then easily adsorb onto dry surfaces.

5.2.1.2 SO3 and Sulfuric Acid
A small fraction (less than 5%) of the sulfur in the fuel will be oxidized to SO3 in the
combustion zone or somewhere upstream of any emissions sampling. SO3 is a gas at
high temperature, but it is highly hygroscopic. As it cools below about 600°F, it will
combine with water to form sulfuric acid.
SO3 + H2O → H2SO4
As the exhaust gas cools further, the acid continues to attract water, so that at ambient
conditions it is present as dilute sulfuric acid. Figure 5-1 is a schematic progression
of sulfur from fuel to sulfuric acid.
Sulfur Oxidation
Fuel S
Combustion Zone
O2 SO2 (gas) 98% to the stack

2% of fuel sulfur
O2 SO3 (gas)
Cooling below -600oF
H2O H2SO4 (gas)

Cooling below -260oF
H2O H2SO4 • nH20 (liquid)

Cooling to ambient temperature
H2O dilute H2SO4
Figure 5-1. Fuel Sulfur Conversion to Pollutants
SO3 and H2SO4 are different chemicals, but in the context of a combustion source,
they are effectively the same thing. Once SO3 is formed it will appear as sulfuric acid
(diluted more or less) when it is emitted – unless it reacts with something to form a
salt. In this text the term “sulfuric acid” includes hydrated and non-hydrated forms.
Example 5-2. Conversion of fuel sulfur to SO3
It is estimated that 2% of the sulfur in oil that contains 1.5% sulfur will be converted
to SO3, the precursor to sulfuric acid. What will the flue gas concentration be,
expressed in units of ppm corrected to 3% O2? Residual oil has a higher heating
value of about 18,500 BTU/lb.

Solution. The following three-step calculation is one of several approaches.
a. Determine the emission rate in units of lb/mmBTU. Divide the fuel sulfur content by
the fuel heating value, correct the weight for the oxidation of sulfur to SO3 and for the
% sulfur conversion.
⎛ lb S ⎞
0.015 ⎜ ⎟
⎝ lb oil ⎠ × 80 ⎛ lb SO3 ⎞ × 2% = 0.0405 ⎛ lb SO3 ⎞
⎜ ⎟ ⎜ mmBTU ⎟
⎛ mmBTU ⎞ 32 ⎝ lb S ⎠ ⎝ ⎠
0.0185 ⎜ ⎟
⎝ lb oil ⎠
b. Convert this to ppm by volume, which requires using the F-factor and the density of
SO3 gas. Use a generic F-factor of 10,500 ft3/mmBTU.
⎛ lb SO3 ⎞
0.0405 ⎜ ⎟
⎝ mmBTU ⎠ × 385 ⎛ ft SO3 ⎞ = 18.6 10-6 = 18.6 ppm SO3
3
⎜
⎛ ft 3 fluegas ⎞ 80 ⎝ lb SO3 ⎠
⎟ ( )
10,500 ⎜ ⎟
⎝ mmBTU ⎠
where 80/385 is the density of SO3 gas in lb/ft3
c. The f-factor gives a volume with zero excess air. The goal is to dilute this concentration
with enough air to achieve 3% O2. Use the excess air correction factor as follows
20.9% − 3%
18.6 ppm × = 15.8 ppm SO3 @ 3% O2
20.9%
The amount of fuel sulfur converted in a combustion source to sulfuric acid is always
small and the actual amount of acid emitted is difficult to predict using analytical
methods. Simple chemical equilibrium models are not adequate primarily because of
the catalytic influences of other inorganic fuel contaminants. For example, vanadium
pentoxide (V2O5) is a catalyst that is used commercially to make sulfuric acid from
SO2. Vanadium is a principal ash constituent of residual oil so any source fired on
residual oil is coated internally with dust and deposits containing various vanadium
compounds. The amount or extent of these deposits varies with time and the extent of
boiler cleaning activity. Catalytic activity depends on temperature and a typical
utility boiler has surfaces at temperatures ranging from 250°F to 1200°F. So
predicting the amount of SO2 that will be oxidized is not analytically practical. On
sources fired with coal or other solid fuels, the ash may contain substantial amounts
of calcium or other chemically basic species. Ash with a high pH will tend to
combine with sulfuric acid, neutralizing it to form solid sulfate salt. Given all these
variables, simple prediction of sulfuric acid emissions has to be based on experience
or actual measurements. Figure 5-2 is one set of data showing how fuel sulfur and
exhaust gas oxygen affect the emissions of sulfuric acid.

Flue gas levels of sulfuric acid (sometimes termed SO3) are generally less than 25 ppm.
While the numbers are small, these emissions can have disproportionate impacts.
Ambient impacts are usually most serious in areas close to the source, most notably in a
plume down wash area. Sulfuric acid has a much stronger odor and is more reactive than
SO2 so even though there is much more SO2 than acid, it is the acid that is noticeable and
has the primary health impact. Note that most of the SO2 is eventually converted to
sulfuric acid at some point downwind of the stack. So, in terms of regional impacts, it
may make little difference how much of the fuel sulfur is emitted in the form of sulfuric
acid.
4% S
18
2% S
16
Sulfur in Fuel
14
12
0.5% S
10
0.25% S
8
0.25% S
4
0
0.8 1.6 2.4 3.2 4.0 4.8
Oxygen in Exhaust Gas, % by volume
Figure 5-2. Conversion of Fuel Sulfur to SO3

Adapted from data in the Handbook of Energy Systems engineering, Leslie C. Wilbur,
Wiley-Interscience, 1985

Sulfuric acid formed upstream of the stack has two significant effects. First, if the
stack concentration of SO3 is above 5-10 ppm, it usually generates a visible plume.
This is an ephemeral plume that varies in appearance depending on time of day and
weather conditions. SO3 emissions produce a classic condensation plume, usually
thin or invisible at the stack exit, it growing larger and denser as it moves down wind.
The plume is usually bluish white, but if the sun is low it can be reddish brown. The
plume visibility (Method 9 opacity) is heavily influenced by background haze or
clouds. On utility sources it can be heavy in the morning, during and after a load
increase, then disappear by noon. The pollutants are usually invisible to an in-stack
opacity monitor because most of the condensation occurs after the exhaust leaves the
stack. So sulfuric acid can be the cause of high plume opacity readings that are hard
to measure, hard to predict, and hard to control.
Second, if SO3 condenses on any surfaces before leaving the stack (which will happen
on any surface below about 260°F) it causes corrosive damage. However, the
temperature at which acid condenses depends on the concentration. Figure 5-3 is the
theoretical dew-point/temperature curve. In practice, the condensation temperature
falls between 250°F and 280°F for typical coal- and oil-fired combustion sources.
This temperature range is the main reason that designed stack temperature on most
sources are seldom much below 300°F. Corrosion-generated holes in ducts tend to
grow larger because the cooling effect of air leaks accelerates acid condensation and
therefore accelerates corrosion. Damage to a dust collector will also eventually degrade
its performance, leading to higher particulate emissions.
100
10
6% H20
1 10% H20
0.1
0.01
200 220 240 260 280 300
o
Dew Point Temperature ( F)
Figure 5-3. Sulfuric Acid Dew Point
Acid condensed on the surface of a duct generates deposits of fly ash, acid and metal
corrosion products. These corrosion deposits eventually flake off, blow out the stack
and can cause nuisance particulate fallout adjacent to the stack. The particles are
large enough to fall out close to the stack rather than being carried long distances in
the plume. The fallout material is easily washed away, but usually not before the acid
has permanently stained the surface of a car, boat or sidewalk. The material rarely

shows up in a Method 5 particulate test, making it impossible to regulate in a
quantitative manner. But it has been enough of a problem at some locations that the
plant routinely paid for car washes, or even installed a car wash for its employees.
So a small amount of sulfuric acid in the exhaust can create significant environmental
problems even though it is the SO2 emission impacts have been the main driving
force behind sulfur related regulations.
5.2.1.3 Ash Retention of Sulfur Oxides

Sulfur oxides, particularly SO3, react with the particulate matter derived from the fuel
ash. Some types of coal have a sufficiently basic ash that it absorbs nearly all the acid
produced during combustion. In these cases, emissions consist entirely of SO2 with
little or no acid emissions. The result is fly ash with significant amounts of sulfate
salts, such as sodium sulfate or calcium sulfate. Some of the SO2 may also react with
fuel ash to form sulfate. Absorption of sulfur oxides by fuel ash usually accounts for
less than 10% of the fuel sulfur. However, there are sources that deliberately add
basic compounds to the combustion zone resulting in the conversion of most of the
fuel sulfur to solids that can be removed by the dust collector. Note that sulfate,
whether in the form of sulfuric acid or a sulfate salt, can add significantly to the
amount of particulate in the exhaust gas.
5.2.2 Hydrochloric Acid

Much of the chloride in fuel will convert to HCl (hydrochloric acid) in the
combustion zone. One may generally assume that all of the organic chlorides will be
converted to HCl. Compounds like NaCl (sodium chloride) may pass through the
combustion zone unchanged. In theory, a fuel that is low in hydrogen (coke or
anthracite coal) can suppress HCl formation and drive the chloride to form Cl2
(chlorine gas). However, in practice, there are not many sources that fire low
hydrogen fuels with significant amounts of chloride. Cement kilns can fire low
hydrogen fuels like pulverized coke, but when they burn hazardous wastes with
significant chlorides, there is enough hydrogen in the wastes that the chloride is
emitted as HCl rather than Cl2. Sources that emit significant amounts of chlorine gas
appear to be a special case that is left to other references. Chlorine gas in small
amounts can be a factor in dioxin formation.
5.3 PARTICULATE MATTER

Particulate matter emissions are composed of fuel ash, products of incomplete
combustion (PIC) and sulfates. PIC are usually a small part of the particulate
emissions except for certain sources such as diesel engines. Ash is any inorganic
substance in the fuel that is solid at typical stack temperatures and that survives, or is
formed, at combustion temperatures. Combustion-generated particulate matter falls
into two groups:

• Large particles – those larger than 1-2µm that are descended primarily from large
fuel particles and,
• Fine particles – those smaller than about 1µm that are composed primarily of
material that was in vapor form in the combustion zone and condensed as the
exhaust gases cooled.
Large particles derive almost entirely from the breakdown of even larger particles of
fuel or ash. As a general rule, large particles are responsible for most of the weight of
measured particulate emissions (Method 5). Fine particles derive primarily from the
agglomeration or condensation of vapors or large molecules – the opposite process
from the formation of large particles. Fine particles are frequently responsible for
most of the visible emissions (Method 9). This is not to say that there is no relation
between opacity and particulate emissions, but the connection between these two
measurements can be very inconsistent.
Typically, the primary constituents of combustion particulate from stationary sources

are silicon, calcium, metals, a variety of crustal elements, and some sulfate salts.
Most particulate matter is composed of inorganic elements, but it can include carbon
or small amounts of high molecular weight hydrocarbons. Diesel engines emit very
fine particulate composed primarily of organic material. Although the amounts are
relatively small, the impact of exhaust outlets near ground level, notably mobile
sources, can be very significant.
5.3.1 Particle Formation – Large Particles

There are basically two mechanisms that form large particulate matter relating to the
ash and the carbon in the fuel. Fuel ash constituents, such as silicon and most metals,
may melt briefly in the combustion zone, but they tend to pass through the
combustion zone unaltered chemically. Usually the main physical alteration is the
removal of organic (fuel) and volatile species or the change in the crystal form of
certain minerals. Simply stated: whatever rocks went into the furnace, will also come
out of the furnace. Pulverized coal is a good example. If coal with 10% ash is
pulverized to 100µm particles and combustion burns away all the organic material,
then the remaining ash particle will be about 50µm diameter if the ash has the same
density as the original coal. In reality, however, the resulting particle may be a little
smaller if the ash is denser than the original coal particles or may be considerably
larger if it is very porous. But in either case, its composition will be very similar to
the original coal ash and its size proportional to the size of the original coal particle.
Figure 5-4 illustrates the formation of large particles in the combustion zone.

Volatiles Burn Coke Burns
Coal Particle or Coke (Carbon) Ash Particle

Oil Droplet Evaporates
Figure 5-4. Formation of Coke and Ash Particulate
5.3.2 Particle Formation – Fine Particles

Fine particles are formed “from the bottom up”; that is, they start as atoms or
molecules that condense or agglomerate into particles. These particles can grow to be
large particles, but their growth rate slows by the time they reach about 0.5µm in size.
This size is a kind of temporary ceiling in particle growth. The fundamental
explanation for this limit is that 0.5µm is about the size where particles become too
big to undergo Brownian motion. They stop bouncing around, so they stop running
into other particles to combine with; therefore, further growth is much slower.
Although some fine particulate can be formed by the breakup of larger particles, for
the most part, fine particles are composed of chemicals that are in a gaseous or vapor
phase in the combustion zone. Obviously this includes carbon, sulfur oxides, and any
heavy hydrocarbons that survive combustion. It also includes any metals and
minerals that are heated to liquid or vapor phase during combustion.
As the exhaust gases leave the combustion zone and begin to cool, some vapor
species begin to condense on existing particles. These vapors prefer to condense on
existing particles (nuclei) rather than to start new particles from scratch (nucleation).
When there are no nuclei already present, and the vapors are cooled to a
supersaturated condition, nucleation can occur, resulting in new particles. When this
happens, a very large number of very small particles are formed. If, for whatever
reason, there are already significant numbers of small “seed” particles available, the
vapors will not supersaturate and will tend to condense on the seed particles. The
result is usually a smaller number of somewhat larger particles. However, the subject
of nucleation and particle formation is well beyond the scope of this text.

5.3.2.1 Pulverized Coal Particulate
Pulverized coal is ground finely enough that most of the particles are smaller than
about 150µm diameter. This dust has a large surface area and it burns very rapidly
after it is blown into the combustion zone. Once all the fuel is burned away there
remains an ash particle – the fly ash that leaves the furnace. The small particles burn
up faster than the big particles. If any of the coal particles are too large, they may not
be in the combustion zone long enough to burn completely. A large particle of
pulverized coal, passing through a typical combustion zone with residence time of
one or two seconds will not burn completely leaving the residual particle with carbon
as well as ash. The larger the original particle, the more carbon will remain in the
particle. As a general rule, any coal particles larger than 150-200µm will not burn
completely and the resulting fly ash will contain some carbon.
Example 5-3. Amount of particulate generated by pulverized coal combustion.
What will the particulate emissions be at the exit of a furnace (before the dust
collector) fired with pulverized coal? The coal properties are 9% ash and 12,500
BTU/lb higher heating value. Assume that 10% of the ash drops out into furnace
hoppers and express the answer in units of lb/mmBTU.
Solution.
Divide the coal ash content by the HHV and multiply by 90% to account for 10% of
the ash that stays in the furnace.
⎛ lb ash ⎞ ⎛ mmBTU ⎞ ⎛ lb particulate ⎞

0.09 ⎜ ⎟ ÷ 0.0125 ⎜ ⎟ × 90% = 6.48 ⎜ mmBTU ⎟
⎝ lb coal ⎠ ⎝ lb coal ⎠ ⎝ ⎠
Note that the potential emissions are much larger than typical regulatory emission
limits. If, for example, the regulatory limit was 0.05 lb/mmBTU, the dust collector
could only allow 0.05/6.48 = .0077 or less than 0.8% of the total particulate to escape.
5.3.2.2 Particulate from No. 6 Oil

No. 6 oil contains less than 0.2% ash, so the particulate emissions resulting from fuel
ash is small. However, with No. 6 oil there are two other particulate formation
mechanisms that are usually more significant than fuel ash. One is sulfate formation,
discussed in a subsequent section, and the other is coke. Droplets of No. 6 oil in the
combustion zone behave quite similar to particles of coal. As they are heated, volatile
material is driven off leaving behind a carbon particle. If the particle is small enough,
it burns up in the one-second or so of available combustion time. If it is larger than
about 200µm, it takes longer than one second to burn and the residual carbon particle
will leave the furnace with the flue gas. Thus the amount of particulate formed in a
No. 6 oil flame depends on the amount of coke, which in turn depends on the size

distribution of droplets in the fuel spray. If the oil atomizers generate a spray with a
mean droplet diameter less than about 150µm, there should be little or no particulate
coke emissions. Coke emissions increase with spray droplet size for mean diameters
greater than 150µm and can amount to 0.3 lb/mmBTU or more. These particles are
usually large enough to be visible without a magnifying glass – 0.1 mm or more.
They are called “cenospheres” and when viewed with a scanning electron microscope
they look like sponge balls (see Figure 5-5). They are composed mostly of carbon
plus the ash that was present in the original fuel droplet. They usually also contain
some sulfate or condensed sulfuric acid.
Figure 5-5. Coke Cenospheres from No. 6 Oil Combustion
5.3.2.3 Black Smoke

Coke carbon is the residue of fuel particles or droplets. Particulate carbon can also be
formed from the condensation of carbon atoms or extremely small carbon particles
generated during the combustion of vaporized fuel. These particles grow from the
vapor phase; whereas, coke particles derive from the break down of larger clumps of

fuel. Fine particulate carbon is the main constituent of soot and of black smoke. This
is also the main reason that a flame is white to red in color. Flame visibility is due to
incandescent carbon particles, most of which are much less than 0.5µm in diameter.
Fine particulate carbon forms in the flame when combustion occurs, and the
hydrocarbon fuel molecules break down, but only in regions where there is not
enough oxygen available to burn all the fuel. Typically both hydrogen (H+) and CO
are more aggressive in capturing oxygen than is atomic carbon.
Whenever the air-fuel ratio is below about 0.3 times stoichiometric (very fuel rich),
the carbon concentration is high enough that carbon atoms agglomerate into very
small particles. When liquid fuel mixes with air in the flame there is always a point
early on in the process where combustion has started, but the mixture is below the
critical air-fuel ratio; therefore, particulate carbon is formed and the flame becomes a
bright visible yellow. Once the fine particulate carbon is formed, it does not burn up
unless there is oxygen available. If enough air is not mixed in with the flame before it
cools below combustion temperatures, some of this carbon (soot) will be emitted as
visible black smoke. So if a visible flame is quenched, or the flame goes out before
enough air mixes with the fuel, the same carbon that makes the flame visible now
makes a visible black plume. This principle can be illustrated by sticking a metal
spoon into a candle flame. The cold metal will quench combustion and the unburned
carbon will leave a black smudge on the spoon.
Note that natural gas flames are not normally visible, except perhaps for a faint blue
glow. The reason is that natural gas has low carbon content and it mixes very rapidly
with air. Therefore, the air-fuel equivalence ratio is usually below 0.3 before
combustion frees enough carbon atoms to create particles. There is no fine particulate
carbon to create either a visible flame or black smoke. If the flame does not get
enough air, the result will usually be CO or unburned CH4, not black smoke. The
faint blue glow of a natural gas flame is the radiation from CO2 and H2O molecules,
the radiation produced from electron transitions in the molecules, not the
incandescence of hot solid particles.
5.3.2.4 Sulfuric Acid and Sulfate Particulate

Most of the fuel sulfur burns to SO2 and remains in gas phase long after it is emitted
from the stack. But some of the fuel sulfur is oxidized to SO3, most of which shows
up as some form of sulfate (SO4=) in the particulate. On a coal-fired furnace, where
the uncontrolled particulate emissions are on the order of 10 lb/mmBTU, the
additional sulfate is not significant to the particulate mass emissions. But on a No. 6
oil-fired source where there is almost no ash, sulfate constitutes most of the
particulate. If 2% of the sulfur in oil with 1% sulfur content is oxidized to SO3, then
sulfate will contribute about 0.03 lb/mmBTU to the emissions. In practice, on No. 6
oil-fired sources, the amount of sulfate captured in particulate emission tests varies
from about 0.015 to more than 0.06 lb/mmBTU. The amount depends on a number of
factors including:
• the amount of sulfur in the oil

• the amount of vanadium in the oil
• excess air levels in the combustion zone and in subsequent gas passes, and
• the amount of ash deposits throughout the boiler.
Although these factors are known to influence sulfate formation, there is no consistent
quantitative relationship to the amount of sulfate formed. In short, sulfate is usually the
primary contributor to particulate emissions from oil-fired sources, but the amount
cannot be predicted except through experience with the facility in question.
Example 5-4. Sulfate emissions from a source fired on No. 6 oil.
A source is fired with No. 6 oil containing 1.4% sulfur and a higher heating value of
18,500 BTU/lb. If 2% of this sulfur is oxidized to sulfate, which is normally a solid
or condensable liquid, what is the potential contribution of sulfates to the particulate
emissions in units of lb/mmBTU?
Solution:
Divide the fuel sulfur content by the HHV to get the sulfur emissions in units of
lb/mmBTU. Adjust the sulfur emissions by the ratio of molecular weights of sulfate
and sulfur, and then multiply by 2%.
⎛ lb S ⎞ ⎛ mmBTU ⎞ 96 ⎛ lb SO4 ⎞ ⎛ lb particulate ⎞

0.014 ⎜ ⎟ ÷ 0.0185 ⎜ ⎟× ⎜ ⎟ × 2% = 0.045 ⎜ mmBTU ⎟
⎝ lb oil ⎠ ⎝ lb oil ⎠ 32 ⎝ lb S ⎠ ⎝ ⎠
5.3.3 Enrichment of Some Chemicals in Fine Particles

Fine particles are formed primarily from volatile species, which means their chemical
composition tends to be quite different from that of large particles. Volatile
substances formed during combustion begin to condense as the exhaust gases cool.
As they condense, they deposit on any available surface – which includes the surface
of fly ash particles. Large ash particles have much less surface area relative to their
volume (or mass) than do small particles. Since the amount of a volatile chemical
deposited is proportional to the available surface area, the amount deposited per unit
volume of particle is inversely proportional to the size of the particle. The layer of
deposits on the surface of a large particle is relatively thin compared to the size of the
original particle. In other words, small particles gather a disproportionate share of the
volatile chemicals regardless of the core composition of the particles. If the large
particles in coal fly ash are analyzed, their composition looks like the original coal
ash. Analyze the small particles and higher concentrations of volatile chemicals are
found as compared to the amounts in the original fuel ash.
Table 5-2 lists some elements that are likely to be enriched in the fine particles of fly
ash of a coal-fired boiler. Most of these elements are present in the ash as oxides,
hydroxides or sulfates. The list will vary considerably depending on which elements

are present in the fuel and on what compounds they are formed during combustion.
The availability of sulfur, chlorine and other reactive elements can result in the
formation of volatile compounds from elements that are not normally considered
volatile.
Table 5-2. Some Elements Enriched in Fine Coal Fly Ash
Antimony Gallium Sodium
Arsenic Lead Thallium
Beryllium Molybdenum Uranium
Cadmium Nickel Vanadium
Chromium Potassium Zinc
Copper Selenium
5.3.4 Mass Emission Transients from Soot Blowing

Most of the emissions from combustion sources have very short residence times –
from the time fuel enters the combustion zone to the time the resulting pollutants are
emitted into the atmosphere. In fact the only pollutants that accumulate in almost any
combustion source are particulates. Ash, especially from solid fuel fired sources,
accumulates on heat transfer surfaces and has to be removed to maintain the heat
transfer or load capacity of the system. Boilers are designed with “soot blowers” that
remove ash/dust deposits in order to maintain performance. Operators routinely use
soot blowers to maintain normal operating temperatures throughout the system. Soot
blowing can be continuous based on the discretion of the operator, or it may be
performed periodically (once a shift, once a day, etc.). If an operator sequentially
activates all the soot blowers on a big utility boiler, the complete cycle can take more
than two hours. On a small factory boiler, manual operation of a single soot blower
may take only 30 seconds. Whenever the soot blowers are operated, accumulated dust
is released into the exhaust gas stream and particulate emissions usually increase.
Depending on where the soot blowers are located and on the performance of the dust
collector, the resulting emissions may increase several fold or not at all.
5.4 METALS
With the exception of mercury, all metals are in the form of particulates when emitted
from combustion sources. Mercury compounds are nearly always in vapor form in
combustion exhaust gases. Usually heavy metals or toxic elements are a small
fraction of the total particulate. However, their classification as hazardous air
pollutants (HAPs) means they are a concern independent of their contribution to total
particulate mass emissions. In addition, many of the heavy metal species are volatile

in the combustion zone and they make a disproportionate contribution to fine
particulate (PM2.5) emissions. Fine particulate matter is more difficult to capture in a
dust collector than large particulate matter. In addition, fine particulates are a greater
health hazard than large particulates. Hence heavy metals merit special attention in
this course.
5.4.1 Volatility of Metals and Compounds

The fuel fed to waste combustors contains a broad spectrum of chemical compounds
and sometimes contain significant amounts of lead and other heavy metals.
Although they may be present in trace amounts, other fuels such as coal may contain
an equally wide array of metals and trace elements. In a preceding section we
showed that volatile compounds are preferentially concentrated in the smallest
particles. When the volatile compounds are HAPs, this means the fine particles may
be more toxic that the larger particles. Table 5-3 lists some elements, their associated
combustion products, and the temperature that indicates their volatility.
Table 5-3. Vaporization of Metal Compounds
Metal Melting Temperature Main Vapor Species Temp. (°F)

(°F) for one ppm
Antimony 1166 Sb2O3 1220
Arsenic 1135 As2O3 90
Barium 1337 Ba(OH)2 1560
Beryllium 2332 Be(OH)2 1930
Cadmium 610 Cd 417
Chromium 3375 CrO2 or CrO3 2935
Gallium 86 unknown unknown
Lead 622 Pb 1160
Molybdenum 4730 unknown unknown
Mercury -38 Hg 57
Nickel 2651 Ni(OH)2 2210
Selenium 423 SeO2 605
Silver 1764 Ag 1660
Thallium 578 Tl2O3 1330
Vanadium 3434 unknown unknown
Zinc 787 unknown unknown

The last column of Table 5-3 is the approximate temperature at which the compound
will evaporate to a concentration of 1 ppm in the exhaust gas. This is also an
indication of the combustion zone temperature required to begin to evaporate the
species and cause it to concentrate in small particles. Although combustion zone
temperatures can momentarily reach temperatures above 3000°F, many of these
compounds require a longer time period or higher temperature to evaporate
significantly. But clearly some of them are likely to evaporate during combustion and
will thus appear as fine particulate at elevated levels.
This table does not list compounds that might be present if there is some form of
chlorine in the fuel. Chloride compounds tend to have low melting temperatures, so
chloride can cause more species to appear in the fine particulate. Obviously, the array
of elements that will predominate in the fine particulate depends on how all the
various inorganic species interact in the combustion zone – a subject that goes far
beyond the scope of this course.
5.4.2 Mercury
Any mercury entering a combustor is vaporized and leaves in the exhaust gas. The
pollutant all goes up the stack unless some of it is captured by carbon in the fly ash.
So unless a combustion system includes a mercury control system, we normally
assume that 100% of the fuel mercury is emitted into the atmosphere.
Public concern over mercury in the environment has risen steadily since about the
time of the 1990 Clean Air Act Amendments. At one time municipal waste
combustors were believed to be primary combustion source of mercury emissions.
Waste control and recycling combined with air pollution control devices have
substantially reduced mercury emissions from waste combustors. Coal-fired boilers
were not an immediate focal point, because mercury levels in coal are very low –
typically 0.15 ppm by weight. However, considering the amount of coal burned and
the fact that nearly all the mercury in the coal is emitted, the air emissions from a
typical power plant are on the order of one or two pounds a day. Thus power plants
are coming under increased pressure to control mercury emissions.
Mercury emissions can be any or all of several chemical forms that include Hg, HgO
and HgCl. The predominant form depends on temperature, excess air and the amount
of chlorine in the fuel. In all cases, mercury appears in the vapor phase in typical
combustion exhausts; although, it can be adsorbed onto ash particulate, particularly if
the ash has a significant amount of carbon. Carbon particulate from any solid fuel
fired source will be captured in a dust collector. So if a source generates considerable
carbon, normally an undesirable waste of fuel, mercury in the fly ash increases and
the air emissions of mercury may be reduced. Control of mercury by this, and other
related measures, is discussed in Chapter 6.

Example 5-5. Mercury emissions rate.
Determine the average daily mercury emissions for an 800-megawatt power plant.
The plant has a heat rate of 9900 BTU/kw-hr, and a 60% utilization factor. It burns
coal with an average mercury content of 0.13 ppm and a higher heating value of
11,900 BTU/lb.
Solution.
Determine the emissions rate by dividing the mercury content by the HHV.
⎛ lb Hg ⎞ ⎛ mmBTU ⎞ ⎛ lb Hg ⎞
( )
0.13 10 −6 ⎜ ⎟
⎝ lb coal ⎠
÷ 0.0119 ⎜ ⎟
⎝ lb coal ⎠
( )
= 10.9 10 −6 ⎜ ⎟
⎝ mmBTU ⎠
Determine the power plant fuel energy use by multiplying the rated load by the heat
rate, adjusting for utilization and multiplying by 24 hours.
⎛ BTU ⎞ ⎛ hr ⎞ ⎛ mmBTU ⎞
800,000(kw) × 9900 ⎜ ⎟ × 60% × 24 ⎜ ⎟ = 114,048 ⎜ ⎟
⎝ kw - hr ⎠ ⎝ day ⎠ ⎝ day ⎠
Finally multiply the emissions by the daily energy use to get the emissions.
10.9 (10-6) x 114,048 = 1.24 lb/day Hg emissions
So even though the amount of mercury in the coal is minuscule, the cumulative
emissions are significant.
Although the chemical form of mercury emissions does not change the amount
emitted, it does affect the method of measurement, which can alter the reported
emission rate. Mercury tends to be scrubbed from the atmosphere by precipitation,
which then transports it to water bodies where it can move up the food chain. Once in
the environment it can be transformed to methyl or dimethyl mercury that is even
more toxic than the inorganic forms.
5.5 NITROGEN OXIDES

5.5.1 Overview of NOx
NOx is an acronym for the sum of two compounds: nitric oxide (NO) and nitrogen
dioxide (NO2).
NOx = NO2 + NO
Both compounds are formed in a combustion zone, but NO usually accounts for
95% or more of the total. However, once emitted into the atmosphere, NO promptly

oxidizes to NO2. Thus from a regulatory perspective it does not matter which
compound is emitted because it will all show up as NO2 in the atmosphere. This has
led to the regulation of the sum of the two or NOx. When determining the mass of
emissions we need to assign the molecular weight – in this case the number is 46,
which is the molecular weight of NO2.
Although both NO and NO2 are toxic, atmospheric concentrations threaten air quality
standards in only a few locations. Exceedances have been recorded in some heavily
polluted urban areas and in the downwash from poorly designed combustion source
stacks. The main reason NOx is regulated is that it is a key contributor to the
formation of ozone, which is not emitted from combustion sources. Ozone is formed
in the atmosphere from reactions between NOx and volatile organic compounds
(VOCs) in the presence of sunlight (a so-called “photochemical reaction”). In
addition to its role in ozone formation, NO2 has a red or brown color and is a small
contributor to reduced visibility – usually in or around urban areas.
There are other oxides of nitrogen, notably nitrous oxide, N2O, which does not appear
to be involved in ozone formation and is not currently regulated. N2O is a stable
compound that is usually present in small amounts in the atmosphere. But it is a
powerful greenhouse gas. Like methane and CO2, it can be formed in some selected
types of combustion systems, usually in small amounts. N2O has been used as an
anesthetic (laughing gas).
NOx is the one pollutant emitted by nearly all combustion sources independent of what
is in the fuel or the configuration of the combustor. The reason is that even when there
is no nitrogen in the fuel, NOx is created from the breakdown of atmospheric N2 and O2
in the combustion zone and the subsequent reaction to form NO.
N2 + O2 ↔ 2NO (5-3)
5.5.1.1 Control Approaches

Control of NOx emissions is discussed in detail in Chapter 6, but it is useful at this
point to know that control approaches fall into one of two categories:
• Combustion strategies and/or devices that control NOx formation, and
• “Back end” controls that destroy NOx after it is formed.
As a general rule combustion controls are relatively inexpensive to install and

operate, but in most cases they have limited effectiveness. Back end controls are
comparatively expensive, but they are applicable to most types of sources and they
can achieve high control efficiencies. Most of the control efforts on stationary
sources in the U.S. from 1970 to 1990 focused on combustion system modifications.
The need for further emissions reduction is resulting in combined back end and
combustion controls on both new and existing sources.

5.5.1.2 Formation Mechanisms
In order to understand combustion based NOx control technologies, one needs an
appreciation of the basic NOx formation mechanisms and their relationship to
different types of fuel and types of combustion systems. NOx is generated in all
combustion sources, but the amount depends on:
• temperature and the flame structure including the amount of oxygen available in
the peak temperature regions of the combustion zone (see Equation 5-11), and
• the amount of nitrogen in the fuel (see Eqn 5-1) in combination with the flame
structure.
These two mechanisms, centered around flame temperature and fuel nitrogen, lead to
the terms “thermal NOx“ and “fuel NOx”. This classification is useful in
understanding what types of “low NOx “ control methods will be effective on a given
combustion source.
There is a third NOx formation mechanism that we will mention; although, it is not
significant in most combustion systems. Small amounts of “prompt NOx” can be
formed at the flame front. Prompt NOx is formed very fast and generally in small
quantities. The fact that prompt NOx formation appears to differ from the models
developed by researchers for fuel and thermal NOx, is not enough to justify further
exploration in this text.
5.5.1.3 Typical Emission Rates

This section presents a general overview of the primary NOx emission formation
mechanisms, which is needed to understand NOx control methods. Actual or typical
emission rates are not easy to predict analytically, so one must rely on experience.
The EPA Compilation of Emission Factors (AP-42) is the most comprehensive
compilation of air emissions experience. Note that these are average emissions for
the available data and emissions from any individual source can be very different.
Table 5-4 is a very limited set of data from AP-42 for uncontrolled NOx emissions.

Table 5-4. AP-42 data for Uncontrolled NOx Emissions
Combustion Source AP-42 Units Heat Input Units
Combustion Turbine (gas/diesel) 0.32/0.98 lb/mmBTU
Large Diesel Engine 3.2 lb/mmBTU
Utility Boiler, 6 Oil with. 0.4% N1 62 lb/1000 gal. fuel 0.41 lb/mmBTU
Boiler Firing No. 2 Oil 24 lb/1000 gal fuel 0.17 lb/mmBTU
Pulverized Coal Boiler 12 - 31 lb/ton coal 0.46 – 1.10 lb/mmBTU
Wood Fireplace 1.8 lb/ton wood 0.15 lb/mmBTU

(1) Residual oil NOx emissions vary depending on fuel nitrogen content.
Note here the large difference between No. 6 and No. 2 oil, basically because of the
nitrogen content of heavy oil. Note also that diesel engines produce far more NOx
than any of the other sources. Some of the reasons for these emission rates are
discussed in this chapter. Chapter 6 discusses control technologies for these sources.
5.5.2 Thermal NOx Formation

Thermal NOx formation by the “Zeldovich” mechanism (see Equation 5-3) requires
very high temperatures. Both N2 and O2 are extremely stable molecules. O2 molecules
are broken apart by combustion chemical reactions and high combustion temperatures
are required to break apart the nitrogen molecules. Figure 5-6 shows the theoretical
equilibrium NO concentration versus temperature. Equilibrium conditions are not
usually achieved in typical combustion zones, but this curve shows that temperatures
approaching 3000°F are required just to push Equation 5-3 toward NO formation. In
practice nearly all the NOx is formed in the hottest regions of the flame. In the hottest
regions the high equilibrium level and the fast reaction rate form so much NOx that
contributions from cooler regions are relatively insignificant. Figure 5-6 suggests that
as the combustion gases cool, NO should decompose. While this may actually happen,
the chemical reaction rate also slows as the gases cool, effectively freezing the amount
of NO formed at higher temperature. In general, NOx is formed rapidly in the hottest
regions of the flame and there is no further creation or destruction as the exhaust gases
cool and are emitted to the atmosphere.

10000
2000
1000
o
200 F
100
3000 3500 4000
Temperature (Degrees F)
Figure 5-6. Theoretical NO vs. Temperature
Natural gas and distillate (No. 2) oil have very little nitrogen, so NOx emissions from
these sources are generated entirely by flame temperature (thermal NOx). This
means that NOx control concepts that reduce flame temperature are quite effective.
Residual (No. 6) oil and all solid fuels have a significant amount of nitrogen and, as a
general rule, more than half of the emissions from these sources originates from fuel
nitrogen (fuel NOx). A control concept that reduces flame temperature will be much
less effective at reducing NOx on these sources. So the type of fuel used by a source
becomes a key factor in selecting the emissions control technology.
There are no simple analytical tools for predicting NOx emission levels. It is not easy
to calculate either the peak temperature or the amount of flame that reaches the peak
temperature; so analytical prediction of thermal NOx formation requires very
sophisticated modeling. Nitrogen in the fuel can be converted either to N2 or NOx
during combustion, but the proportions depend on a number of flame details, so here
again the NOx emissions caused by fuel nitrogen cannot be predicted analytically.
NOx emission levels must be measured or estimated from past experience.

5.5.3 NOx Formation from Fuel Nitrogen
During combustion every nitrogen atom in the fuel will do one of two things:
• react with oxygen to form NO, or
• react with another N atom to form N2
Chapter 6 provides a substantial discussion on NOx reduction technologies that

attempt to drive nitrogen atoms to form molecular nitrogen during combustion. Fuel
nitrogen is not an issue for most engines and combustion turbines because they do not
normally use fuel that contains nitrogen, so this discussion is restricted to boilers or
furnaces. When no low-NOx technology is applied, typical burners convert about
half the fuel nitrogen to NOx. The exact amount depends on combustion zone excess
air and burner design. Low NOx burner systems in boilers can reduce fuel nitrogen
conversion to the range of 15% to 25%.
Example 5-6. Fuel Nitrogen Conversion to NOx
A power plant fires coal with 1.3% nitrogen and a higher heating value of 13,200
BTU/lb. If all of the fuel nitrogen were converted to NOx, what would the NOx
emissions rate be?
Solution:
Use Eqn 5-1 and divide the fuel nitrogen content by the fuel HHV to obtain the fuel
nitrogen emissions. Then multiply ratio of molecular weights of NO2 and N to obtain
the NOx emissions.
⎛ lb N ⎞ ⎛ mmBTU ⎞ 46 ⎛ NO2 ⎞ ⎛ lb NOx ⎞

0.013 ⎜ ⎟ ÷ 0.0132 ⎜ ⎟ × ⎜ ⎟ = 3.2 ⎜ ⎟
⎝ lb coal ⎠ ⎝ lb coal ⎠ 14 ⎝ N ⎠ ⎝ mmBTU ⎠
This is roughly three times the uncontrolled emission level of a typical coal-fired
boiler, so we know that less than a third of the fuel nitrogen is normally converted to
NOx in these systems.
Figure 5-7 is based on a nitrogen mass balance for residual oil and shows the
relationship between fuel nitrogen and fuel NOx emissions. To get total NOx
emissions you must add the thermal NOx, roughly 0.15 lb/mmBTU to the data in
Figure 5-7. A similar figure could be drawn for bituminous coal and would show
emissions about 50% higher because of the lower heating value of coal.

0.7
0.6
50%
0.5
Typical Range
Uncontrolled Burners
0.4
30%
0.3
25%
Typical Range
0.2 Low NOx Burners
15%
0.1
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Fuel Nitrogen, (% by wt)
Figure 5-7. Fuel NOx Emissions vs. Fuel Nitrogen Content
Fuel nitrogen content depends on the source of the fuel and it varies considerably.
Although fuel with a specific limit on the sulfur content can be purchased, it is not yet
practical to purchase fuel with a limit on nitrogen content.
Diluting residual oil with enough No. 2 oil to meet a nitrogen content target can
create low nitrogen oil. Table 5-5 shows typical ranges for fossil fuel nitrogen
contents. Natural gas may contain nitrogen, but the nitrogen in natural gas is
molecular nitrogen, not organic nitrogen that could contribute to NOx formation.
Table 5-5. Typical Fuel Nitrogen Content
Fuel Nitrogen (% by wt.)
Coal 0.5 - 2
Residual Oil 0.3 - 0.6
No. 2 Oil < 0.1
Nitrogen in residual oil appears to have increased over a period of years as indicated
by the data in Figure 5-8. This data is from utility plants in the eastern and
northeastern U.S. One possible explanation for the increase is that the recovery
methods used to draw oil from diminished reserves are pulling up crude oil that has
more nitrogen. This would correlate with anecdotal reports of heavier (higher
specific gravity) oil trends during the same time period.

0.7
Percent Sulfur
Half
0.6 One
Two
0.5
0.4
0.3
0.2
0.1
0.0
70 75 80 85 90 95
Figure 5-8. Residual Oil Nitrogen Content vs. Time
5.5.4 Premixed and Diffusion Combustion

Most combustors inject the fuel and air separately into the combustion zone where
they mix and burn simultaneously; this is a diffusion flame. A few, such as natural
gas-fired engines, premix the fuel and air before entering the combustion chamber
where the mixture is ignited giving a premixed flame. In most cases, diffusion
combustors are easier to design and operate. Only gas fuel or vaporized liquid fuels
can be used in real premixed combustors. Burners that premix pulverized coal or
atomized No. 6 oil could be conceptually built, but they would not be truly premixed.
Only a gas can be fully premixed with air.
The difference between diffusion and premixed combustion was described briefly in
association with Figure 4-5 (p. 4-29) and flame temperatures. When fuel and air are
premixed, the maximum flame temperature is equal to the average flame temperature
because the mixture burns at a uniform temperature. Figure 4-5 and the premixed
excess air or stoichiometric ratio can be used to determine the maximum flame
temperature. So the more excess air that is present, the lower the temperature will be
of a premixed flame. (Refer to the right side of Figure 4-5.) Therefore, lean
premixed combustion can dramatically reduce thermal NOx formation.
Now consider the diffusion flame in which fuel (pulverized coal, an oil spray or a jet
of gas) is injected into the flame along with air. As each little parcel of fuel mixes
with air and burns, its air-fuel ratio goes from zero (pure fuel) to whatever the final
air-fuel ratio is at the exit of the combustion zone. Figure 4-5 indicates that as it
passes through an equivalence ratio = 1.0, its temperature could reach the adiabatic

maximum. The fuel passes through and creates a hot spot before it mixes with the
excess air and cools off to the average combustor exit temperature. So diffusion
flames are locally hotter than premixed flame and thus they make more thermal NOx.
There is another factor that plays a role in NOx formation in diffusion flames, and
that is the amount of oxygen or excess air that is present in various regions of the
flame. NOx is not created when there is little or no oxygen available. When the air-
fuel ratio is fuel rich, the left side of Figure 4-5, NOx formation is suppressed because
there is not any available oxygen; and, for the same reason, combustion is not
completed either. NOx formation only starts after the air-fuel mixture passes the
stoichiometric point and oxygen becomes available for NOx formation. Both fuel
NOx and thermal NOx formation become active only on the right side of
stoichiometric in Figure 4-5. The conceptual basis for most boiler NOx control
devices is to accomplish most of the combustion on the rich side of stoichiometric,
then add enough excess air to finish combustion without creating a lot of NOx.
At the present time the only combustion devices that premix fuel and air are gas or
gasoline-fired reciprocating engines and gas-fired combustion turbines with dry low
NOx combustors. Attempts to vaporize No. 2 oil have been demonstrated in
combustion turbines, but these systems run into durability problems. They tend to
develop coke deposits in the process of continuously vaporizing fuel for long periods of
time. Residual oil and all solid fuels can only be fired in diffusion flame combustors.
5.5.5 NOx from Typical Combustion Systems

Most of the NOx control efforts of stationary sources prior to about 1990 involve
changes to combustion systems to reduce NOx formation. While back end NOx
controls are becoming increasingly important, the design of low NOx combustion
systems will continue to be an important element in NOx control. There are many
types and configurations of combustion systems and low NOx systems are site
specific. One aspect of combustion and NOx formation that is common to all systems
is the difference between premixed and diffusion flames.
Combustors can be divided into three general categories:
• Furnaces & boilers

• Reciprocating engines
• Combustion turbines
These combustion systems are all very different as are the NOx control strategies. In
all cases NOx is formed in the combustion zone, then emitted into the atmosphere
through an exhaust pipe.
All large combustion systems operate continuously over some load range and most
have continuous operator supervision. They are not “off-on” devices like a
residential furnace or water heater. This means that emissions may be an issue not

just at full load, but also over the load range. In fact, emissions during start-up or
shut-down, which can be lengthy, may also be an issue.
5.5.5.1 NOx from Boilers and Furnaces

Boilers burn a wide variety of fuels at atmospheric pressure. They range in size from
home furnaces to electric utility boilers that stand 200 feet high and burn oil at a rate
of 800 gallons/minute. A boiler consists of a firebox where the fuel is burned
followed by banks of tubes where heat from the combustion gases is transferred to
water or steam. Combustion temperatures are in the range of 3500°F and the flue
gases are cooled to about 300°F at the stack.
Boilers have separate systems to supply fuel and air to the combustor and there are
automatic controls driving these systems. Fuel is supplied at a sufficient rate to
satisfy boiler load demand. Air is supplied at a rate that matches the fuel flow. If the
controls supply insufficient air, combustion will be incomplete and the boiler will
smoke. So the controls always supply a bit more air than it takes to prevent the boiler
from smoking. Air flow to the boiler is generally characterized by the amount of
excess air, which is the amount above and beyond that which is theoretically required
for complete combustion. It is measured by the amount of O2 in the flue gas. This O2
is associated with the air that was not consumed during combustion, so by definition
it is “excess”. Figure 5-9 shows the general relationship between excess air (or stack
O2) and emissions of NOx and CO. The figure illustrates the general form of the data
if one were to systematically vary the excess air while measuring NOx, CO and O2 in
the flue gases. This same type of behavior will be observed on all boilers and
furnaces. It would also hold true for combustion turbines and many reciprocating
engines, except that it is not practical to systematically vary the excess air levels on
those sources.
CO & Smoke
Excess Air or O2
Figure 5-9. Typical NOx and CO vs. Excess Air
Figure 5-9 shows NOx increasing with excess air and this behavior is true on all
suspension-fired furnaces and boilers, regardless of fuel type. We know from
Figure 4-5 (p. 4-29) that the average combustor temperature decreases with increasing

excess air. That should mean that NOx also decreases, but Figure 5-9 shows the
opposite. The reason that NOx increases with excess air is that as we increase excess
air, more and more of the combustion is occurring in a region with extra oxygen. An
increasing amount of the flame volume is to the right of stoichiometric in Figure 4-5,
so more of the flame can generate NOx. Since all boilers operate with diffusion
flames, all the fuel has to pass through stoichiometric as it mixes with air. As we
increase excess air we simply move more of the flame from fuel rich to air rich and
NOx increases proportionately.
Figure 5-9 illustrates several important points. First note that CO emissions (black
smoke will behave similarly to CO) are negligible at high excess air levels. As excess
air is reduced, CO appears abruptly and increases rapidly; this is commonly referred
to as the knee in the curve. NOx, on the other hand, decreases almost linearly as
excess air is reduced. There is an optimum excess air operating point near the knee in
the curve where NOx is minimized, but where CO (and/or smoke) is not excessive.
Any further reduction in excess air would reduce NOx at the expense of significantly
increased CO or smoke. Any increase in excess air will increase NOx without
reducing CO. In general, on nearly all combustion systems, emissions control will be
a compromise between NOx and CO, or smoke. To some degree, one must trade one
type of emission for the other.
This highlights the need to control air flow precisely on any combustion system that
is attempting to minimize NOx. Air flow needs to be maintained near the minimum
practical (not quite smoking) level. The automatic air flow control system needs to
do this continuously as the boiler load changes. Precise control of excess air is the
first and most important step in controlling emissions from boilers and furnaces. This
can only be achieved by using an O2 monitor tied into an intelligent (microprocessor
based) control system.
5.5.5.2 Fuel Effects on Boiler Emissions

It is useful, from a NOx perspective, to divide fuels into three categories:
Category No. 1: Clean fuels (natural gas and distillate oils)

Category No. 2: Residual oil, pulverized (powder) coal, and sander dust, and
Category No. 3: Solid fuels
Clean fuels contain very little nitrogen, which means that NOx from these sources
is generated only by the heat of combustion (thermal NOx). Generally, simply
lowering the flame temperature can reduce NOx. Note that most engines and all
combustion turbines only burn clean fuels. Fuels in categories No. 2 and No. 3
contain enough nitrogen that it accounts for much or most of the NOx emissions.
Flame temperature based controls are much less effective with these fuels, so effective
NOx reduction requires some form of staged combustion (low NOx burners) or back
end controls.
We have separated solid fuels (coal, wood, etc) into two categories because there are
two very different types of combustors. When fuel is divided into fine particles (a

powder or atomized spray), it burns very rapidly; typical residence times are less than
½ second. However, by contrast, a small lump of coal, a wood chip, or material in
solid waste will take a minute or more to burn. Category No.2 fuels are sprayed into
the combustion chamber and burn in suspension with no significant settling of
particulate matter. Category No. 3 fuels settle to the floor of the furnace where they
burn on a grate, similar to the fire in a home fireplace.
From the perspective of NOx control, suspension burning (Category No. 2) offers the
opportunity to control the fuel and air mixing in the flame in a way that affects NOx
formation. Grate burning (Category No. 3) offers very little ability to control the
combustion. Hence stoker furnaces (stoker coal, wood chips, and solid waste) offer a
very limited opportunity for combustion-based NOx controls. They are amenable to
the relatively new technology of reburning, but for the most part NOx reduction from
these units is achieved by back end controls.
Although boilers come in all sizes, many of the basic features are common to all of
them. Size does have some influence on NOx emissions and NOx controls. First,
small boilers usually burn clean fuels, while larger boilers burn the least expensive
fuel available, which usually means higher NOx emissions. Second, while industrial
boilers generally have only a single burner, large utility boilers have multiple burners,
anywhere from 12 to several dozen. This has both advantages and disadvantages
when it comes to NOx control. The advantage is that low NOx staged combustion
firing can frequently be achieved by taking one or more burners out of service, which
typically does not reduce the full load capability. However, the disadvantage is that
low NOx operation requires that each burner get the same amount of fuel and air;
each one needs to operate at the same air-fuel ratio. While this seems like a
straightforward requirement, it can prove very difficult to achieve in practice.
5.5.5.3 NOx from Reciprocating Engines

Reciprocating engines, similar to a car or truck engine, are used for stationary power
sources. Land based engines fire either natural gas or No. 2 (or diesel) oil – clean
fuels. Combustion occurs at very high pressure (several hundred psi) and at high
temperature. Temperatures are particularly high because the compression
temperature, before ignition, is on the order of 1000°F. As a consequence of very
high temperatures, NOx emissions are high, roughly 10 times the emissions from a
boiler burning the same amount of fuel. Note that transportation vehicles generate
more NOx in the U.S. than do stationary sources.
Reciprocating engines typically are designed to run on either gas or distillate oil.
Although marine diesels commonly use residual oil, this is rare for land based
engines. In a diesel engine, the diesel oil is injected into the cylinder where the
compression temperature is high enough to ignite the fuel. Injection is normally
timed for maximum power, which means just before the piston reaches the top of its
stroke. Delaying (retarding) the injection timing reduces both power and NOx
emissions. Excess air levels on diesel engines vary from near zero when under heavy
load to very high levels when load is reduced. The reason this occurs is that air flow

is determined by engine speed while fuel flow is determined by load. Air and fuel
flow are controlled independently, and excess air varies with speed and load.
When operating on gas, the fuel is premixed with the incoming air and is ignited with
an electric spark from the spark plug. The mixture has to be fairly close to
stoichiometric to be ignitable. If the mixture is too lean, the engine stops running; if it
is too rich, it smokes. Engines are characterized as lean burn or rich burn depending
on whether they are designed to run with an exhaust O2 level more or less than 1% O2,
(i.e., about 5% excess air). Figure 5-10 shows the theoretical relationship between air-
fuel ratio and emissions. Rich burning invariably generates CO, hydrocarbons and
smoke and slightly reduced power. Lean burn engines operate slightly above
stoichiometric where NOx emissions are high, but PIC are low Figure 5-10 helps to
illustrate why NOx emissions are high from all types of reciprocating engines.
Standard engines require an air-fuel ratio near a stoichiometric ratio of one in order to
ignite the mixture. So engines won’t run in the low emission region on the right side
of Figure 5-10. Viable concepts for dramatically reducing emissions from both rich
burn and lean burn engines are discussed in Chapter 6. However, the news is much
less promising for diesel engines where diffusion flames thwart low NOx concepts.
Reciprocating engines are factory built as opposed to many boilers that are built on
site. In contrast to boilers, their combustion systems and operating controls are built-
in and not amenable to simple adjustment or alteration. Any changes or alterations to
the engine are usually provided by the manufacturer or developed in conjunction with
the manufacturer. This is in contrast to boiler burners that are readily adjusted,
modified or changed completely. Thus engines generally have much more
predictable NOx emissions than boilers.
LOW EMISSION
RICH-BURN LEAN-BURN COMBUSTION
2000
1800
1600
1400
1200
1000
NOx
800
CO
600
400
NMHC
200
0
14 16 18 20 22 24 26 28 30
A/F RATIO
STOICHIOMETRIC A/F
Figure 5-10. Emissions vs. Excess Air for Gas-Fired Engines

Environmental Protection Agency. Alternative Control Techniques Document: NOx Emissions from Stationary
Reciprocating Internal Combustion Engines, July 1993, EPA Report 453/R-93-032, NTIS PB94-104494.

5.5.5.4 Combustion Turbines
Combustion turbines are aircraft “jet” engines adapted for stationary power sources.
Combustion in turbines is steady state; starting the engine corresponds to establishing
a stable fire in the combustor. The new “dry low NOx” turbine combustors premix
fuel and air, but all earlier combustors have a hot diffusion flame that generates
relatively high NOx emissions.
Combustion turbines can, in concept, burn any liquid or gaseous fuel. But in practice
they only burn clean fuels, and the cleaner the better. Any contaminants in the fuel
will eventually deposit on the turbine blades leading to frequent and expensive
maintenance. So, for the most part, NOx is the only significant emission from gas
turbines. Uncontrolled NOx emissions from turbines are higher than from boilers, but
water injection has kept NOx emissions from diffusion-flame units relatively low
beginning in the mid 1970s.
Like reciprocating engines, the NOx emissions from gas turbines are very predictable
because there are no “non-factory” variations from one engine to another. Given the
engine load and ambient conditions, NOx emissions can typically be predicted about
as accurately as they can be measured by an emissions monitoring system. NOx
emissions will vary predictably with ambient conditions unless there is a NOx control
device that is programmed to compensate. It is common practice to integrate NOx
control with the turbine fuel control. Any increase in ambient absolute (not relative)
humidity will reduce NOx emissions. The reason is that water vapor has a very high
specific heat, so as high humidity cools the flame, NOx emissions decrease. The
effect can be significant in areas with humid summers or if the engine inlet has a
fogging system designed to increase engine power on hot days.
5.6 SMOKE, CARBON MONOXIDE AND ORGANIC

COMPOUNDS
If combustion is complete there will be no products of incomplete combustion (PIC)
and no emission of black smoke, CO or organic compounds. Most stationary sources
come close to this ideal most of the time so emission of PIC are not a major concern.
For many sources, small or moderate amounts of PIC are emitted during initial start-
up and drop to negligible levels as soon as the combustor is warmed up.
As a general rule, black smoke and CO are the primary PIC [normally black smoke is
composed of fine particulate carbon]. The amount of organic compounds is nearly
always less than the amount of CO – in most boilers organic compounds are at non-
detectable levels even when there is significant CO and/or visible smoke. The
amount of organic compounds from combustion sources is isn’t normally sufficient to
create a visible mist or white plume, except during start-up. So a visible white plume
usually represents water vapor or inorganic species, not PIC.

5.6.1 Complete Combustion and Fuel-Air Mixing
When fuel is burned there is a complex series of chemical reactions leading to the
final products of H2O and CO2. The last reaction in the string is normally the
oxidation of CO to CO2. Thus when there is a low level of CO in the exhaust it is a
strong indication that all the hydrocarbons have been consumed. Viewed in another
way, boiler combustion zone exit temperatures are typically 1800°F or above.
Practical experience shows that when there is CO remaining at the furnace exit, it
does not burn up and will appear at the stack; in effect it stops burning at about
1800°F. Virtually all hydrocarbons break down at temperatures in the range of
1400°F to 1600°F. So when CO is low, one is virtually assured that combustion is
complete and that hydrocarbons are destroyed. In practice, the minimum CO levels
from boilers are in the range of 5 to 50 ppm. Hydrocarbons are highly unlikely to be
present when CO is less than 200 to 300 ppm and, depending on the source, may not
be measurable even at CO levels of 2000 to 3000 ppm.
The key elements of complete combustion are:
• An effective air-fuel mixing pattern. This is mostly a matter of burner geometry

to control the injection flow pattern of fuel and air. Pulverized coal must be finely
ground and residual oil must be finely atomized. In a stoker, the fuel must be
evenly distributed on the grate and the overfired air pattern must provide effective
mixing.
• Sufficient O2, meaning precise operational control of excess air.
• The absence of cool combustion quenching regions in the flame. This is mostly a
matter of combustion chamber geometry, but the cold walls of a reciprocating
engine combustion chamber are unavoidable. Too much air in the wrong place
can also quench the flame.
Some texts offer a rather simple criteria for good combustion, touting the need for
adequate time, temperature and turbulence. – the “3 T’s of combustion” While these
are certainly basic combustion parameters, they are seldom helpful in assessing
combustor performance or in resolving problems. Given a good flame pattern (e.g.
mixing pattern), the trouble with these parameters is:
• Combustion is normally very fast, tenths of a second or less, and simply

increasing residence time in a combustion chamber will not solve an underlying
problem of ineffective mixing caused by poor combustor design. Residence time
is relevant to basic combustion design but is seldom a useful consideration when
evaluating an existing combustion system.
• Adequate temperature to drive the combustion chemistry is an automatic result of

good combustion in most combustors. Incinerators are the only devices where
adequate temperature is an issue. Good thermal oxidizers achieve essentially

100% destruction of organic compounds at about 1400°F. Solid waste
incinerators and stoker furnaces require higher average temperatures because of
unavoidable temperature stratification and cold spots. In general, raising the
temperature on these devices is an expensive way to mask an underlying problem
of an ineffective air-fuel mixing pattern.
• High turbulence is generated by high velocity air fed through the burner, which
results in a high-pressure drop. Pressure drop does not need to be large in a well-
designed burner, as witnessed by most of the burners in existence. Gas turbine
combustors, with very high intensity combustion, require a pressure drop of less
than 3% of the total pressure. Boiler burners operate well at a 1% pressure drop,
about 4 inches w.g. High turbulence, or pressure drop, is not a fundamental
requirement for good combustion although it may be required to achieve rapid
mixing. Turbulence or pressure drop alone tells little about how well the fuel and
air are mixing.
5.6.2 Burner Geometry

Burner geometry, as discussed in Chapter 2, governs air-fuel mixing. The basic
objective is to completely mix, at a molecular level, all the fuel with all the air.
Mixing has to be substantially completed within the flame zone – meaning that it has
to be fast. Outside of the flame zone temperatures drop, chemical reactions slow
down, and PIC emissions may result. The flow pattern in the flame is controlled
almost entirely by the air flow pattern emerging from the burner (e.g., it is governed
by the burner air distributor geometry). Air-fuel mixing is controlled primarily by the
fuel injection pattern and the size distribution of fuel particles. A secondary factor is
the flow and turbulence pattern established by the air. Thus the various geometric
aspects of the burner govern air-fuel mixing. A well-designed and adjusted burner
will give complete combustion and negligible PIC emissions, provided the firing rate
and the excess air levels are within the normal range of the burner.
Having emphasized rapid mixing, low NOx burners introduce another twist. Low
NOx burners are explored in detail in Chapter 6, but a key element is to control the
mixing very precisely in a way that seems to contradict traditional good combustion.
Combustion is staged and the flame is lengthened. PIC may increase slightly, but
NOx is substantially reduced.
5.6.3 Excess Air

Excess air was discussed qualitatively in Chapter 2 and quantitatively in Chapter 4.
All combustors need a certain minimum amount of excess air to minimize CO and
smoke. A well-designed burner will operate at very low excess air levels without
producing PIC. In fact, burner performance is characterized by the amount of excess
air required to minimize CO and smoke. The best boiler burners will operate with
less than 5% excess air or about 1% O2. Reciprocating engines fired on gas or
gasoline also operate in this excess air range.

In the days when the main objective of combustor operation was to minimize PIC,
the only requirement for air flow control was to maintain sufficient excess air to
eliminate smoke. With the advent of NOx control requirements, air flow must be
controlled much more precisely; it must continuously be just enough to minimize PIC
and minimize NOx emissions.
Note that while insufficient excess air is the most common cause of PIC emissions,
they can also be generated by very high excess air levels, which can reduce average
temperatures to the point of local flame quenching.
5.6.4 Incinerator Temperatures

As noted in Chapter 2, most well designed thermal oxidizers achieve complete
hydrocarbon destruction at temperatures of about 1400°F. When incinerators and
oxidizers require higher average or exit temperatures to destroy organic species, the
reason is usually that the temperature is not uniform and the higher average
temperature warms up the cool spots. In a thermal oxidizer, which destroys organic
gases, good performance can be obtained by adjusting the combustor geometry to
achieve effective mixing and thus uniform temperatures.
Figure 5-11 shows the theoretical destruction efficiency of certain hazardous organic
materials as a function of residence time and temperature. The curves were drawn for
typical solvent materials. Note that there is a fairly sharp temperature demarcation
between little effect and complete destruction. Clearly there is no advantage to
operating at temperatures far above the destruction temperature. Note also that there
is not much difference between a residence time of about 0.5 and 2.0 seconds – the
range of practical interest. Long residence time reduces the required temperature by a
very small amount. The curves tend to reinforce the point that a well-designed
thermal oxidizer does not need either a very high temperature or a long residence time
to work well.
100
1 sec.
.1
80 0.01
0.001
60
Increasing Residence Time
40
20
0
800 1000 1200 1400 1600 1800 2000
TEMPERATURE, oF
Figure 5-11. Organic Material Destruction vs. Temperature and Time

Solid waste incinerators will inherently have substantial temperature variations in the
combustion zone. Good design of the fuel feed and overfired air systems can
minimize these variations, but average exit temperatures above 1600°F to 1800°F are
typically necessary to achieve complete combustion in the coolest areas. If a solid
waste combustor is poorly designed or improperly operated, exit temperatures as high
as 2000°F may not be sufficient to achieve complete combustion. Note that when
operating a solid waste incinerator, excess air levels of much more than 100% (10.5%
O2) will drop the average exit temperature to the point that some flame quenching and
incomplete combustion are likely (see Figure 4-5).
5.6.5 Dioxin-Furan Formation

Polychlorinated-dibenzo-dioxins (PCDD) and polychlorinated-dibenzo-furans
(PCDF) are very toxic compounds that are chemically stable and environmentally
persistent. Their toxicity combined with their ability to accumulate in the food chain
has led to regulation at levels measured in nanograms per cubic meter of exhaust gas.
They are produced almost exclusively by combustion systems when the fuel contains
chlorine. In addition, there must be some level of incomplete combustion because
organic species cannot be created once all the carbon has oxidized to CO or CO2.
Solid waste incinerators are the sources most likely to meet these conditions and,
indeed they have been a significant source of PCDD and PCDF. Attempts to
reconcile the amount of PCDD in the environment with the amount emitted from
various sources have led to conflicting conclusions about which sources are the
largest contributors. Other sources include forest fires and automobiles – where the
amount of chlorine in the fuel is very small, but the amount of fuel burned is very
large. Utility boilers probably emit some PCDD, but the amount of organic material
in their exhausts is undetectable by normal test methods and PCDD would have to be
generated from carbon.
Dioxin-furan formation can be eliminated by completely burning the fuel. The most
efficient method of generating dioxins is to incompletely burn chlorinated aromatic
compounds such as chlorinated phenols once used as coolants in electric
transformers. Any time there is a fire in an old transformer, downwind or adjacent
areas are heavily contaminated with dioxins. Perhaps the next best way to emit
dioxins is to incompletely burn any chlorine containing waste – such as PVC plastic.
Municipal and hospital waste incinerators built before 1980 frequently fit in this
category. Most of these sources generated substantial amounts of dioxins during
combustion, although most of them have been regulated out of existence.
Research since about 1990 has shown that dioxins and furans can be created in the
exhaust gas after the combustion zone – primarily at temperatures in the range of 250-
300°C (480° - 570°F), which can include the normal operating temperature of a dust
collector. The mechanism is a chemical reaction between organic PIC and molecular
chlorine (Cl2) in the presence of fly ash that apparently acts as a catalyst. This
synthesis apparently requires a time span of several seconds. The way to avoid
dioxin formation is (1) to completely burn the fuel, and (2) to cool the gas rapidly

through the critical temperature range of 250°-300°C. This basically means avoiding
the installation of a dust collector or a long duct where the exhaust gas is in the
critical temperature range. A baghouse operating in the critical temperature range can
be particularly troublesome because flue gas passes through the filter cake, which
appears to be a catalyst for dioxin formation.
There is some evidence that dioxins might be formed from carbon; although this
appears to be much less efficient than forming dioxins from organic material. This
possibility reinforces the need for the dust collector to operate well outside of the
250°-300°C temperature range.
5.7 OPACITY
Plume opacity regulations and measurement procedures were first promulgated in the
mid 1970s as a surrogate to particulate test methods. Opacity regulations were based
in part on the reasoning that the amount of particulate should be related to the relative
opacity of the plume. Particulate testing is cumbersome, time consuming and usually
gives the source ample opportunity to prepare for the test. The resulting data shows
the capabilities of the source, but may have little bearing on its normal performance.
Opacity monitoring by Method 9, on the other hand, can be performed at almost any
time and at a considerable distance from the source. So opacity monitoring by trained
observers was intended to provide an indication of how well a source was complying
with a required particulate emission rate. However, since there is not a clear
correlation between opacity and particulate emissions, the opacity requirements are
enforceable separately from the particulate emission requirements.
With certain sources, such as pulverized coal-fired boilers equipped with electrostatic
precipitators, the particulate emissions rate usually correlates fairly well with the
relative opacity. On average, these sources have emissions less than about 0.1
lb/mmBTU when the plume opacity is less than about 20%. However, on other
sources the quality of the correlation can deteriorate rather badly. The first difficulty
with trying to relate opacity to particulate emissions is that opacity depends as much
on particle size as it does on mass of emissions. Figure 5-12 is the theoretical
relationship between particle size and opacity. The ordinate shows the relative
opacity of a fixed amount (mass) of particulate distributed in a fixed volume of air.
For particulate matter larger than 1-2µm, the opacity is inversely proportional to
particle size. 5µm particles will generate 10 times the opacity of the same amount of
material in 50µm particles. As particle size approaches the wavelength of visible
light (0.4 - 0.7µm), the relative opacity depends on particle composition as well as
particle size.
A coal-fired power plant with an electrostatic precipitator will emit particulate with a
fairly constant size distribution. The precipitator controls emissions and the size
distribution does not change dramatically as precipitator performance changes. So
opacity variations generally reflect changes in particulate mass emissions. If the

source has a baghouse, then this correlation breaks down. When emissions increase
because of holes developing in the bags, there is a change in the size distribution of
particles being emitted. Particles are no longer the tiny particles that seep through the
filter. Particles going through holes in the bags have the same size distribution as the
inlet particles. This means the particulate emissions from a baghouse-controlled
source can increase substantially with little impact on plume opacity.
100
10.0
1.0
0.1
0.01 0.1 1.0 10.0
Invisible Invisible
Gases Mean Particle Radius (Microns) Particles
Figure 5-12. Dust Opacity vs. Particle Size

Interpretation of data by Ensor & Pilat, 1975
The correlation between opacity and emissions rate on a residual oil-fired source is
capricious. Black or gray visible plumes, associated with low excess air, are caused
by submicron soot, which has maximum opacity, but relatively little mass. If the
source emits coke from poorly atomized oil, the mass emissions can be large, but
these 100+µm particles are essentially invisible in the plume. So this source can have
a highly visible plume with low emissions or an invisible plume with high emissions.
In short, plume opacity tells almost nothing about the particulate mass emissions from
a residual oil-fired source.
Sulfuric acid in the exhaust introduces further complications. The acid will
contribute significant mass to a Method 5 (filter) test depending, to some extent, on

the operating temperature of the test equipment. Although it is not normally visible
in the stack it may generate a highly visible condensation plume after it leaves the
stack. The plume can be clear at the stack exit, but increase steadily in both
dimension and opacity as it moves downwind. Both the visibility of the plume and its
color depend on the angle of the sun and other atmospheric factors.
Plume opacity observations are an effective way to gauge the visual impact of a
source, but they may or may not be related to the particulate mass emissions rate.
5.8 CONCLUSION
This chapter explores air pollutant formation in combustion sources, while the next
chapter reviews air emission control. For most pollutants this is a logical separation
of topics. However NOx formation and control are closely linked when combustion
controls are used. This linkage is addressed in part by providing a modest overlap in
these two chapters in the area of NOx emissions.
Air emissions from combustion sources are generated by inorganic elements in the
fuel, by incomplete combustion of the fuel, and, in the case of NOx, simply by the
heat of combustion. The maximum possible amount of inorganic emissions from a
source can be bracketed using a simple conservation of mass calculation. However,
this approach does not give a good estimate of actual emissions for most pollutants –
sulfur/SO2 being the main exception. Emission control devices capture several
pollutants, particulates being the best example, so that only a small fraction of the
original fuel contaminant is actually emitted. In the case of NOx, the amount of fuel
nitrogen converted to NOx depends on the design and operation of the combustor. In
addition, a portion of the total NOx emissions is thermal NOx, which is independent
of fuel nitrogen. Many new sources include back-end NOx control devices, in which
case there is little relation between fuel nitrogen content and NOx emissions.
Emissions from incomplete combustion vary from near zero on many sources to
potentially significant amounts of CO, smoke, HAPs and particulates. Although there
are guidelines, such as AP-42, on what to expect from different types of sources,
emissions of various products of incomplete combustion (PIC) are strongly dependent
on combustor design and operating conditions. Dioxins and furans are products of
incomplete combustion, but some secondary formation can occur in the backend of a
boiler – usually involving fly ash in the dust collector as a catalyst.
Particulate emissions can be roughly divided into two groups. Most of the mass
(weight) of emissions is in particles larger than about 1 or 2µm (larger than PM2.5).
Large particles for the most part are direct descendents of fuel particles with
essentially the same composition as the original fuel ash. The second group is fine
particles that are smaller than about 1µm and are derived from gases in the
combustion zone that condense or combine with other species to form solid particles.
Fine particles are composed almost entirely of species that were in the vapor phase at

some time during combustion. Hence the fine particle chemistry includes sulfates
and various metals like lead, arsenic and beryllium. The fine particles are generally
more water soluble and more toxic than the large particles. The relative mass of fine
particulate may be small, but this material is frequently responsible for most of the
plume opacity from a combustion source. Hence fine particulate emissions are just as
environmentally significant as large particle emissions even though they may
constitute a small fraction of the actual or potential mass emissions.

Review Exercises
1. For which of the following pollutants can the uncontrolled emission rates not be predicted
by a fuel analysis?
a. SO2
b. Dioxins and furans
c. Particulate from pulverized coal firing
d. HCl
2. Calculate the SO2 emissions for an uncontrolled source firing 1.0% sulfur oil. Assume the
oil higher heating value is 18,500 BTU/lb.
a. 0.54 lb/mmBTU
b. 0.93 lb/mmBTU
c. 1.08 lb/mmBTU
d. Insufficient data to determine the emission rate.
3. Which of the following fuels cannot generate pollutants other than NOx and CO in
significant quantities.
a. Landfill gas
b. Distillate oil
c. No. 6 oil
d. Natural gas
4. Sulfur oxide emissions exhibit which of the following characteristics?

a. Nearly all fuel sulfur is converted to SO2 in the combustion zone.
b. Any SO3 formed during combustion will be in the form liquid sulfuric acid when
cooled to ambient temperature.
c. Small amounts of sulfuric acid can form a highly visible plume downwind of a stack.
d. All of the above
5. Which of the following statements is not true?

a. Any particle larger than 5 µm was probably formed by the breakup of a larger
particle.
b. Any particle smaller than 1µm is composed largely of condensed gas-phase species.
c. The chemical composition of fine particulate (<1µm) is similar to that of large
particles.
d. Fine particulate is much more visible (higher opacity) than an equal mass of large
particulate.

6. What is the maximum possible emission rate for particulate leaving a pulverized coal
furnace when the coal ash content is 12.6% and the higher heating value is 10,900 BTU/lb?
a. 11.6 lb/mmBTU
b. 5.8 lb/mmBTU
c. 0.12 lb/mmBTU
d. 10.4 lb/mmBTU
7. In a typical large boiler, particles of coal or drops of No. 6 oil larger than 200 µm will do
which of the following?
a. Burn completely because of their long residence time in the furnace.
b. Burn incompletely and created carbon particle emissions.
c. Fail to ignite
d. Require a very low excess air level to allow sufficient residence time for combustion
8. Estimate the sulfate contribution to the particulate emissions from a No. 6 oil-fired source
if the oil has 0.9% sulfur, a heating value of 18,100 BTU/lb and 2.5% of the sulfur is
oxidized to sulfate?
a. 0.037 lb/mmBTU
b. 0.0124 lb/mmBTU
c. 1.49 lb/mmBTU
d. None of the above
9. Which statement is not true of mercury emissions from coal-fired sources?

a. Most of the mercury in the exhaust is in the vapor phase.
b. The low concentration of mercury in the typical stack is a result of mercury capture in
the dust collector
c. Although coal contains only ppm levels of mercury, the emissions from an electric
utility can amount to several pounds each day.
d. Air emission of mercury will be reduced if there is a substantial amount of carbon in
the fly ash.
10. Which of the following statements is true of NOx emissions?

a. NOx emissions from a furnace will increase as excess air increases.
b. If fuel and air are premixed before combustion, NOx can be reduced by reducing
excess air.
c. Most combustion modifications that reduce NOx also reduce CO and PIC emissions.
d. Fuel NOx formation can be reduced by reducing excess air
11. Reciprocating engine NOx emissions are ________.

a. high because of the amount of nitrogen in the fuel
b high because of high temperatures during combustion
c. high because engines operate at high excess air levels
d. not much higher than other sources fired on similar fuels

12. Complete combustion generally requires which of the following?
a. Effective air-fuel mixing and adequate excess air.
b. Adequate amounts of time, temperature and turbulence.
c. An absence of cold surfaces in the combustion zone.
d. A and C above
13. Which statement is not true of the destruction of organic gases in a thermal oxidizer ________.
a. improves dramatically when residence time is increased beyond 0.5 seconds
b. is relatively independent of residence time in actual (as opposed to theoretical)
oxidizers.
c. is typically near 100%, then drops rapid if temperature drops below a critical levels
d. the minimum temperature is about the same for destruction most organic gases
14. Dioxins and furans ________.

a. can be formed by catalytic reactions on fly ash particles in a dust collector at a
temperature of 500°F
b. will not be formed if there is complete combustion of organic species in the
combustion zone
c. cannot be formed if there is no chlorine in the fuel
d. All of the above
15. Which of the following statements is true regarding opacity?

a. Sulfuric acid can form a mist of particles too fine to be visible.
b. Large particles emitted from a coal or residual oil-fired boiler have so much mass that
the plume is always clearly visible.
c. The opacity of a given mass of particles increases linearly as the particle size is
decreased down to about the wave length of visible light.
d. Emissions of black soot emissions are visible primarily because of the large mass
emission rate.

Review Answers
1. b. Dioxins and furans
2. c. 1.08 lb/mmBTU.
3. d. Natural gas
4. d. All of the above.
5. c. The chemical composition of small and large particles is similar.
6. a. 11.6 lb/mmBTU
7. b. Burn incompletely and create carbon particle emissions.
8. a. 0.037 lb/mmBTU
9. b. The low concentration of mercury in the typical stack is a result of mercury capture in
the dust collector
10. a. NOx emissions from a furnace will increase as excess air increases.
11. b high because of high temperatures during combustion
12. d. A and C above
13. a. improves dramatically when residence time is increased beyond 0.5 seconds
15. c. The opacity of a given mass of particles increases linearly as the particle size is
decreased down to about the wave length of visible light.

References
Buonicore, Anthony J. and Davis, Wayne T., (Editors). Air Pollution Engineering
Manual, (Air & Waste Management Association, Van Nostrand Reinhold, 1992).
This book has sections on combustion sources as well as various emission control
technologies.
California Air Resources Board. Sources and Control of Oxides of Nitrogen

Emissions, August 1997. This gives a relatively current summary of NOx formation
and control technology for all types of sources. Summaries of California rules
illustrate NOx levels that can be achieved by state-of-the-art technology.
Clement, Raymond and Kagel, Ronald. Emissions from Combustion Processes:

Origin, Measurement, Control, (Lewis Publishers, 1990). The book is a collection of
31 articles that focus primarily on organic pollutants.
waste combustion in rotating kilns and covers the formation and control of organic
pollutants.

NOx Emissions from Industrial/Commercial/Institutional (ICI) Boilers, March 1994,
EPA Report 453/R-94-022, NTIS PB94-177177. This weighty document is a fairly a
comprehensive description of small to mid size boilers and current NOx technology.
It includes waste-fired boilers.

database included.


NOx Emissions from Utility Boilers, March 1994, EPA Report 453/R-94-023, NTIS
PB94-184165. This weighty document is a fairly a comprehensive description of
large boiler types and NOx technology with a significant database included.

U.S. Environmental Protection Agency. Compilation of Air Pollution Emission Factors
(AP-42), available from NTIS, Federal Book Stores, or you may download the latest
version from the EPA TTN BBS. This ever-evolving document is organized by source
type and has credible descriptions of each one. It is generally the "bible" for estimating
emission rates whenever actual measurements are not available. There are lengthy
background documents for each section of AP-42.
U.S. Environmental Protection Agency. New Source Performance Standards,

Subpart Da – Technical Support for Proposed Revisions to NOx Standard, June 1997,
EPA Report 453/R-94-012. This inch thick report is narrowly focused on utility
boilers, but covers current NOx technology in great depth.
U.S. Environmental Protection Agency. Phase II NOx Controls for the MARAMA
and NESCAUM Regions, November 1995, EPA Report 453/R-96-002. This report
provides an in-depth coverage of NOx technology on utility boilers.
U.S. Environmental Protection Agency. The US EPA-DOE-EPRI Combined Power

Plant Air Pollution Control Symposium: The Mega Symposium and the A&WMA
Specialty Conference on Mercury Emissions: Fate, Effects and Control, 2001,
proceedings available on CD-ROM from www.awma.org. This and other conference
proceedings provide current information on air pollutant formation and control.


Chapter 6: Air Pollution Control Systems
LESSON GOAL
Demonstrate an understanding of how various devices and methods are used to
control air pollution emissions from combustion sources through successful
completion of the chapter review exercises.
STUDENT OBJECTIVES
1. Describe the basic particle collection mechanisms of impaction, filtration, and

electrostatic attraction.
2. Describe one or more types of inertial particle collectors and the size range of
particles applicable to each.
3. List the main advantages of particulate scrubbers and discuss how they work.
4. Identify the advantages and operating limitations of baghouses and discuss their
operating principles.
5. List the advantages and limitations of electrostatic precipitators and discuss how
they work.
6. Describe the most effective means of monitoring the performance of each major type
of particle collector.
7. Identify how SO2 emissions can be controlled by fuel switching, as well as dry or
wet scrubbing.
8. Describe why SO3 emissions can be difficult to control.
9. List three general methods of reducing NOx formation from boilers or furnaces.
10. List two predominant temperature NOx reduction methods used in boilers and
combustion turbines.
11. Define the meaning of FGR and the limitations when using it on boilers that fire a
sulfur bearing fuel.

12. Discuss what stratified combustion is and how it is applied in several ways to
practical combustion systems.
13. Describe how reburning works to reduce NOx and the types of sources on which it is
most effective.
14. Identify what constitutes a low NOx burner.
15. Describe the difference between diffusion and premixed flames and how NOx
control works in premixed flames.
16. Discuss the basic concepts behind low NOx burners used in gas turbines and
reciprocating engines.
17. Identify the basic elements of SNCR, SCR and NSCR NOx control technologies.

Chapter 6: Air Pollution Control Systems
6.1 INTRODUCTION
Pollutants emitted from combustion sources are formed in the combustion zone and,
unless they are captured by a control device, they will be emitted to the atmosphere
with the exhaust gases. Small amounts of two pollutants, sulfuric acid and
dioxins/furans, can be created in cooler areas downstream of the combustion zone.
For all practical purposes all the other pollutants are not significantly created nor
destroyed once they pass beyond the flame.
Control of air emissions may be done in the combustion zone, with a control device,
or with a combination of the two. In many cases, even if a combination of controls is
not used, the performance of the control device can be strongly affected by the
operation of the combustion system. Thus, air emissions control should always be
viewed as a synergistic process involving both the combustion zone and the control
device. This chapter provides a brief description of control technologies for most
pollutants with a focus more on the relation to the combustion source than on the
details of various controls. A partial exception is NOx control where much of the
actual control is the design and operation of the combustion system.
Control devices can be combined, they may act synergistically, or more than one
pollutant may be controlled with a single device. For example, if a caustic dust is
blown into the exhaust gas upstream of a baghouse, the baghouse will collect acid
gases as well as particulate. Carbon dust on the baghouse filters will reduce mercury
emissions. If coal fly ash is collected ahead of a lime SO2 scrubber, then both the fly
ash and the scrubber residue may have commercial value/uses. There are numerous
ways in which control devices may be integrated or combined, depending on the fuel,
the source, the environmental requirements and potential reuse/recycle markets.
Emissions are almost always affected by the choice of fuel and emissions control, in
some cases, is achieved entirely by choosing the fuel or specifying fuel properties. The
pollutant that is most commonly the driving force for switching fuels is SO2, where the
emissions rate is directly related to fuel sulfur content. But NOx emissions can also be
strongly affected by the type of fuel – especially when choosing between gas and
residual oil or coal. Fuel properties can directly affect the performance of emission
control devices. For example, the performance of an electrostatic precipitator may be

influenced by the amount of alkali in coal ash. And, the durability of a catalyst used to
control NOx can be affected by sulfur and trace elements in the fuel.
The choice between different types of fuel, say between oil and gas, or between oil
and coal, will obviously involve the capabilities of the hardware, relative fuel costs,
emission controls/limits and several other factors. But even a choice between two
types of coal can be sharply limited by combustion system hardware capabilities. For
example, switching from an eastern bituminous coal to a Powder River Basin low
sulfur coal can trigger serious ash/slag problems around the burners or problems with
electrostatic precipitator performance. While the general nature of such problems
may be anticipated, the magnitude and details are specific to the combustion system
and the particular fuel. So a source might potentially achieve its emission reduction
objectives with a fuel change, but there are usually significant side effects, which can
be critical to the feasibility or viability of the fuel choice.
6.2 PARTICULATE MATTER AND METAL EMISSIONS

CONTROL
Particulate matter emissions are typically controlled with some type of dust collector.
The type of dust collector depends upon the required collection efficiency, the size of
the particulate and to a lesser extent on the particle properties. When high collection
efficiency is required, either an electrostatic precipitator (ESP) or a baghouse is
usually required. If there is no need for high efficiency collection of fine particulate,
then an inertial (cyclone) collector may be adequate. If the particulate tends to be
sticky, or if there is a severe threat of a dust collector fire, then a scrubber or a wet
ESP may be the best choice.
Particle collectors are based on three basic mechanisms with several variations and
combinations.
• Inertial Collectors. Inertial collectors work by accelerating or curving the gas

flow so that particles cross the streamlines and impact or agglomerate on a
surface. A particulate scrubber is a special type of inertial collector.
• Filters. Filters create passages that are too small for particles to pass through
although ultimately filters also work by impaction.
• Electrostatic Collectors. Electrostatic collectors use an electric field to draw

charged particles to a collection surface.
The performance of all collectors is governed in large part by the behavior of

different size particles. Figure 6-1 illustrates the movement of different size particles
relative to streamlines; the illustration is not drawn to scale.

Streamline
Fine <.05 microns
Small 1-2 microns
Large 5-10 microns
Figure 6-1. Particle Motion vs. Gas Streamlines
The movement of particles larger that about 5µm to 10µm is substantially influenced
by inertia. When the flow goes around a bend, these particles tend to go straight; in
other words, they cross the streamlines. So when the flow is forced in a circular path,
as it is in a cyclone, these particles tend to migrate to the walls where they agglomerate
and can be collected.
Mid-size particles from about 0.5µm to 5µm tend to follow the gas streamlines very
closely. They are essentially locked or frozen into the flow so that when the flow
curves around an obstacle, such as a small drop of water in a scrubber, the particles
follow the streamlines and miss the obstacle. In general, the smaller the obstacle, the
sharper the flow bend and the smaller the particle that will impact obstacles in the
flow. Impactors, used to measure particle size distribution, pass the flow through a
series of smaller and smaller orifices to classify the particulate into different sizes.
Small cyclone collectors will capture smaller particles better than large cyclone
collectors.
Particles smaller than about 0.5µm basically follow the streamlines, but they undergo
additional random motion. They are small enough that the gas molecule collisions
are strong enough to move them. They undergo Brownian motion, which means they
cover more distance and have an increased likelihood of impacting other particles or a
solid surface.
6.2.1 Particle Collectors

Particle collectors can be divided into four categories, although some devices are a
combination of categories.
• Inertial or mechanical collectors

• Scrubbers (a form of inertial collector
• Filters
• Electrostatic precipitators

The basic operation of each type of collector will be discussed, followed by a brief
explanation of some innovative combinations.
6.2.1.1 Inertial Collectors

Inertial or mechanical collectors work by accelerating the flow (a curved flow path
accelerates the flow in a radial direction), which causes large particles to migrate
across the streamlines to a collecting surface. Cyclones and multiclones are the most
common inertial collectors. These mechanical collectors are simple and relatively
small. In addition, they require little maintenance or operator attention and have
proven to be very reliable. However, for the most part, they do not efficiently collect
particles smaller than about 5-10µm. So their use in air pollution applications is
limited to cases where most of the particulate is relatively large and where low to
moderate collection efficiency is acceptable. Figure 6-2 shows two typical cyclone
collectors. The basic concept is to centrifuge the particulate to the chamber wall
where it agglomerates into larger particles that drop down into the hopper. They are
typically designed for a pressure drop of about 4 in. w.g. Moderate variations in flow
or pressure drop usually have little influence on the collection efficiency, but they
lose efficiency at high and at low flow rates.
Clean Flow
Outlet
Inlet
Dirty Gas
Solids
Cyclone Collector
Solids
Multiclone Collector
Figure 6-2. Basic Inertial Collector
Cyclone-type collectors are commonly used in industry to separate solid materials

like plastic pellets from pneumatic transport systems. In these applications, they may
be essentially 100% efficient. Air pollution control collection efficiency for particles

larger than 50µm usually exceeds 90%. Efficiency for smaller particles depends in
part on the size of the cyclone – smaller cyclones will collect smaller particles. A
sawmill may use a 6-8 foot diameter cyclone to capture chips from a saw, while the
wood-fired boiler will use a multiple set of 8" diameter cyclones to capture most of
the particulate down to 10µm or somewhat smaller. Overall collection efficiency for
a multiclone system on a solid fuel-fired boiler seldom is much above 80%. Most
stoker boilers use a multiclone as a primary collector to catch larger particles, some of
which may still be burning. Newer units will also have a high efficiency secondary
collector to meet air emission requirements.
Because inertial collectors are not efficient when collecting submicron particulate,
they cannot be used to control particles formed by condensation such as fume or
black smoke. In fact, their ability to control opacity is limited to those sources where
the opacity is caused by the emission of relatively large amounts of large particulate
matter. When an inertial collector is the only control device on a coal- or wood-fired
stoker, the plume is usually visible even when there is no black smoke.
Inertial collectors are small compared to a baghouse or precipitator. A multiclone

collector may be integrated with a boiler or it may appear as no more than an enlarged
section of the duct with hoppers below. An inertial collector is a simple device with
no external adjustments/controls and almost no moving parts. Thus, its performance
tends to be very consistent and degraded performance tends to be associated with one
of the following:
• Wear. Erosion cuts a hole in an internal component, resulting in gas bypassing

the collector.
• Ash Build Up. Ash builds up in the hopper or flow path to the point that it
compromises the operation.
• Air Leaks. An air leak in the hopper will blow collected dust into the outlet gas
stream.
Although there are some monitoring or diagnostic procedures that may detect these
problems, periodic inspections and maintenance are probably the keys to consistent
collector performance.
6.2.1.2 Particulate Scrubbers

Scrubbers use water to “wash” the exhaust gas. They are commonly used to control
the emission of acid gases and less commonly used for particulate control. Scrubbers
have difficulty collecting fine particulate; however, they are more efficient than
mechanical collectors. Because of this tendency toward low collection efficiency
they are not normally used as the only particulate control device on combustion
sources. Some of the features of particulate scrubbers in comparison with other
particulate control devices are:

• Size. A scrubber is physically smaller than a precipitator or baghouse.
• Cost. Gas pressure drop across a particulate scrubber is high, so fan power
becomes a significant operating expense. Scrubbers are also considerably cheaper
to build than a baghouse.
• Hazards. A scrubber may be attractive if the dust is flammable and there is a high
probability of a collector fire.
• Supplemental Equipment. Some sort of water treatment facility is required before

waste water can be discharged.
Thus, scrubbers usually are used in combination with other control devices and/or on
unique combustion sources.
The basic concept of a particulate scrubber is to spray water into the flue gas so the
particles are captured by water drops, then collect the water. Capture of the particles
occurs when they impact water droplets, so collection depends on the velocity
difference between the particles and the water drops. Once the particulate and the
water drops are combined, the water must be collected and managed for soluble and
insoluble particulate matter. While simply spraying enough water into the gas stream
will collect much of the particulate, small particle collection requires considerable
energy to create movement between the water drops and the gas streamlines. Figure 6-
3 shows the basic elements of a particulate scrubber. Accelerating and decelerating the
flow using a venturi or restriction in the flow is used to achieve mixing. The higher the
venturi velocity (scrubber pressure drop), the smaller the particles will be that are
collected. Pressure drops of at least 10" w.g. up to more than 30" w.g. are common.
But particles smaller than about 0.5 to 1µm cannot be efficiently collected regardless of
pressure drop. The reason is that fine particles follow gas streamlines around the water
drops even when the water drops are moving fast.
Once the water has captured the particulate, the water needs to be collected, which
means the venturi is followed by one or more mechanical collectors that impact the
water drops on a surface that drains to the scrubber sump. The final water collector is
called a mist eliminator and usually takes the form of a multi-layer mesh or other
tortuous path device.
Exhaust gas can enter a scrubber at any temperature where it is cooled by evaporation
and heating of the water. At the scrubber exit, the exhaust will usually be at the
saturation temperature, meaning some number below 200°F. Enough water must be
supplied to provide evaporative cooling and to scrub out the particulate. The exhaust
will be saturated with water and may carry some liquid mist as well, so any
downstream control device as well as any exhaust sampling system needs to be
designed accordingly.

The collection performance of a particulate scrubber is directly dependent on two
parameters: (1) gas pressure drop (delta p) and (2) water flow rate. Any plan to
monitor scrubber performance needs to include these as primary parameters.
Outlet
Inlet Temperature
Quench
Spray
Mist Eliminator
Venturi
with Dust
Collector
Sprays Possible
Packed Bed
Scrubber
Possible
Cyclone Separator
Configuration
Drain
To Water Treatment
and/or Recycle
Figure 6-3. Elements of a Scrubber
Example 6-1. Scrubber Water Flow
A scrubber is known to require about water at a flow rate of 15 gal/min per 1000 cfm
inlet flow rate. Stack measurements at the scrubber outlet show 22,000 acfm @ 155°F.
The scrubber inlet conditions are 435°F and 25 inches w.g. draft pressure. How much
water is required?
Solution:
Determine the inlet gas flow by correcting the stack flow to inlet conditions using Eqn 4-
14).
435 + 460 29.92

22,000 acfm × × = 30,163 acfm inlet
155 + 460 29.92 + 25
13.6
Calculate the water flow rate

⎛ gal ⎞
15 ⎜ min ⎟ × 30,163 (cfm) = 452 ⎛ gal ⎞
⎜⎜ 1000 cfm ⎟⎟ ⎜ min ⎟
⎝ ⎠
⎝ ⎠
The scrubber will require a water flow of about 452 gpm to operate properly.
6.2.1.3 Fabric Filters

A fabric filter or baghouse is a common, simple and very effective particulate control
device. When new, the collection efficiency approaches 100% and the outlet gas will
be nearly void of particles. In addition, baghouses can, depending on the chemistry
of the dust, remove gases such as SO2 or HCl, although this efficiency is usually less
than 90%-95%.
A baghouse is functionally similar to a vacuum cleaner. Gas is drawn through a filter

and the particulate accumulates on the surface. Filter bags are typically 6" to 8" in
diameter and can be 30' long. A large baghouse can have hundreds of bags.
Particulate can be collected on either the inside or the outside of the bags, depending
on the configuration. Figure 6-4 shows the basic elements of a baghouse. Gas flows
into the bag chamber, through the bags and out the top. Periodically the accumulated
dust cake is removed from the bags and falls into the hopper below.
Pulsed Outlet
High Pressure
Air to Clean Bags
Cloth Bags
Inlet
Dust
Figure 6-4. Fabric Filter Components

The bags must be cleaned automatically while the baghouse is in service. The two
most common methods of cleaning dust from the bags are:
• Pulse jet. Dust is collected on the outside of the bags with the open (clean) end
upward. A blast of air into the open end momentarily reverses the gas flow
through the bag and flexes the bag outward. This dislodges the dust cake, which
then drops into the hopper.
• Reverse flow. The baghouse is built with a number of compartments, each with
isolation dampers. When the bags are to be cleaned in one of compartments, it is
isolated and clean air is blown backwards through the bags knocking the dust
cake from the bag surface.
As dust builds into a layer on the bag surface, the pressure required to force gas
through the bag increases. Cleaning the bag reduces the pressure drop, so over a
period of a few hours, the pressure drop across a baghouse has a saw tooth pattern.
There may also be a long-term increase in pressure drop, but the average pressure
drop should reach a steady value after a period of only a few weeks.
Filters catch particulate more effectively than one would surmise from the size of the
openings between the fibers. Some particles are too large to pass through the filter
and are captured this way. But as soon as a layer of dust begins to form on the
surface, the fine particulate begins to be captured before it reaches the filter. Thus,
the filter cake becomes the primary collection media in which very fine particulate is
captured. The result is particulate capture efficiency approaching 100% for nearly all
particle sizes. The particles near 0.5µm diameter that follow the streamlines, can slip
through, but even with these, the collection efficiency tends to be 90% or better.
There are a variety of materials used for baghouse filters. The choice depends on
factors such as peak operating temperature, characteristics of the dust, and cleaning
method. In current practice, the maximum operating temperatures are about 550°F for
fiberglass bags. Although higher temperatures are possible, the associated materials
tend to be less durable. If the exhaust temperature significantly exceeds the bag’s
operating limit, damage to the bag causes either holes, or plugging (blinding), or both.
Baghouse Monitoring
A baghouse has two principal failure modes:
• Bag Wear. Holes or tears in the bags allow dust to pass through the holes and
bypass the fabric filter. Since bags normally last 3 years or more, holes are not
something one expects to find on a weekly or even monthly basis until the bags
approach the end of their life.
• Dust Accumulation. Dust accumulation that plugs up the bags resulting in an

increased pressure drop or reduction in flow rate, or both. There are two
subcategories: (a) temporary accumulations that can be removed by the normal

cleaning cycle, and (b) permanent "blinding" of the fabric that can only be fixed
by washing or replacing the bags.
Excess emissions are usually caused by holes or leaks in the bags; some of the gas
passes through the holes carrying its dust burden with it. So a primary objective of
monitoring (such as might be required for compliance assurance) should be to detect
leaks. The most sensitive monitoring technology in this case is tribo electric sensors.
A tribo electric probe generates a current (micro amperes) that is roughly proportional
to the mass of particulate striking it. These devices are sensitive enough to give a
baseline current reading when the bags are new (no leaks), so any increase in mass
loading can be detected as an increase in current. These devices are relatively
inexpensive (less than $2000 each) and they give a continuous output.
One might think that an opacity meter would provide an effective indication of bag
leaks, but in fact opacity is much less sensitive than a tribo-electric meter. The reason
is apparent once you realize that the opacity at the outlet of a leak-free baghouse is
probably on the order of 0.1%. Emissions would have to increase by about an order of
magnitude before the opacity meter detects an increase; it is simply much less sensitive
than the tribo-electric meter at particulate concentrations typical of the baghouse outlet.
Of course if the bags have holes and the opacity is already 5% or more, then any further
deterioration in the bags will cause a measurable increase in opacity.
Inlet-outlet pressure drop indicates the operational status of a baghouse, but is

insensitive to bag leaks, short of sensing a massive failure. For example, if holes
cause 1% of the flow to bypass the filter fabric, the emissions will increase by one to
two orders of magnitude, corresponding to 1% of the dust leaking through rather than
0.1% or 0.01% with no leaks. The holes reduce flow through the bag fabric by 1%,
which will reduce the pressure drop about 2% (Bernoulli's equation for fluid flow).
That change in pressure drop is much less than that resulting from bag cleaning and
fluctuations in gas flow, so it is not detectable. Emissions can increase by a factor of
10 with no measurable change in pressure drop.
Example 6-2. Effect of Baghouse Leaks
A new baghouse has a measured collection efficiency of 99.95%. Over a period of

time the bags develop leaks that allow 0.7% of the gas flow to bypass the fabric.
How much will the emissions increase and how much will the ∆p decrease as a result
of this leakage?
Solution:
Initially 0.05% of the particulate (100% – 99.95%) gets through the fabric filters. The
air leakage of 0.7% allows an additional 0.7% of the particulate to get through. Thus
the emissions increase from 0.05% to 0.7 + 0.05 = 0.75% of the inlet particulate. The
emissions increase by a factor of 0.75/0.05 = 15, or 15 times as much emissions.

The gas flow through the baghouse does not change, but with the leaks only 99.3% of
the gas (100% - 0.7%) is passing through the fabric. The pressure drop will decrease
in proportion to the square of the flow rate, which has decreased to 99.3%
of the original.
0.9932 = 0.986
The pressure drops to 0.986 of the original or a 1.4% decrease. Thus leaks would
increase emissions by a factor of 15 while only changing the pressure drop by 1.4%.
Although a decrease in pressure drop will not indicate anything less than a major bag
failure, a systematic increase in pressure drop can be used to show when the bags are
plugging up. As a new baghouse develops a steady dust accumulation, a modest
increase in pressure drop is expected. But an increase of more than 30%-50% is
cause for concern. Plugged bags will not immediately increase emissions but they are
likely to reduce flow rate, and plugging portends accelerated bag failure. When there
is a high-pressure drop across the bags, the fabric is eroded rapidly around any
pinhole, which then grows larger. So a high-pressure drop is a sign that remedial
action is needed to prevent accelerated failure.
Gas temperature has little affect on baghouse performance unless (a) high
temperature exceeds the fabric limit, or (b) low temperatures allow water or acid to
condense in the fabric. In the second case, the bag can become permanently plugged
(blinded) such that it requires bag replacement or washing. Severe temperature
excursions signal catastrophe while temperature anywhere in the normal bag
operating range has little emissions significance.
6.2.1.4 Electrostatic Precipitators

Precipitators are commonly used on large combustion sources and typically have
collection efficiencies nearly as high as a baghouse – well in excess of 99%. They
have two advantages over baghouses: (1) the gas pressure drop is negligible, and (2)
they are not easily damaged by modest temperature excursions or small dust fires.
But there are numerous problems that can degrade performance.
Precipitators work by electrically charging the particles, then drawing them to a

collector plate with a high voltage (electrostatic) field. Accumulated dust cake is
then removed from the plates by mechanical rappers that periodically bang the plates
or by a water wash. The physical hardware consists of parallel plates that form gas
passages typically 9" to 18" wide. In the middle of each passage is a set of wires or
rods maintained at 40 to 60 kv negative electric charge. A corona of electrons flows
from the electrodes to the plates. The electrons hit particles, making them negatively
charged. The DC electric field draws charged particles to the collector plates where
they accumulate into a dust layer and lose their charge to the plate. Because particles
get charged regardless of size, precipitators, like baghouses, can collect very fine
particles.

All the internal components are steel or ceramic, so they normally last many years
before major components need to be replaced. Operating limitations can be generated
either during the charging process or in the discharge/dust removal process. Most of
the charging problems result from reduced corona current that results when arcing
forces the power supply to reduce the operating voltage. Once on the plates, the
particles must discharge before the dust cake will release from the plate. Thus, high
resistivity dust is hard to collect. The chemistry of the dust, the temperature of the
flue gas, the moisture content of the flue gas, and the flue gas SO3 content all affect
resistivity. The preceding discussion merely introduces a subject so complex that it
could fill several volumes of books.
There are two basic types of precipitators, wet and dry. In a wet precipitator, the
collection surface is cleaned by a steady or intermittent flow of water. Currently, wet
precipitators have been used in special applications where the dust is sticky or
flammable; however, they have not seen common use on combustion sources.
Nonetheless, they are purported to be better than dry precipitators at catching
extremely fine particulate. Tests and initial trials are being conducted to use them for
final flue gas clean up on utility boilers. Based on the early success of these tests,
these devices may see increased use in the future to control opacity and fine particle
emissions from conventional large sources.
Dry precipitators are by far the most common type. Depending on their location in
the gas path they are classified as “hot side” or “cold side” precipitators. Hot side
precipitators are installed upstream of the air heater in utility boilers and operate at
temperatures in the range of 600° to 700°F. Cold side precipitators operate in the
range of 300° to 400°F. The advantage of hot side collection lies in the properties of
the dust, which can be much easier to collect than when it is cooled to 300°F. Hot
side precipitators must be much larger than cold side precipitators in order to handle
the larger volume of flue gas. They also operate at lower collection voltage because
hot gas is more easily ionized. These disadvantages tend to out weigh the advantages
and few, if any, hot side precipitators have been built in recent years.
Precipitators are built with a number of electrical sections in the direction of flow and
there is a separate power control cabinet for each section. The controls continuously
seek the highest practical operating voltage and they interrupt power momentarily to
suppress large sparks or arcs. Most precipitators have sufficient operating margin
that they can meet performance requirements with one or more sections out of service
and/or with reduced power levels. This has lead to debate in a few cases about
whether a large precipitator should be operated at its maximum collection efficiency,
or simply operated well enough to keep emissions in compliance. Figure 6-5 shows
the basic elements of an electrostatic precipitator with four collection fields in the
direction of flow.
Many precipitators are built as two units, side by side, each one taking half the flow.
High voltage transformer-rectifier sets are located on top of the box and the electrical
cabinets that control them are normally at ground level. Rappers for cleaning the

plates are not shown in Figure 6-5. Precipitator collector plates in a large unit are
typically about 35 feet high. The high-tension electrode grids built prior to about
1980 were made of wires with weights at the bottom. A rigid frame in most current
designs has replaced this design.
In order for a precipitator to work well, the gas must move slowly enough to give
sufficient residence time for dust collection. Local (or general) high velocities will
sweep gas out before it is fully treated; there is a critical gas velocity, above which
performance degrades rapidly. So precipitator performance is very sensitive to
distortions in gas distribution and to excessive gas flow rate.
Precipitator Monitoring
Perhaps the first thing to note about precipitator monitoring is that opacity is a more
effective tool on a precipitator than it is on a baghouse. The reason is that a
precipitator preferentially captures the largest particles first. Small particles penetrate
further and when performance deteriorates the finest particles are the first to slip
High Voltage
Rectifier-Transformers
Inlet Outlet
Dust Dust Dust Dust Control Cabinets
Figure 6-5. Elements of an Electrostatic Precipitator

through. Opacity is inversely proportional to particle size, so the fine particles
generate significant opacity while contributing very little to mass loading. When dust
begins to penetrate the precipitator, it is the fine particles with the highest opacity and
least mass that appear first. So an opacity increase is an early indicator that mass
emissions are starting to increase.
Although there are a number of potential causes for reduced collection performance,
nearly all of them can be divided into two categories:
• problems with the flow geometry, and
• problems reflected in reduced power levels.
Flow geometry problems are a consequence of the internal configuration. They

cannot be diagnosed by external measurements or tests and they cannot be corrected
by any operational control. When the configuration of the precipitator inlet duct
generates an uneven flow distribution, the precipitator can achieve its design
efficiency only when the highest flow area is at or below the design velocity. So the
total flow has to be reduced to accommodate the worst section. Most collectors are
built with some excess capacity or operating margin. A poor flow distribution
consumes this margin leaving the unit vulnerable to emissions excursions as it
approaches full load.
Most operating problems are reflected in reduced voltage or current to the affected
section. Problems that interfere with precipitator collection performance include:
thick dust accumulations due to a change in dust characteristics or inadequate rapping,
hopper overflows resulting from failed fly ash transport, and short circuits or electrical
control failure. Note that dust collection performance can be significantly affected by
variations in gas temperature, gas moisture content, chemical composition of the dust,
particle size and other source controlled parameters. To the extent that they affect
emissions, most of the variables that affect the ongoing performance of a precipitator
can be tracked by monitoring the (DC or secondary) current and voltage levels at the
control cabinets that power each of the sections. Current is the more useful of the two
readings, because it drops when there are dust layer buildups that interfere with the
electric field (which is responsible for particle collection). The collection performance
of any section in the precipitator is almost directly reflected in the current and voltage
readings. By tracking these readings over time, one can immediately see the effects of
dust accumulations or changes in the dust characteristics. These readings tend to be
fairly constant because changes, when they occur, take place over days or weeks.
Abrupt changes are usually a result of process changes or mechanical malfunctions.
6.2.1.5 Collector Combinations

There are numerous ways to combine the basic collection concept, but this section will
touch on only a few of them. First, it is possible to combine electrostatic precipitation
with other collector types, such as a baghouse. The advantage is that charged particles
tend to stick to the collecting surface. This can boost the performance of an inertial

collector or a scrubber, which normally cannot collect fine particulate. Used in
combination with a baghouse it changes the character of the filter cake making it more
porous. This allows the design to use higher gas velocities resulting in a smaller, less
expensive, baghouse.
Combining control systems for particulate and gaseous emissions is not a new
concept, but it has seen increased use as more and more large boilers are required to
control SO2 emissions as well as particulate. Either a precipitator or a baghouse can
be placed downstream of certain types of scrubbers to catch either fine particulate that
passes through a particle scrubber or the dust that results from a dry scrubbing
process (see Section 6.3.1.2) controls. Although dry chemistry is generally less
effective than wet chemistry, acid gases will tend to react with basic dust. So SO2
will react with dry Ca(OH)2 to form calcium sulfate; although the reaction is much
slower than when the two compounds are in liquid solution. Thus both SO2 and HCl
will, to some extent, react with available compounds in the fly ash to remove the
gases from the exhaust and create particulate matter. This mechanism is much more
effective in a bag house than in an electrostatic precipitator.
6.2.2 Dust Collector Fires

Dust collector fires are an unusual but significant concern because the damage can
take weeks to repair. When there is a fire in a baghouse the bags in the affected area
will almost certainly have to be replaced. A substantial fire in a precipitator will warp
the plates, which reduces the electrical spacing and hence the operating voltage. In
most cases, operators are not aware of fires until the damage is discovered at a later
date. Fires, for the most part, are preventable because it is always possible to
eliminate one of the three essential fire elements: (1) fuel, (2) oxygen, and (3) ignition.
Unfortunately, there are two possible fire scenarios that are not altogether avoidable:
• A rare, but possible, occurrence is an explosion from the ignition of combustible

gases mixed with air in the dust collector.
• The most common scenario occurs when combustible dust accumulates in the
hoppers or on the plates and ignites when air is introduced through leaks, or a
process upset.
Combustible gas fires are rare because it is very unusual to have sufficient
combustible gas and air present simultaneously. Under normal operating conditions,
a combustion device only generates combustible gas when there is no excess air, so
gas explosions in the dust collector cannot occur. Almost the only way to get a gas
explosion is for the fire to go out and fuel flow continue – as can happen during
startup if there is a failure in the flame safety system. During startup, it is normal
practice to have a precipitator shut off so there is no ignition source to trigger an
explosion.

Dust collector fires can be ignited either by the sparks in an electrostatic precipitator
or by burning embers generated in a biofuel-fired stoker. Most dust collector fires
occur in the hoppers where they are sustained by an air leak. However, maintaining
the mechanical integrity of the hopper and the dust evacuation system can usually
prevent air leaks. Fires are also possible in a precipitator if there are significant layers
of combustible dust on the plates and the excess air level is allowed to rise above
12–15% O2 – which happens during a combustion upset or when the combustor is
shut down. Dust collector fires are most common on wood-fired stoker boilers
because of the amount of carbon typically in the fly ash. However, most solid and
liquid fuel combustors are, under certain operating conditions, capable of generating
combustible ash.
As a general rule, dust collector fires occur during or immediately following a

combustion process upset or shut down. Eliminating one or more of the three critical
elements – fuel, air and ignition – can prevent fire. Contingency planning to establish
operator response or control system actions during an upset is an effective way to
prevent fires.
6.2.3 Oil Fired Particulate

Particulate emissions from oil-fired sources are relatively low because there is very
little ash, even in the heaviest residual oil. Most of these sources do not have dust
collectors and stack emissions consist of all the particulate formed in the combustion
process, as discussed in the preceding chapter, not the emissions remaining after a
dust collector. Emissions are controlled by maintaining good fuel atomizer
performance to limit carbon emissions and, in some cases, by using low excess air
levels to limit sulfate formation. Using low excess air levels generally results in
emission rates below 0.1 lb/mmBTU and in most cases below 0.05 lb/mmBTU.
Operation of dust collectors on oil-fired sources is challenging because most of the

ash is submicron size particulate and a substantial fraction of it may be sulfuric acid.
This makes the ash difficult to collect, corrosive and, unless it is kept warmer than the
sulfuric acid dew point, there is enough liquid to turn the ash into a sticky paste. The
only way to deal with this troublesome ash is to keep the dust collector, the hoppers
and the ash handling system hot enough to limit acid condensation – typically above
280°F. This translates into a lot of hardware to maintain and service, so an
electrostatic precipitator on a large utility boiler can require one full time operator just
to monitor and maintain the system and all its heaters. A baghouse is probably not a
feasible particulate control device for an oil-fired source because under some
operating conditions the ash will tend to permanently plug the bags.

6.3 SULFUR OXIDES AND HYDROGEN CHLORIDE
CONTROLS
Limiting the amount of sulfur or chlorine in the fuel can effectively control the
emissions of both SO2 and HCl. However, fuel limitations are not always feasible or
economical, in which case some sort of back-end control is required.
6.3.1 Approaches to SOx Control

The CAAA of 1990 imposed SO2 emission caps on most utility boilers that required a
large reduction in emissions. Economic analysis performed prior to enactment of the
new legislation indicated that compliance would be achieved by the extensive use of
scrubbers to capture SO2 emissions. However, the forecasts were inaccurate and most
utilities complied by changing to lower sulfur fuel (which was coal in most cases).
There were also cases in which the cost and availability of natural gas allowed some
utilities to replace some high sulfur coal or oil with gas. These choices reflected
market conditions more than they reflected uncertainty about the cost and practical
implications of scrubbers. However, at the present time, much or most of the possible
fuel switches already have taken place, so that future SO2 emission reductions will
have to be met primarily by installing scrubbers.
We noted earlier that virtually all fuel sulfur burns to SO2 and will be emitted from
the stack unless it is captured by a control device. A few percent of the fuel sulfur
will be oxidized to SO3 (a gas) or a solid sulfate salt. Since SO2 constitutes 95% or
more of the total sulfur emissions, control devices focus on this gas. However SO3,
although small in quantity, can be noxious in several ways, so most SO2 control
devices must deal with SO3 in some manner.
6.3.1.1 Fuel Switching

SOx emissions can be reduced by reducing the fuel sulfur content, as with low sulfur
coal, or by reducing the amount of sulfur-bearing fuel used, as by firing gas in place
of coal. If fuel sulfur is reduced as a result of a continuous emission limit, this
establishes a limit on the fuel sulfur content, as a restriction to firing oil with less than
1.0% sulfur. If, however, the restriction is an emissions cap – such as a limit of 100
tons of SO2 per year – then the source can fire any fuel it chooses, so long as the
emissions do not exceed the cap. In this case, the source could fire sulfur-bearing
fuel part of the year and use natural gas during a season when natural gas is less
expensive. Furthermore, if there is a market for emission allowances, a source can
buy or sell allowances in the course of meeting its requirement to keep total emissions
within the capped amount. Overall, emission caps provide a source with considerable
flexibility in their approach to emissions control.
Fuels can be changed only within the limits and capabilities of the combustion
hardware. These restrictions go beyond the simple requirements for appropriate
burners and fuel handling systems. For example, natural gas (with virtually no sulfur)

can be fired in boilers designed for coal, wood or oil. However, gas flames are much
less radiant than the flames from pulverized coal or oil, so less heat is transferred by
radiation. This means that in a boiler designed for oil or coal, natural gas fires carry
more heat out of the fire box and into the convective sections of the boiler. For power
boilers this means more energy flows into the superheat and reheat sections, which
translates into higher steam temperatures when fired on natural gas. In some units
excessive temperatures can mean the maximum boiler load is reduced.
Changing to a coal with lower sulfur content usually means that the coal comes from
a different mine and that it has different characteristics than the original coal. There
are a number of important coal properties, but some of the most significant relate to
the behavior of the coal ash in a pulverized coal boiler. To some extent, laboratory
tests can be used to determine whether a proposed coal will work in a given boiler.
However, test firing is normally used to determine how well a proposed coal will
work. An inappropriate coal can generate molten slag on the boiler walls, or massive
slag build-ups on the walls or in the convective passes. This defeats heat transfer and
can force reduced capacity and/or availability
6.3.1.2 Flue Gas Desulfurization

Sulfur oxide gases can be converted to solid oxides such as sodium sulfate or calcium
sulfate during the combustion process, or they can be removed from the exhaust gas
with a scrubber. Most of the control methods use an alkali, such as calcium, to
convert the sulfur to a solid salt. The amount of alkali and the efficiency of sulfur
capture depend on the process and operating variables. Scrubbers, as the word
implies, use water, but sulfur oxide scrubbers can be either wet or dry. APTI Course
415, Control of Gaseous Emissions, discusses control technologies in some detail, so
the following discussion only reviews the highlights.
As with most systems, the resulting solids are a waste material, but a few systems
generate commercial gypsum or other sulfur products. Calcium is the most common
cation, but sodium, magnesium and combinations are also used. The choice of
cation, and associated system, is based in part on availability and cost.
Wet Scrubbers
SO2 is a water-soluble gas that dissolves into the drops of water in a scrubber as long
as the water does not become acidic. So a wet scrubber sprays water with a high pH
into the flue gas, cations in the water react with the acid gases, and the liquid effluent
carries the sulfate salt either in suspension or in solution. Hence, a wet scrubber has
two systems: (1) the spray system, or contactor, to mix the exhaust gas with the
scrubber liquid and then collect the water, and (2) a water management system to
maintain sufficient pH, remove the solids from the water and manage the solids. The
contactor usually includes some kind of packed bed as shown in Figure 6-3 with a
large wetted surface area. Sometimes waste water can be evaporated into the exhaust
stream in the scrubber, so there is no water discharged from the plant.

Wet scrubbers are conceptually similar to particulate scrubbers in that they spray a
large amount of water through the exhaust gas to achieve complete contact between
the liquid and the gas. However, they do not require a high gas pressure drop (unless
they are also intended to collect particulate) and they may use an even larger water
flow rate. Water circulates in a loop from the spray to the scrubber sump to a
recirculation tank and back to the spray. The exhaust gas leaving the scrubber is
saturated and it will carry any fine particulate or mist that is not captured in the mist
eliminator. Mist escaping from a wet scrubber will include most of the SO3 (sulfuric
acid), which, though highly soluble, condenses into submicron particles that seldom
impact the scrubber water drops. So wet scrubbers do a poor job of controlling the
emissions of sulfuric acid, mercury or any submicron particulate.
Semi-Dry Scrubbers
Semi-dry or spray drying scrubbers begin the scrubbing process similarly to wet
scrubbers, but they spray much less water into the exhaust gas. The amount of water
is controlled so that it evaporates completely, leaving behind dry salt particles that
can be captured in a dust collector. The exhaust temperature is typically 15° to 30°F
above saturation temperature. The semi-dry scrubber must be followed by either a
baghouse or an electrostatic precipitator to collect the sulfate dust created in the
scrubber. In some systems, this is the only dust collector. When a baghouse is used,
there is some secondary scrubbing of the gases as they pass through the filter cake on
the surface of the bags. This can be a key factor when very high collection efficiency
is required, or if the system is also required to remove mercury from the exhaust gas.
A baghouse also assures high collection efficiency for fine particles and mists, such
as sulfuric acid. Semi-dry scrubbers can remove SO2 with 90%-95% efficiency.
Semi-dry scrubbers with baghouses have seen wide application on municipal waste
combustors.
Dry Scrubbers
Dry scrubbing is less efficient at removing SO2 than either wet or semi-dry scrubbers.
However, it can significantly reduce SO2 emissions when high efficiency is not
required. This method blows powdered reagent directly into the exhaust gas or the
combustion zone where some of it reacts with the acid gases. In most cases, the
amount of reagent must be much larger than the amount calculated from a chemical
balance – sometimes 5 to 10 times as much reagent. A fluid bed combustor can work
as a dry scrubber if limestone is added to the bed.
Several factors affect the performance of a dry scrubber. First, the reagent must be
highly reactive with SO2 in order to be effective. Dry scrubbing is typically more
efficient in capturing SO3 (sulfuric acid) than in capturing SO2. Second, the reagent
needs to have a large surface area – meaning it must be in very small particles
(5-10µm), which must be fairly porous. Third, it needs to be injected beyond any
temperature zone that could fuse the particles or destroy their chemical reactivity.

Finally, dry scrubbing benefits greatly from reactions between the exhaust gas and the
filter cake in a baghouse. An electrostatic precipitator is usually much less effective
in a dry scrubbing application than a baghouse.
6.3.2 SO3 Control

SO3 emissions are problematic for several reasons:
• SO3 emissions are not significantly controlled by scrubbers or electrostatic

precipitators. In fact wet scrubbers simply condense the SO3 acid, exacerbating
downstream problems. Control with a baghouse requires an alkali dust in the
filter cake.
• Relatively small amounts (10 ppm or less from a large utility boiler) can condense
downstream of the stack to create a highly visible plume that is difficult to
control.
• SO3, if allowed to condense as sulfuric acid, seriously damages ductwork and any
unprotected pollution control devices.
• If an SO3 bearing plume touches down near the source, it is a highly toxic gas
with ambient impacts that include a pungent odor and corrosion.
Sulfuric acid exists in different forms, depending on the temperature. Above 500°–
600°F, SO3 is a gas. As it cools, it combines with increasing amounts of water and
below about 250°F it condenses to liquid sulfuric acid. Condensation can occur on
cool surfaces or it can form a mist or cloud of submicron size drops. When
combustion exhaust gases with SO3 are vented to the atmosphere, the condensed acid
can form a visible cloud. Either a wet or semi-dry scrubber will drop the temperature
below 250°F and cause any acid present to condense into submicron drops. As
described in the discussion of particulate scrubbers, the acid drops are so small that
they do not impact the scrubber drops. So the scrubber causes the acid to condense,
and that makes it impossible for the scrubber to capture most of the acid mist.
However, if the scrubber is followed by a good dust collector, it will capture most of
the acid mist. In fact, a baghouse will capture nearly all the acid mist.
If the SO3 is not condensed, it can be captured by a dry scrubber in which an alkaline
powder is mixed with the flue gas. SO3 in the gas phase will react with dry alkaline
materials, which then can be captured by a dust collector. However, this method of
SO3 capture tends to be relatively inefficient – requiring several times as much alkali
reagent as is theoretically necessary. However, since the amount of SO3 is typically
in the range of 10 ppm, supplying excess reagent may not be a problem. If the system
is an oil-fired boiler, with no dust collector, adding alkali dust to the flue gas can
easily increase the particulate emissions to unacceptable levels.

6.3.3 HCL Control
Hydrochloric acid can be collected in the same way as SO2. However, the biggest
difference is that HCl is more water soluble than SO2, so it is easier to collect and, in
the same scrubber, HCl collection efficiency will generally be much higher than SO2
collection efficiency.
6.4 NITROGEN OXIDES CONTROL
6.4.1 Approaches to NOx Control

NOx control is driven by the need to control ambient ozone levels. Although NO2 is
yellow or brown gas, it is rarely present in sufficient concentrations to be visible in
the plume from combustion sources (most of the NOx emissions are NO). On a
regional scale, NO2 may be a small contributor to haze and visibility impairment.
NOx control follows a two-tiered approach. The first level is to minimize the amount
of NOx formed by adjusting the combustor design and operation. When further
reduction is required, a system is installed to remove NOx from the flue gas – similar
in function to a dust collector or a scrubber. Combustion controls are generally
inexpensive to build and operate although many of them provide only modest NOx
reductions. Back-end controls incur significant expenses, but they typically remove
90% or more of the NOx remaining after implementation of low NOx combustion.
As the science of NOx controls evolves, we are seeing continued advances in both
combustion and back-end control technologies. These advances are closely coupled
to ever lower expectations for NOx emission levels on new sources. One should
remember that the very low NOx emissions from many new major sources are only
possible when firing natural gas. When these sources fire distillate oil emissions are
usually much higher than when firing gas.
6.4.2 Combustion Modifications

In general, there are three levels of combustion modifications that are used to reduce
NOx formation.
• Excess air control. Any effort to reduce NOx emissions begins with precise air
flow control. It is the first step taken to reduce NOx and it is an integral part of
any subsequent combustion based control effort.
• Reduce peak flame temperature. Modest reductions in peak flame temperatures

will dramatically reduce the formation of thermal NOx with little or no effect on
fuel NOx.

• Staged combustion. Staging the combustion will control the formation of both
thermal NOx and fuel NOx, but it requires considerable sophistication to achieve
significant NOx reductions without incurring unacceptable emissions of black
smoke or CO.
The order here reflects both the difficulty of implementation and the potential
effectiveness of NOx reduction from most sources.
6.4.2.1 Excess Air Control

Figures 5-9 and 5-10 show that the amount of excess air in the combustion zone
affects NOx emissions. Reducing excess air reduces NOx emissions on most types of
sources, including virtually all boilers and furnaces. Thus, the first step toward NOx
reduction on boilers and furnaces is to reduce the excess air as far as possible. All
sources require some excess air because, as you can see from Figure 5-9 and 5-10,
when the excess air is too low, there is a sharp increase in emissions of CO, smoke and
other products of incomplete combustion (PIC). Thus, there is an optimum excess air
level where NOx is minimized and PIC are not excessive. This means that the air flow
must be precisely controlled in order to continuously minimize NOx. On newer
boilers good air flow control is a matter of operating practice and does not require
additional hardware. On older systems the controls may need to be upgraded with the
installation of an O2 meter linked to the automatic controls.
There is one NOx control method that combines precise excess air control with a
catalytic reaction. It is called non-selective catalytic reduction (NSCR) and is used on
reciprocating engines. The concept is presented here and outlined again in Section
6.4.4 on back end control methods. NSCR uses a catalyst to make NOx react with
PIC to destroy both species. The general chemical reaction is:
NOx + CO → N2 + CO2
or
NOx + HC → N2 + H2O + CO2
As written here, these reactions have not been balanced, but it should be obvious that
the amount of unburned fuel needs to be closely matched to the amount of NOx in
order to destroy all of both types of pollutant. If there is any oxygen present, it burns
the VOC and CO, leaving the NOx untouched. This balance is achieved by precisely
controlling the excess air level. As illustrated in Figure 5-9 and 5-10, at very low
excess air levels any change in the amount of excess air will change PIC and NOx in
opposite directions. So the NSCR reactions can be balanced by operating at exactly
the right excess air level.
Although this is a simple concept, its implementation is daunting because of the

precision required. In practice, NSCR is accomplished with a sensitive O2 or CO
meter linked to the air-fuel control device of the engine. We will not get into the
details of this control technology here; however, it is worth noting that this
technology requires not only the right catalyst, but also a very good gas concentration

instrument linked to very good controls. The fact that gasoline powered automobiles
operate very well for many years without major repair or adjustment is testament to
the successful development of this technology. This technology has been used
primarily on gas fired reciprocating engines, but it is potentially applicable on boilers
and furnaces.
6.4.2.2 Temperature Reduction

NOx is formed during combustion because the high temperatures break apart a few
N2 molecules making N atoms available to react with oxygen. N2 is very stable and
only decomposes at very high temperatures, so NOx is formed primarily in the hottest
regions of the flame – typically at temperatures above 3000°F. Any change in
combustion that reduces peak flame temperature will reduce the formation of thermal
NOx. Temperature reduction approaches can be very effective in reducing NOx
when there is no nitrogen in the fuel, in some cases reducing emissions by 90% or
more. Even when there is nitrogen in the fuel, flame temperature reduction is
normally part of a NOx reduction effort.
The theoretical equilibrium curve in Figure 5-6 suggests that a 200°F temperature
reduction will reduce NOx by 50%. Although equilibrium is not achieved in most
combustion systems, the effect of peak temperature on NOx levels in a real system is
probably similar to equilibrium theory. It should be noted that NO is quite stable and
once formed in the combustion zone, it does not decompose as the flue gas cools off.
If NOx formation can be suppressed in the flame zone, it will not be formed by
secondary reactions downstream.
There are a number of methods of cooling the flame including:
• Spraying water or steam into the flame.

• Cooling the combustion air supply.
• Recirculating cool exhaust gas into the flame.
• Retarding the ignition of an internal combustion engine.
• Raising the excess air level of a premixed flame.
None of these methods are particularly advantageous to combustion, aside from

reducing NOx, but some incur less efficiency penalty or fewer operating problems
than others. The methods that have become common practice for different types of
sources are discussed in the following sections.
Water Injection
Water injection has been the accepted method to reduce emissions from combustion
turbines for about 25 years. The flame cooling effect of spraying water in with the
fuel is obvious. Figure 6-6 is an example of the effect of water or steam injection on
emissions. The process requires very clean water, but the penalty to engine efficiency
is quite small for two reasons. First, water provides cooling that would otherwise
have to be provided with air because the maximum allowable turbine inlet

temperature is far below the peak flame temperature in the primary zone. Second,
injected water turns to steam, which provides added mass flow through the turbine,
thus adding power to compensate for the energy lost to evaporating water. The water
mass flow rate into a turbine combustor can be as high as the fuel flow rate. The
resulting NOx emissions can be reduced by more than 70% without dramatically
increasing the emissions of CO and hydrocarbons. Water injection is not commonly
used on boilers because there are other less expensive alternatives for reducing NOx.
Water injection has been tried on reciprocating engines, but it significantly reduces
engine efficiency and tends to cause operating problems.
80
Water Injection
60
Steam Injection
40
Firing Natural Gas

20
Firing Distillate Oil
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Water-to-Fuel Ratio (lb/lb)
Figure 6-6. Effect of Water Injection on Combustion Turbine Emissions.

Source: Schreiber, 1991
Flue Gas Recirculation
Flue gas recirculation (FGR) or exhaust gas recirculation (EGR) works by mixing
some flue gas with the incoming combustion air. This increases the mass flow
through the combustion zone and decreases the concentration of O2 available for
combustion. Increasing the gas flow in the combustion zone decreases the
temperature because the same amount of energy is distributed to a larger thermal
mass. Figure 6-7 shows the basic elements of a FGR system. Note that FGR operates
without changing the inlet air flow, the stack exhaust flow or the fuel flow. The only
flow that increases is the flow through the combustor and through most of the heat
transfer elements.
Flue gas recirculation has been used both on boilers and on reciprocating engines.
The penalty on boiler efficiency and performance is almost negligible with
recirculation rates up to 20%-25%. When there is no fuel nitrogen present, NOx

can be reduced by 75% or more as indicated in Figure 6-8. FGR is typically a part of
any effort to substantially reduce NOx emissions from a boiler.
Exhaust gas recirculation can reduce NOx emissions on reciprocating engines, but it
requires considerable development and tuning to avoid significant degradation of
performance and efficiency.
Stack
Induced or Passive FGR

Natural Gas Fired
Air In
60°F
Air Heater
Active FGR
Sulfur Bearing Fuels
O2 = 17%-20%
Figure 6-7. Flue Gas Recirculation
Two possible FGR systems are shown in Figure 6-7. In most boiler FGR systems a
fan is required to mix hot gas from the exhaust side, at low pressure, with hot
combustion air at higher pressure. Electric power to operate the gas recirculation fan
is a significant operating cost because the fan moves a substantial volume of hot gas
across a significant pressure differential. It is also possible to recirculate gas from the
stack to the inlet of the FD fan, which does not require a separate fan (called passive
or induced FGR). However, this is only practical if the fuel is free of sulfur.
Whenever there is sulfur in the fuel, there will be a significant amount of sulfuric acid
in the exhaust gas. If acid vapor is introduced into the combustion air supply, it will

condense and corrode in the inlet air duct and the burner. Passive or induced FGR is
suitable only on combustors where natural gas or sulfur free petroleum is the only
fuel. In other cases, the FGR system must be entirely on the hot side of an air heater
in order to prevent acid corrosion problems. This limits its application on industrial
boilers, many of which do not have air heaters.
100
75
50
25
0 10 20 30 40 50
% Recirculation
Figure 6-8. Effect of Flue Gas Recirculation on Emissions
The FGR systems illustrated in Figure 6-7 imply that the recirculated exhaust gas is
uniformly mixed with the combustion air, and many systems do exactly that.
However, the FGR system can be built to deliver the gas to a specific region of the
burner so that one part of the flame is affected more than another. This approach can
amplify the NOx reducing effect of FGR, but if it is not carefully developed, it can
also decrease flame stability or increase CO emissions. In summary, FGR is an
effective method of reducing NOx emissions especially on natural gas-fired sources,
but its effectiveness depends in part on design details.
Ignition Retard
The combustion process in reciprocating engines is linked to crankshaft rotation and

is normally timed to generate maximum pressure in the cylinder, which corresponds
with maximum temperature and, hence, maximum NOx. If the ignition is delayed by

a few degrees of crankshaft rotation, then pressure, temperature and NOx are all
reduced as is engine power and efficiency. On most engines, ignition – or fuel
injection – timing is a relatively simple adjustment so it can be used to reduce NOx
emissions on existing engines. The resulting NOx reduction is considerably more
than the resulting loss in engine efficiency, so there is a net decrease in NOx
emissions. NOx reductions of 20-25% can be achieved
6.4.2.3 Off-Stoichiometric Furnace Combustion

When boilers fire coal or No. 6 oil, more than half of the total NOx comes from fuel
nitrogen, so thermal approaches like flue gas recirculation are much less effective
than when firing clean fuels. When firing nitrogen-bearing fuels, combustion based
approaches to NOx reduction use some form of staged combustion. Staged
combustion is capable of controlling both thermal and fuel NOx. There are several
fuel staging concepts, but two are noteworthy:
• Stratified combustion. Fuel and an insufficient amount of air mix and burn in a
primary combustion zone. With insufficient oxygen available for complete
combustion, all the O2 is consumed by carbon and hydrogen, leaving none available
to form NOx. The fuel nitrogen combines to form N2. During the few hundredths
of a second it takes for combustion to occur, the flame cools slightly. Once this
cooling has occurred, the rest of the air is added to complete combustion. Since the
fuel nitrogen radicals have disappeared, and the flame is too cool to generate a lot
of thermal NOx, relatively little NOx will be formed in this secondary combustion
zone. This is the conceptual basis for low NOx burners and for burners-out-of-
service or over-fired air applied to a large furnace. The trick has been to get this
concept to actually work consistently in real furnaces. To be successful it requires
precise control of fuel-air mixing throughout the combustion zone. Stratified
combustion is being used with considerable success on suspension-fired boilers; it
is not applicable to stokers. An inevitable side effect of staged combustion is
longer, less brilliant fires.
• Reburning. A normal boiler combustion zone is operated with minimum practical

NOx emissions. Then a second combustion zone is added downstream after the flue
gases have cooled significantly. This second zone is staged by injecting fuel –
usually natural gas – with no additional air. With insufficient oxygen, this fuel
creates a reducing zone, which actually consumes NOx formed in the initial
combustion zone. Then further downstream additional air is injected to complete
combustion. No additional NOx is formed because the temperatures are too low.
While this approach works (conceptually) on any furnace with sufficient space, it is
particularly attractive on systems like stoker boilers where no other combustion NOx
control can be implemented.
With this introduction, each of the staged combustion concepts can be discussed in
greater detail.

Stratified Combustion in Large Furnaces
Early efforts to reduce NOx from utility boilers in Southern California paved the way
for technologies still used today. Utility boilers have many burners on multiple levels
and the fuel can be shut off from a few of the burners without affecting the boiler
load. The remaining burners pick up the slack. If fuel is shut off to a few of the top-
level burners without shutting the burner air dampers, the bottom of the furnace will
be fuel rich. Air flow through the burners out of service (BOOS) will make up the air
deficit without generating much additional NOx. Figure 6-9 illustrates this concept.
Typically, this approach can achieve a 15% to 30% NOx reduction without seriously
affecting boiler performance or generating excessive smoke. This is not always as
simple as it sounds and there will be one or more particular combinations of burners-
out-of-service which are more effective than others. However, the BOOS approach is
still used on some boilers.
Removing burners from service frequently incurs some sort of operating restriction
on a large boiler. To get around this, utilities began installing overfired air (OFA)
ports above the top row of burners as shown in Figure 6-9. OFA ports have become a
fairly standard component of a low NOx system on large utility boilers.
One consequence of staged combustion in any form is that flames are longer and
generally less brilliant. In fact, some low NOx fires have an alarmingly dark and
dirty appearance, which is counter to the traditional operating philosophy of
optimizing performance with short bright flames.
2000oF
Burn
Out OFA Ports
Zone
BOOS (No Fuel; Air Only)
Fuel
Rich Normal Air
Zone Extra Fuel
Figure 6-9. Burners-Out-of Service (BOOS) and OFA Ports

U.S. Environmental Protection Agency. Alternative Control Techniques Document: NOx Emissions
from Utility Boilers, March 1994, EPA Report 453/R-94-023, NTIS PB94-184165.

Reburning
Reburning is a relatively recent NOx reduction technique that is particularly attractive

for existing boilers when traditional combustion control technologies are not
applicable. Reburning is not restricted to large utility boilers, but there are geometric
limitations. Figure 6-10 is a general schematic of reburning in a large utility boiler.
The reburning section is installed just downstream of the main combustion zone
where the amount of NOx present will be typical of uncontrolled emission levels.
Therefore, it can be used on stoker furnaces and cyclone-fired boilers.
Reburning seeks to reduce (literally and chemically) the NOx by injecting more fuel
with no accompanying air. This creates a fuel rich zone where no additional NOx
will be formed because of (1) the lack of oxygen and (2) temperatures are below that
required for NOx formation, but are still hot enough to ignite the supplemental fuel.
In this fuel-rich zone, hungry fuel species take oxygen away from the NOx, reducing
it to N2. Of course the gases leaving this reducing zone are full of PIC, so additional
air is now required to complete combustion. Overfired air can be added without
creating much NOx because the gases beyond the primary flame zone have lost heat
to the furnace walls and temperatures are too low to generate thermal NOx.
Reheater/Superheater
Burnout Zone
Overfire Air
• Normal Excess Air
Reburn Fuel - Gas Reburn Zone

~20% • Slightly Fuel Rich
• NOx Reduced to N2
Primary Fuel - Coal Primary Combustion Zone

~80% • Reduced Firing Rate
• Low Excess Air
• Lower NOx
Figure 6-10. Reburning on a Utility Boiler

U.S. Environmental Protection Agency. Control of NOx Emissions by Reburning,
February 1996, EPA Report 625/R-96/001.
NOx reduction levels using reburning depend on how much fuel is burned in the
reburn zone. Full-scale demonstration projects firing up to 25% of the total heat input
has reduced NOx by about 50%. Since this scheme works by creating a fuel rich
zone, the amount of reburn fuel must, in theory, be more than enough to consume all

the excess air present in the original combustion zone. In practice, nonuniform
excess air distribution and nonuniform mixing mean that any amount of reburn fuel
will have some effect, but no amount will destroy all the NOx.
Typically, the injected fuel is natural gas because it is relatively easy to inject and
mix. At least one demonstration project has shown than “micronized” (very finely
pulverized) coal will also work. Although any fuel that can be thoroughly mixed with
the flue gases should work, reburning fuel will probably be limited to natural gas for
practical reasons.
6.4.2.4 Low NOx Burners

Once the success of BOOS and OFA ports in large furnaces was demonstrated, it was
natural to try to extend the concept to individual burners. The BOOS concept cannot
be used on smaller boilers with only one or two burners because there are insufficient
burners to remove from service and still maintain load. In addition, these boilers
frequently lack any location for installation of OFA ports. And if low NOx burners
can be combined with OFA ports in large boilers, there should be additional NOx
reduction. All the companies that manufacture boiler burners have developed low
NOx burners with varying degrees of success.
Combustion Zone
Wind Box
Sliding Air Damper
Fuel Rich
Primary
Zone
Primary Secondary Basic Features

Swirl Swirl
Vanes Vanes • Low Emissions
• Air Shut Off Damper
• Fixed Swirl Vanes
• Fixed Burner Geometry
Figure 6-11. Low NOx Burner Concept
Figure 6-11 illustrates the low NOx burner concept. The burner creates a primary
zone in which all the fuel burns without quite enough air. Then, a few feet further
into the furnace, additional air is mixed into complete combustion. However, the
concept is much easier to illustrate than it is to execute in practice.
The first challenge is to design a burner that will actually create two mixing zones –
one following the other. A number of vendors with combustion laboratories have

been able to develop and demonstrate full-scale burners, but a major challenge has
been to get these burners to perform as well in the field as they do in the laboratory.
Several problems arise in practice, particularly on multiple burner boilers. First, air
flow and fuel flow may not be identical to all burners. If one burner operates with
lower excess air than the rest, then the whole furnace has to be fired with enough air
to keep that rich burner from smoking. Air velocity in the wind box upstream of the
burner may cause more air to enter one side of a burner than the other. A similar
problem can appear in the fuel distribution of a pulverized coal burner. The
powdered coal is transported pneumatically and may not enter the combustion zone in
a circumferentially uniform pattern. A critical element of combustion NOx control
on large boilers is to get all the burners to behave in exactly the same manner. This
has frequently taken a great deal of tuning and troubleshooting.
A question that is frequently asked is, “What is a low NOx burner?” The best answer
appears to be: “A burner with low NOx emissions”. There is no hardware specification
that assures low emissions. Most low NOx burners have a couple of features worth
noting:
• Manufacturer Presets. All adjustable features are configured so the boiler

operators will not routinely alter them. Adjustments should be made when the
system is initially tuned, and then left alone. Boiler operators are accustomed to
adjusting burners to get a nice looking fire, and this kind of tinkering will defeat
low NOx performance.
• Separate Air Flow and Direction Dampers. Air registers that control swirl must
be separate from those than control flow. Burners used to have dampers that
combined swirl and air flow control. With a low NOx burner you need to be able
to shut off the burner and its air flow without adjusting the swirl (flow pattern).
• Precise Air Flow Control. The system will have air flow controls that can
maintain the air flow at the minimum level practical without generating excess
CO or smoke.
Note that low NOx burners will only give low NOx performance if the whole furnace
installation works as intended. Low NOx burners alone will not necessarily give low
NOx performance.
6.4.3 Premixed vs. Diffusion Flames

With the exception of reciprocating engines that burn gasoline or natural gas, most
combustion sources use diffusion flames so the fuel and air mix as they burn. These
burners are generally easier to build and operate than burners based on premixed fuel
and air. Premixed combustion is impossible with solid fuel (e.g., coal, etc.) or
residual oil and has yet to be achieved commercially with No. 2 oil. So premixed
combustion is limited to natural gas and very volatile liquids such as gasoline.
Premixed burner designs offer a unique opportunity to reduce NOx emissions because

(1) there is no fuel nitrogen to make fuel NOx, and (2) ultra lean combustion, where it
can be achieved, is cool combustion that makes very little thermal NOx.
6.4.3.1 Theoretical Differences

The difference between premixed and diffusion combustion in real systems becomes
apparent in NOx emission levels and in the choice of NOx control strategies. As
previously noted, the peak flame temperature in premixed flames is equal to the adiabatic
flame temperature; in other words, it decreases as excess air increases. Figure 6-12 is
similar to Figure 4-5 and it shows NOx emissions versus excess air for both premixed and
diffusion flames. The NOx curves are drawn to illustrate a concept and are not the result
Fuel Rich Fuel Lean
0 0.5 1.0 1.5 2.0

Air-Fuel Ratio
0 5 7 9 10
Exhaust O2 (%)
Figure 6-12. Flame Temperature and Emissions versus Air-fuel Ratio
of any specific analytical model. Note that the premixed flame NOx emission rate peaks
at a point slightly to the right of the line of equivalence ratio equals one, and then falls
rapidly as excess air increases. By contrast, diffusion flame NOx emissions steadily
increase as excess air increases.
In a premixed flame all the combustion occurs at a fixed air-fuel ratio – the same
excess air level as measured at the combustor exit. Thus, nowhere in the flame can the
temperature exceed the adiabatic temperature for that air-fuel ratio. The higher the

excess air level, the lower the flame temperature and the lower the NOx, except at very
low excess air levels where the lack of oxygen inhibits NOx formation. A diffusion
flame is different because all of the fuel passes through the mixture point where the
air-fuel ratio = 1.0 at some point in the flame. So diffusion flame temperatures can
locally approach the peak adiabatic temperature regardless of the exhaust excess air
level (assuming the excess air level is greater than zero). Diffusion flame NOx levels
depend on how much of the combustion occurs in regions of the flame where there is a
finite amount of oxygen available (e.g., where not all the oxygen has been consumed
by fuel combustion). The higher the overall excess air level, the more of the flame
that has excess oxygen available for NOx formation and the higher the NOx. So
diffusion flame NOx levels increase with increased excess air.
If one can operate a premixed burner at a fairly high air-fuel ratio (on the right side of
Figure 6-12) both the NOx and CO emissions will be low. The challenge has been to
get these burners to operate reliably. Very lean flames are hard to light and they blow
out easily. Flames that blow out and engines that quit are unacceptable regardless of
their exceptional emission characteristics.
6.4.3.2 Combustion Turbine Burner

Emissions from combustion turbines have been controlled to relatively low levels for
many years by the use of water injected into the combustion chamber with the fuel.
Water injection incurs a relatively small engine performance penalty because the
steam helps to drive the turbine and offsets most of the losses associated with
evaporating the water. The main drawback is the need for substantial amounts of
ultra pure water. Any contamination in the water gradually builds up on the turbine
blades and degrades engine performance. Combustion turbine (CT) power plants
have had to operate large water treatment facilities to supply water for the turbines.
This is a significant burden in remote facilities, especially where water is scarce.
It has taken many years to develop, but the major CT vendors are now offering “dry
low NOx” or “lean-premix” combustors for their gas-fired turbines. The basic
concept is to premix the fuel and air, then operate the combustor very lean – on the
far right side of the curves in Figures 4-5 and 6-12.
The challenge is to get the engine started. Figure 6-13 shows the sequence involved
in starting the engine and bringing it up into its operating range. The emissions
associated with this operation are shown in Figure 6-14.
Normal operation is shown in the last illustration. The first three figures show how
the engine is started and reaches its normal operating mode. Initial light off and low
load operation is in a conventional combustion mode with characteristic high NOx
emissions. Then the engine fuel controls take the combustion process through a
transition (lower two figures) in which fuel is shut off to the primary zone to extinguish
the flame followed by reintroducing most of the fuel into the primary zone. Here the
fuel is premixed with air before being ignited in the secondary zone. The result is lean,
cool combustion that generates very low NOx emissions. There is no engine efficiency

penalty and the hardware is only moderately more expensive than a traditional
combustor. The biggest cost is the years of engineering work required to develop this
type of combustor. These costs must be amortized in the price of the new hardware.
“Dry” low NOx turbine engines have very low emissions when operated above the
critical transition point – 40% to 50% load in the example shown here. However, the
emissions are high at low load, so extensive operation at low load must be avoided.
Fortunately, these engines by themselves can start and come up to load quite rapidly;
typically less than 30 minutes from a cold start. However, if the engine is feeding a
waste heat boiler, a load ramp limit may be imposed by the boiler. Boilers generally
take many hours to heat up.
Fuel Primary Operation Fuel Lean-Lean Operation

100% Ignition to 20% Load 70% 20% to 40% Load
30%
Fuel Fuel
100% 83%
17%
Second-Stage Burning Premixed Operation

Transient During 40% to 100% Load
Transfer to Premixed
Figure 6-13. Dry Low NOx Combustion Concept – Gas Firing

Marvin Schorr, GE Industrial & Power Systems, 1993
Although gas turbines are conceptually simple, the systems can get complex and they
are still evolving. Here are several issues to consider when evaluating emissions from
a CT system.
Maximum power from a combustion turbine is controlled by the amount of air that is
pumped through the engine. This means that on a hot day or at altitude – when the air
is thin (less dense) – engine power diminishes. In many cases, there can be a 20%
power difference between summer and winter. NOx emissions are also sensitive to
the amount of moisture in the air. Emissions are higher on a dry day than on a warm,
moist day. The effect of ambient conditions on power and on engine emissions is
highly predictable for each engine design. The dependence of emissions on ambient
conditions is engine specific and each manufacturer has developed a correlation for
each engine model. There is a default correlation listed in Subpart GG of 40CFR63
to correct NOx emissions to standard conditions, but the manufacturer’s correlation
should always be used when it is available.

240 100
200
NOx 80
160
60
120
40
80 NOx
CO 20
40
CO
0 10 20 30 40 50 60 70 80 90 100
Turbine Load (%)
Figure 6-14. Dry Low NOx CT Emissions – Gas Fired

Marvin Schorr, GE Industrial & Power Systems, 1993
One way to compensate for warm ambient conditions and to get more power from the
engine is to cool the inlet air. Spraying a fine mist of water into the air at the engine
inlet cools the air by evaporation and adding water lowers NOx emissions. Thus, the
maximum engine power and emissions are both significantly influenced by inlet
evaporative coolers – also called inlet fogging.
A gas turbine has a device called a fuel control that correlates fuel flow with air flow,
engine speed and load demand. Originally, the fuel control was a complex
mechanical device, but the advent of microcomputers allows software to absorb the
complexity. Computers also allow the fuel control to serve an emissions control role.
Emissions from a gas turbine are highly predictable, so the amount of water injection
required either in the combustor or in the form of inlet fogging can be programmed
into the engine’s control system. So, to a substantial degree, a turbine engine can be
programmed to meet an emission limit over its load range and over a range of
ambient conditions without relying on any actual emission measurement.
6.4.3.3 Catalytic Turbine Combustion

A recent development in gas-fired combustion turbine operation is the advent of the
catalytic combustor. In place of a combustor with a hot flame followed by dilution
air, most of the air and the fuel is premixed, and then passed through a catalyst where
combustion occurs at essentially the same temperature as the turbine inlet – much
lower than typical flame temperatures. This virtually eliminates NOx formation and
typical emission levels are in the range of one to two ppm. The technology is
patented and is marketed under the name XONON (no NOx spelled backwards) by
Catalytica Systems, Inc. of Mountain View, CA. Figure 6-15 is shows the concepts.

Xonon Operates with Same Turbine Inlet Temperature (TIT)
Conventional
Combustor Fuel
Xonon
Combustor Fuel
High temperature in
flame produces NOx
Conventional
Temperature
Xonon
Same TIT
Figure 6-15. Xonon Catalytic Turbine Combustor

Courtesy of Chuck Solt of Catalytica Inc
In place of a normal primary zone there is an air fuel mixing zone. This includes an
ignition or warm up burner that consumes a small fraction of the fuel to maintain
adequate inlet temperature to the catalyst. Most of the fuel burns in or downstream of
the catalyst where the high excess air level reduces flame temperature and little or no
NOx is formed.
The technology appears to be broadly applicable to gas turbines, but is new enough
that the durability of the catalyst is not currently known, although it is guaranteed for
one year. One drawback is that an engine fitted with a XONON combustor cannot
use liquid fuel as a backup. Like all other turbine combustors, the XONON
combustor must be specifically developed for each type of engine; there is no such
thing as a generic, one type fits all, catalytic combustor. And, like most combustors,
one of the biggest challenges has been the development of the start-up procedure to
get the engine running. A normal (non-catalytic) flame is necessary for ignition and
to warm the catalyst to operating temperature before converting to catalytic operation.

6.4.3.4 Reciprocating Engines
NOx emissions from reciprocating engines have been reduced by several methods
that include: ignition retard, modifying the air-fuel ratio, exhaust gas recirculation,
and combustion chamber modifications. Exhaust gas recirculation has not been
widely used on stationary engines. Ignition retard amounts to detuning the engine
because it reduces power and/or fuel efficiency. By delaying ignition, combustion
occurs at lower pressure and lower temperature, which reduces NOx formation and
may increase PIC emissions. This approach is available on nearly all reciprocating
engines, but the effectiveness in reducing NOx and the effect on engine performance
varies from one engine model to another. On gas fired and dual fuel engines, ignition
retard has largely been superceded by combustion chamber modification. On diesel
engines ignition retard continues to be the primary tool available for NOx reduction.
While the effectiveness varies, reductions of 40% have been observed (see AP-42).
Air-fuel ratio adjustments, as indicated in Figure 6-12 clearly affects NOx and PIC
emissions. This approach to emissions control is viable on gas fired and dual-fuel
engines, but not on diesel engines. On gas and dual-fuel fired engines the viable
range of air-fuel ratios depends on the engine design. Lean mixtures are more
difficult to ignite and richer mixtures generate PIC. These issues are discussed in the
following paragraphs
There is a fundamental difference in the combustion process between diesel fuel and
natural gas or gasoline, which evaporates before ignition occurs. Natural gas or
gasoline is premixed with the air before combustion starts. Diesel fuel is a liquid that
mixes with the air and evaporates as it is burning. The result is that diesel
combustion occurs at near stoichiometric conditions even though the final exhaust
mixture may have a lot of excess air. The difference between premixed and diffusion
combustion is a key factor in the ability to control NOx emissions because premixed
combustion allows engine designs with low peak flame temperatures. As a result,
NOx control efforts have been much more successful on gas-fired engines than on
diesel-fueled engines.
Figures 5-10 and 6-12 illustrate the effect of the air-fuel ratio on emissions.
Reciprocating engines without emission controls are usually lean burn engines; in
other words, they operate with sufficient air to minimize smoke and CO emissions.
Diesel engines, on the other hand, approach rich burn conditions at full throttle. Rich
burn engines operate with low enough air that there are some PIC. In general, rich-
burning is associated with more power and increased emissions while lean-burning
implies less power and lower emissions. Extremely lean burning appears, in Figure 6-
12, to be the desired operating range for low emissions, but conventional gas or
gasoline engines will not operate in this range because electric spark plugs will not
ignite a very lean fuel mixture.
NOx reduction efforts on gas-fired engines have made impressive progress since
about 1990. By fitting the combustion chamber with a “pre-chamber”, NOx

emissions can be reduced by 90% or more. Figure 6-16 shows this concept. The pre-
chamber is a small combustion chamber off the main chamber that operates at an
ignitable air-fuel ratio. When it is ignited it shoots a flame into the main chamber that
is strong enough to ignite a very lean air-fuel mixture – it acts like a very powerful
spark plug. This allows the fuel mixture in the main chamber to be very lean – on the
far right side of Figure 6-16 where both NOx and CO emissions are low. In this case
there is no significant degradation in engine performance, nor any increase in CO
emissions, which is in stark contrast to other NOx control technologies. In some
cases, prechamber-equipped engines actually run better than their predecessors. The
major engine manufacturers have made retrofit hardware available for some of the
existing population of engines.
OPEN CHAMBER PRE-CHAMBER
8-16 gm/HP-hr 1-2 gm/HP-hr

(7-14 lb/mmBTU) (1-2 lb/mmBTU)
Figure 6-16. Low Emission Combustion Chamber Concept

U.S. Environmental Protection Agency. Alternative Control Techniques Document: NOx
Emissions from Stationary Reciprocating Internal Combustion Engines, July 1993, EPA Report
453/R-93-032, NTIS PB94-104494.
Diesel engines have always been high NOx emitters and this continues to be the case.
Modest reductions can be achieved by retarding the timing, but this reduces engine
efficiency and peak power level. The exhaust temperature from diesel engines is in
the right range for use of selective catalytic reduction (SCR), so this may be the
appropriate control technology when substantial NOx reductions are necessary.
Table 6-1 shows some typical NOx emissions levels for the different types of engines
with uncontrolled and low emission combustion technology. In this table,
adjustments refer primarily to the effect of retarding the timing of ignition or fuel
injection.

Table 6-1. Reciprocating Engine NOx Emissions - lb/mmBTU
Concept Uncontrolled Adjustments Low Emission
Rich Burn,
4.64 3.5± 0.6
Spark Ignition
Lean Burn,
5.13 No Change 0.6
Spark Ignition
Diesel 3.95 2.7 NA
Dual Fuel 2.72 1.9 0.6
Source: U.S. Environmental Protection Agency. Alternative Control Techniques Document: NOx
Emissions from Stationary Reciprocating Internal Combustion Engines, July 1993, EPA Report
6.4.4 Add-On or Back End NOx Control Systems

The combustion-based NOx controls discussed so far use techniques specific to each
type of combustor. By contrast, back end controls work (with some limitations) on
any type of NOx source. In general, they are more expensive to install and operate,
but can be highly effective. Thus, their predominant use to date has been in those
areas where large NOx reductions were required.
Most of the NOx emitted from combustion sources is NO, which is not particularly
soluble in water, so water-based scrubbers have little effect on NOx emissions. Most
back end controls use a chemical to react with and destroy the NOx. The most common
chemical used to destroy NOx is ammonia. Simply stated, the desired reaction is:
NOx + NH3 → N2 + H2O
Although it is not written as a balanced equation, it is clear that the amount of ammonia
must be matched to the amount of NOx. One drawback to this approach is that
ammonia is highly toxic and a large tank of ammonia raises safety issues. Systems that
use urea (NH2-CO-NH2) are also in use – the general form of the chemical reaction is
similar. Urea is a solid that can be dissolved in water, but it does not pose the same
health hazard as ammonia. However, it is considerably more expensive.
Before delving into this in further detail, note that the amount of ammonia or urea
must be controlled to match the amount of NOx. Too little reagent and some NOx
will be left over, and too much reagent means ammonia emissions (ammonia slip).
So nearly all of the back-end approaches need a system that measures NOx levels and
controls the reagent injection rate – a cost and maintenance concern.

While increased PIC emissions are normally undesirable, these emissions are a necessary
element for one type of catalytic air pollution control – non-selective catalytic reduction
system (NSCR). Modern automobile engines operate in a rich burn mode in order for
this air pollution control catalyst to work (see Sections 6.4.2.1 and 6.4.4.3).
6.4.4.1 Selective Non-Catalytic Reduction – SNCR

NOx reduction with either ammonia or urea will occur naturally, but only within a fairly
narrow temperature window. For ammonia, that window is roughly 1650°-1800°F; the
urea window is only slightly wider. If the temperature is too low, the reaction is too slow
to be useful. If the temperature is too high, the reagent breaks down and forms additional
NOx. This temperature window is inaccessible on engines and combustion turbines, but it
does exist on most boilers. As the combustion gases pass from a boiler fire box through
the convective heating section, the gases pass through the critical temperature range at
some point in the boiler. However, on most existing boilers, the window is located in the
middle of a bank of tubes and there is not enough space to insert an ammonia distribution
grid.
Only a small amount of reagent is required, but it must be distributed evenly across a
large duct. The resulting need for a distribution grid and a few feet of duct length for
mixing/reaction means the system needs an empty area several feet long in the direction
of flow. This space can be designed into a new system, but is unlikely to be present on
an existing system, so retrofit may be impossible. There are several boilers that were
built originally with ammonia based SNCR systems. Exxon held the original patents for
this system under the name Thermal Denox.
One of the challenges in using SNCR, either with ammonia or urea, is that the
appropriate temperature location in the boiler moves around depending on the boiler load.
The lower the load, the lower the furnace temperatures and the closer the SNCR window
is to the fire box. As a result, SNCR systems usually need two or three injection grids.
The injection point is moved from one grid to another, depending on the load. Figure
6-17 shows the general configuration of an SNCR system.
Using urea dissolved in water gets around some of the problems with ammonia based
SNCR. The urea solution can be sprayed into the right place in the upper furnace and by
the time the water evaporates, the reagent has mixed with the flue gas and arrived at the
appropriate temperature regime. This requires careful planning and placement of water
sprays, which can be done using a fluid mechanic model of the boiler supported by actual
temperature measurements. SNCR has been implemented on a number of existing
furnaces, but is not possible on all furnaces. The disadvantage of urea compared to
ammonia is that urea is considerably more expensive. Nalco held the original patents for
urea-based SNCR and has licensed several companies to implement the technology.
SNCR application is limited to boilers and specifically to those with appropriate

geometry and temperature windows

Reheater
Super Heater
Ammonia/Urea Delivery System

Ammonia/Urea
Injection
Ports
Super Heater
Economizer
To Air Heater Air Compressor
Figure 6-17. SNCR General Configuration

Source: Johnson & Overduin, 1991
6.4.4.2 Selective Catalytic Reduction – SCR

Selective catalytic reduction (SCR) systems are routinely used to reduce NOx
emissions from combustion turbines and reciprocating engines. Systems have been
demonstrated on coal-fired boilers, but there are very few in operation in the U.S.
The main reason there are so few boiler SCR systems is that no new large steam
electric plants have been built in this country in the past 15 years and, until quite
recently, the rules have not required existing plants to achieve large NOx reductions.
However, it appears that the “NOx SIP call” will require existing utility plants to
begin implementing major NOx reductions. Since most of the combustion based
NOx reductions have been implemented, large additional reductions are likely to
mean the installation of SCR systems on many utility boilers.
SCR uses the same reagents as SNCR, but by flowing the flue gas and reagent
through an appropriate catalyst; the reaction occurs at a much lower temperature.
SCR systems can operate at temperatures ranging from about 600° to 1100°F,
depending on the catalyst. Many combustion sources have exhausts in this
temperature range, including engines and combustion turbines. The temperature
window is wide enough that most systems do not require any temperature controls, so
SCR has been successfully applied to many sources of various types. NOx reduction
efficiencies are typically in the range of about 90%. Systems are now being built
with outlet emissions guaranteed at 3 ppm. The primary disadvantage of an SCR
system, is the cost of installing and operating a system that includes the catalyst,
injection controls and reagent. Figure 6-18 shows the general configuration of an
SCR system.

Many factors go into the design and operation of an SCR system – this text briefly
explores some of the major elements. There are three primary types of catalyst and
each manufacturer has its own proprietary variations for each.
• Precious metal (platinum) 450° – 550°F

• Vanadium/titanium catalysts 550° – 800°F
• Iron-Zeolite catalysts 800° – 1000°F
Outside of these temperature windows, the catalysts are less effective and less
durable. Large boilers typically have temperatures at the exit of the economizer in
the range of 600°–800°F where vanadium/titanium catalysts are appropriate.
Combustion turbine exhausts are typically above 1000°F where zeolite catalysts have
marginal durability. However, most new power plants are combined cycle units so
space can be designed in the middle of the boiler where temperatures are appropriate
for one of the currently available catalysts.
Selective Catalytic Reduction
NH3 Air Injection
Air
Economizer SCR Reactor
SCR Catalyst Layer

Inlet 600-700°F
Damper
SCR
Steam Bypass
Generator
Air Heater
Evaporator
Combustion Air
Ammonia Storage Tank
Figure 6-18. SCR General Configuration

Cochran, Gregory & Rummenhohl, presented at Power-Gen in Dallas, TX 1993

A catalyst typically lasts several years before degrading to the point that it needs
to be replaced. When the exhaust gas is nearly free of contaminants, as is the case
with natural gas, a catalyst may last for many years. Vendors have developed
catalysts with considerable tolerance for the primary exhaust contaminants, but
eventually performance deteriorates and the catalyst must be replaced. Catalysts on
coal-fired boilers are expected to last about three years, but there is wide variation in
the contaminants in different coals, so actual catalyst lifetimes should be based on
experience. The reference by Ron Heck provides extensive information on air
pollution catalysts.
Catalysts take the form of a material coated on a support structure, with the gas
flowing through. The support is typically a honeycomb grid or a stack of waffle
plates. The amount of catalytic surface area is proportional to the gas flow rate.
When the exhaust is clean, as with an engine, the openings in the honeycomb or grid
can be small and a fairly large surface area can be achieved in a relatively small
volume. However, a dirty exhaust, as from a coal-fired boiler, requires large holes in
the support structure that will not clog with ash. This means the size of the catalyst
must be much larger in order to keep the same total surface area. Thus, catalysts on a
coal-fired boiler are nearly the size of a dust collector, while a combined cycle
catalyst can be neatly packaged within the boiler, similar to the heat exchange
sections.
6.4.4.3 NSCR with Rich Burning

Non-selective catalytic reduction (NSCR is introduced in Section 6.4.2.1) is a way to
destroy NOx using chemicals naturally present in the combustion exhaust – namely
CO and hydrocarbons (PIC). The catalysts developed to date are reliable and active
over a fairly wide temperature range. The hard part is to get just the right amount of
PIC to match the amount of NOx. As discussed in Section 6.4.2.1, PIC can be
generated by reducing the excess air, but very precise air control is required to keep
the amount of PIC within an acceptable range. Reciprocating engines are the only
sources where NSCR is currently being applied; boiler applications may be possible.
NSCR is not applicable to combustion turbines because of their inherently high
exhaust oxygen content. In order to apply it to a diesel engine, an air flow control
would be required, which would be a significant alteration of existing engine
technology. However, when NSCR is properly implemented on reciprocating engines,
NOx can be reduced by more than 90% while PIC are held to acceptable levels.
6.4.5 Combined NOx Controls

A variety of combustion sources and NOx control methods have been discussed with
a focus on the prevailing practice and the most effective technologies. In the
following table, the applicability of different technologies to the various types of
sources is summarized. In the instances listed as “maybe” the applicability of a
technology depends on site-specific conditions, or this technology normally would

not be used in favor of a more effective technology. The use of “??” indicates the
application is conceivable, but unlikely, or the technology may not yet exist.
Table 6-2. Combinations of NOx Control Technologies
Utility Package Stoker Combust. Gas-fired Diesel

Boiler Boiler Boiler Turbine Engine Engine
Excess Air Control yes yes ?? na no no
Low NOx Burner yes yes na yes yes ??
Overfire Air yes maybe ?? na na na
Flue Gas Recirc yes maybe ?? na maybe yes
Reburning yes ?? yes na na na
Water Injection ?? ?? no yes maybe maybe
Ignition Retard na na na na yes yes
NSCR maybe maybe no no yes no
SNCR maybe maybe maybe na na na
SCR yes yes yes yes yes yes
Most NOx reduction technologies can be combined and the resulting reduction will
be approximately the sum of the reductions due to each technology. As an example,
a low NOx emissions system installed on a utility boiler will probably include all of
the following:
• very good excess air controls

• low NOx burners
• flue gas recirculation
• over-fired air ports
The first two items can be installed on almost any size and boiler configuration. The
last two can usually be installed on large utility boilers, but may not be applicable to
industrial boilers.
When back end controls – SCR or SNCR – are used, it is customary to apply as many
combustion controls as practical in combination with the back end control. Thus, a
combustion turbine that requires maximum NOx reduction will use water injection or
a dry low NOx combustor in combination with SCR on the back end.
Controls that are not combined include:

• SNCR and SCR
• Water injection and dry low NOx combustor on a turbine
• Detuning and a low NOx combustor on a gas-fired engine

6.5 CARBON MONOXIDE AND ORGANIC EMISSIONS
The emissions of PIC divide naturally into two groups of sources:
• Sources where combustion is the vehicle for generating heat or power, and where
PIC might be created.
• Sources where combustion is used to control PIC or organic emissions from other
processes.
In regard to the second group, we have discussed incinerators and related pollution
control devices in previous chapters. However, with regard to the first group we
discussed the formation of PIC in Chapter 5 and provided an understanding of how
controlling PIC formation is the basis for prevention. Most stationary sources do not
generate significant amounts of CO or hydrocarbons. Boiler CO levels are normally
less than 100 or 200 ppm and hydrocarbons are rarely detectable. Engine PIC levels
are generally higher than boiler PIC emissions. Diesel emissions tend to include
hazardous or toxic organic compounds and organic particulate matter emissions that
are rare from boilers.
6.5.1 Trade-Offs with NOx

With few exceptions, combustion-based NOx reduction technologies tend to increase
PIC. Most low NOx flames are cooler and have slower fuel-air mixing than
conventional flames, which are the opposite of approaches used to minimize PIC.
Figure 5-9 illustrates these opposite trends for excess air – the most important
combustion control parameter.
With a few notable exceptions, PIC are typically not a serious concern from
stationary combustion sources. However, PIC will usually increase when NOx
controls are implemented. Thus, it is prudent for regulatory agencies to place a limit
on PIC emissions in any permit where NOx is a significant issue. This is particularly
relevant on gas-fired boilers where NOx compliance might be achieved simply by
reducing the excess air without regard for CO emissions. Natural gas boilers do not
normally smoke and CO emissions will not be detected unless they are measured.
This does not necessarily mean that CO needs to be continuously monitored, but
rather that there should be reasonable assurance that NOx compliance will be
achieved without excessive CO emissions.
6.5.2 Catalytic Control Systems

PIC emissions can commonly be controlled using an oxidation catalyst in the exhaust.
These catalysts are frequently used on engine and turbine exhausts – either alone or
directly downstream of an SCR catalyst. These catalysts operate well over a wide

range of temperatures; beginning at a few hundred degrees Fahrenheit they oxidize
CO and most hydrocarbons. Destruction efficiency is generally very high although
refractory hydrocarbons like methane require somewhat higher temperatures than
other compounds. In order to work, there must be some oxygen in the exhaust – a
requirement that will be met in nearly all cases with the possible exception of an
exhaust where an NSCR system is in operation.
6.5.3 Hydrocarbon Capture

Control of hydrocarbons in combustion exhausts is normally achieved by adjusting
the combustion system to prevent hydrocarbon formation, which is sufficient to
reduce organic emissions to less than 1 ppm. No further reduction is required for
most pollutants. However, certain highly toxic compounds such as dioxins and furans
need to be reduced to much lower levels. Even at very low levels dioxins can usually
be controlled by good combustion practices, but passing flue gas through activated
carbon will capture essentially all of the residual hydrocarbon species as well as
mercury vapor and some other compounds. The temperature of the carbon has to be
reasonably low, usually less than about 200°F. The cooler the carbon, the greater is
its capacity to absorb various pollutants.
Carbon capture can be achieved in one of several ways. Fixed carbon beds or
canisters are common on small sources. When the carbon approaches saturation it
can be replaced or regenerated in place. Regeneration means heating the carbon to
drive off the collected contaminants (and presumably capturing them for disposal or
destroying them). Doing it in place requires a double system with valving to switch
from one carbon bed to the other when one is being regenerated. A common
approach to carbon adsorption on larger sources is to inject powdered carbon into the
exhaust gas, then collect it in a baghouse. The coating of carbon on the bags ensures
that high contact will occur between the gas and the carbon so that high capture rates
can be achieved. If the same approach is used with an electrostatic precipitator,
capture rates typically drop to 50% or less. The contaminated carbon is discarded
with the collected dust.

6.6 CONCLUSION
Control of pollutant emissions usually starts in the combustion zone. The first step in
SO2 control is to use fuel with lower sulfur content. All NOx controls depend at least
partially on combustion modifications or precise combustion control. On boilers and
furnaces CO and hydrocarbon control is totally dependent on the design and operation
of the combustion system. Reciprocating engines and turbines frequently rely on
catalyst systems for CO and PIC control.
Post combustion or “back end” controls usually come into play when major emission
reductions are required and frequently emissions control is the sum of combustion
and back end controls. Most NOx controls are in this category, while many back end
control technologies are coupled to the combustion system. NSCR is an extreme
example in that it relies on exact control of the combustion air-fuel ratio. In all
cases, the design of a back end control device must be matched to the combustion
source both in terms of the amount of gas to be treated and the character of the
pollutants. For example, most dust collectors have very specific maximum gas flow
rates and their collection efficiency and durability depend on the
physical/chemical/electrical characteristics of the particles. The contaminated carbon
is discarded.
Air pollution control from combustion sources involves synergy between the
combustor and control devices as well as between the controls for different types of
pollutants.

Review Exercises
1. Which of the following statements is true?
a. The choice of fuel has little effect on most of the emissions from a typical source.
b. Combustion system design and adjustment can substantially influence the
performance of a pollution control device.
c. NOx emissions are only marginally influenced by the choice of fuel.
d. All air pollutants are created entirely in the combustion zone.
2. Which of the following statements is false?

a. Particles that follow stream lines exactly are the most difficult to collect
b. The performance of all particle collectors is affected by particle size.
c. Particles much smaller than 0.5µm are the most difficult to collect because the follow
streamlines exactly.
d. Particles larger than 50µm can be collected by any type of collector.
3. A particulate scrubber ________.

a. is quite effective at collecting submicron size fume particulate.
b. will capture nearly all the sulfuric acid mist in the gas because it is water soluble.
c. will not efficiently catch particulate smaller than about 1µm unless the gas pressure
drop is very high.
d. is not commonly used to collect flammable dust.
4. Which of the following statements is true?

a. Fabric filters can capture particles of any size with high efficiency.
b. Fabric filters can be permanently plugged if the dust is sticky or wet when collected.
c. Baghouse pressure drop is a very insensitive indicator of changes in emissions.
d. All of the above
5. If the collection efficiency of a new dust collector is 99.6% and leaks subsequently develop
that allow 1.3% of the exhaust to bypass the collector, the emissions will increase by a
factor of ________.
a. 4.25
b. 3.25
c. 2
d. Cannot determine from the information given.
6. Which of the following statements is not true about electrostatic precipitators?

a. Dust collection efficiency of an electrostatic precipitator is sensitive to the amount of
gas flowing through the unit.
b. Electrostatic precipitators can be designed to collect more than 99% of the particulate.
c. Precipitator collection performance can be monitored by tracking the secondary
amperage to each field.
d. All of the above are true.

7. Fires can be a problem in dust collectors on some types of sources. The best approach to
dealing with this problem is to ________.
a. install a wet scrubber even if the emissions may frequently exceed the standard
b. do whatever is necessary to assure an incombustible dust, including switching to a
more expensive fuel
c. operate the source in such a way that there will never be a simultaneous occurrence of
combustible dust, high O2 level, and an ignition source in the collector
d. never operate the source with an O2 level above about 12%
8. Switching fuel is a practical method of reducing SO2 emissions provided ________.

a. the sulfur content and heating value of the new fuel are lower than the existing fuel.
b. the increased cost of lower sulfur fuel is attractive compared with the cost of
installing and operating a scrubber for a comparable SO2 reduction.
c. the cost of changes to the facility required to burn the new fuel are reasonable.
d. B and C
9. If a baghouse is installed downstream of a dry or semi-dry SO2 scrubber one would expect
that ________.
a. all of the SO2 leaving the scrubber will go up the stack
b. if there is particulate carbon present at the scrubber exit, a significant amount of any
mercury in the gas will be captured in the baghouse
c. the baghouse will require special bag fabrics because moisture in the exhaust will
tend to plug up the bags
d. no sulfate particulate will reach the baghouse because all the sulfur is removed in the
scrubber
10. SO3 emissions are hard to control because ________.

a. at temperatures above 270°F they are a vapor that usually is not effectively captured
in a dust collector
b. a scrubber condenses the acid to submicron size particles that are not collected by a
scrubber
c. it is difficult to prevent SO3 formation in most combustion systems
d. All of the above
11. When trying to progressively reduce NOx from an existing boiler, one would implement
control strategies in which of the following sequences?
a. SCR, OFA, Low NOx burners.
b. Excess air control, OFA , low NOx burners, SCR
c. FGR, SCR, excess air control, low NOx burners
d. Burners out of service, SNCR, reburning, OFA

12. Flame temperature reduction will ________.
a. reduce the amount of NOx formed from fuel nitrogen
b. never increase the amount of CO or VOC emissions
c. be implemented most frequently with either water sprays or flue gas recirculation
d. always incur a major penalty in fuel efficiency
13. As a general rule, combustion modifications to reduce NOx will ________.

a. tend to increase emissions of CO, carbon and VOC
b. probably lengthen the flame so it fills more of the furnace
c. require good air flow control regardless of what techniques are utilized
d. All of the above
14. Which of the following is not an element of a low NOx burner?

a. Features that encourage the operators to tune the burner for best flame appearance.
b. Separate controls for air flow rate and air direction or distribution.
c. Low NOx emissions.
d. Factory tuning of the whole system after the burners are installed.
15. NOx emission control for combustion turbines ________.

a. never used combustion water spray because it would put out the flame
b. are using premixed burners in the latest technology
c. are able to use No. 2 oil in the new dry low NOx burners operating in the low NOx
mode
d. allows the turbines to achieve very low emissions during start up
16. Which statement is not true of low NOx combustion chambers for reciprocating engines?
a. Low NOx combustion chambers can reduce NOx emissions by about 90% from
uncontrolled levels.
b. The new low NOx chambers allow diesel-fueled engines to achieve very low
NOx emissions.
c. Low NOx combustion chambers only work when firing gas fuel.
d. In some cases, the low NOx chambers exhibit lower CO emissions and better engine
performance than the original chambers.
17. In regard to back end controls, which statement is incorrect?

a. Most systems rely on ammonia or urea to react with NOx and destroy it.
b. Both SCR and SNCR systems require accurately controlled amounts of reagent well
mixed with the exhaust gas.
c. SNCR systems can be retrofitted to a wide range of existing sources.
d. The reaction of ammonia with SOx will produce sticky compounds that can plug up
heat transfer equipment.

Review Answers
1. b. Combustion system design and adjustment can substantially influence the
performance of a pollution control device
2. c. Particles much smaller than 0.5µm are the most difficult to collect because the follow
streamlines exactly.
3. c. will not efficiently catch particulate smaller than about 1µm unless the gas pressure
drop is very high
5. a. 4.25
6. d. All of the above are true.
7. c. operate the source in such a way that there will never be a simultaneous occurrence of
combustible dust, high O2 level, and an ignition source in the collector
8. d. B and C
9. b. if there is particulate carbon present at the scrubber exit, a significant amount of any
mercury in the gas will be captured in the baghouse
11. b. Excess air control, OFA , low NOx burners, SCR
12. c. is most frequently implemented either with water sprays or flue gas recirculation
14. a. Features that encourage the operators to tune the burner for best flame appearance.
15. b. are using premixed burners in the latest technology
16. b. The new low NOx chambers allow diesel-fueled engines to achieve very low
NOx emissions
17. c. SNCR systems can be retrofitted to a wide range or existing sources

References
Buonicore, Anthony J. and Davis, Wayne T. (Editors). Air Pollution Engineering

Manual, (Air & Waste Management Association, Van Nostrand Reinhold, 1992).
This book has sections on combustion sources as well as various emission control
technologies.
California Air Resources Board. Sources and Control of Oxides of Nitrogen

Emissions, August 1997. This gives a relatively current summary of NOx formation
and control technology for all types of sources. Summaries of California rules
illustrate NOx levels that can be achieved by state-of-the-art technology.
Clement, Raymond and Kagel, Ronald, (Editors). Emissions from Combustion

Processes: Origin, Measurement, Control, (Lewis Publishers, 1990). The book is a
collection of 31 articles that focus primarily on organic pollutants.
waste combustion in rotating kilns and covers the formation and control of organic
pollutants.
Heck, Ronald M. and Farrauto, Robert J. Catalytic Air Pollution Control

(Commercial Technology, John Wiley & Sons, Inc., 1995). A 200-page book that is
probably the best available reference on this subject.
Institute of Clean Air Companies. Selective Catalytic Reduction (SCR) Control of

NOx Emissions, November 1997. A ‘white paper’ by the Institute of Clean Air
Companies.

NOx Emissions from Industrial/Commercial/Institutional (ICI) Boilers, March 1994,
EPA Report 453/R-94-022, NTIS PB94-177177. This weighty document is a fairly a
comprehensive description of small to mid size boilers and current NOx technology.
It includes waste fired boilers.

database included.



NOx Emissions from Utility Boilers, March 1994, EPA Report 453/R-94-023, NTIS
PB94-184165. This weighty document is a fairly a comprehensive description of
large boiler types and NOx technology with a significant database included.
U.S. Environmental Protection Agency. Compilation of Air Pollution Emission

Factors (AP-42), available from NTIS, Federal Book Stores, or available online from
the EPA TTN BBS. This ever-evolving document is organized by source type and
has credible descriptions of each one. Generally considered the "bible" for estimating
emission rates whenever actual measurements are not available.
U.S. Environmental Protection Agency. John Richards. Control of Gaseous

Emissions, APTI Course 415, Student Manual, NC State University, 1995. This book
has credible sections on incineration and NOx control as well as covering areas not
related to combustion sources.
U.S. Environmental Protection Agency. Control of NOx Emissions by Reburning,

February 1996, EPA Report 625/R-96/001. This appears to be a definitive document
on reburning, which, in practice, is only applicable to large boilers.
U.S. Environmental Protection Agency. New Source Performance Standards,

Subpart Da – Technical Support for Proposed Revisions to NOx Standard, June 1997,
EPA Report 453/R-94-012. This inch thick report is narrowly focused on utility
boilers, but covers current NOx technology in great depth.
U.S. Environmental Protection Agency. Phase II NOx Controls for the MARAMA
and NESCAUM Regions, November 1995, EPA Report 453/R-96-002. This report is
an in depth coverage of NOx technology on utility boilers.
U.S. Environmental Protection Agency. Campbell, Lisa M., Stone, Diana K., and
Shareef, Gunseli S. Sourcebook: NOx Control Technology Data, (Radian
Corporation, 1991), EPA Report 600/2-91-029, NTIS PB91-217364. A report for the
Air and Energy Engineering Research Laboratory of EPA. A broad, but very shallow
update and overview of NOx technology. The appendix contains a considerable
amount of data from existing sources on control technologies in use and permit limits.
U.S. Environmental Protection Agency. Summary of NOx Control Technologies and

Their Availability and Extent of Application, Midwest Research Institute, 1992, EPA
Report 450/3-92-004, NTIS PB92-158260. This appears to be a readable and fairly
comprehensive review of NOx emissions and combustion control technology for
boilers. There is minimal data for SCR and SNCR.

U.S. Environmental Protection Agency. The U.S. EPA-DOE-EPRI Combined Power
Plant Air Pollution Control Symposium: The Mega Symposium and the A&WMA
Specialty Conference on Mercury Emissions: Fate, Effects and Control, 2001,
proceedings available on CD-ROM from www.awma.org. This and other conference
proceedings provide current information on air pollutant formation and control.
Vatavuk, William M. Estimating Costs of Air Pollution Control, (Lewis Publishers,

1990). In addition to costs this book provides a lot of information about control
systems, because costs cannot be estimated until the system is defined.

Chapter 7: Combustion Source Emissions
LESSON GOAL
Demonstrate an understanding of the relationship between energy efficiency and
CO2 emissions. Understand the units used in emissions measurement through
successful completion of the chapter review exercises.
STUDENT OBJECTIVES
1. Describe the difference between combustion efficiency and thermal efficiency.

2. Describe the general levels of thermal efficiency expected from typical combustion
sources.
3. Calculate the heat loss from incomplete combustion by using measured CO
emissions data or fly ash carbon content.
4. Compute stack thermal heat losses from stack exhaust measurements.
5. Compute the CO2 emissions from sources fired on commercial fuels.
6. List the causes of variations in emission rates.
7. Describe the common types of continuous emission monitors, Reference Test
Methods, and indirect or parametric emission monitoring.
8. Correct an exhaust gas volume or a sample concentration to a standard level of air
dilution.
9. Convert between emission units of lb/mmBTU and ppm, corrected.
10. List two methods of determining an emission rate in units of lb/hr.

Chapter 7: Combustion Source Emissions
7.1 INTRODUCTION
This chapter presents several subject areas that do not fit neatly into any one of the
preceding chapters. The discussion will also address several topics that combine
elements from more than one of the preceding chapters. While this chapter is not
intended to serve as a conclusion for Combustion Source Evaluation, it does address
much of the material that is not sufficiently discussed in the preceding chapters.
7.2 ENERGY USE AND CO2 EMISSIONS
7.2.1 Efficiency of Various Systems

Table 7-1. Typical Efficiency Levels
System Combustion Efficiency Thermal Efficiency
Boiler, gas-fired 100% 82%
Steam-Elect. Coal Power

99% 34%
Plant
Simple Cycle Gas Turbine 100% 38%
Combined Cycle Gas 100% 55%

Turbine
Cogeneration System 100% 50%-80%
Efficiency is defined as the ratio of useful energy to total available energy, where the
total energy is that available from complete combustion of the fuel. Table 7-1 lists
two efficiencies for several typical combustion sources.
Combustion efficiency shows how much of the fuel was completely burned and
converted to thermal energy. For most combustion sources, combustion efficiency is
essentially 100%. None of the fuel remains unburned, so there is little air pollution in

the form of CO, black smoke or organic material (PIC). However, one exception is
coal-fired boilers that may lose up to one percent of the fuel in the form of carbon or
coke collected with the fly ash in the dust collector.
Thermal efficiency is entirely different. There is no way, even theoretically, to

capture and use all the heat that is released by combustion and carried away by the
flue gas. Boilers are fairly efficient because the only heat loss is the warm flue gas
going up the stack; nearly all the remaining energy is present in the steam. However,
devices that generate mechanical or electric power are much less efficient overall than
boilers and furnaces. For example, note that about 40% of the energy going into a
steam electric power plant is dumped from the steam condenser, either into an
adjacent body of water or into the atmosphere via a cooling tower.
Most of the electric power plants being built in this country have a combustion
turbine at their core. These plants provide an interesting example of thermal
efficiency. A combustion turbine (CT) – although it does not generate any steam that
must be condensed with the associated energy loss – does have a very large and hot
exhaust flow. Most of the input fuel energy to a CT is lost when hot gases escape up
the stack. A combined cycle power plant (see Figure 2-4, p. 2-9) captures most of the
turbine exhaust energy and uses it to heat a boiler that powers a steam turbine. The
net heat loss is diminished, but is still a major part of the input fuel energy. Another
alternative is to take the waste heat either from an engine or from a steam turbine
condenser and use it to heat a building or an industrial process – a procedure called
cogeneration. Cogeneration systems can potentially utilize 80% to 90% of the fuel
energy and still generate mechanical or electric power. However, in order to
minimize heat loss, there must be a good match between power demand and heat
demand – an objective that is often hard to achieve. Peak electric demand usually
occurs at a different time of the day or year than peak heat demand. So if the
cogeneration system is run to satisfy the electric demand, the heat output will not be
coordinated with the heat demand, or visa versa. Overall efficiency depends on how
well the two energy demands can be matched.
7.2.2 Combustion Efficiency

The simplest way to measure combustion efficiency is to measure the inefficiency,
or measure the amount of unburned fuel and then relate that to the total amount of
fuel. While there are many possible pollutant species, in regard to efficiency there are
only two of practical interest: carbon monoxide and particulate carbon. CO can be
emitted from any combustion source, but among stationary sources only gas-fired
sources can emit large amounts of CO without smoking. Carbon can also be emitted
from any source, but only sources that generate fly ash, such as coal or wood fired
boilers, can create significant amounts of carbon.
CO is a combustible gas with a heating value of 315 BTU/std ft3, so its presence
in the exhaust signals lost energy. The CO concentration is directly related to the
fractional heat loss.

⎛ ppm CO ⎞ ⎛ std ft CO ⎞ ⎛ 315 ⎞ ⎛ mmBTU ⎞ ⎛ std ft 3 E.G. ⎞
3
20.0
% Energy Loss = ⎜ ⎟⎜ ⎟ ⎜ 6 ⎟⎜
* ⎟ * 8, 710 ⎜ ⎟ *100 *
⎝ 10
6
⎠ ⎝ std ft E.G. ⎠ ⎝ 10 ⎠ ⎝ std ft CO ⎠
3 3
⎝ mmBTU ⎠ ( 20.9 - % O2 )
20.0
= 0.00027 * ppm CO * (7-1)
( 20.9 - % O2 )
Where: ppm CO = CO concentration in E.G., dry basis
% O2 = Oxygen concentration in E.G., % by volume, dry basis
8,750 = dry F-factor (Fd) of natural gas – see Table 4-5,
Example 7-1. Fractional Heat Loss
How much heat is being lost out the stack of a natural gas-fired source, due to CO,
where measured CO = 800 ppm and O2 = 4.2% by volume, dry basis?
Solution:
Insert the values for CO and O2 into Equation 7-1.
0.00027 * 800 * 20.9/(20.9-4.2) = 0.27%
So 800 ppm of CO emissions represents slightly more than one quarter of one percent
(of total heat input) energy loss.
Coal or any solid fueled furnace can leave carbon in the ash. When the ash is divided
between several streams, as between bottom ash and dust collector ash, calculating
losses starts to become complicated. Take, for example, the simple case of a pulverized
coal boiler, where the dust collector captures nearly all the fuel ash. If the fly ash is
analyzed using a test called Loss on Ignition or LOI, it gives us the percent carbon in
the ash. Analysis of the coal gives us the heating value and the amount of ash in the
coal. This is enough information to determine how much carbon is being emitted
relative to the amount of coal being fired. The equation for the heat loss associated
with unburned carbon in the fly ash is:
% Energy Loss = %A * %C/100 * 14100/HHV (7-2)
Where: %A = coal ash content, % by wt.

%C = fly ash carbon content (LOI), % by wt.
14,100 = heating value of pure carbon, BTU/lb
HHV = heating value of the coal, BTU/lb
This gives an energy loss estimate that can be determined from readily available data.

Example 7-2.
Determine the approximate energy loss from a pulverized coal source where the fly
ash has 4% combustibles and the coal properties are 11% ash and 12,700 BTU/lb.
Solution:
Insert the data in Equation 7-2.
11% * 4% /100 * 14100/12,700 = 0.49%
So 4% combustibles in the fly ash translate to about one half of one percent energy loss.
This calculation of energy loss works well for pulverized coal where the ash content
is well known and most of it comes out of the dust collector. However, this method
could give misleading results on a wood-fired stoker where the ash content fluctuates
between 1% and 3%, and where the ash emerges from several points in the boiler
system. The method also will not work well on a No. 6 oil-fired boiler where most of
the particulate is sulfates, not fuel ash.
Combustion losses are generally small and usually can be ignored when system
thermal efficiencies are calculated. Air emissions of CO or ash carbon content are
usually much more important issues than combustion efficiency losses. The bottom
line is that large PIC emission rates only represent small combustion efficiency
losses.
7.2.3 Thermal Efficiency

Thermal efficiency for a combustion source starts with the energy supplied by the
fuel and ends where we choose. For example, on a steam-electric power plant, we
can look at the efficiency of the boiler producing hot steam, or we can look at the
efficiency at the turbine shaft, or at the generator electrical output, or even the total
system efficiency where the end point is the electric power delivered to the user. This
course will only deal with the initial combustion source and not get into the details of
various thermodynamic cycles. This discussion will look at boilers but not the
associated steam turbines, nor the electric generator. The following discussion also
applies to combustion turbines. Reciprocating engines are more complicated simply
because of heat lost through cylinder cooling water. A schematic illustration of the
principal energy flows of a combustion source is shown in Figure 7-1.

Exhaust
Stack Loss:
• Hot Gas
Useful • Moisture
Energy Out • PIC (CO, Organics, Carbon)
Air
Combustion
Fuel Radiation or Cooling
Energy Water Losses
In
Figure 7-1. Combustion Energy Flows
Thermal efficiency can be expressed either as the ratio of useful energy output divided
by the total energy input, or as 100% minus the energy losses (expressed as percent)
% Thermal Efficiency = Useful Output Energy / Energy Input (7-3a)
% Thermal Efficiency = 100% - % Energy Losses (7-3b)
The first equation can be problematic to use in practice because either or both of the
energy terms can be difficult to measure. For many combustion sources, the second
equation can be disarmingly simple to use. Frequently equation (7-3b) allows a very
good approximate efficiency to be determined from measurements of stack temperature,
O2 and some knowledge of the fuel. Small changes in combustion efficiency can usually
be determined from simple stack data very accurately with this equation.
Losses from combustion sources fall into one of the following categories:
• Category 1. Sensible losses of hot exhaust gases venting to atmosphere.
• Category 2. Heat of vaporization losses from venting uncondensed water to

atmosphere.
• Category 3. Unburned fuel in either the exhaust gases or discarded ashes.
• Category 4. Radiation/convection losses from the outside walls of the furnace and
cooling water losses (reciprocating engines).
• Category 5. Miscellaneous small losses such as the energy in hot ashes that are
discarded.
If reciprocating engines are excluded for the moment, then in most cases the first two
are significant and the remaining three are each on the order of 1% or less. We have
already shown ways to determine the most common losses in Category 3. Category 4

can be determined by standard methods, such as those provided by the ASME Test
Codes – assuming there are no cooling water losses.
Category 2, water vapor losses, requires knowledge of the amount of water vapor in
the exhaust gases. Table 4-6 provides approximate values for uncondensed water
vapor losses for three common fuels. The associated discussion shows how to
calculate water vapor losses when exhaust water vapor content is measured.
Category 1, sensible heat losses, were discussed in Section 4.3.3 and quantified in
Equation 4-24. This loss can be determined fairly accurately from just two
measurements: stack temperature and stack O2 concentration.
Thus, for boilers, furnaces, and combustion turbines, the thermal efficiency is
approximately:
Efficiency = 100% - sensible loss (% QS ) - water vapor loss (% QLH )

= 100% - % QS - % QLH
Inserting Equations 4-21 and 4-24 for the two loss terms on a coal, oil or gas-fired
system gives:
⎡ ∆T ⎛ 20.9 ⎞ ⎛ 0.49 * % H2O ⎞ ⎛ 20.9 ⎞⎤

Efficiency = 100 ⎢1 − *⎜ ⎟−⎜ ⎟ *⎜ ⎟⎥
⎣ 4200 ⎝ 20.9 - % O2 ⎠ ⎝ 100 ⎠ ⎝ 20.9 - % O 2 ⎠ ⎦
⎡ ⎛ 20.9 ⎞ ⎧⎛ ∆T ⎞ ⎛ 0.49 * % H2O ⎞⎫⎤

= 100 ⎢1 − ⎜ ⎟ ⎨⎜ ⎟ −⎜ ⎟ ⎬⎥ (7-4)
⎣ ⎝ 20.9 - % O2 ⎠ ⎩⎝ 4200 ⎠ ⎝ 100 ⎠ ⎭⎦
Where: ∆T = Exhaust temperature minus ambient temperature

%H2O and %O2 are measured exhaust concentrations after the last
heat exchanger.
In this equation, the water vapor term, which includes the H2O and O2 variables, is
actually a constant for any given fuel. So when %H2O data is not available, one can
use an estimated loss such as the values listed in the last column of Table 4-6 (p. 4-34).
Typical water vapor energy losses are between 4% and 10%.
Using Equation 7-4 with measured stack data gives an efficiency that is slightly high
because it ignored radiation losses, but it will give values that are generally within 2%
of the values that could be achieved by a more thorough analysis. Equation 7-4 is a
simplified version of an ASME Performance Test Code and is a powerful engineering
tool. This equation is particularly valuable to air emission professionals because of the
ability to determine thermal efficiency fairly accurately without the need for any data
on fuel flow or the useful energy output.

Example 7-3.
Determine the approximate efficiency of an oil-fired boiler where the stack

temperature is 350°F, ambient temperature is 50°F, stack O2 is 5% and stack water
vapor content is estimated at 11% by volume.
Solution:
Entering these values in Equation 7-4 gives:
⎛ 350 − 50 20.9 11 20.9 ⎞

100 × ⎜ 1 − × − 0.49 × × ⎟=
⎝ 4200 20.9 − 5 100 20.9 − 5 ⎠
100 − 9.39 − 7.08 = 83.5%
So about 7% of the fuel energy is carried in uncondensed water vapor and 9.4% is in
the form of hot gas. This leaves 83.5% that is used to produce steam.
It was demonstrated in Chapter 4 that the firing rate of a source could be determined
from measured stack flow and O2 data. Here it has been shown that stack temperature
and O2 are the main measurements needed to determine the thermal efficiency of
most sources. This gives air emissions analysts the ability to quantitatively
characterize a source using stack data alone.
Equation 7-4 is particularly useful in determining how much the source efficiency
changes as a result of changes to the hardware or to the operation. If, through cleaning
or other improvements, the exhaust temperature is decreased to 300°F in the previous
example, the thermal loss term drops to 7.82% and efficiency rises to 85.1%. If,
through some combustion tuning, the exhaust O2 is decreased to 4% (with the stack at
350°F), the thermal losses decrease to 8.83% from 9.39% and the efficiency rises to
84.1%. So measuring exhaust temperature and O2 in conjunction with Equation 7-4
provides a very sensitive way to track thermal efficiency changes.
One must remember that combustion efficiency is inherently high in nearly all
systems and that only a tiny fraction of the fuel is left unburned, creating emissions.
However, the thermal efficiency of combustion devices varies widely, depending on
the thermodynamic cycle. The common misuse of the term “combustion efficiency,”
to characterize overall thermal efficiency, should not be interpreted to mean that
unburned fuel is going up the stack. Standard stack emission measurements can be
used to determine emissions, thermal losses, and the source firing rate.

7.2.4 Power Plant Efficiency
The traditional steam-electric power plant provides an illustration of how efficiency
relates to fuel consumption. The generating units in these plants consist of a very big
high-pressure boiler connected to a steam turbine that is coupled to an electric generator.
Typical stack parameters on these boilers are 300° to 350°F and 5% O2, similar to the
last example. So, based on the discussion in the preceding Section, we know that the
efficiency of the boiler in converting fuel energy to steam energy is in the range of 80-
85%. [Boiler efficiency varies with fuel type – coal is more efficient and natural gas
less efficient because of the amount of water generated during combustion.] So if the
boiler is 84% efficient, then 100 mmBTU of input energy will produce 84 mmBTU of
steam energy. Energy loss at this stage of the process is only 16%, which is relatively
small.
The overall efficiency of electric power plants is characterized by the heat rate, which
is the number of BTU required to produce a kilowatt-hour of electric power. In
typical steam-electric power plants, this number is around 10,000 BTU/kw-hr; the
most efficient plants reach about 9,000 BTU/kw-hr.
Note that if the power generating process were 100% efficient, then heat rate would
be 3410 BTU/kw-hr. So the typical power plant efficiency is
3410/10,000 = 0.341
or 34.1% efficient. If the overall generating system is about 34% efficient and the
boiler is over 80% efficient, there must be a huge loss somewhere. Most of the loss is
in the condenser, which dumps about 40% of the input energy into the air or into an
adjacent body of water. Steam drives the turbine that drives the electric generator. A
few percent of the steam energy is lost in both of these components. However, the
largest loss is in the cool low-pressure steam after it leaves the final turbine stage.
This spent steam passes through a condenser that converts it back to water to feed the
boiler and in the process, gives up about 1000 BTU per pound of steam to the
condenser
Example 7-4
How much energy is required to operate a 150-megawatt (MW) power plant with a
heat rate of 10,200 BTU/kw-hr? What is the overall thermal efficiency?
Solution:
Multiply the heat rate by the load (using consistent units) to get energy use.
150,000 kw * 10,200 BTU/kw-hr = 1530 mmBTU/hr

The efficiency follows from the ratio the ideal energy to actual energy use:
3410/10,200 = 0.334 = 33.4%
Knowing the energy input can be used to determine the amount of fuel used or to
estimate the total emissions.
The newest combustion turbine power plants boast heat rates in excess of 7500
BTU/kw-hr. However, engine manufacturers compute heat rate using the lower
heating value of the fuel (usually natural gas). By using the lower heating value the
manufacturers can claim higher efficiency. Using the data from Table 3-4 (Fuel
Heating Values), one may convert to a higher heating value basis, which gives about
8300 BTU/kw-hr. This can be compared directly to the heat rates reported by steam-
electric power plants. New combined cycle power plants are considerably more
efficient than simple cycle engines or steam power plants and some are now claiming
thermal efficiencies above 60%, based on the LHV of the fuel.
7.2.5 CO2 Emissions

Carbon is the principal constituent of all our common fuels and the resulting carbon
dioxide is a greenhouse gas. The CO2 emission rate depends on the type of fuel, and
is directly proportional to the amount of heat released. The amount of CO2 generated
by burning fuel can be calculated in the same way as pollutant emissions (see
Equation 5-1).
lb CO2 % fuel C/100 44

= × (7-5)
mmBTU HHV/106 12
Where: % fuel C is the fuel % carbon content (ultimate analysis)

44/12 is the molecular weight ration of CO2/C
Table 7-2 lists some typical emission rates for various fuels.
Table 7-2. CO2 Typical Emission Rates

Fuel CO2 lb/mmBTU
Natural Gas 120
No. 2 Oil, Diesel 165
No. 6 Oil 180
Bituminous Coal 185
Lignite 300+
Carbon 260

Clearly, natural gas is the best fuel for minimizing CO2 emissions because more than
half its energy comes from burning hydrogen. Heavy oil or coal increases the
emissions by 50%. Lignite has relatively high CO2 emissions because, not only does
most of the energy derives from burning carbon, but some energy must be used to
evaporate all the water that is typical of lignite fuel. While natural gas is desirable, if
used efficiently, any substantial gas leaks could offset the advantage because methane
is a far more potent greenhouse gas than CO2.
To compute the actual emissions, multiply the firing rate by the CO2 emissions rate
from Table 7-2.
Example 7-5.
From the last example of a 150 MW power plant that consumes 1530 mmBTU/hr of
energy, assuming it is fired on No. 6 oil, what is the CO2 emissions rate?
Solution:
Take the CO2 emission rate for No. 6 oil from Table 7-2 and multiply by the heat
input.
1530 mmBTU/hr * 180 lb CO2/mmBTU = 275,400 lb CO2/hr
So this 150 MW power plant at full load will emit about 138 tons/hr of CO2, or as
much as 3300 tons/day. Switching the fuel to natural gas would reduce the potential
emissions to about 2000 tons/day – still a substantial quantity.
7.3 EMISSIONS MONITORING
7.3.1 Emissions Variability

Source emission rates will vary for a variety of reasons. If mechanical breakdown
and long term wear are excluded, other potential causes include:
• Load changes
• Operator implemented changes
• Natural short term fluctuations – particularly excess air level
• Changes in atmospheric conditions – specifically temperature and humidity
• Start up of cold equipment
• Fluctuations in fuel properties
Each of these factors has a different interpretation depending on the type of source
and the type of emissions. However, the following discussion is only intended as an

overview. Table 7-3 attempts to organize the subject and provide a general indication
of time scales and magnitudes.
Table 7-3. Typical Variation in Emission Rates

Source and Cause of Variation, Time Scale NOx PIC
Boiler - operator instigated changes (1 hr) ±15% 0 to Excessive
Boiler - natural draft fluctuations (15 sec.) ±5% ±50%
Reciprocating engine - atmospheric change (12 hr) ±5% ±10%
Gas Turbine - atmospheric changes (12 hr) ±10% ±10%
Waste combustor - waste properties (15 min) ±15% ±75%
The numbers here are illustrative and not to be used to predict the variations from any
given source.
Note in Table 7-3, in nearly all cases, the fluctuation in NOx emissions is less than
the fluctuation in PIC emissions. The reason for this is evident by examining Figure 5-9.
NOx varies gradually and steadily as excess air changes. CO and smoke are either near
zero or, they rise dramatically as excess air decreases. So small fluctuations in excess air
can trigger large changes in PIC emissions, but only small changes in NOx emissions. In
boilers and furnaces, the air flow is never perfectly steady, but fluctuates naturally by one
or two percent. When the average excess air level is near the lowest practical level, air
flow fluctuations are enough to generate spikes of CO and/or smoke. This is the reason
that gray smoke from boilers is emitted in unsteady puffs and the trace from the opacity
meter shows either nothing when there is plenty of excess air, or a series of spikes when
the excess air is low. On boilers, the operators directly control the excess air level, so
emissions can vary over a wide range depending on how the operator is running the boiler.
Engine emissions are usually very stable. Engine air flows experience much smaller and
more rapid fluctuations than boilers, so excess air does not drive noticeable changes in
engine emissions. A significant increase in air humidity will cause a noticeable decrease
in NOx formation on either reciprocating or turbine engines. Ambient temperature has a
strong effect on maximum engine load for either reciprocating or turbine engines and this
can affect peak emissions.
Load always affects the total mass emissions (lb/hr) and usually affects the emission rate
(concentration in units of ppm or lb/mmBTU) on all sources. Over the normal operating
range, an increase in load generally means an increase in NOx emission rate. PIC are
usually low and they may go up or down with load. Low or minimum load on many
sources (boilers and engines) can be a higher emission point for both NOx and PIC.
Boilers frequently have higher excess air levels at low load that increase NOx emissions.
At idle, both reciprocating and turbine engines can be operating outside of their optimum
envelope and both NOx and PIC emission rates can be higher. Start up generally means

higher emissions for all types of combustion sources. Start up times vary from a few
minutes for some engines to many hours for large boilers.
7.3.2 Measurement Methods

This section is a brief introduction to emissions monitoring. There are three general
methods of measuring emissions: (1) continuous emission monitors (CEMs), (2) federal
reference method (FRM or RM) sampling trains, and (3) indirect or parametric emission
monitoring (PEM). CEMs are used for most gas phase emissions. FRM sampling trains
are required for particulate mass emissions and other pollutants that cannot be measured
by continuous instruments, either because the appropriate instrument does not exist or the
concentration levels are below the range of available instruments. Indirect (PEM)
monitoring infers emissions on a specific source from other data such as fuel flow and
exhaust temperature. An indirect monitoring system must be calibrated initially by one
of the direct measurement methods. Indirect monitoring is generally less expensive than
direct measurements and in some cases gives equivalent or better data.
Direct measurement methods determine the concentration of the pollutant, either in

volume based units such as percent or parts per million (ppm), or in weight based
units such as mg/m3 or grains/ft3. A direct measuring system must also record
supporting data including one or more of the following:
• O2 or CO2 concentration
• Exhaust flow rate
• Exhaust moisture content
• Fuel flow rate or power output
• Ambient conditions (temperature, atmospheric pressure, humidity)
Some of this supplemental data is necessary to calculate and report data in units that
may be compared to permit levels. Additional data may include ambient conditions,
exhaust temperature and pressure, etc. that may be necessary for data reduction or to
meet other permit requirements.
7.3.2.1 Continuous Emission Monitoring Systems (CEM)

The basic elements of a CEM system are shown in Figure 7-2. A gas sample is sucked
from a probe in the stack, through an umbilical tube to an analysis cabinet. The analysis
cabinet houses a sample gas pumping, drying and filtering system to supply clean dry gas
to the analyzers. Instruments in the analysis cabinet measure the concentration of various
species and report the results to a data acquisition system (DAS). Periodically,
calibration gases from cylinders are sent through the system and the instruments are
calibrated. These systems can either be set up temporarily and operated manually
(similar to reference method test methods of 40 CFR 60, Appendix A), or they can be
permanently installed in which case they operate automatically. On permanent CEM

systems there are one or more computers to (1) operate and calibrate the sampling system
and (2) process and store the data.
Stack
Probe and Umbillical Cord DAS

with Sample Line, Calibration Gas Line,
Electric Power and Control
Air conditioned
weather proof
Sample enclosure
Analysis
Calibration,
etc.
Figure 7-2. Elements of a Continuous Monitoring System
There are several types of CEM systems, each designed around a different method to
handle one universal concern – water condensation in the sampling system. Liquid
water in the sample stops or damages the instrument analyzers and absorbs some of
the pollutants, such as NO2 and SO2. Flue gas moisture levels are typically 10% or
more, which means the water condenses when the CEM sample gas temperature
drops below about 120°-130°F. Drying the sample (by removing water or by dilution
with dry air) changes its volume and alters the concentration of all pollutants. Water
vapor in the flue gas introduces complex issues for both measurement and data
reduction in the CEM. A brief description is provided about how the two most
common systems work.
A dry extractive CEM sucks the gas sample through a heated system to the analysis
cabinet. The gas is kept above the dew point temperature until it passes through a
condenser that removes water with minimum contact between liquid water and the
sample gas. The sample gas is analyzed and all the raw concentration data is obtained
in units of ppm or percent concentration in dry exhaust gas (ppm or % on a dry basis).
However, many of the regulations require measurement of total mass emissions –
lb/hr. This typically requires stack flow rate data that is easy to measure, but reflects
the flow of wet (total) flue gas. In order to combine the dry concentration data with
the wet flow rate data, the system needs the exhaust gas moisture content, which
requires additional data.
The dilution extraction CEM was developed to overcome the problem of correctly
matching dry and wet data. As the exhaust sample is drawn from the stack it is diluted
with a constant flow of dry air to give a sample with a known dilution ratio. Dilution

ratios of 100:1 are common. This reduces the water vapor concentration and drops the
dew point temperature well below the ambient temperature. The diluted gas sample
can be passed through the instruments without any water condensation. Since the
sample contains the water and all the original pollutants it is still considered a “wet”
sample. Calibration gases are also passed through the dilution system so the
instruments are calibrated to stack concentration levels even though the sample
concentrations at the instrument are a fraction of stack levels. So, for example, if the
dilution rate was 100:1 and 50 ppm cal gas was passed through the system, the
instrument would see 0.5 ppm. The data system would interpret the instrument reading
as 50 ppm. Concentration data from the instruments in a dilution extraction system can
be combined with stack flow data without the need for a water vapor correction.
Perhaps the main drawback to the dilution extraction system emerges on certain new
sources with NOx emission levels of a few ppm, resulting in extremely low
concentrations reaching the instruments.
7.3.2.2 Federal Reference Method Sampling Trains

Reference method (RM) sampling trains are used to measure particulate emissions
and other pollutants where acceptable continuous analyzers are not available. The
sampling system is set up adjacent to the port(s) on the stack or duct that are used for
sample extraction. Typically a sampling train draws sample for one hour and the
resulting measurement is an average for that one-hour period. Usually three one-hour
test runs are made and results are reported as the average. Figure 7-3 shows the
components of a Method 5 particulate sampling train. Here particulate matter is
collected on a filter and the volume of gas sampled is measured with a dry gas meter.
The filter is weighed before and after the test giving a weight that can be combined
with the sample volume to give a concentration – grams/m3 or lb/ft3. Impingers in an
ice bath condense the water vapor, which is subsequently measured. The impingers
can also be used to capture condensable particulate or certain gases, such as
ammonia, formaldehyde, HCl, etc. Most reference method sample trains are similar
to that shown in Figure 7-3. The difference is that the filter would be replaced by
something like a canister of resin adsorbent, or there might be additional collection
devices between or after the impingers. The collection devices are taken to a lab to
analyze for specific substances.

Stack
wall
Thermometer
Temperature
sensor Thermometer
Check
Filter valve
holder
Heated
S - type area Vacuum
pitot tube line
Pitot Thermometers Impingers Ice bath
manometer By-pass Vacuum
Orifice valve gauge
Gas exit
Main
valve
Dry
gas
meter
Orifice Air-tight
manometer pump
Figure 7-3. EPA Method 5 Sampling Train

Source: 40CFR60 Appendix A
7.3.2.3 Indirect or Parametric Emission Monitoring (PEM)

Indirect (PEM) monitoring is attractive primarily because it is much ess expensive
than a CEM and it gives continuous data. It is technically feasible to implement PEM
on any source where the emissions can be accurately predicted from readily available
data such as fuel flow, exhaust temperature, etc. PEM has been applied quite
successfully to monitor NOx emissions from combustion turbines and reciprocating
engines. It may also be used to monitor NOx emissions from small to medium size
industrial boilers. Engine sources equipped with a SCR system are also candidates;
although, catalyst deterioration would not be detected by a PEM system.
PEM systems are much less accurate when used to monitor CO emissions. As noted
earlier, CO emissions are very sensitive to small changes in excess air and other
combustor parameters, so CO does not lend itself well to indirect monitoring.
PEM systems usually rely on fuel flow or load and one or two other parameters as the
primary basis for determining emissions. The algorithms used to convert this data to
emissions rate can be set up to report emissions in pounds per hour or lb/mmBTU or
both. Hence the issue of how to deal with exhaust moisture content is moot. PEM
systems generally do not report either exhaust flow rate or moisture; although they
could be programmed to do so.
7.3.3 Calculating Emissions

CEM instruments measure concentration in units of parts per million by volume.
Depending on whether it is a dry extraction or a dilution extraction system the measured
data could be ppmd (ppm by vol., dry basis) or ppmw (ppm by vol., wet basis).

A CEM system processes the measured data and typically reports emissions in units of:
• lb/hr,
• lb/mmBTU, or
• ppmd corrected for dilution
From these units, one can generally convert to other units using simple formulas and
possibly some supporting data.
7.3.3.1 Correcting for Dilution

Exhaust gas is composed of the products of combustion plus a certain amount of
excess air. Excess air in the stack gas may originate either in the combustion zone or
from leaks in the ducts. Regardless of its origin, it dilutes the combustion gases and
is present in the sample analyzed by the CEM. If emission standards were based on
concentration as measured, then any source could reduce its emissions by simply
diluting its flue gas with air. Sources with leaky ducts would have an advantage over
sources with tight ducts. So to avoid this problem all emission concentrations are
corrected for dilution and reported in forms such as:
• ppmvd corrected to 3% O2, or • ppmvd corrected to 12% CO2
Measured O2 or CO2 are used to determine the amount of dilution air present in the
sampled gas. The formula for correcting the measured emissions concentration is:
20.9 − Y
ppm @ Y%O2 = ppm(meas.) × (7-6a)
20.9 − % O2 (meas.)
x%
ppm corrected to X% CO2 = ppm(meas.) * (7-6b)
% CO2 (meas.)
Where: X and Y are percentages specified by the applicable emission
standard for the source
Example 7-6.
If the measured data is NOx = 135 ppm, O2 = 4.7%, what is the NOx concentration
when corrected to 3% O2?
Solution:
Inserting the data in equation (7-6a):
20.9 − 3
135ppm × = 149ppm
20.9 − 4.7
3%O2 was used for this example; a number of other reference values are used in the
Federal regulations, but the values are typical of the sources to which the standards

apply. Note that O2 or CO2 must be measured at the same time and location as the
emissions; standard practice is to divide the sample flow, sending part of it to each
instrument.
Note that this is only a correction for air dilution of the flue gas and not a correction
for the effect that the excess air in the combustion zone has on the amounts of
pollutants formed. Changing combustion zone excess air will alter emission levels as
well as changing the amount of air dilution.
7.3.3.2 Emissions in lb/mmBTU

One of the most common, and most useful, emission units is pounds per million BTU
(lb/mmBTU). One particularly attractive feature of this dimension is that it is not
necessary to measure the firing rate of the source to determine the emissions. All the
necessary data can be acquired at the stack sampling location; the key measured
parameters are the pollutant concentration and O2 concentration. This means the
reported emissions are not dependent on data, such as fuel flow rate, provided by the
source operator. In fact, except for the use of the F-factor, lb/mmBTU uses the same
data as ppm (corrected). The formula is:
⎛ lb ⎞ A (ppmdv) MWA 20.9

A ⎜ ⎟ = × × Fd × (7-7)
⎝ mmBTU ⎠ 1,000,000 385 20.9 − %O2
Where: ppmdv A = measured concentration of air pollutant A

MWA = molecular weight of A, 46 for NO2, 64 for SO2, etc.
%O2 = measured oxygen concentration, % by vol., dry basis
385 = std ft3 / lb-mole of ideal gas
Fd = dry F-factor, std ft3 / mmBTU
bituminous coal 9780
oil 9190
natural gas 8710
wood 9240
Note that the measured values (ppm of pollutant and % O2) are both based on dry flue
gas. Neither gas flow rate nor source firing rate are needed. Since essentially the
same data is used to calculate lb/mmBTU and ppm corrected, it makes little
difference which dimensions are used unless the source switches fuels. Units of
lb/mmBTU may be easier to use for computing emission inventories, etc.
7.3.3.3 Emissions in lb/hr

Mass emission rates in pounds per hour (lb/hr) can be determined either by
multiplying the source firing rate (mmBTU/hr) by the emission concentration
(lb/mmBTU), or it can be determined entirely from stack data. The basic formula for
computing lb/hr from measured data is:

A lb( hr ) = exhaust flow ( scf hr ) × ppmw of A MW
1,000,000
×
385
A
(7-8)
Where: A = pollutant species

MWA = molecular weight of species A
exhaust flow is the total (wet) flue gas flow in standard cubic
feet per hour
A is measured in a wet (not dried) sample
385 = the number of standard cubic feet of gas in a pound
mole @ 68°F. (MW/385 = gas density in lb/3)
Note that there is no dilution correction here because the total flue gas flow is measured.
However, the pollutant must be measured in a wet gas sample (dilution extraction
system) or a dry measurement must be corrected back to its wet concentration. This
same equation could be written with dry measurements; however, the trouble is that
there is no simple instrument to measure dry flue gas flow rate.
As an alternative, mass emission rate can be determined without measuring stack flow if
the firing rate of the source is known. The formula is:
Emissions lb ( hr ) = Emissions (lb mmBTU) × Firing Rate (mmBTU hr ) (7-9)
In this case, the firing rate (mmBTU/hr) would be determined from fuel flow measure-
ments and the emissions in lb/mmBTU from stack measurements and Equation 7-7.
This method is applicable to gas and oil-fired sources where the fuel heating value is
consistent and the fuel flow can be measured accurately. On a coal-fired source the fuel
heating value can fluctuate over a considerable range, so this approach to emissions
calculation is tenuous. On a solid waste-fired source the fuel analysis varies over such
a wide range that this method of calculating emissions is not applicable.
7.4 RECIPROCATING ENGINES AND COMBUSTION

TURBINES
7.4.1 Characteristics of Reciprocating Engines and

Turbines
Combustion turbines are the prime mover in most new electric power plants in the
U.S. The price and availability of natural gas is one of the reasons for this
dominance. Most of these units run entirely on natural gas; although many can use
distillate oil (No. 2, diesel or kerosene) as a backup fuel if the gas supply is shut off.

Many new installations are combined cycle or cogeneration systems that use the waste
heat from the turbine or reciprocating engine to make hot water or steam for other uses.
Without waste heat recovery turbines and reciprocating engines have about the same
thermal efficiency as a steam electric power plant – about 35 -40%. The latest large
combustion turbines are somewhat higher; however, small turbines may be
considerably lower. Nonetheless, with waste heat recovery, all of these systems
generally can exceed 50% in overall thermal efficiency. High efficiency, combined
with the fact that they burn natural gas, means they emit considerably less carbon
dioxide and almost no sulfur compared to traditional coal-fired steam-electric power
plants.
7.4.2 Emissions and Control

Since most turbines and reciprocating engines run on natural gas or occasionally on
clean oil, the only emission of serious concern is NOx. PIC are rarely high and in most
new installations they are further controlled to very low levels with an oxidation
catalyst. Low NOx combustion systems, based on premixing natural gas and air, have
been developed for both turbines and reciprocating engines that reduce NOx
dramatically from previous uncontrolled levels. In NOx sensitive areas, low NOx
combustors are supplemented with SCR systems.
Uncontrolled NOx emissions from nearly all reciprocating engines are very high – in
the range of 4 lb/mmBTU. On gas-fired engines, emissions can be reduced by about
90% if low NOx combustion chambers are available. Also on gas-fired engines
emissions can be reduced dramatically by installing a non-selective catalytic reduction
(NSCR) system and controlling the engine air flow to keep the air fuel-air ratio precisely
stoichiometric. No low NOx combustion systems are yet available for diesel engines,
but SCR controls can be installed on any reciprocating engine. Technology
development is continuing on the development of emission controls for both gas and oil-
fired engines.
On new installations it is fairly common practice to put an SCR NOx control system
on the exhaust of either a reciprocating engine or a combustion turbine. At the present
time the upper temperature limit for zeolite catalysts used in this application is about
1000°F. Higher temperatures decrease the life of the catalyst. Most turbines have
exhaust temperatures that exceed 1000°F at full load, so the temperature needs to be
reduced. When the turbine exhausts to a boiler (combined cycle or cogeneration), this
is accomplished by locating the catalyst in the middle of the boiler after the gas
temperature has been reduced. Figure 7-4 shows the general arrangement. The first
section of the boiler reduces the temperature to a level acceptable to the catalyst.

Bypass
Turbine NOx
Exhaust
Bypass
Damper
NH3 Vaporizer
Air Heater
Figure 7-4. SCR Installation on a Combined Cycle System
Emissions during the startup of combustion turbines are higher than full load emissions,
especially if the turbine is equipped with a low NOx combustor. Even large turbines
can start up quickly and generally can achieve full load in about 20 minutes, so high
startup emissions might not be a serious problem. However, if the turbine feeds a
boiler (heat recovery steam generator or HRSG), the boiler requires several hours to
warm up. Many of these systems are equipped with a bypass duct that allows the
turbine to produce load while the boiler is warming up using perhaps 20% of the
turbine exhaust. However, the exhaust flow through the bypass does not go through the
SCR system. So even if the turbine load is high enough to put it into its low NOx
operating mode, the additional NOx reduction of the SCR system will not be achieved
until the boiler warms up and all the exhaust passes through the boiler.
CO and organic vapor emissions are relatively low, but sometimes significant from
either combustion turbines or reciprocating engines. A simple way to deal with PIC
emissions is to install an oxidation catalyst after the SCR system. CO and organic
vapor emissions can be reduced to levels of a few ppm.
7.5 NATURAL GAS, NO. 2 AND NO. 4 OIL FIRED BOILERS
7.5.1 Characteristics of Boilers

Boilers are the most populous stationary source. They range in size from residential
furnaces to utility boilers fired at thousands of mmBTU/hr. The focus of this course is
with industrial and utility boilers that have firing rates higher than about 10 mmBTU/hr.
Boilers can be designed to burn virtually any fuel from natural gas to low-grade coal
mine waste. The emissions depend on the fuels used and on the design of air pollution
control equipment. Small natural gas-fired boilers have little or no emissions control,
while large coal- or waste-fired boilers have extensive air pollution control hardware.

Emissions from boilers, depending on the fuel, can include almost any air pollutant, but
with rare exceptions the emissions do not include significant levels of organic
compounds. Boiler CO emissions are generally quite low.
Boilers typically have long lives. They are easy to repair and boiler technology has not
changed very much over the past 40 to 50 years. So many boilers built after about 1950
are not obsolete; they have not worn out and they are still operating. Thus much of the
current boiler population predates all environmental regulation – other than rules to
avoid excessive smoke. These boilers have been repaired and the burners and controls
have probably been upgraded. But most of these upgrades have focused on improved
reliability or efficiency, not on reduced emissions. So the current population of boilers
include (1) some newer boilers built to meet varying levels of emission standards, and
(2) many older boilers, with varying levels of combustion upgrade.
7.5.2 Nitrogen Oxides Control

At the beginning of the 21st Century, NOx is controlled on existing boilers largely by
means of combustion or operational adjustments to reduce NOx formation. There are
a few SNCR systems in service – mostly on municipal waste to energy boilers.
Combustion NOx controls consist of various combinations of the following elements:
• operating controls including low excess air and various burner adjustments
• staged combustion on large furnaces using specific combinations of burners
• low NOx burners
• overfired air
• flue gas recirculation.
• natural gas reburning
• switching fuel to natural gas or distillate oil
Aggressive use of combustion hardware adjustment/changes (the first five elements)

can reduce NOx by 50% on most boilers and by 80%–90% on natural gas or distillate
oil-fired boilers. The main cost of these changes is the initial capital cost of hardware
changes. Fuel switching or natural gas reburning generally triggers significant
operating costs because distillate oil is always more expensive than No. 6 oil, and
natural gas can be more expensive than other fuels, depending on location and time of
year. Fuel switching can be incremental, such as blending distillate and No. 6 oil, or
seasonal such as switching to natural gas during summer months, or a combination.
NOx emissions are controlled to reduce ozone, which is a seasonal air pollution issue.
Thus, depending on the jurisdiction, NOx reductions may be more important in the
summer than in the winter. Most of the combustion hardware based controls can be
operated year-round with little or no cost penalty, but fuel switching can be expensive and
so it may be reserved for those times when additional NOx emission control is necessary.

Low NOx burner development is a continuing process and some further reduction
from existing emission levels is likely. However, the largest reductions have
probably already been achieved. Progress on large, multi-burner utility boilers is
likely to be site specific rather than the result of new, low NOx burner technology.
Except for a few SNCR installations, primarily on municipal waste-to-energy boilers,

SNCR or SCR controls have not been implemented on many boilers in the U.S. SCR
systems may begin to appear on boilers as a result of regulatory pressure to reduce
ozone levels in the eastern U.S. and some other areas.
7.5.3 Sulfur Oxides Control

Natural gas contains almost no sulfur and No. 2 oil contains low levels of sulfur –
usually much less than the 0.5% permitted in ASTM specifications. So sulfur oxide
emissions are an issue for boilers fired on No. 6 oil, coal, and some wastes. There are
a few sulfur oxide scrubbers installed on coal-fired boilers in the U.S., but most sulfur
reductions have been achieved by changing fuels. Coal and oil are sold on the basis of
fuel heating value and fuel sulfur content, making it practical to control emissions by
controlling the sulfur content of purchased fuel. When sulfur oxide scrubbers were
first installed on a few large utility boilers in the 1980s there was little prior experience
and these installations encountered a number of serious problems. In the intervening
years, there have been significant advances in scrubber technology and the utility
industry has accumulated operating experience. So the options for SO2 emissions
control can be evaluated with much more certainty about their cost and impact.
SO3 (sulfuric acid) emissions are not directly controlled by regulations in most areas,
but the emissions are indirectly controlled by prohibitions against visible plumes from
the stack. When vapor phase sulfuric acid exceeds about 5 ppm, it can form a visible
condensation plume that dissipates very slowly. This plume is most prevalent on No. 6
oil-fired boilers that use oil with 1% or higher sulfur content. Plumes have been
observed from boilers firing 0.7% or even 0.5% sulfur oil, so reducing fuel sulfur
content is not a sure path to invisible plumes. Coal-fired boilers usually don’t emit
visible acid plumes – usually because the fly ash absorbs enough acid to keep the
amount of acid below the visible threshold. One or more of several approaches is used
to reduce visible acid plumes from oil-fired boilers:
• reducing oil sulfur content may reduce the plume to acceptable levels
• keep the air heater and back end temperatures high enough to prevent condensed
acid accumulation during low load operation
• operate at sufficiently low excess air levels that acid formation is reduced
• find a basic fuel additive, such as magnesium oxide, that reduces the amount of acid
Adequate control of visible acid plumes always requires some level of trial and error
even when preceded by a perceptive diagnostic program to evaluate the causes.

7.5.4 Particulate Emissions
Particulate emissions should not be an issue with gas-fired boilers, although measurable
emissions may be created as artifacts of the test procedure. Particulate emissions
should not be significant from distillate oil-fired boilers unless the fuel has enough
sulfur to form sulfuric acid and the test is run at low enough temperatures to capture the
acid. Particulate emissions from No. 6 oil-fired boilers are significant and, as discussed
earlier, they can exceed regulated levels if the oil atomization is poor or if the boiler
generates a lot of sulfate and the emission limit is less than about 0.1 lb/mmBTU.
Pulverized coal generates huge amounts of particulate and a high efficiency (>99%)
dust collector is always required. Solid fuel stokers for wood or other fuels generate
substantial amounts of particulate and some form of dust collector, either internal or
external to the boiler, is always used. All new stoker boilers have a high efficiency dust
collector in addition to an internal cyclone or multiclone collector.
Electrostatic precipitators have been added to a few large No. 6 oil-fired boilers with
mixed results. They do reduce particulate emissions, but at considerable cost. Most
large No. 6 oil-fired boilers can meet a particulate emission standard of about 0.05
lb/mmBTU in the absence of any dust collector. So the necessity and benefit of a
high efficiency dust collector is often questionable. Oil-fired boiler ash levels are low
enough and acid levels are high enough that the collected material becomes damp,
sticky and highly corrosive unless the dust collector and its hoppers are maintained
well above the acid dew point of about 260°F. This requires an adequate initial
design and a concerted effort on the part of the operating staff. Then there is the
problem of what to do with the collected ash. Some facilities inject all the ash back
into the boiler where any carbon burns and the incombustible portion gradually
accumulates in the boiler hopper(s). Some utilities are able to sell this bottom ash for
its vanadium content. Other disposal routes generally must treat the ash as a
hazardous waste because it is acidic and contains a number of water-soluble metallic
compounds.
7.6 COAL FIRED BOILERS

Most of the preceding discussion of gas- and oil-fired boilers applies to pulverized
coal boilers as well. We will limit the discussion here to additional features that are
specific to coal-fired systems.
7.6.1 Characteristics of Pulverized Coal Boilers

Boilers designed for pulverized coal (PC) can also be fired on natural gas or oil,
assuming the appropriate burner and fuel-handling hardware is in place. However,
the converse is not true: boilers designed for oil or gas cannot fire pulverized coal
without major changes and a substantial derating. PC boilers normally can fire only
small amounts of gas or light oil to light off and warm up the boiler before the first

coal mill is started. The difference between PC and other types of boilers originates
in the ash content of the coal, which is typically about 10%. So a power plant that
burns several thousand tons of coal a day has to deal with several hundred tons of
powdered rock passing through the boiler and duct work up to the dust collector.
This has several affects on the design and operation of the boiler:
• Ash/dust accumulation on surfaces in the boiler reduces heat transfer, so a PC

boiler needs more total heat transfer area.
• Different amounts of ash deposited in various boiler sections results in

unbalanced heat transfer rates, and imbalances between steam generation and
superheat and/or reheat, etc. Internal soot blowers are provided to allow operators
to selectively clean sections and try to maintain balanced temperatures.
• If the dust is heated to its melting point, it will tend to agglomerate and will tend
to stick tenaciously to surfaces. Dry ash (not molten) can be removed with soot
blowers. However, slag accumulations in the boiler can block gas passageways
or grow big enough to pose a physical hazard.
• The firebox in a PC boiler is large in order to (1) reduce wall temperatures, (2)
provide more combustion residence time and (3) provide enough heat transfer
area when the surfaces are partially fouled with ash.
• Operators use excess air and specific burner arrangements to control firebox
temperatures and the tendency for ash to form slag accumulations.
So any combustion adjustments or changes on a PC boiler to improve emissions can

have other affects that may be critically important to boiler operation.
Overall, PC boiler emissions are not substantially above those from a comparable
residual oil-fired boiler. Combustion controlled NOx emissions are higher than oil-
fired emissions because of the high nitrogen content of coal. Particulate emissions
after a dust collector are equivalent to or lower than uncontrolled oil-fired particulate
emissions. The particulate that is emitted from a coal boiler tends to be relatively
insoluble silica and metal oxides, while that from an oil-fired boiler is mostly soluble
sulfates with vanadium and other metals. Sulfuric acid emissions and visible plumes
are less prevalent from properly operated PC boilers than from comparable residual
oil-fired boilers.

Combustion based NOx controls on a PC boiler are similar to those on an oil-fired
boiler. One of the critical factors in NOx control on large boilers in general, and PC
boilers in particular, is to achieve a uniform distribution of fuel and air to each burner.
The burner-to-burner fuel and air distribution is never perfectly uniform, but the
extent of the problem on individual boilers was never realized until attempts were

made to reduce NOx emissions. Engineers discovered that multiple low NOx burners
did not reduce NOx emissions anywhere near as much as had been observed on single
burners in a test furnace. Measuring fuel and air distribution to individual burners is
both difficult and time consuming. But research on various boilers showed that PC
blown from the coal grinding mills was not evenly divided between all the burners
and that the coal flow could be highly non-uniform even within the coal pipe to an
individual burner. Once these problems were identified and resolved, low NOx
performance was usually achieved, and some of the problems with slag accumulation
and non-uniform temperatures were resolved simultaneously. The point here is that
the whole system has to be fine tuned to achieve low NOx performance, and some of
the operating problems that might be attributed to new low NOx hardware could be
existing problems that are brought to light by low NOx development work.
7.6.3 Sulfur Oxides and Particulate Matter

Reduced SO2 emissions have frequently been achieved by switching to lower sulfur
fuel. Any fuel change in a PC boiler requires some test firing and demonstration that
the new coal will perform acceptably – mostly to be sure that the ash and slag do not
trigger unmanageable accumulations in the boiler. On individual boilers the question
comes down to this: Which low sulfur coals are acceptable?
Low sulfur coal can also trigger problems in an electrostatic precipitator. Low sulfur
coal from the Powder River Basin has a very basic ash that combines with sulfuric
acid. Eliminating the acid in the exhaust gas reduces corrosion problems, but it also
raises the resistivity of the ash to the point that it is difficult to collect in an
electrostatic precipitator. In some cases, this has been resolved by adding a few ppm
of SO3 to the exhaust to lower the dust resistivity and recover the performance of the
precipitator.
There are many potential issues associated with changing coal to achieve an
emissions goal. Anyone dealing with coal-fired boilers needs to be aware that they
are complex and changes to one parameter can have a peripheral impact on a number
of other parameters.
7.7 WOOD FIRING AND STOKER FURNACES

Up until about 1940 nearly all furnaces and boilers were stoker fired, mostly with
coal and a few with wood. Today almost the only large stoker boilers are operating at
forest products industries, burning waste wood, and at municipal waste-to-energy
facilities. Most of the smaller boilers have been replaced by units that fire either
natural gas or oil. But this is a relatively recent phenomenon. Residential coal
delivery was still in vogue as late as the early 1940s in many major U.S. cities.

7.7.1 Characteristics of Stoker Furnaces
Combustion in a stoker occurs in two regions: (1) combustion of solid fuel in a fuel
bed, and (2) combustion of gases above the bed. Bed combustion can be controlled
by setting the thickness of the bed and by the amount of undergrate air fired through
the bed. Above the bed combustion is controlled by the design of the overfire air
ports and by the amount of overfire air. This arrangement limits the ability to tune
the combustion in the same way that suspension-fired systems (i.e., PC boilers) are
adjusted to control air emissions. A further limitation on older stokers with air-cooled
grates (most of them) is that the undergrate air flow must be relatively high to keep
the grate from overheating. This contributes to a high excess air requirement,
characteristic of all stokers, and it limits the amount of air that can be blown through
the overfire air ports to control emissions of smoke and CO. Overall, a stoker offers
relatively few options for combustion tuning to reduce air pollutant formation.
7.7.2 Particulate Matter Emissions

Most of the fuel in a stoker burns on the grate and the resulting ash is then deposited,
manually or automatically, in a bottom ash hopper. So most of the potential
particulate emissions are kept in the boiler rather than being carried out by the
exhaust gases. Retention can be anywhere from 50% to 90%. Ash retention depends
primarily on two factors: (1) the size distribution of the fuel, and (2) feeder design –
over-grate spreaders generate higher emissions.
Stokers will tend to emit any ash and fuel particles smaller than one to two millimeters
that become airborne in the furnace. Most stokers include a multi-clone collector on
the boiler to capture most of these particles. On older units, this is the only collector
and the resulting emissions can be in the range of 0.3 to 0.7 lb/mmBTU. On newer
units with a good dust collector, the multi-clone catches most of the burning particles
and helps to protect the dust collector against fire.

Nearly all the NOx from a stoker is formed during combustion in the fuel bed where
there is no way to modify fuel mixing or stage combustion. Hence the combustion
modifications used on PC or suspension-fired systems are inapplicable or relatively
ineffective on a stoker. Flue gas recirculation (FGR) into the undergrate air may
provide a modest NOx reduction. Reburning with natural gas is effective and it can
be used simultaneously to improve a stoker’s load response. NOx reductions in the
range of 70% to 80% appear to be feasible. Several municipal waste-to-energy
boilers have been designed and built with SNCR ammonia injection systems that
reduce uncontrolled emissions by about 50%. A number of existing municipal waste-
to-energy boilers have been retrofitted with SNCR urea injection systems that reduce
NOx by somewhat more than 50%.

7.7.4 PIC and Dioxin-furans
Many stokers blow a significant amount of burning carbon particulate out of the
combustion zone. This material can either be reinjected or discarded. On wood-fired
boilers this can amount to several percent of the fuel input. For the most part this
material is pure carbon with little or no organic content.
Combustion in the fuel bed on a stoker grate releases CO and organic gases that burn
above the bed in the over-fired region. Most of the time the over-bed temperatures
are high enough to destroy or decompose the organic gases, leaving CO and perhaps
methane or other low molecular weight species. If the overfired air system is well
designed and operated, it will burn up these combustible gases leaving only low
levels of CO in the exhaust. Thus emissions of PIC gases depend almost entirely on
the operation of the overfired air system. Tuning this system can usually get rid of
smoke, but CO levels of a few hundred ppm are common on older stokers.
Older municipal waste-fired stokers were notorious for high emissions of dioxins,
furans and other polycyclic organic species. This appears to be a consequence of
several factors:
• grate and waste feeder designs that were inappropriate for municipal waste
• furnace temperatures that locally dipped to 1600°F or less
• inadequate overfired air design
Newer waste-to-energy facilities resolved these problems and the emissions of

hazardous organic species from these units is very low.

Review Exercises
1. Which statement is false?
a. When combustion efficiency is nearly 100%, PIC are near zero.
b. Thermodynamics do not permit a thermal efficiency of 100%.
c. Most electric power plants have thermal efficiencies less than 50%
d. Very efficient combustion is the key to high overall (thermal) efficiency.
2. The energy loss associate with CO emissions of 320 ppm in a stack with O2 = 6.2% is
________.
a. 0.05%
b. 0.12%
c. 0.24%
d. 1.15%
3. How much energy is wasted on a pulverized coal boiler when the fly ash contains an
average of 7.1% carbon? The coal has 13.2% ash and HHV = 11,800 BTU/lb.
a. 0.11%
b. 1.00%
c. 1.12%
d. 2.24%
4. Most of the energy lost up a combustion source stack is in the form of ________.
a. hot gas and water vapor
b. smoke and unburned carbon
c. carbon monoxide
d. kinetic energy of the high velocity gas
5. Heat losses from uncondensed water vapor could be ignored in calculating thermal
efficiency if the ________.
a. moisture content of the fuel is very high.
b. stack temperature is above the water dew point.
c. fuel is dry and the lower heating value of the fuel is used in calculations.
d. stack temperature is below 212°F.
6. What is the approximate thermal efficiency of a gas-fired boiler when stack measurements
show T = 290°F, O2 = 2.9%, water vapor = 14% by volume? Ambient temperature is 48°F.
a. 78.1%
b. 85.3%
c. 88.6%
d. 90.5%

7. What is the firing rate or fuel energy use of a 480-mw power plant with a heat rate of 9750
BTU/kw-hr?
a. 4680 mmBTU/hr
b. 4860 mmBTU/hr
c. 5108 mmBTU/hr
d. 5522 mmBTU/hr
8. What are the daily CO2 emissions from a source that fires coal at a rate of 375 mmBTU/hr?
a. 524,000 lb/day
b. 1,450,000 lb/day
c. 1,665,000 lb/day
d. 69,400 lb/day
9. The dilution extraction CEM came into use as a result of ________.

a. the need to measure pollutant concentration without drying the sample gas
b. the inconsistency between measuring total gas flow and gas concentrations measured
after water vapor is removed from the gas sample
c. the need to measure and report SO2 and NOx emissions in units of lb or tons/day
d. All of the above
10. CO emissions are measured at 225 ppm in an exhaust gas with 5.3% oxygen. What would
the concentration be if air dilution was reduced so that O2 = 3%?
a. 258 ppm
b. 186 ppm
c. 248 ppm
d. Insufficient data to determine the answer.

Review Answers
1. d. Very efficient combustion is the key to high overall (thermal) efficiency
2. b. 0.12% energy loss
3. c. 1.12% energy loss
4. a. hot gas and water vapor
5. c. fuel is dry and the lower heating value of the fuel is used in calculations
(see discussion in Section 3.3.2.1)
6. b. 85.3%
7. a. 4680 mmBTU/hr
8. c. 1,665,000 lb/day
10. a. 258 ppm

References
Kitto, J.B. Steam; Its Generation and Use, (Babcock and Wilcox Company,
Barberton, Ohio, 1992). A huge book with in depth coverage of large boiler furnaces,
particularly coal-fired boilers. It also covers fluidized beds and refuse firing. The
book has grown over the years and the material covered has changed.
Pritchard, R., Guy, J.J., and Conner, N.E. Handbook of Industrial Gas Utilization,
(Van Nostrand Reinhold Co., 1977). The authors are/were at the University of
Salford (GB) and the detail/clarity are typically British. The book covers all aspects
of combustion hardware and furnace design for any system that fires gas or gas in
combination with other fuels.

(North American Manufacturing Company, Cleveland, OH, 1993). A good basic
reference for small to medium size gas and oil-fired combustion systems.

Volume II, (North American Manufacturing Company, Cleveland, OH, 1995). The
book provides good detail on burner types, control systems, emissions, heat recovery
and other special topics.
U.S. Code of Federal Regulations, Title 40, Part 60, Appendix A (Test Methods).
These are the “reference” test methods that are the accepted procedures for virtually
all emissions testing.
U.S. Environmental Protection Agency. John Richards. Control of Gaseous

Emissions, APTI Course 415, Student Manual, NC State University, 1995. This
manual has credible sections on incineration and NOx control as well as covering
areas not related to combustion sources.
American Society of Mechanical Engineers. Power Test Codes, PTC 4.1, Steam
Generating Units This gives methods, accepted by industry, for measuring boiler
efficiency.

Combustion Source Evaluation: Student Manual

Uploaded by

Copyright:

Available Formats

Combustion Source Evaluation: Student Manual

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Combustion Source Evaluation: Student Manual

Uploaded by

Copyright:

Available Formats

United States Air Pollution Training Institute (APTI) June 2003

Environmental Protection Environmental Research Center, MD 17

This publication is available, free of charge, to schools or governmental

This consent does not extend to copying or transmission by any means

Printed on recycled paper in the United States of America.

• Combustion sources burning • Combustion principles

• Combustion sources burning • Design and operational

• Combustion engines (diesel • Selected fundamental

• Pollution control devices • Fuels

How to Use This Workbook

Each chapter provides a lesson goal, instructional objectives, subject

LIST OF FIGURES .....................................................................................................................xiv

1.0 Overview of Combustion Sources 1-1

1.1 Significance of Combustion Sources................................................................ 1-2

Review Exercises 1-14

2.0 General Types and Characteristics of Combustion Sources 2-1

2.1 Combustion Systems .......................................................................................... 2-3

2.2.2 Boilers...... .......................................................................................... 2-10

Review Exercises 2-45

3.1 Introduction.......................................................................................................... 3-3

3.2 Fuel Characteristics and Usage Trends .............................................................. 3-3

3.2.1 General Types ...................................................................................... 3-4

Review Exercises 3-45

4.0 Fundamentals of Combustion 4-1

4.1 Introduction.......................................................................................................... 4-3

Review Exercises 4-40

5.0 Air Pollution Formation 5-1

Review Exercises 5-42

6.0 Air Pollution Control Systems 6-1

Review Exercises 6-50

7.0 Combustion Source Emissions 7-1

Review Exercises 7-29

Figure 1-1 Global Carbon Cycle 1-3

Table 1-1 Examples of sources and fuels 1-2

BACT Best Available Control Technology

BIF Boilers and Industrial Furnaces - refers to federal waste fuel

CAAA Clean Air Act Amendments

CEM Continuous Emission Monitors

DAS Data Acquisition System

DP Differential Pressure or Delta P

DSCFM Dry standard cubic feet per minute

engine Either a combustion turbine or a reciprocating engine

EPA Environmental Protection Agency

ESP Electrostatic Precipitator

Fw-factor Similar to Fd except water vapor is included. Value is about

FGC Flue gas conditioning – adding something such as SO3 to the

FGD Flue gas desulfurization

FGR Flue gas recirculation - technology used to reduce the

FID Flame ionization detector

HAP Hazardous Air Pollutants

HHV Higher Heating Value (of a fuel) - a number obtained by a lab

incineration Combustion for purposes of destruction/disposal; usually

LEL Lower Explosive Level – the volume concentration of a gas in

LHV Lower Heating Value (of a fuel) – a number in BTU/lb obtained

MACT Maximum Achievable Control Technology

NAAQS National Ambient Air Quality Standards

NDIR Non-dispensive infrared

NESHAP National Emission Standards for Hazardous Air Pollutants

NSCR Non Selective Catalytic Reduction – NOx destruction by