&EPA
United States
Environmental Protection
Agency
Office of
Research and Development
Washington, DC 20460
EPA/600/8-91/049aA
August 1991
External Review Draft
Air Quality
Criteria for
Oxides of
Nitrogen
Review
Draft
Do not
Cite or Quote
Volume I of III
NOTICE
This document is a preliminary draft It has not been formally
released by EPA and should not at this stage be construed to
represent Agency policy It is being circulated for comment on its
technical accuracy and policy implications
-------
(Do not Cite or Quote) EPA 600/8-91/049aA
August 1991
External Review Draft
Air Quality Criteria for
Oxides of Nitrogen
Volume I of III
NOTICE
This document is a preliminary draft. It has not been formally released by
EPA and should not at this stage be construed to represent Agency policy.
It is being circulated for comment on its technical accuracy and policy
implications.
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Printed on Recycled Paper
-------
Air Quality Criteria for Oxides of Nitrogen
TABLE OF CONTENTS (cont'd)
Volume HI Page
13. STUDIES OF THE EFFECTS OF NITROGEN COMPOUNDS
ON ANIMALS .............. . 13-1
14. EPIDEMIOLOGY STUDIES OF OXIDES OF NITROGEN . 14-1
15. CONTROLLED HUMAN EXPOSURE STUDIES OF OXIDES
OF NITROGEN 15-1
16. HEALTH EFFECTS ASSOCIATED WITH EXPOSURE TO
NITROGEN DIOXIDE ..-....,.'. . 16-1
APPENDIX A: GLOSSARY OF TERMS AND SYMBOLS A-l
August 1991 I-iv DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS
TABLES I-xiii
FIGURES I-xvii
AUTHORS I-xxi
CONTRIBUTORS AND REVIEWERS I-xxiii
1. SUMMARY OF EFFECTS OF OXIDES OF NITROGEN AND
RELATED COMPOUNDS ON HUMAN HEALTH AND
WELFARE . . . 1-1
1.1 INTRODUCTION . 1-1
1.1.1 Critical Issues . . 1-2
1.1.2 Organization of the Document 1-3
1.2 CHEMISTRY, SOURCES, TRANSPORT AND
TRANSFORMATION, SAMPLING, AMBIENT AND INDOOR
LEVELS, AND EXPOSURE OF OXIDES OF NITROGEN .... 1-4
1.2.1 Chemical and Physical Properties of Oxides of
Nitrogen 1-4
1.2.2 Sources of Nitrogen Oxides Influencing Ambient and
Indoor Air Quality 1-5
1.2.3 Transport and Transformation 1-5
1.2.3.1 Ozone Production 1-5
1.2.3.2 Production of Odd Nitrogen Species 1-6
1.2.3.3 Transport 1-7
1.2.3.4 Oxides of Nitrogen and the Greenhouse
Effect 1-9
1.2.3.5 Deposition of Nitrogen Oxides 1-10
1.2.4 Sampling and Analysis of Oxides of Nitrogen 1-10
1.2.5 Ambient and Indoor Concentration of Oxides of
Nitrogen 1-12
1.2.5.1 Ambient Concentration 1-12
1.2.5.2 Indoor Concentration 1-13
1.2.6 Assessing Total Human Exposure to Nitrogen Dioxide . . 1-15
1.3 EFFECTS OF NITROGEN OXIDES ON VEGETATION,
ECOSYSTEMS, VISIBILITY, AND MATERIALS 1-16
1.3.1 Effects of Nitrogen Oxides on Vegetation 1-16
1.3.1.1 Introduction . . . . 1-16
1.3.1.2 Nitrogen Dioxide 1-17
1.3.1.3 Nitric Oxide 1-23
1.3.1.4 Pollutant Combinations 1-23
1.3.2 The Effects of Nitrogen Oxides on Natural Ecosystems
and Their Components 1-24
1.3.2.1 Ecosystems: Structure, Function, Response . . 1-24
1.3.2.2 Nitrogen Deposition 1-27
1.3.2.3 Effect of Deposited Nitrogen on Forest
Vegetation and Soils . 1-28
August 1991 I-v DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS (eont'd)
1.3.2.4 Effects of Nitrogen on Sensitive Terrestrial
Vegetation 1-31
1.3.2.5 Nitrogen Saturation, Critical Loads, and
Current Deposition 1-32
1.3.2.6 Effects of Nitrogen on Wetlands and Bogs . . . 1-36
1.3.2.7 Effects of Nitrogen on Aquatic Systems . 1-38
1.3.3 Effects of Nitrogen Oxides on Visibility 1-47
1.3.4 Effects of Nitrogen Oxides on Materials 1-50
1.4 HEALTH EFFECTS OF OXIDES OF NITROGEN 1-52
1.4.1 Animal Toxicology 1-52
1.4.1.1 Animal-to-Human Dosimetric Estimates .... 1-53
1.4.1.2 Biochemical and Cellular Mechanisms 1-54
1.4.1.3 Effects on Host Defenses 1-55
1.4.1.4 Effects of Chronic Exposure on the
Development of Chronic Lung Disease 1-57
1.4.1,5 Potential Carcinogenic or Co-Carcinogenic
Effects 1-59
1.4.1.6 Extrapulmonary Effects 1-59
1.4.1.7 Susceptibility of Subpopulations 1-60
1.4.1.8 Influence of Exposure Patterns 1-61
1.4.1.9 Interactions with Other Pollutants 1-61
1.4.2 Epidemiology Studies of Oxides of Nitrogen 1-62
1.4.3 Controlled Human Exposure Studies of Oxides of
Nitrogen , . 1-67
1.4.4 Discussion of Health Effects Associated With Exposure
to Nitrogen Oxide 1-73
1.4.4.1 Airway Reactivity in Asthmatics and Exposure
to NO2 1-73
1.4.4.2 Respiratory Morbidity in Children Associated
with Exposure to NO2 1-74
1.4.4.3 Biological Basis Relating NO2 Exposure to
Respiratory Morbidity: Effects of NQ^ on
the Respiratory Host Defense System 1-75
1.4.4.4. Emphysema and Exposure to NO2 1-78
1.4.4.5 Subpopulations Potentially Susceptible to NO2
Exposure 1-80
TfrTyTy^pT^'I'iTI*"* 'EC* " 1 _S*?
2. INTRODUCTION 2-1
3. GENERAL CHEMICAL AND PHYSICAL OF NOX
AND NOX-DERIVED POLLUTANTS 3-1
3.1 INTRODUCTION AND OVERVIEW 3-1
August 1991 I-vi DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS (cont'd)
3.2 NITROGEN OXIDES . . . . 3-4
3.2.1 Nitric Oxide (NO) 3-6
3.2.2 Nitrogen Dioxide (NO2) 3-9
: 3.2.3 Nitrous Oxide (N2O) . . . 3-10
3.2.4 Nitrogen Trioxide (NO3) . . . . 3-11
3.2.5 Dinitrogen Trioxide (N2O3) (Also Known as Nitrogen
Sesquioxide) ....................... . . ... 3-12
: 3.2.6 Dinitrogen Tetroxide (N2O4) (Also Known as Nitrogen
Tetroxide •'••' 3-12
3.3 NITRATES, NITRITES, AND NITROGEN ACIDS .......... 3-13
3.4 AMMONIA (NH3) 3-14
3.5 N-NITROSO COMPOUNDS ..... 3-15
3.6 SUMMARY . . ........ 3-16
3.6.1 Nitrogen Oxides 3-17
3.6.2 Nitrates, Nitrites, and Nitrogen Acids 3-18
3.6.3 N-Nitroso Compounds 3-18
REFERENCES 3-19
4. SOURCES OF NITROGEN OXIDES INFLUENCING AMBIENT
AND INDOOR AIR QUALITY 4-1
4.1 INTRODUCTION . . .......... 4-1
4.2 AMBIENT SOURCES OF NITROGEN OXIDES . . . ....... 4-1
4.2.1 Combustion Generated NOX. Emissions : ...... . . ... 4-2
4.2.1.1 Generation of NOX from Biomass Burning . . . 4-2
4.2.1.2 Generation of NOX From Lightning . ...... 4-4
4.2.1.3 Generation of NOX From Soils . . .;.". . . ••;'. . 4-4
4.2.1.4 Generation of NOX From the Ocean . . . •... . 4-5
4.2.2 Removal of NOX From the Atmosphere . : 4-6
4.2.3 Global Budgets for NOX . ,.,..• :. . . . . 4-9
4.2.4 Major Sources of NOX Emissions in the United States ... 4-9
4.2.5 Conclusions . .". . ...... i . . . 4-11
4.3 SOURCES OF NITROGEN OXIDES INFLUENCING
INDOOR AIR QUALITY 4-12
4.3.1 Introduction . 4-12
4.3.2 Formation of Nitrogen Oxides in Combustion in
Gas-fueled Household Appliances 4-14
4.3.3 Gas Stoves Used for Cooking ..•'.- 4-20
4.3.3.1 Study of Himmel and Dewerth (1974) 4-20
4.3.3.2 Study of Cote et al. (1974) 4-24
4.3.3.3 Study by the Massachusetts Institute of
Technology (1976) 4-24
4.3.3.4 Study of Traynor et al. (1982b) ......... 4-24
4.3.3.5 Studies of Cole et al. (1983) and
Moschandreas et al. (1985) . • '..-'.. 4-24
August 1991 I-vii DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS (cont'd)
Page
4.3.3.6 Studies of Fortmann et al. (1984) and
Borrazzo et al. (1987) 4-28
4.3.3.7 Study of Cole and ZawacM (1985) ........ 4-28
4.3.3.8 Study of Tikalsky et al. (1987) 4-28
4.3.3.9 Summary of Emissions From Gas Stoves .... 4-29
4.3.4 Unvented Space Heaters Fueled with Natural Gas
and Propane 4-33
4.3.4.1 Study of Thrasher and Dewerth (1979) 4-33
4.3.4.2 Study of Traynor et al. '(1983a) . . . 4-34
4.3.4.3 Study of Traynor et al. (1984) 4-34
4.3.4.4 Studies of Billick et al. (1984),
Moschandreas et al. (1985), and Zawacki
etal. (1984) . . 4-37
4.3.4.5 Study of Zawacki et al. (1986) .......... 4-38
4.3.4.6 GATC Task on Environmental Control 4-38
4.3.4.7 Summary of Emissions from Unvented
Gas Space Heaters 4-38
4.3.5 Kerosene Heaters . 4-41
4.3.5.1 Study of Yamanaka et al. (1979) 4-41
4.3.5.2 Study of Leaderer (1982) 4-41
4.3.5.3 Study of Traynor et al. (19835) 4-42
4.3.5.4 Study of Apte and Traynor (1986) 4-43
4.3.5.5 Summary of Emissions from Kerosene
Heaters 4-43
4.3.6 Wood Stoves 4-44
4.3.7 Tobacco Products 4-45
4.3.8 Comparison of Emissions from Sources Influencing
Indoor Air Quality 4-45
4.4 SUMMARY OF EMISSIONS OF NOX FROM AMBIENT
AND INDOOR SOURCES 4-48
REFERENCES 4-49
5. TRANSPORT AND TRANSFORMATION OF NITROGEN
OXIDES 5-1
5.1 BACKGROUND 5-1
5.2 THE ROLE OF NOX IN OZONE PRODUCTION 5-3
5.2.1 NOx-Rich Chemistry . 5-8
5.2.2 Ozone Production in NOx-Poor Environments 5-10
5.3 ODD NITROGEN SPECIES 5-19
5.3.1 Nitric Acid . 5-19
5.3.2 Nitrous Acid 5-21
5.3.3 Peroxynitric Acid 5-21
5.3.4 Peroxyacylnitrates 5-22
August 1991 I-viii DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS (cont'd)
Page
5.3.5 Nitrate Radical 5-23
5.3.6 Dinitrogen Pentoxide 5-25
5.3.7 Summary of NOy Species 5-26
5.3.8 Amines, Nitrosamines, and Nitramines 5-27
5.4 TRANSPORT 5-31
5.4.1 Transport of Reactive Nitrogen Species in Urban
Plumes .". . . . 5-33
5.4.2 Transport and Chemistry in NOx-Rich Plumes ....... 5-35
5.4.3 Regional Transport . 5-37
5.5 OXIDES OF NITROGEN AND THE GREENHOUSE
EFFECT 5-41
5.5.1 NO_-Related Ozone Greenhouse Effects 5-41
A
5.5.2 Nitrous Oxide Greenhouse Contributions 5-43
5.6 STRATOSPHERIC OZONE DEPLETION BY OXIDES
OF NITROGEN . . . 5-44
5.7 DEPOSITION OF NITROGEN OXIDES . 5-49
5.7.1 Dry Deposition of Nitrogen Oxides 5-49
5.7.2 Methods For Determining Vd 5-50
5.7.2.1 Eddy Correlation 5-50
5.7.2.2 Vertical Gradient Methods . , 5-51
5.7.2.3 Chamber Methods 5-51
5.7.3 NOX Deposition 5-52
5.7.4 HNO3 Deposition 5-52
5.7.5 PAN Deposition 5-52
5.7.6 Wet Deposition of Nitrogen Oxides 5-53
5.8 SUMMARY AND CONCLUSIONS ..'..... 5-53
5.8.1 Ozone Production . . 5-53
5.8.2 Production of Odd Nitrogen Species ............. 5-54
5.8.3 Transport 5-55
5.8.4 Oxides of Nitrogen and the Greenhouse Effect 5-57
5.8.5 Deposition of Nitrogen Oxides 5-58
REFERENCES 5-59
6. SAMPLING AND ANALYSIS FOR OXIDES OF NITROGEN
AND RELATED SPECIES '..'..' 6-1
6.1 INTRODUCTION . 6-1
6.2 NITRIC OXIDE (NO) 6-2
6.2.1 Chemiluminescence 6-2
6.2.2 Laser-Induced Fluorescence , 6-5
6.2.3 Absorption Spectroscopy 6-6
6.2.4 Passive Samplers 6-8
6.2.5 Calibration^. 6-9
6.2.6 Intercomparisons 6-10
August 1991 J-ix DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS (cont'd)
Page
6.2.7 Sampling Considerations for NO and Other
Nitrogen-Containing Species 6-12
6.3 NITROGEN DIOXIDE (NO^ 6-12
6.3.1 Chemiluminescence, NO + O3 (CLM) 6-13
6.3.2 Chemiluminescence (CLM), Luminol. 6-16
6.3.3 Photofragmentation/Two Photon-Laser-Induced
Fluorescence (PF/TP-LBF) 6-17
6.3.4 Absorption Spectroscopy 6-18
6.3.5 Wet Chemical Methods :..... 6-21
6.3.5.1 Griess-Saltzman Method 6-21
6.3.5.2 Continuous Saltzman Method 6-21
6.3.5.3 Alkaline Guiacol Method 6-22
6.3.5.4 Jacobs-Hochheiser Method '. 6-22
6.3.5.5 Sodium Arsenite Method (Manual and
Continuous) . . ; ; 6-23
6.3.5.6 Triethanolamine-Guaiacol-Sulfite (TGS)
Method 6-23
6.3.5.7 Triethanolamine (TEA) Method 6-24
6.3.6 Other Active Methods 6-24
6.3.7 Passive Samplers 6-26
6.3.8 Calibration 6-30
6.3.9 Intercomparisons 6-31
6.3.10 Designated Methods 6-34
6.4 OXIDES OF NITROGEN (NOX) . 6-38
6.5 TOTAL REACTIVE ODD NITROGEN (NO ) 6-40
6.6 PEROXYACETYL NITRATE (PAN) 6-41
6.6.1 Gas Chromatography-Electron Capture Detection
(GC-ECD) . . . 6-41
6.6.2 Alkaline Hydrolysis . 6-42
6.6.3 Gas Chromatography (GC)-Alternate Detectors ....... 6-43
6.6.4 Peroxyacetyl Nitrate Stability 6-43
6.6.5 Calibration 6-44
6.6.6 Other Organic Nitrates 6-46
6.7 NITRIC ACID (HNO3) . 6-47
6.7.1 Filtration 6-47
6.7.2 Denuders 6-48
6.7.3 Chemiluminescence (CLM) . . . . 6-51
6.7.4 Absorption Spectroscopy 6-51
6.7.5 Calibration 6-52
6.7.6 Intercomparisons 6-53
6.8 NITROUS ACID (HNO^ 6-55
6.8.1 Denuders .;..:. 6-55
6.8.2 Chemiluminescence (CLM) 6-56
August 1991 I-x DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS
6.8.3 Photofiragmentation/Laser-Induced Fluorescence
(PF/LIF) ....... 6-56
6.8.4 Absorption Spectroscopy 6-56
6.8.5 Calibration . . . * ..... 6-57
6.9 DIMTROGEN PENTOXTOE (N2O5) AND NITRATE '
RADICALS (NOg) 6-57
6.10 PARTICULATE NITRATE (PN) *....... 6-58
6.10.1 Filtration . ..... 6-59
6.10.2 DenudersMhration 6-62
6.10.3 Analysis . 4'... . . 6-63
6.11 NITROUS OXIDE (N2O) 6-67
6.12 SUMMARY ........... 6-68
REFERENCES . . . . . ... 6-70
7. AMBIENT AND INDOOR CONCENTRATIONS OF NITROGEN
DIOXIDE . . . .... .-.;•.•-. . - . 7-1
7.1 INTRODUCTION . 7-1
7.2 AMBIENT AIR CONCENTRATIONS OF NITROGEN
DIOXIDE . . ........ 7-1
7.2.1 The National Air Monitoring Network ...... ..,. .... 7-2
7.2.2 Peak Annual NO2 Averages in Metropolitan Statistical
Areas, 1988-89 ...... . . 7-3
7.2.3 Trends in Ambient NO2 Concentrations . . . ...... . . 7-3
7.2.4 Patterns in Ambient NO2 Concentrations ......,,..: 7-3
7.2.4.1 Seasonal Pattens .... ...-. .... . . .... ... 7-5
7.2.4.2 Diurnal Patterns ........... : ;..:......".•'• 7-7
7.2.4.3 Distributional Patterns ............... 7-12
7.3 INDOOR AIR CONCENTRATIONS OF NITROGEN
OXIDES .-.- , ...... . , 7-14
7.3.1 Background . . . •. . ... . . . 7-14
7.3.2 Residences Without Indoor Sources ........ .••.-.•. . . 7-17
7.3.3 Residences With Gas Appliances ^ ....... 7-23
7.3.3.1 Average Indoor Concentrations and .
Estimated Contributions ........ ... . . . 7-25
7.3.3.2 Spatial and Temporal Distributions ....... 7-31
7.3.3.3 Peak Indoor Concentrations , . .. 7-32
7.3.4 Unvented Space Heaters . ....... i ......... 7-36
7.3.4.1 Unvented Kerosene Space Heaters, ........ 7-37
7.3.5 Other Sources . . .... 7-43
7.3.6 Modeling of Indoor Concentrations ..... k *....... 7-43
7.3.6.1 Physical/Chemical Models .......... ;.. . 7-44
7.3.6.2 Statistical/Empirical Models ....... V. ;.•:;,.. . . 7-46
August 1991 I-xi DRAFT-DO NOT QUOTE OR CITE
-------
CONTENTS (cont'd)
Page
7.3.7 Predictions of Indoor and Outdoor Annual Averages . . . 7-48
7.3.8 Reactive Decay Rate of NO2 Indoors 7-51
7.4 NITRIC AND NITROUS ACIDS CONCENTRATIONS 7-54
7.5 SUMMARY 7-56
7.5.1 Ambient Nitrogen Dioxide Levels . 7-56
7.5.2 Indoor Nitrogen Dioxide Levels 7-57
REFERENCES , . 7-60
8. ASSESSING TOTAL HUMAN EXPOSURE TO NITROGEN
DIOXIDE . 8-1
8.1 INTRODUCTION 8-1
8.2 DIRECT METHODS 8-3
8.2.1 Biomarkers 8-3
8.2.2 Personal Monitoring 8-4
8.3 INDIRECT METHODS . . . 8-15
8.3.1 Personal Exposure Models 8-17
8.4 SUMMARY . 8-21
REFERENCES 8-22
August 1991 I-xii DRAFT-DO NOT QUOTE OR CITE
-------
TABLES
Number Page
1-1 Measurements of Various Forms of Annual Nitrogen Deposition to
North American and European Ecosystems . 1-29
1-2 Three Nitrogen Budgets for the Chesapeake Bay . 1-46
3-1 Theoretical Concentrations of Nitrogen Oxides and Nitrogen ...'.'
Acids which Would be Present at Equilibrium with Molecular
Nitrogen, Molecular Oxygen, and Water in Air at 25 °C, 1 atm,
50% Relative Humidity '3-3
3-2 Some Physical and Thermodynamic Properties of the Nitrogen.
Oxides 3-5
3-3 Theoretical Equilibrium Concentrations of Nitric Oxide and '
Nitrogen Dioxide in Air (50% Relative Humidity) at Various , .
Temperatures , 3-8
*',!"' • -
3-4 Theoretical Concentrations of Dinitrogen Trioxide and Dinitrogen
Tetroxide in Equilibrium with Various Levels of Gaseous Nitric
Oxide and Nitrogen Dioxide in Air at 25 °C 3-13
4-1 Global NOX Emissions from the Burning of Fossil Fuels and
Biomass 4-3
4-2 Global and North America Natural Emissions of NOX from
Lightning, Soils and Oceans 4-5
4-3 Estimates of Nitrous Oxide (N2O) and Ammonia (NH4+)
Emissions to the Troposphere (106 Metric Tons n/yr) 4-6
4-4 Global Budget of Nitrogen Oxides in the Troposphere 4-8
4-5 Estimates of Anthropogenic NOX Emissions in the United States
(1985) Expressed in Millions of Metric Tons 4-10
4-6 Estimates of NOX Emissions from Anthropogenic and Natural
Sources in the United States and Canada (Millions of Metric
Tons/Year of NO2 Equivalent Emissions) 4-13
4-7 Emission Factors for NO and NO2 from Burners on Gas Stoves,
after Himmel and Dewerth (1974) . 4-22
4-8 Emission Factors for NO and NO2 from Pilot Lights on Gas
Stoves, after Himmel and Dewerth (1974) . 4-23
August 1991 I-xiii DRAFT-DO NOT QUOTE OR CITE
-------
TABLES (cont'd)
Number Page
4-9 Emission Factors for NO, NO^, and NOX for Top Burners on Gas
Stoves Measured with a Sampling Hood and with a Chamber,
after Cole et al. (1983) and Mosdhandreas et al. (1985) ....... 4-26
4-10 Emission Factors for NO, NO2, and NOX for Ovens, after Cole
et al. (1983) and Moschandreas et al. (1985) ............. 4-27
4-11 Emission Factors for NO and NO2 from Pilot Lights on Gas
Stoves, after Moschandreas et al. (1985) ................ 4-27
4-12 Emission Factors for NO and NQ-j from Range-Top Burners of
Improved Design, after Cole and ZawacM (1985) .......... 4-29
4-13 Emission Factors for NO2 from Ten Gas Stoves in Use in
Residence, Measured Independently by Research Groups
{TMsky et al., 1987) ........... ................ 4-30
4-14 Average Emission Factors for NO, NO2, and NOX from Burners
on Gas Stoves Based on Data Reported in the Literature ...... 4-30
4-15A Data of Thrasher and Dewerth (1979) for Conveetive Heaters
with Drilled Ports Using Natural Gas .................. 4-35
4-15B Data of Traynor et al. (1983a) for Convective Heaters with
Radiant Tfles Using Natural Gas ....... ........... . . . 4-35
4-15C Data of Traynor etal. (1984) for Convective Heaters with
Radiant Tiles, and for Infrared Heaters, Fueled with Natural
Gas and Propane ......... ....... . ....... ...... 4-36
4-1SD Date of Billick et al. (1984), Moschandreas et al. (1985), and
ZawacM et al. (1984) for Heaters wilh Bunsen, Catalytic, and
Ceramic Tile Burners Using Natural Gas ............... 4-36
4-16 Data of ZawacM et al. (1986) for Convective and Infrared
Heaters of Various Designs, Using Natural Gas and Propane ... 4-39
4-17 Data from Cole and ZawacM (1985) Obtained from Ihe GATC
Task on Environmental Control Survey. Both Heaters Have
Flame Retention Screen-Type Steel Burners and Use Natural
Gas .......... ...... ....... ........ . ......
4-18 Average Emission Factors for NO and NO2 from Kerosene
Heaters, after Leaderer (1982) and Traynor et al. (1983b) ..... 4-42
August 1991 I-xiv DRAFT-DO NOT QUOTE OR CITE
-------
TABLES (cont'd)
Number Page
4-19 Average Emission Factors for NO, NO2, and NOX from Various
Sources Based on Data Reported in the Literature 4-46
5-1 Major Reactions in the NO3-N2O5 System at Night 5-36
5-2 Average Afternoon Background Pollutant Concentrations
Measured at Kenosha, WI . 5-39
6-1 Performance Specifications for Nitrogen Dioxide Automated
Methods 6-35
6-2 Comparability Test Specifications for Nitrogen Dioxide ......... 6-36
6-3 Reference and Equivalent Methods for Nitrogen Dioxide
Designated by the EPA . . . . ......... 6-36
6-4 National Precision and Accuracy Probability Limit Values
Expressed as Percent for Continuous and Manual Methods for
NO2 6-38
7-1 Maximum Annual Average NO2 Concentrations Reported in
Metropolitan Statistical Areas - 1988,1989 ............... 7-4
7-2 Hourly Incidence of NO2 Concentrations Greater Than 0.2 ppm
for Stations with More Than One Occurrence, 1988 7-10
7-3 Average Outdoor Concentrations of NO2 (/ig/m3) and Average
Indoor/Outdoor Ratios in Homes without Gas Appliances •
or Unvented Space Heaters from Field Studies of Private
Residences . 7-19
7-4 Indoor and Outdoor Concentrations of NO2 0*g/m3) in Homes
with Gas Appliances, and the Calculated Average Contribution
of those Appliances to Indoor Residential NO2 Levels 7-26
7-5 Summary Statistics for Gas Range NO2 Maxima 0*g/m3) Over
Several Averaging Times 7-35
7-6 Two-Week Average NO2 Levels by Location for Homes in Six
Principle Source Categories -.....- , . , 7-38
7-7 One-Week Average NO2 Levels in Homes in North Central Texas
by Source Category, with and without Unvented Gas Space
Heater (UVGSH) 7-42
August 1991 I-xv DRAFT-DO NOT QUOTE OR CITE
-------
TABLES (cont'd)
Number Page
7-8 Empirical Statistical Models (Regression) for Residential NO2
Concentrations Reported from Field Studies of Indoor Levels. . . . 7-47
7-9 Average Annual NO2 Means (ppm) and Standard Errors of
Estimates for 40 Sites in the U.S. by Annual Average 7-49
r*
7-10 Average Annual NO2 Means (ng/m ) and Standard Errors of
Estimates for 100 Homes Based on Data of Lambert (1991) .... 7-50
7-11 Summary Statistics for Indoor and Outdoor Concentrations
(in ppb) of Gas Species Measured During Summer and Winter
Sampling Periods in Boston 7-55
8-1 Electric-Range Home Least Squares Regression Coefficients and
T-Statistics (in Parentheses) 8-20
8-2 Gas-Range Home Least Squares Regression Coefficients and
T-Statisdcs (in Parentheses) 8-20
August 1991 I-xvi DRAFT-DO NOT QUOTE OR CITE
-------
FIGURES
Number Page
1-1 The relationship between the onset of either foliar lesions or
metabolic and growth effects and NO2 exposure 1-18
1-2 Effects of NO2 on plant productivity 1-22
1-3 Schematic representation of the nitrogen cycle, emphasizing
human activities that affect fluxes of nitrogen 1-26
1-4 Location of acid-sensitive lakes and streams in the northeastern
United States where the importance of NO3" to seasonal
water chemistry can be determined 1-41
1-5 Location of acid-sensitive lakes and streams in the southeastern
United States where the importance of NO3" to seasonal water
chemistry can be determined 1-42
1-6 Location of acid-sensitive lakes and streams in the western
United States where the importance of NO3" to seasonal water
chemistry can be determined 1-43
1-7 U.S. Environmental Protection Agency meta-analysis of
epidemiologic studies of NO2 exposure effects on respiratory
disease in children <12 years old 1-66
3-1 Calculated steady states of the free troposphere as a function
of NOX concentration 3-4
4-1 Historical estimates of NOX emissions from man-made and
natural sources in the United States 4-11
4-2 Density of NOX emission from man-made sources in 1980 4-12
4-3 Laminar blue-flame 4-15
4-4 Radial profiles of NO and temperature at 7.19 cm above base . . . 4-17
4-5 Radial profile of NO2 at 7.19 cm above burner base 4-18
4-6 Emission factors for NO as a function of gas flow rate 4-32
4-7 Emission factors for NO2 as a function of gas flow rate 4-33
5-1 Summary of the gas phase chemistry of NOX in the clean
troposphere 5-2
August 1991 I-xvii DRAFT-DO NOT QUOTE OR CITE
-------
FIGURES (cont'd)
Number
5-2 Major chemical reactions affecting oxygen species in the
troposphere 5-3
5-3 Major chemical reactions affecting hydrogen species (OH, H,
H2O, H2O2) in the troposphere 5-4
5-4 Schematic diagram of the combined reactions of nitrogen,
oxygen, and hydrogen 5-5
5-5 Calculated steady state concentrations in the free
troposphere as a function of NOX 5-6
5-6 Hydrocarbon oxidation in the atmosphere 5-9
5-7 Idealized dependence of O3 production on HC/NOX 5-11
5-8 (a) Summertime (June 1 to August 31) and (b) wintertime
(December 1 to February 28) O3 mixing ratio vs. NOX mixing
ratio during the morning and afternoon 5-15
5-9 Summertime O3 mixing ratio vs. NOX mixing ratio measured
during the afternoon hours 5-16
5-10 Model calculated daytime change in O3 (from sunrise to
4:30 P.M.) for summer clear sky conditions is compared to the
observed difference between the afternoon (2:00-7:00 P.M.) and
the morning (7:00-11:00 A.M.) for clear sky conditions 5-17
5-11 O3 production per unit NOX per day, (AP) from the NMHC-PO
model are plotted as function of NOX mixing ratios 5-18
5-12 NOy shortfall . 5-27
5-13 Formation and decay of diethylnitrosamine in the dark and in
the sunlight from diethylamine (open squares) and from
triethylamine (open circles) 5-29
5-14 Pollutant levels at the Kenosha sampling site before and after
passage of the lake breeze front 5-39
6-1 Absolute error in the NO2 (A NO2) for 10 s in the dark sampling
line 6-13
August 1991 I-xviii DRAFT-DO NOT QUOTE OR CITE
-------
FIGURES (cont'd)
Number . Page
7-1 Distribution of peak annual NO2 averages in 103 MSAs,
1988-89 ................................. . . 7-5
7-2 National tread in the composite annual average nitrogen dioxide
concentrations at both NAMS and all sites with 95% confidence
intervals, 1980-1989 ............................ 7-6
7-3 Metropolitan area trends in the composite annual average
nitrogen dioxide concentration, 1980-1989 ............... 7-7
7-4 Monthly 50ft, 90th and 98th pereentiles of 1-h NO2
concentrations at selected stations, 1986-1989 ............. 7-8
7-5 Annual average NO2 vs second Mgh 1-h concentration at
216 stations— 1988 ............. ................ 7-9
7-6 Hourly relative frequency distributions of 1-h NO2 values at
four selected stations for 1988, with numbers of values greater
than 0.2 ppm ............ ....*............... 7-11
7-7 Percent of 1-h values above 0.03 and 0.05 ppm vs. annual
averages > 0.03 ppm, 1988 ....................... 7-13
7-8 Relative distributions of 1-h NO2 values at selected stations,
1988 ______ ....... . .......... ......,.....;. 7-14
7-9 Cumulative frequency distribution of NO2 concentrations
(one-week sampling period) by location for homes with
no gas appliances for a winter period in Southern
California .... ....... ...... ....... .... ...... 7-22
7-10 Cumulative frequency distribution and arithmetic means of NO2
concentrations (two week sampling period) by location for homes
with no kerosene heater and no gas range for a winter period in
New Haven, CT, area .... ....................... 7-23
7-11 Cumulative frequency distribution of NC^ concentrations (one
week sampling period) by location for homes with no gas
appliances for a summer period in Southern California ....... 7-24
7-12 Indoor/outdoor NO2 concentration ratios (2-week sampling
periods) as a function of time for three homes without
indoor NO2 sources .............. . ........... . . 7-25
August 1991 I-xix DRAFT-DO NOT QUOTE OR CITE
-------
FIGURES (cont'd)
Number , Page
7-13 Verticle distribution of average NO2 concentrations (48-h
sampling periods) measured in nine New York City
apartments . 7-33
7-14 Mean NO2 concentrations (1-week sampling periods) for eight
sampling periods by location in the home and type of cooking
fuel 7-34
7-15 Cumulative frequency distribution and arithmetic means by
location, of average NO2 levels (2-week sampling periods) during
kerosene heater use for residences with one kerosene heater and
no gas range, New Haven, CT, area study, winter 1983 7-40
7-16 Bar graph of NO2 removal rate for various materials evaluated
in a 1.64 m3 test chamber at 50% relative humidity . 7-52
7-17 Concentration distributions (in ppb) for gas phase species in
Boston (A) HONO; (B) NO2; (C) NH3; (D) HNO3 . 7-56
8-1 Average personal NO2 exposure for each household compared with
outdoor concentrations for summer and winter . 8-7
8-2 Average personal NO2 exposure for each home compared
with average indoor concentrations for summer and winter 8-8
8-3 Comparison of the house average two-week NO2 concentrations
with the total personal NO2 levels measured over the same
time period for one adult resident in each house,
New Haven, CT, area, winter 1983 8-10
8-4a Proportion of time spent by women who are full-time homemakers
in indoor, outdoor, and in-transit microenvironments 8-18
8-4b Proportion of time spent by employed persons in indoor, outdoor,
and in-transit microenvironments 8-18
August 1991 , i-xx DRAFT-DO NOT QUOTE OR CITE
-------
AUTHORS
Chapter 3: General Chemical and Physical Properties of NOX and NOx-Derived Pollutants
Dr. Robert W. Elias
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Chapter 4: Sources of Nitrogen Oxides Influencing Ambient and Indoor Air Quality
Dr. Gregory J. McRae Dr. Cliff I. Davidson
702 Woodland Drive Department of Civil Engineering
Ohara Township Carnegie Mellon University
Pittsburgh, PA 15238 Pittsburgh, PA 15213
ChapterS: Transport and Transformation
Dr. Halvor Westberg
Laboratory for Atmospheric Research
Washington State University
Pullman, WA 99164-2730
Chapter 6: Sampling and Analysis for Ambient Oxides of Nitrogen and Related Species
Dr. Joseph Sickles
Atmospheric Research and Exposure Assessment
Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Chapter 7: Ambient and Indoor Concentrations of Nitrogen Oxides
Dr. Brian Leaderer Mr. Tom McMullen
Pierce Foundation Laboratory Environmental Criteria and Assessment
290 Congress Avenue Office
New Haven, CT 06519 U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Robert W. Elias
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
August 1991 I-xxi DRAFT-DO NOT QUOTE OR CITE
-------
AUTHORS (cont'd)
Chapter 8: Exposure
Dr. Brian Leaderer
Keice Foundation Laboratory
290 Congress Avenue
New Haven, CT 06519
August 1991 I-xxii DRAFT-DO NOT QUOTE OR CITE
-------
CONTRIBUTORS AND REVIEWERS
Dr. A. Paul Altshuller
Atmospheric Research and Exposure
Assessment Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Michael Berry
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Irwin H. Billick
Gas Research Institute
8600 West Byrn Mawr Avenue
Chicago, IL 60631
Dr. Steven D. Colome
Integrated Environmental Services
University Tower, Suite 1090
4199 Campus Drive
Irvine, CA 92715
Dr. Cliff Davidson
Department of Civil Engineering
Carnegie Mellon University
Pittsburgh, PA 15213
Dr. Marcia Dodge
Air and Energy Engineering Research
Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Robert Elias
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Reseach Triangle Park, NC 27711
Dr. Don Fox
School of Public Health
ENVR-CBF7400
University of North Carolina
Chapel Hill, NC 27599-7400
Dr. Vic Hasselblad
Center for Health Policy Research
Duke University
Durham, NC 27713
Dr. Brian Heikes
Graduate School of Oceanography
University of Rhode Island
Narrangansett Bay Campus
Narrangansett, M 02882-1197
Ms. Pamela Johnson
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Dennis J. Kotchmar
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Douglas Latimer
Latimer & Associates
505 27th Way #301
Boulder, CO 80303
Dr. Brian Leaderer
Pierce Foundation Laboratory
Yale University School of Medicine
290 Congress Avenue
New Haven, CT 06519
Mr. Frank McElroy
Atmospheric Reseach and Exposure
Assessment Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. J. David Mobley
Air and Energy Engineering Research
Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
August 1991
I-xxiii
DRAFT-DO NOT QUOTE OR CITE
-------
CONTRIBUTORS AND REVIEWERS (cont'd)
Mir, Warren Porter
U.S. Consumer Products Safety Commission
5401 Westbard Avenue
Room 724
Bethesda, MD 20816
Dr. P. Barry Ryan
Department of Environmental Science and
Physiology
Harvard School of Public Health
677 Huntington Avenue
Boston, MA 02115
Dr. Joseph E. Sickles n
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709
Dr. Thomas Stock
University of Texas
School of Public Health
P.O. Box 20186
Houston, TX 77225
Ms. Beverly Tilton
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. John H. Wasser
Air and Energy Engineering Research
laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Hal Westberg
Laboratory for Atmospheric Research
Washington State University
Pullman, WA 99164-2730
Mr. James White
Air and Energy Engineering Research
Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Warren White
6840 Waterman Avenue
St. Louis, Missouri 63130
Dr. Ron Wyzga
Electric Power Research Institute
3412 Hillview Avenue
P.O. Box 10412
Palo Alto, CA 94303
August 1991
I-xxiv
DRAFT-DO NOT QUOTE OR CITE
-------
i 1. SUMMARY OF EFFECTS OF OXIDES OF
2 NITROGEN AND RELATED COMPOUNDS ON
3 HUMAN HEALTH AND WELFARE
4
5
6 1.1 INTRODUCTION
7 This revised Air Quality Criteria for Oxides of Nitrogen reviews and evaluates the
8 scientific information on the health and welfare effects associated with exposure to the
9 concentrations of nitrogen dioxide (NO^ found in ambient air. The purpose of this
10 document is to present air quality criteria for oxides of nitrogen (NOX) in accordance with
11 Sections 108 and 109 of the Clean Air Act (CAA).
12 Section 108 (U.S. Code, 1991) directs the Administrator of the U. S. Environmental
13 Protection Agency (EPA) to list pollutants that may reasonably be anticipated to endanger
14 public health and welfare, and to issue air quality criteria for them. These air quality criteria
15 are to reflect the latest scientific information useful in indicating the kind and extent of all
16 identifiable effects on public health and welfare that may be expected from the presence of
17 the pollutants in the ambient air.
18 Section 109(a,b) (U.S. Code, 1991) directs the EPA Administrator to propose and
19 promulgate "primary" and "secondary" National Ambient Air Quality Standards (NAAQS)
20 for pollutants identified under Section 108. Section 109(b)(l) defines a primary standard as a
21 level of air quality, the attainment and maintenance of which in the judgment of the
22 Administrator, based on the criteria and allowing for an adequate margin of safety, is
23 requisite to protect the public health. Section 109(d) of the Act requires periodic review and,
24 if appropriate, revision of existing criteria and standards. In addition, Section 109(c)
25 specifically requires the Administrator to promulgate a primary standard for NO2 with an
26 averaging time of hot more than 3 h, unless no significant evidence is found that such a
27 short-term standard is required to protect health. Under Section 109(b) of the Clean Air Act,
28 the Administrator must consider available information to set secondary NAAQS that are based
29 on the criteria and are requisite to protect the public welfare from any known or anticipated
30 adverse effects associated with the presence of such pollutants. The welfare effects included
31 in the criteria are effects on vegetation, crops, soils, water, animals, man-made materials,
August 1991 1-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 weather, visibility, and climate, as well as damage to and deterioration of property, hazards
2 to transportation, and effects on economic values, personal comfort, and well-being.
3 A variety of NOX compounds and their transformation products occurs naturally in the
4 environment and also results from human activities. In addition to NO2, nitric oxide (NO),
5 nitrous oxide (N2O), gaseous nitrous acid (HONO), gaseous nitric acid (HNO3), and both
6 nitrite and nitrate particles have all been found in the ambient air. The formation of
7 nitrosamines in the atmosphere by reaction of NOX with amines has also been suggested, but
8 not yet convincingly demonstrated. Available scientific research on the potential health and
9 welfare effects of NOX compounds provides the strongest evidence linking specific adverse
10 effects to near-ambient concentrations of NO2. Therefore, EPA has focused its criteria
11 reviews primarily on health and welfare effects reported to be associated with exposure to
12 NO2. Nitrogen dioxide is an air pollutant generated mainly by the photochemical oxidation
13 of NO, which is emitted from a variety of mobile and stationary sources. At elevated
14 concentrations, NO2 can adversely affect human health, vegetation, materials, and visibility.
15 Nitrogen oxide compounds can also contribute to increased rates of acidic deposition and high
16 ozone concentrations.
17
18 1.1,1 Critical Issues
19 Based on the available scientific evidence several critical questions or issues are
20 addressed in this document. These include several key issues, as follows:
21 * Is the apparent relationship between short-term exposure (1 to 3 h) in the
22 range 0.2 to 0.5 ppm NO2 and increased bronchial reactivity in asthmatics
23 an adequate basis to form criteria for a short-term NO2 ambient standard?
24 How frequently do such short-term ambient levels occur in or above this
25 concentration range? .
26
27 • Is the strength and consistency of the epidemiologic data base and its
28 analysis relating NO2 exposure and an increased rate and/or severity of
29 respiratory disease and symptoms adequate to quantitatively assess whether
30 ambient or near ambient levels pose an increased level of risk?
31
32 * Is the host defense data base adequate to provide a plausible biological basis
33 for relationships seen in epidemiologic studies between respiratory disease
34 and exposure to NO2?
35
August 1991 1-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 * What is the emphysematous potential in humans from exposure to long-
2 term chronic ambient levels of NO2?
3
4 * What subpopulation groups ase likely to be more susceptible (ban others to
5 effects from ambient NO2 exposure and potentially at heightened risk?
6
7 * Is the data base concerning the ecological effects of critical nitrogen loading
8 adequate to allow conclusions to be determined? ,
9 ••••••>..
.0
.1 1.1.2 Organization of the Document
12 The document consists of 16 chapters. This Summary chapter for the entire document
3 is followed by a general introduction in Chapter 2. Chapters 3 through 8 provide background
14 information on physical and chemical properties of NO2 and related compounds; sources and
15 emissions; atmospheric transport, transformation and fate of NO2; methods for the collection
16 and measurement of NO2; and ambient air concentrations and factors affecting exposure of
17 the general population. Chapter 9 evaluates NO^ effects on crops and natural vegetation,
L8 while Chapter 10 discusses effects on terrestrial and,aquatic ecosystems. Chapter 11
19 describes effects on visibility, and Chapter 12 describes damage to materials attributable to
!0 NO2. Chapters 13 through 16 evaluate information concerning the health effects of NO2.
Jl More specifically, Chapter 13 discusses respiratory tract deposition of NO2 and information
12 derived from experimental toxicological studies of animals. Chapter 14 discusses
13 epidemiological studies, and Chapter 15 discusses clinical studies. Chapter 16 smnmarizes
14 and integrates information on key critical health issues arising from the three preceding
15 chapters.
16 This document consists of the review and evaluation of relevant literature on NOX
17 through early 1991. The material selected for review and comment in the text generally
!8 comes from the more recent literature, with emphasis on studies conducted at or near
19 pollutant concentrations found ia ambient air. Older literature that was cited in the previous
10 criteria document for NOX (U.S. Environmental Protection Agency, 1982a) is generally not
51 discussed, with the exception of some allusions in the text to certain older studies judged to
J2 be significant because of their potential usefulness in deriving an NAAQS. The newer
S3 information on NOX now available, in some instances, makes possible a better understanding
H of earlier studies, such that a more detailed and comprehensive picture of health effects is
August 1991 1-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 emerging. Information drawn from key literature is the main focus of textual discussions and
2 is presented in tables as well; whereas reports of lesser importance for purposes of this
3 document are mainly summarized only in tables.
4
5
6 1.2 CHEMISTRY, SOURCES, TRANSPORT AND
7 TRANSFORMATION, SAMPLING, AMBIENT AND INDOOR
8 LEVELS, AND EXPOSURE OF OXIDES OF NITROGEN
9 1.2.1 Chemical and Physical Properties of Oxides of Nitrogen
10 There are eight nitrogen oxides that may be present in the ambient air; nitric oxide
11 (NO), nitrogen dioxide (NO^, nitrous oxide (N2O), unsymmetrical nitrogen trioxide
12 (OONO), symmetrical nitrogen trioxide (O-N(O)-O), dinitrogen trioxide (N2O3), dinitrogen
13 tetroxide (^O^, and dinitrogen pentoxide (N2O5).
14 Of these, NO and NO2 are generally considered the most important in the lower
15 troposphere because they may be present in significant concentrations in polluted
16 atmospheres. Their interconvertibility in photochemical smog reactions has frequently
17 resulted in their being grouped together under the designation NOX, although analytic
18 techniques can distinguish clearly between them. Of the two, NO2 is the more toxic and
19 irritating compound.
20 Nitrous oxide is ubiquitous even in the absence of anthropogenic sources since it is a
21 product of natural biologic processes in soil. It is not known, however, to be involved in any
22 photochemical smog reactions. Although N2O is not generally considered to be an air
23 pollutant, it is a principal reactant in upper atmospheric reactions involving the ozone layer.
24 While OONO, O-N(O)-O, N2O3, N2O4, and N2O5 may play a role in atmospheric
25 chemical reactions leading to the transformation, transport, and ultimate removal of nitrogen
26 compounds from ambient air, they are present only in very low concentrations, even in
27 polluted environments.
28 Ammonia (NH3) originates on a global scale during the decomposition of nitrogenous
29 matter in natural ecosystems but it may also be produced locally by human activities such as
30 the maintenance of dense animal populations. Some researchers have suggested conversion of
31 NH3 to NOX in the atmosphere.
August 1991 1-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 1.2.2 Sources of Nitrogen Oxides Influencing Ambient and Indoor Air
2 Quality
3 Quantitative estimates of the total amount of NOX emitted to the ambient global
4 atmosphere are available. These estimates suggest that 40-50 x 106 metric tons of NOX are
5 emitted annually, with about 18-19 X 106 metric tons in the United States alone. These
6 emissions have a direct impact on visibility, human health, and natural ecosystem processes,
:" • • i •;*,-•, .'•.'",.
7 as well as an indirect impact on atmospheric processes that may contribute to acidic
8 deposition, global warming and stratospheric ozone depletion.
9 In addition to ambient emissions of NOX, indoor NOX emissions also contribute to total
.0 human exposure to NOX compounds and consequent human health effects. These two factors,
. 1 ambient and indoor emissions, determine air concentrations and exposure in the human
.2 environment. The important indoor source;? of NOX are gas stoves, unvented space heaters,
.3 kerosene heaters, wood stoves, and tobacco products. Total emissions and the ratio of
.4 NO/NO2 from gas stoves and space heaters differ according to fuel flow rate and flame
.5 adjustment. Additonal factors, such as the load (e.g. cold pot of .water), heater type
.6 (convective vs. radiant) and fuel type (natural gas, propane or kerosene) may also be
.7 important. Only limited information is available for wood stoves and tobacco products.
.8 . . ..,..,..; .,..'.
,9 1.2.3 Transport and Transformation
JO Nitrogen oxides are important chemical species in the planetary boundary layer, as well
11 as in the free troposphere and the stratosphere. Nitrogen oxides play important roles (1) in
12 the control of concentrations of radicals in the clean troposphere; (2) in the production of
13 tropospheric ozone (O3); and (3) directly or indirectly, in the production and deposition of
14 acidic species.
15 . . ._.......
16 1.2.3.1 Ozone Production
17 Combustion processes emit a variety of nitrogen compounds, but chiefly NO, which is
!8 rapidly oxidized to NO2 in ambient air, primarily by O3. Photolytic decomposition of NO2
19 then leads to regeneration of NO, producing also an excited oxygen atom, O(3P), that reacts
50 with molecular oxygen to form O3. In the absence of competing reactions, NO, NO2, and
51 O3 reach an equilibrium described by the steady-state equation.
August 1991 1-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 Competing reactions exist, however, so that free radicals (hydroperoxy, organic peroxy)
2 generated from the the oxidative degradation of volatile organic compounds (VOCs) oxidize
3 NO to NO2 without destroying O3. Thus, the amount of O3 formed in ambient air is
4 dependent upon the concentration of NOX present as well as the concentrations and
5 reactivities of VOC species.
6 • . '.-;,.
7 1.2.3.2 Production of Odd Nitrogen Species
8 Photochemical processes that include the coupled reactions of NOX, oxygen species, and
9 free radicals produce not only O3 but nitrogen^xmtaining products as well. These oxidation
10 products, NOy, include HNO3, HO2NO2, and HNO2; RC(O)O2NO2; N2O5; and inorganic
11 and organic nitrates.
12 Nitric acid is a major sink for active nitrogen and is a contributor to acidic deposition.
13 It has been estimated to account for approximately one-third of the total acidity deposited in
14 the eastern United States (National Researcli Council, 1983). Potential-physical and chemical
15 sinks for HNO3 include wet and dry deposition, photolysis, reaction with OH radicals, and
16 neutralization by gaseous ammonia.
17 Peroxyacyl nitrates are formed from the combination of organic peroxy radicals with
18 NO2. Peroxyacetyl nitrate is the most abundant member in the lower troposphere of this
19 homologous series of compounds. It can serve in the troposphere as a temporary reservoir
20 for reactive nitrogen species and can be regionally transported; but it cannot function as a
21 true sink in the lower troposphere because of its thermal instability. In the upper .
22 troposphere, where temperatures are colder, the lifetime of PAN is longer but is only about
23 3 months, since PAN is photolyzed and also reacts with OH radical. :
24 The NO3 radical is a short-lived NOX that is formed in the troposphere primarily by the
25 reaction of NO2 with O3. In daylight, NO3 undergoes rapid photolysis or reaction with NO.
26 After sunset, accumulation of NO3 can occur and is expected to be controlled by the
27 availability of NO2 and O3 plus chemical destruction mechanisms involving the formation of
28 N2O5 and HNO3.
29 DMtrogen pentoxide, the anhydride of HNO3, is primarily a nighttime constituent of
30 ambient air since it is formed from the reaction of NO3 (itself a nighttiine species) and NO2.
31 Dinitrogen pentoxide is thermally unstable, but at the lower temperatures of the upper
August 1991 1-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 troposphere it can serve as a temporary reservoir of NO3. In ambient air, N2O5 reacts
2 heterogeneously with water to form HNO3, which in turn is deposited out; thus, N2O5
3 provides an important removal mechanism for HNO3.
4 Amines, nitrosamines, and nitramines are thought to exist in ambient air but at low
5 concentrations. Both nitrosamines and nitramines have short lifetimes in ambient air because
'6 they are photolytically decomposed (nitrosamines) and/or react with OH radical and O3
7 (nitramines and nitrosamines).
8
9 1.2,3.3 Transport
10 General Features
[ 1 The transport and dispersion of the various nitrogenous species are dependent on both
12 meteorological and chemical parameters. Advection, diffusion, and chemical transformations
13 combine to dictate the atmospheric residence time of a particular trace gas. In turn,
14 atmospheric residence times help determine the geographic extent of transport of given
15 species. Surface emissions are dispersed vertically and horizontally through the atmosphere
16 by turbulent mixing processes that are dependent to a large extent on the vertical temperature
17 structure and wind speed.
18 As the result of meteorological processes, NOX emitted in the early morning hours in an
19 urban area will disperse vertically and move downwind as the day progresses. On sunny
20 summer days, most of the NOX will have been converted to HNO3 and PAN by sunset.
21 Much of the HNO3 is removed by deposition as the air mass is transported, but HNO3 and
22 PAN carried in layers aloft (above the nighttime inversion layer but below a higher
23 subsidence inversion) can potentially be transported long distances.
24
25 Transport of Reactive Nitrogen Species in Urban Plumes
26 Studies of the fate of reactive nitrogen species in daytime urban plumes indicate removal
27 rates ranging from 0.04 h"1 in Los Angeles (Chang et al., 1979), to 0.1 h"1 in Detroit (Kelly,
28 1987), to 0.14 to 0.24 h'1 (for 4 different, nonconseeutive days) in Boston (Spicer, 1982).
29 In the Detroit study, HNO3 accounted for 67 to 84% of the nitrogenous transformation
30 products, but still fell short of predicted HNO3 levels. Removal by incorporation into coarse
August 1991 1-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 atmospheric aerosol was postulated as a major sink for HNO3 and the cause of the
2 discrepancy between measured and predicted levels (Kelly, 1987).
3 The nighttime chemistry of NOy is poorly understood. Nighttime concentrations of
4 NO3 show a typical pattern of increase until O3 is no longer available, followed by a
5 decrease as NO emissions cannot be oxidized by O3 to NO2 but react instead with the NO3.
6
7 Transport and Chemistry in NOx-Rich Plumes
8 Ozone build up in power plant plumes appears to be the result of mixing of VOCs into
9 the plume as it moves downwind, since the VOC/NOX ratio in these plumes is quite low. An
10 O3 build up is not found in all power plant plumes, however (e.g., Hegg et al., 1977; Ogren
11 et al., 1977; White, 1977); and the most important factor in the in-plume formation of ozone
12 appears to be the availability of reactive VOC in the dilution air.
13 Little information is available on the fate of reactive nitrogen species in NOx-rieh
14 plumes. Aircraft measurements have shown little increase in inorganic and paniculate NO3
15 concentrations in power plant plumes (Hegg and Hobbs, 1979). Chamber and modeling
16 studies indicate that in NO^-rich but VOC-poor plumes the NOX lifetime will be long enough
17 to allow NOX to be incorporated into.regional air masses (Spicer et al., 1981).
18
19 Regional Transport
20 Transport of reactive NOX in regional air masses can occur via several mechanisms:
21 (1) mesoscale phenomena, such as mountain-valley wind flow or land-sea breeze circulations
22 (transport for 10's to 100's of km ); (2) synoptic weather systems such as the migratory highs
23 that cross the eastern United States in the summertime (transport for many 100's of km); and
24 (3) mesoscale phenomena coupled with slow-moving high-pressure systems having weak
25 pressure gradients. M the latter interrelated phenomena, mountain-valley or land-water
26 breezes can govern pollutant transport in the immediate vicinity of sources but the ultimate
27 fate of reactive NOX species will be distribution into the synoptic system.
28 Information remains sparse on NOX species and their concentrations in synoptic
29 transport systems. Calculated fluxes for the northeastern, Atlantic coast area (Luke and
30 Dickerson, 1987; Galloway et al., 1984; Galloway and Whelpdale, 1987) correspond to about
August 1991 1-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 25% of the NOX emitted to the atmosphere of eastern North America, using Logan's emission
2 estimate of 4.5 x 109 g/year (Logan, 1983).
3
4 1.2.3.4 Oxides of Nitrogen and the Greenhouse Effect
5 Except for N2O, the reactive nitrogen species comprising the NOX and NO families in
6 the atmosphere do not absorb infrared radiation and therefore do not contribute directly to
7 radiative "greenhouse" forcing. They can, however, contribute indirectly to greenhouse
8 processes through the photochemical production of O3 in the troposphere. Ozone absorbs
9 infrared radiation about 2,000 times more effectively per mole than CO2 does, but at its
0 present tropospheric levels O3 contributes only about 8% of the total theorized greenhouse
1 effect (Rodhe, 1990).
2 Nitrous oxide, which is chemically inert in the troposphere, readily absorbs infrared
3 radiation and is among the more significant non-carbon dioxide greenhouse gases.
4 Absorption of visible radiation by N2O could make this compound a possible source of other
5 climatic influences if atmospheric concentrations become sufficiently large (Wuebbles, 1989).
6
7 Nitrous Oxide Greenhouse Contributions
8 Nitrous oxide, on a per mole basis, is thought to be 200 times more effective than CO2
9 as an absorber of heat radiation (Wuebbles, 1989; Rodhe, 1990). It is estimated at present
0 levels to be responsible for about 4 to 5% of the theorized greenhouse effect (Hansen et al.,
1 1989; Rodhe, 1990). Assuming a 0.2% per year increase, Levander (1990) predicted that the
2 increased atmospheric mixing ratio of N2O after 50 years would result in an even greater
3 efficiency (350 times) in infrared radiation absorption compared to expected CO2 levels.
4
5 Stratospheric Ozone Depletion by Oxides of Nitrogen
6 In mid-latitudes and lower to mid-stratospheric regions, cyclic reactions initiated by the
7 oxidation of NO by O3 lead to the net destruction of two molecules of O3. Among the
8 stratospheric O3-depletion mechanisms that have been proposed, however, are much more
9 important reactions involving the dimerization of CIO in the presence of a third body, M, and
0 subsequent sequences in which the monomer is regenerated and two O3 molecules are
1 destroyed (Fahey et al., 1989; Molina and Molina, 1987). McElroy et al. (1986) also
August 1991 1-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 proposed Cl and Br oxidation cycles as an important stratospheric O3-depletion mechanism.
2 In this mechanism, the dimerizaton and regeneration of CIO, coupled with the oxidation of Br
3 and Cl by O3, are limited by the reaction of CIO and BrO with NO2 to form the unreactive
4 C1NO3 or BrNO3 (McElroy et aL, 1986a). Reactions of CIO and BrO with NO2 are
5 important sinks for the halogen oxides but are insignificant sinks for NO2 (McElroy and
6 Salawitch, 1989).
7 Sequestration of reactive NOy by heterogeneous reactions on the ice surfaces of polar
8 stratospheric clouds has been proposed as a means of removing NO2. Its removal allows
9 other O3-depleting cycles to proceed (Molina et al., 1987; Tolbert et aL, 1987; Tolbert et aL,
10 1988a). Dinitrogen pentoxide has been implicated in these heterogeneous reactions.
11
12 1.2.3.5 Deposition of Nitrogen Oxides
13 Both wet and dry deposition of NOX and other nitrogen species occur, but wet
14 deposition is not a significant removal mechanism for NO or NO2, since both gases are
15 minimally soluble in water. Transformation to more highly oxidized forms is necessary for
16 effective wet depositon of NOX; and the reaction of NO2 with the OH radical to form HNO3
17 appears to be the main source of NO3" in precipitation. About one-third of the emissions of
18 NOX in the United States is estimated to be removed by wet deposition (Hicks et al., 1990).
19 Dry deposition fluxes for NOX are highly uncertain, mainly because of analytical
20 problems and the simultaneous occurrence of emission and deposition of NOX. Data available
21 indicate, however, that NO emissions exceed NO deposition and that NO2 deposition exceeds
22 NO deposition. Reported Vd values for respective nitrogen species are: < 0.1 to
23 ~0.2 cm s"1 for NO; 0.3 to 0.8 cm s"1 for NO2; 0.5 to 3.0 cm s"1 for HNO3 over land and
24 0.3 to 0.7 cm s"1 for HNO3 over water (Huebert and Robert, 1985). The few data that exist
25 show deposition rates for PAN of 0.01 cm s"1 over water (Andreae et al., 1988) and
26 0.25 cm s"1 (Garland and Penkett, 1976) to 2 cm s"1 (Jacob and Wofsy, 1988) over land.
27
28 1.2.4 Sampling and Analysis of Oxides of Nitrogen
29 Since the publication in 1971 of the original version of Air Qualify Criteria for Nitrogen
30 Oxides, changes have occurred in the technology associated with the sampling and analysis
31 for ambient NOX and related species. During the 1970s, roughly the period between
August 1991 i-io DRAFT-DO NOT QUOTE OR CITE
-------
1 publication of the original NOX Criteria Document and its first update and revision, several
2 events occurred that focused on the determination of NO2 in ambient air. In 1973, the
3 original Reference Method was withdrawn because of unresolvable technical difficulties.
4 Major methods development efforts over the next three to four years yielded both automated
5 and manual methods that were suitable for the determination of NO2 in ambient air. As a
6 result, EPA designated a new Reference Method and Equivalent Methods for NO2. The
7 Reference Method specifies a measurement principle and calibration procedures, namely gas
8 phase chemiluminescence (GP-CLM) with calibration using either gas phase titration of NO
9 with O3 or a NO2 permeation device. The Sodium Arsenite Method in both the manual and
0 continuous forms and the TGS Method were also designated as Equivalent Methods.
1 Subsequently, commercial GP-CLM instruments were designated as Reference Methods. The
2 sensitivity of these devices was in the low ppb range and while the GP CLM instruments
3 were recognized as being susceptible to interferences by other nitroxy species, it was believed
4 that the atmospheric concentrations of these compounds were generally low relative to NO2.
5 In the 1980s, since the first update and revision of the Criteria Document, additional
6 developments have occurred. Information from air quality monitoring networks is now
7 readily available and has shown the GP CLM instruments to have nominal precision of
8 +10 to 15% and accuracy of 20% and to have replaced manual methods to a large extent in
9 network applications. Heightened interest in the research community on the speciation of
0 atmospheric trace gases and specifically nitrogen-containing species has prompted a new wave
1 of methods development. While the basic design and performance of the commercial
2 instruments have remained essentially unchanged, researchers have improved GP CLM
3 measurement technology and refined other instrumental methods to permit the determination
4 of NO, NO2, and NOy in the low ppt range. Although GP CLM NO detectors coupled with
5 catalytic NO2-to-NO converters are still not specific for NO2, they have proven useful for
6 measuring NOy, and GP CLM NO detectors coupled with photolytic NO2-to-NO converters
.7 have shown improved specificity for NO2.
.8 A continuous liquid phase CLM device for sensitively detecting NO2 has been
.9 developed and may be suitable to measure NO2 if interference problems can be overcome.
'0 Passive samplers for NO2 have been used, primarily for workplace and indoor applications
August 1991 1-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 but hold promise for ambient measurements as well. Gas ehromatography with electron
2 capture detection is useful in the determination of PAN, other organic nitrates, and N2O.
3 Laser induced fluorescence has been introduced to detect NO, NO2, and HNO2 with
4 high sensitivity and specificity. Tunable-diode laser spectroscopy has been used to detect
5 NO, NO2, and HNO3. Long path spectroscopy has also been used to detect NO, NO2,
6 HNO2> and NO3. These spectroscopic methods are research tools and are not yet easily or
7 economically suited for routine monitoring.
8 Interest in acidification of the environment has resulted in the development of methods
9 for HNO2 and HNO3. Integrative methods using denuders have been introduced to permit
10 sensitive determination of these and other species. In recent years the potential for artifacts
11 in using filters for sampling particulate matter and specifically paniculate nitrate has been
12 recognized. This has given rise to careful characterization of filter media for potential
13 artifacts and the use of combinations of denuders and filters to permit more specific
14 determination of nitrogen-containing gases and particulate nitrates in ambient air.
15
16 1.2.5 Ambient and Indoor Concentration of Oxides of Nitrogen
17 1.2.5.1 Ambient Concentration
18 In urban areas, hourly patterns at fixed-site, ambient air monitors often follow a
19 biomodal pattern of morning and evening peaks, related to motor vehicular traffic patterns.
20 Sites affected by large stationary sources of NO2 (or NO that rapidly converts to NO^ are
21 often characterized by short episodes at relatively high concentrations.
22 The highest hourly and annual NO2 levels are reported from air monitoring stations in
23 California. The seasonal patterns at California stations are usually quite marked and reach
24 their highest levels through the fall and winter months, while stations elsewhere in the U.S.
25 usually have less prominent seasonal patterns and may peak in the winter, in the summer, or
26 contain little discernable variation.
27 Since 1980, the average level among reporting NO2 stations has been below 0.03 ppm,
28 with no significant trend evident. For 103 Metropolitan Statistical Areas reporting a valid
29 year's data for at least one station in 1988 and/or 1989, peak annual averages ranged from
30 0.007 to 0.061 ppm. The only recently measured exceedances of the annual standard,
31 0.053 ppm, have occurred at stations in' southern California.
August 1991 1-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 One-hour NO2 values can exceed 0.2 ppm, but, in J988, only 16 stations (12 of which
2 are in California) reported an apparently credible second high 1-h value greater than 0.2 ppm.
3 Since at least 98% of 1-h values at most stations are below 0.1 ppm, these values above
4 0.2 ppm are quite rare excursions.
5 .••--...... ..-.-••
6 1.2.5.2 Indoor Concentration
7 Indoor concentrations of NO2 are a function of outdoor concentrations, indoor sources
8 (source type, condition of source, source use, etc.), infiltration/ventilation, air mixing within
9 and between rooms, reactive decay by interior surfaces and air cleaning or source venting. In
10 homes without indoor sources of NO2» concentrations will be below outdoor levels due to
11 removal by the building envelope and interior surfaces, thus providing some degree of
12 protection from outdoor concentrations. Indoor/outdoor ratios for homes without sources
13 vary considerably by season of the year with the lowest ratios occurring in the winter and
i
14 highest occurring during the summer. The differences are probably due to seasonal
15 differences in infiltration rates, NO2 reactivity rates, the penetration factor and outdoor
16 concentrations. The indoor/outdoor ratio for these homes does not appear to vary by
17 geographic area, housing type or outdoor concentration.
18 Gas appliances (gas range/oven, water heater, etc.) represent the major indoor source
19 category for indoor residential NO2 by virtue of the number of homes with such sources
20 (approximately 45% of all homes in the U.S.). NO2 levels in homes with gas appliances are
21 higher than those without such appliances. Within this catagory, the gas range/oven is by far
22 the major contributor, espically when used as a supplemntal heat source. Average indoor
23 concentrations (a one to two week measurement period) in excess of 100 jig/m3 have been
24 measured in some homes with gas ranges. Homes with gas ranges with pilot lights will have
25 higher NO2 levels than homes that have gas ranges without pilot lights. Average NO2
26 concentrations in homes with gas ranges/ovens exhibit a spatial gradient within and between
27 rooms. Kitchen levels are higher than other rooms and a steep vertical concentration gradient
28 in the kitchen has been observed in some homes, with concentrations being highest nearest
29 the ceiling. Average NO2 concentrations are highest during the winter months and lowest
30 during the summer months. This seasonal temporal gradient is probably related to seasonal
31 differences in infiltration, source use, NO2 reactivity rates indoors and outdoor
August 1991 1-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations. The impact of gas appliance use on indoor NO2 levels may be superimposed
2 upon the background level resulting from outdoor concentrations. The results of field studies
3 of the impact of gas ranges on indoor NO2 levels are fairly consistent. Once corrected for
4 the contribution of outdoor concentrations, the average contribution of gas ranges to NO2
5 levels indoors is similar by locations within homes and by season across the studies. Only
6 very limited data exist on short-term average (3 h or less) indoor concentrations of NO2
7 associated with gas appliance use. The limited data suggest that short-term averages of NO2
8 are several times the longer-term average concentrations measured.
9 Unvented kerosene and gas space heaters are important sources of NO2 in homes, both
10 because of the NO2 production rate of the heaters and the length of time they are used. Field
11 studies indicate that average residential concentrations (one week or two week average levels)
12 exhibit a wide distribution varying primarily with the amount of heater use and type of.
13 heater. Under similar operating conditions unvented gas space heaters appear to be associated
14 with higher indoor NO2 levels than kerosene heaters. Average concentrations in homes using
15 unvented kerosene heaters have been measured well in excess of 100 /ig/m3. In one study
16 calculations of NO2 residential levels during heater use (in homes without gas appliances)
17 indicate that approximately 50% of the homes have concentrations above 100 /ng/m3 and
18 8% above 480 ftg/m3. Peak NO2 levels of 847 jug/m3 over a 1-h period in a home during
19 use of a kerosene heater have been measured, A large field study of indoor NO2
20 concentrations in homes using unvented gas space heaters (most also had gas ranges) found
21 that approximately 70% of the homes had average concentrations in excess of 100 /ig/m3 and
22 20% in excess of 480 j«g/m3. This study found that the indoor/outdoor temperature
23 difference was the best indicator of indoor NO2 levels during the colder winter periods when
24 heating demands are greatest. No data have yet been published that provide concentrations
25 during heater use nor are any short-term peak indoor concentrations yet published for homes
26 with unvented gas space heaters. No spatial gradient of NO2 was found in homes with
27 unvented kerosene space heaters, contrary to the strong spatial gradient noted for homes with
28 gas appliances. This is probably due to the strong convective heat output and the long
29 operating hours of the heaters, which result in rapid mixing within the homes. Published
30 data are not yet available on spatial concentrations during the use of unvented gas space
31 heaters.
August 1991 1-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 Improper use of gas appliances (e.g., using a gas range to heat a living space) and
2 improperly operating gas appliances or vented heating systems (e.g., out-of-repair gas range
3 or improper operation of a gas wall or floor furnace) can be important contributors to indoor
4 NO2 concentrations, but little data are available to assess the extent of that contribution.
5 Field data do not exist that would allow for an assessment of the contributions of wood or
6 coal burning stoves or fireplaces to indoor NO2 concentrations, but such a contribution would
7 be expected to be small. Cigarette smoking is expected to add little NO2 to homes^
8 Efforts to model indoor NO2 levels have employed both physical/chemical and
9 empirical/statistical models. Physical/chemical models have largely been applied to test
10 house data or to small samples of homes where detailed data on the factor impacting the
11 concentrations have been measured. Empirical/statistical models have been developed from
12 large field study data bases. These employ questionnaire responses and measured physical
13 data (house volume, etc.) as independent variables and have met with only moderate success.
14 The removal of NO2 indoors by surfaces (reactive decay) is often equal to or greater
15 than infiltration in removing NO2. NO2 reactive decay rates vary sharply as a function of the
16 type of material and surface area of the material. The degree of mixing or turbulence in a
17 space is also important. The role of relative humidity is not clear and some materials can
18 , become NO2 saturated.
19
20 1.2.6 Assessing Total Human Exposure to Nitrogen Dioxide
21 Exposure to NO2 occurs across a number of micorenvironments or settings. An
22 individual's integrated exposure is the sum of all of the individual NO2 exposures over all
23 time intervals for all microenvironments, weighted by the time in each microenvironment.
24 Accurate assessments of total NO2 exposure and of the environments in which exposures take
25 place are essential to minimize misclassification errors in epidemiologic studies, in defining
26 population exposure distributions in risk assessment, and in developing effective mitigation
27 measures in risk management. Personal NO2 exposures can be assessed by direct and indirect
28 measures. Direct measures include biomarkers and use of personal monitors. No validated
29 biomarkers for exposure are available for NO2. A limited number of studies have been
30 conducted in which personal exposures to NO2 were measured using passive monitors. These
31 studies generally indicate that the outdoor levels of NO2, while related to both personal levels
August 1991 1-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 and indoor concentrations, are poor predictors of personal exposures for most populations.
2 Average indoor residential concentrations (e.g., whole house average or bedroom level) tend
3 to be the best predictor of personal exposure, typically explaining 50 to 60% of the variation
4 in personal exposures. In selected populations, the indoor residential environment may not be
5 a good predictor of total exposure because of the higher percent of time spent in different
6 environments and/or the potential for unusual NO2 concentrations.
7 Indirect measures of personal exposure to NO2 employ various degrees of
8 microenvironmental monitoring and questionnaires to estimate an individual's or population's
9 total exposure. One such model, developed from an extensive monitoring and questionnaire
10 data base, can estimate population exposure distributions from easily obtained data on outdoor
11 NO2 concentrations, housing characteristics, and time-activity patterns. This model is
12 proposed for use in evaluating the impact of various NO2 mitigation measures. The model is
13 promising but it has not yet been validated and the uncertainty associated with it has not been
14 characterized.
15
16
17 1.3 EFFECTS OF NITROGEN OXIDES ON VEGETATION,
18 ECOSYSTEMS, VISIBILITY, AND MATERIALS
19 1.3.1 Effects of Nitrogen Oxides on Vegetation
20 1.3.1.1 Introduction
21 To protect vegetation from the deleterious effects of nitrogen oxides, it is important to
22 determine the impact of respective concentrations and durations of exposure to NOX on
23 vegetation. In Chapter 9, the biochemistry and physiology of individual plants and uniculture
24 crops are discussed in relation to the types of injury induced by such exposures and in
25 relation to how the plant may be protected, in part, either by exclusion or detoxification of
26 oxides of nitrogen.
27 Plants require reduced nitrogen compounds to form proteins, nucleic acids, and many
28 secondary products in order to survive and grow. Under most circumstances, nitrogen enters
29 the plant through roots in three modes:
30 (1) absorption of ammonia (and ammonium),
31 (2) absorption of nitrate (and nitrite), and
32 (3) nitrogen fixation by symbiotic organisms.
August 1991 1-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 Thus, any pollutant that can be converted chemically or biologically into nitrate, nitrite
2 or ammonia can be used by the plant. Nitrogen oxides that faU upon the soil can potentially
3 be easily converted by microbial or chemical action and so can be readily adsorbed by the
4 roots. Ground-deposited NOX can enter the cellular metabolic processes easily through the
5 soil/root interface, although the deposition can overload the soil/plant systems (see
6 Chapter 10). Gaseous NOX that enters through the leaf can likewise be converted through
7 enzyme systems which can handle the derived compounds.
8 This update of the scientific literature has established several new findings since the
9 preparation of the last criteria document (U.S. Environmental Protection Agency, 1982a).
10 • NO and NO2 interact differently within the plant. Thus, the effects of NOX
11 must be categorized according to NOX species.
12 '
13 • NO2 is water soluble and can be incorporated into normal plant nitrogen
14 metabolism; upon entry into the cell it is hydrated to form nitric and
15 nitrous acids that exist as anions in the aqueous milieu of the cell.
16 Despite the fact that NO2" and NO3~ are normal anions in the plant, too
17 much nitrogen can be toxic. The conversion of the biochemical species
18 can overwhelm the stepped metabolic process so that the concentration
19 can rise to detrimental levels.
20
21 • NO is a water insoluble compound and induces free radical reactions.
22 While the exact sequence of reactions is still unknown, it is clear that NO
23 behaves differently than NO2. When both NO and NO2 are present, NO
24 seems, like NO2, to be converted into nitrite and nitrate. Metabolic
25 incorporation leads to detoxification of most of the species of NOX and to
26 making the potentially toxic compounds not only harmless to the plant but
27 important to its normal growth.
28
29 • The third category contains the remainder of nitrogen oxides, which are
30 not well defined and the reactions for which are poorly understood.
31
32 1.3.1.2 Nitrogen Dioxide
33 NO2 exposure can overload the nitrogen metabolism pathways. Unfortunately, most
34 studies of the effects of NO2 on vegetation have not traced the inhibition or stimulation of
35 these pathways, but rather have examined visible injury or changes of gross productivity
36 (measured by several possible methods). A summary of data from such investigations is
37 represented by the curves in Figure 1-1, which is reproduced from the previous NO2 criteria
38 document (U.S. Environmental Protection Agency, 1982a). The curves on the figure
39 represent envelopes of the studies in which either (A) metabolic and growth effects or
. August 1991 1-17 DRAFT-DO NOT QUOTE OR CITE
-------
0.01
0.1
Days
1.0
10
100
1000-=
I
100-=
i »-"
0.1
Death
Metabolic and
growth
effects
=-1000
=-100
Threshold for
foliar lesions
I limil| I llll!ll| I lllil!l| I MI!III| I I III
0.1 1.0 10 100 1000 10,000
Duration of exposure, hours
Figure 1-1. The relationship between the onset of either foliar lesions or metabolic and
growth effects and NO}, exposure. The curves contain data points of plant
exposures above which effects were observed.
Source: U.S. Environmental Protection Agency (1982a)
1 (B) visible injury patterns (threshold for foliar lesions) were noted for a given duration of
2 NO2 (abscissa) at a given concentration (ordinate). The lowest curve on the plot indicates
3 where major alterations in plant metabolism occur (although largely undefined, most studies
4 used inhibition of photosynthesis as the marker). The region of the figure below this curve is
5 where NO2 does not affect plant metabolism. A second region in the figure, between this
6 curve and the next higher curve, is the region in which disturbances in metabolism and
7 growth occur (the plant is not normal) but tissue death is not observed. Exposures at levels
August 1991
1-18
DRAFT-DO NOT QUOTE OR CITE
-------
1 and durations in a third region above this curve ("threshold for foliar lesion") result in cell or
2 tissue death (foliar lesions). For very short durations and at very high exposure
3 concentrations, plant death occurs. Although not shown on this curve, there is a poorly
4 defined region where growth stimulation for some plants may occur with NO2 exposure under
5 some conditions (see next section). It is important to note that the NO2 concentration
6 necessary to induce any changes is non-linearly dependent upon the duration of exposure.
7 It is useful to examine Figure 1-1 in more detail to understand the relationships between
i
8 concentration and duration of exposure and the induction of toxic effects such as altered
9 metabolism and foliar injury. The curves can be separated into two sections in which the
10 relationship between duration and concentration is nearly linear. To provide an example,
11 only the curve which marks the beginning of threshold foliar injury will be examined. The
12 first section (Section A) extends from about 0.13 to 0.78 h (8 to 47 min) and has a very steep
13 slope. The second section (Section B) extends from about 3 h to 14 days and has a relatively
14 shallow slope.
15 As discussed in Chapter 9, these two sections can be separately fitted to a power-law
16 relationship, such as;'
17 Cn X T = D0
18 where C is the external concentration in ppm, T is the time in hours, and D0 and n are
19 constants. This formula is fitted to the curves and the following values for each section for
20 the constants are found to be: .
21 Section Time Region n (Power) P.
** " \J --=».*«—
22 A 15-50 min 0.30 1.6
23 B 3 h - 14 days 2.90 55.4
24
25
26 Section A represents very high levels of NO2, which occur infrequently in nature; as
27 such, then, discussion of that section, does not aid in understanding of problems that occur
28 under ambient levels of NO2. At very high levels of NO2 the rate at which the NO2 can
29 enter the tissue water and be converted into nitrate/nitrite is very limited. In section A, then,
30 the concentration of internal NO2 would be expected to be very near that of the outside. In
31 other words, the stomates are probably not limiting the reaction rates unless they are closed.
August 1991 s . 1-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 For longer time periods at concentrations presently occurring within the environment
2 (Section B), the flow of NO2 into the cells is high enough to lower the internal NO2
3 concentrations (relative to the external value). The ability of the plant to utilize the nitrite
4 and ammonia formed would be the governing mechanism of detoxification within this time
5 scale of days to weeks. For time periods of hours or greater, the flow through the stomate
6 and pools of metabolites should have stabilized to a nearly steady-state level. Also the
7 activity of inducible reductase enzymes should have begun to rise. The major question then
8 becomes whether the plant can handle the total increased flow of nitrogen. For much lower
9 exposures and longer durations, however, the question of limitations becomes whether the
10 plant can find some method to use the accumulated nitrogen (now converted to amino acids
11 and proteins).
12 The most obvious sign that NO2 exposure is exceeding the ability of the plant to
13 assimilate extra nitrogen is the appearance of visible injury on the leaf surface. For the most
14 part, visible injury patterns consist of localized chlorotic spots, which in the presence of light
15 and with time develop into necrotic section between the veins. The tissue next to the larger
16 veins remains apparently untouched until much of the leaf is destroyed, possibly due to the
17 ability of the plant to export the excess nitrogen through the veins to other portions. Other
18 evidence of injury is early senescence or leaf drop, as if the aging processes within the leaf
19 have been accelerated.
20 Data presented (Figure 1-1) clearly show that the concept of dose (concentration x
21 duration) will not work, as the effects of NO2 are decidedly nonlinear. Most of the exposure
22 data are discussed in Table 9-5 (Chapter 9). The extensive narrative in that table makes it
23 difficult to summarize the effects easily. However, the majority of the observed effects fall
24 into three categories: (1) no change or effect; (2) a slight increase in mass of the plant or
25 portions of the plant; and (3) a decrease in mass of the plant or portions of the plant.
26 Those plants for which no effects are noted must be tolerant of or able to exclude the
27 excess nitrogen from NO2 exposure. Plants which increase in mass are often those which are
28 suffering from a nitrogen deficiency and so, not surprisingly, grow better under conditions
29 closer to their nitrogen optimum. The more important category is lowered productivity.
30 Productivity loss is generally due to a loss of carbon fixation if the other nutrients are
31 abundant (Sinn and Pell, 1984).
August 1991 1-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 A simplistic, but useful, approach to providing a quantitative assessment of NO2 effects
2 is to transform the exposure information and the narrative in Table 9-3 (Chapter 9) into gross
3 quantitative measures of (a) no effect, (b) decrease, or (c) increase in some measure of
4 productivity, without regard to the actual type of measurement. This approach loses
5 information but it is beneficial in that it permits the tabulation of effects to determine whether
6 there are definite levels of exposure that will lead to toxic injury. It must be noted that, even
7 if the details in the table are examined, there are too many variables, such as humidity, light
8 intensity, soil water potential, and tissue or soil nitrogen, to allow a coherent detailed
9 understanding of the influence of these on the effects.
0 Diagrams of the data tabulations, analogous to Figure 1-1, are presented in Figure 1-2,
1 as log (concentration) versus log (duration). The data indicating a decline in some measure
2 of productivity are shown in Figure 1-2A as a "scattergram". There are several points .of
3 interest. The data seem to indicate that as the duration of exposure lengthens the
4 concentration required to cause some decrease in productivity declines. Hence, exposure for
5 a day to 1 ppm NO2 is somewhat equivalent to 0.1 ppm for a month. The figure shows a
6 non-linear dependence of dose (time x concentration). Furthermore, the line below the axis
7 label shows the lowest measured concentration within the varied time intervals for which a
8 decline in productivity was noted. For durations of a day or longer a decline is noted for
9 NO2 concentrations of 0.02 to 0.1 ppm.
:0 Data for which no observed effect was noted are given in Figure 1-2B and show less
1 dependence upon concentration. Here the maximum concentration for which there was no
:2 effect is shown below the x-axis. For durations of exposure above a day, concentrations as
3 high as 2.1 ppm NO2 have been used without an effect being observed.
14 Data for which a stimulation of some measure of productivity exist are presented in
'5 Figure 1-2C. While there are fewer examples from the literature, the minimum values here
'.6 indicate that exposures of a day to 2 weeks to 0.1 ppm NO2 can cause an increase while for
',7 longer exposures (greater than one month), increases can be induced by as low as 0.024 ppm.
'8 There are a few tentative concepts which can be stated from these data sets
•9
•0 (1) While there are no absolute limits, for the most part lower concentrations
il will cause some reduction in productivity (higher or lower) with longer
12 times of exposure.
13
August 1991 1-21 DRAFT-DO NOT QUOTE OR CITE
-------
Decline In Productivity
No Changs In Productivity
-1
1hr,
0 Ugjdays]1
mot*. calc.
1 diy 2 wto 1 mo.
0.2 0.6 1 \A 14 22
Log [days]
D mass, oakx
0,6 ppm
0.02 ppm 0.1 ppm 0.02 ppm
Increase In Productivity
lag [days]
o meas. cafe.
Mgure 1-2. Effects of NO2 on plant productivity. This figure is similar to that in
Figure 1-1; however, separate experiments are shown by individual symbols
as a function of log (concentration of NO2) and log(duration of exposure).
The data are from Table 9-1. All the data in each sub-figure were fitted to
a linear curve by least squares. Numbers below the curve are minimum
values of concentration reported in the indicated time interval. The three
separate figures are for: (A, top) Decline in productivity (exponent=2.7
± 0.2; ^=0.200; u=87); (B, middle) No observed effect upon productivity
(exponent=1.7 ± 0.2; 1^=0.200; n=87); (C, bottom) Increase in
productivity (exponent=2.1 ± 0.4; 1^=0.200; n—87). The measure of
productivity ranged from leaf and root growth and early senescence to
flower/seed production.
August 1991
1-22 DRAFT-DO NOT QUOTE OR CITE
-------
1 (2) The concept of a strict dose (concentration x time) does not work. The
2 effects are decidedly nonlinear; the slopes of the Figures 1-1 and 1-2
3 suggest that it may be power of 2 to 3.
4
5 (3) Under varied circumstances within the range of NO2 exposure given in
6 Figure 1-2, a given species will be either affected or not affected by NO2.
7 Not enough is known to determine precisely when a plant will be altered
8 by the exposure.
9
.0 (4) The majority of the data in Figure 1-2A and 1-2B suggest that
. 1 concentrations below 0.1 ppm for days to a month have little effect upon
.2 productivity. The data are less clear for very long exposures; it may be
.3 that very low concentrations over a year's exposure may be enough to
.4 cause ecological problems. The lack of data makes any conclusion
.5 premature.
.6
.7
8 1.3.1.3 Nitric Oxide
.9 The solubility of NO indicates that it does not react rapidly with water to form nitrite.
50 Most of the chemical studies of NO which indicate that it reacts rapidly with free radicals
51 have been carried out in non-polar solvents, and are therefore suspect. To be sure, there are
52 many biochemical reactions that occur via free radicals, and NO could easily react with free
53 radicals and alter normal metabolism. Yet under most conditions these critical free radical
14 reactions are heavily protected or tightly bound within enzymes. It may be that only at high
55 levels is there enough free NO present to initiate these damaging reactions. It is hard to
56 calculate what the level of NO would have to be in the atmosphere to build reactive
57 conditions of NO within the cell water since there are so many unknowns.
58 Like nitrite, NO can alter photosynthesis. The threshold exposure for inhibiting
59 photosynthesis in pat (Avena satavia) and alfalfa (Medicago sativd) was 0.6 ppm (NO) for
50 90 min. NO inhibited photosynthesis more rapidly than NO2, but the rate of recovery was
51 also faster. Data showing effects on productivity are sparse, which limits possible
52 conclusions. However, publications have shown that growth of several species/cultivars can
53 be impaired when exposed at concentrations of 0.3-4.0 ppm NO for 30 days or more.
54 ' " " ' " ' " •"•..'•-"••.•=•.••••••• •-
55 1.3.1.4 Pollutant Combinations
56 The exposure regime is an important consideration in evaluating studies in which plants
57 are exposed to pollutant combinations. The evaluation must consider not only the reported
August 1991 1-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 biological impact but also must determine whether the pollutant concentrations and their
2 individual and joint occurrences are reasonable in relation to concentrations and frequency of
3 occurrence monitored in the ambient air. Limited analyses of ambient air monitoring data
4 have evaluated the frequency of pollutant (NO2/SO2 and NO2/O3) co-occurrence (Lefohn and
5 Tingey, 1984; Lane and Bell, 1984; Jacobson and McManus, 1985; Lefohn et al., 1987). In
6 general, the studies have concluded that (a) the co-occurrence of two-pollutant mixtures lasted
7 only a few hours per episode, (b) the time between episodes is generally large (weeks,
8 sometimes months), and (c) the periods of co-occurrence represent a very small portion of the
9 potential plant growing period.
10 Several studies of NO2 in combination with other pollutants have reported that the
11 injury/growth threshold was significantly lowered by the pollutant combinations. The effects
12 of the combined gases were at least additive and in some cases greater-than-additive. The
13 experiments have been done under a very wide range of conditions, but, for the most part,
14 the pollutant concentrations were relatively high and the frequency of pollutant co-occurrence
15 was greater than typically monitored in the ambient air.
16
17 1.3.2 The Effects of Nitrogen Oxides on Natural Ecosystems and Their
18 Components
19 1.3.2.1 Ecosystems: Structure, Function, Response
20 Since the mid-1980's the view has emerged that the atmospheric deposition of inorganic
21 nitrogen has impacted terrestrial and aquatic ecosystems. It is known that in many areas of
22 the United States the atmospheric input of nitrogen compounds is significant (U.S.
23 Environmental Protection Agency, 1982a). Although the evidence linking nitrogen deposition
24 with ecological impacts is tenuous, there is a growing concern that has been magnified by:
25 (1) increased atmospheric concentrations of nitrogen compounds in North America and most
26 European countries, and (2) the possibility that ecosystems formerly limited by nitrogen have
27 become nitrogen saturated via atmospheric deposition (Skeffington and Wilson, 1988).
28 Nitrogen saturation of ecosystems has been defined as occurring: (1) when nitrogen
29 losses from ecosystems exceeds inputs (Brown et al., 1988) and (2) when the availability of
30 ammonium and nitrate exceeds the total combined plant and microbial nutritional demands
31 (Aber et al., 1989). The concerns regarding nitrogen saturation have led to efforts to develop
August 1991 1-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 "critical loads" of nitrogen for various ecosystems. A "critical load" is defined as:
2 "a quantitative estimated of an exposure to one or more pollutants below which significant
3 harmful effects on specified sensitive elements of the environment do not occur according to
4 present knowledge" (Nilsson and Grennfelt, 1988). The possible effects of nitrogen
5 deposition and saturation on various ecosystems are outlined below.
5 Ecosystems are basically energy and nutrient transfer systems. Nitrogen, one of the
7 main constituents of protein molecules essential to all life, is recycled within ecosystems.
B Nutrients and energy move through a system as organic matter. Energy moves
3 unidireetionally and ultimately dissipates into the environment. Nutrients are recycled into
3 the system. Nitrogen is unique among nutrients in that its retention and loss is regulated
1 almost exclusively by biological processes. Most organisms cannot use the molecular
I nitrogen found in the earth's atmosphere until it has been transformed into a utilizable form
3 by terrestrial or aquatic microorganisms. The transformations of nitrogen as it moves
1 through an ecosystem are referred to as the "nitrogen cycle" (Figure 1-3). Mature natural
5 ecosystems are essentially self-sufficient and independent of external additions. Modern
5 technology, by altering amounts of nitrogen moving through the cycle, may be upsetting the
7 relationships that exist among the various components and thus may be altering ecosystem
3 structure and function.
3 Air pollutants are known to alter the structure and diversity of plant communities.
) Because ecosystem components are chemically interrelated, stresses placed on individual
1 components, such as those resulting from nitrogen loading, can produce perturbations that are
I not readily reversed (Guderian et al., 1985). The response of an ecosystem to environmental
3 changes or perturbations is determined by the responses of its constituent organisms.
\ Individual species differ appreciably in their sensitivity to stresses; the changes that occur
5 within plant communities reflect such differences. The responses of populations of organisms
5 to environmental perturbation depends upon the genetic constitution (genotype) of their
7 individual members, their life cycles, and the microhabitats in which the organisms exist.
3 A common response in a community under stress is the elimination of the more sensitive
) populations and an increase in abundance of species that tolerate or are favored by the stress
) (Woodwell, 1970; Guderian et al., 1985). Those organisms able to cope with the stresses
1 survive and reproduce.
August 1991 1-25 DRAFT-DO NOT QUOTE OR CITE
-------
To Stratosphere
-I
»-»
vo
o\
O
3
8
/o
d
o
so
ir Photochemical
[6 Smog
Fixation by
Ughlnlng
Gaseous
Absorption w
Industrial
Fixation
Biological
Rxatlon
,anic x _
* Ammonification
f ^N^*Kfc**^1_
flxatlwt
by Algae
Detritus
3-*
Dertitrirfication
Figure 1-3. Schematic representation of the nitrogen cycle, emphasizing human activities that affect fluxes of nitrogen.
Source: Adopted from National Research Council (1978),
-------
1 1.3.2.2 Nitrogen Deposition
2 The removal (dry deposition) of reactive nitrogen gases from the atmosphere occurs
3 along several pathways leading to foliage, bark, or soil with pathways to foliage being
4 predominant during the growing season. The prevalence of any particular type of deposition
5 is a function of (1) the physicochemical properties of nitrogen compounds, (2) their ambient
6 concentration, and (3) the presence of suitable receptor sites in the landscape (e.g., leaves
7 with open stomata).
8 With the possible exception of nitric acid vapor, deposition characteristics of reactive
9 nitrogen compounds are highly variable and dramatically influenced by environmental
10 conditions that affect stomatal conductance. The close relationship between stomatal
11 conductance and the deposition of NO and NO2 implies that gaseous deposition of reactive
.2 nitrogen oxides is greatly reduced in the dark when stomata close (Hanson et al., 1989; Saxe,
,3 1986; Hutchinson et al., 1972).. Deposition of gaseous nitrogen forms is usually proportional
A to ambient concentrations, but "compensation concentrations" at which no uptake occurs (i.e.,
.5 < 0.003-0.005 ppmv) have been reported for NO2 and NH3. Data for NO, NO2, and HNO3
.6 (Grennfelt et al., 1983; Johansson, 1987; Marshall and Cadle, 1989; Skarby et al., 1981),
.7 during the period vegetation is dormant, show a reduced potential for deposition.
.8 Conversely, paniculate nitrate and ammonium deposition do not appear to be affected by the
.9 season of the year (Gravenhorst et al., 1983; Lovett and Lindberg, 1984).
!0 Daytime rates of nitrogen oxide or ammonia deposition can be approximated from
\l ambient concentrations of the gases (U.S. Environmental Protection Agency, 1982; Hicks
12 et al., 1985). Hanson et al. (1989) predicted NO2-N inputs between 0.04 and 1.9 kg N ha'1
>3 yr"1 for natural forests and inputs up to 12 kg N ha"1 yr"1 for forests in urban environments.
14 For a forested watershed, Grennfelt and Hultberg (1986) calculated the annual deposition of
15 NO2 plus HNO3 to be in the range from 3.6 to 5.1 kg N ha"1 yr"1. Hill (1971) estimated the
',6 removal of NO2 from the atmosphere in Southern California to be approximately 109 kg N
17 ha"1 yr-1. - '
18 Preliminary particle deposition measurements and calculated dry deposition estimates of
'.9 reactive nitrogen gases, indicate significant nitrogen inputs to terrestrial systems. Barrie and
iO Sirois (1986) estimated that dry deposition contributed 21 to 30% of total NO3" deposition in
il eastern Canada. Lovett and Lindberg (1986) concluded that dry deposition of nitrate is the
August 1991 1-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 largest form of inorganic nitrogen deposited to oak-hickory forests in eastern Tennessee.
2 Annual estimates of NH3 deposition have been reported (Cowling and Lockyer, 1981;
3 Sinclair and van Houtte, 1982), but numerous reports of NH3 evolution from foliage under
4 conditions of high soil nitrogen confound simple estimates of annual NH3-N deposition.
5 Lovett (1991) summarized research data for a number of forested sites in North America and
6 Norway and concluded that dry deposition of nitrogen typically occurs at annual rates
7 approximately equal to nitrogen deposited in precipitation.
8 Because gaseous deposition is difficult to measure accurately or continuously at the
9 landscape level of resolution, estimates of dry nitrogen deposition must rely on models.
10 Rigorous models of pollutant deposition have been developed (Hicks et al., 1985; Baldocchi,
11 1988; Baldocchi et al., 1987), and will be needed in the future for accurate determination of
12 reactive nitrogen gas and particle deposition to forest stands and ecosystems.
13 Increased efforts have been made to establish both wet and dry deposition rates of
14 nitrogen to various types of ecosystems. These current deposition data are important as they
15 provide a basis to evaluate potential effects against "suggested critical levels". Although the
16 concept of critical nitrogen loading has not yet been widely adopted in North America (based
17 on amount of published data), a comparison of total nitrogen deposition data for North
18 America and proposed critical loads provides a reasonable comparison of the status of
19 terrestrial systems with respect to changes expected from elevated levels of nitrogen
20 deposition. Table 1-1 summarizes information regarding the total (wet and dry) deposition of
21 nitrogen to a variety of ecosystems/forest types or regional areas in North America and
22 Europe.
23
24 1.3.2.3 Effect of Deposited Nitrogen on Forest Vegetation and Soils
25 The effects of nitrogen deposition upon biological systems must be viewed from the
26 perspective of the amount of nitrogen in the system, the biological demand for nitrogen, and
27 the amount of deposition. If nitrogen is deposited on an nitrogen-deficient ecosystem, a
28 growth increase will likely occur. If nitrogen is deposited on an ecosystem with adequate
29 supplies of nitrogen, nitrate leaching will eventually occur. Nitrate leaching is usually
30 deemed undesireable in that it can contaminate groundwater and lead to soil acidification.
31 Agricultural lands are excluded from this discussion because crops are routinely fertilized
August 1991 1-28 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
1
TABLE 1-1. MEASUREMENTS OF VARIOUS FORMS OF ANNUAL NITROG
DEPOSITION TO NORTH AMERICAN AND EUROPEAN ECOSYSTEMS.
MEASUREMENTS OF TOTAL DEPOSITION DATA THAT DO NOT INCLUDE
BOTH A WET AND DRY ESTIMATE PROBABLY UNDERESTIMATE TOTAL
NITROGEN DEPOSITION AND ARE ENCLOSED IN PARENTHESES
Site
Location/Vegetation
United States
CA, Chaparral
CA, Sierra Nevada
GA, Loblolly pine
NC, Hardwoods
NC, White pine
NY, Red spruce
TN, Oak forest Wl
WA, Douglas fir
U.S. Regions
Adirondacks
Canada
Alberta (southern)
British Columbia
Ontario
Ontario (southern)
Fed. Rep. Germany
Spruce (SE slope)
Netherlands
Oak-birch
Douglas fir
Douglas fir
Norway
Spruce
Forms of N
Wet
Cloud Rain
8.2
_
3.7
4.8
3.7
7.3 6.1
6.0
1
6.3
. - . 7.3
5.5
3.7
•1 -3
16.5
—
._ •
19.3
10.3
Deposition (Kg ha"1 yr"1)
Dry
Particles Gases
..
•
1.0 4.2
0.5
0.9 2.7
0.2 2.3
1.2
_ _
4.7
12.2°
._'
_.
1.4
„,
-
._
95.7°
0.7 0.2
Total
23b
(2)
9
5.3
7
16
7
(1)
11
19.5
(5)
(4)
3.7
16.5
24-56b
17-64b
115
11.2
*—Symbolizes data not available or in the case of cloud deposition not present.
"Total nitrogen deposition was based on bulk deposition and throughfaU measurements and does include
components of wet and dry deposition.
Includes deposition from gaseous forms.
with amounts of nitrogen (100-300 kg/ha) that far exceed pollutant inputs even in the most
heavily polluted areas. Biological competition for atmospherically-deposited nitrogen among
heterotrophie microorganisms, plants, and nitrifying microorganisms combined with the
August 1991
1-29
DRAFT-DO NOT QUOTE OR CITE
-------
reactions between NH4+ and humus in the soil determine the degree to which ,
2 increase in vegetation growth will occur and the degree to which incoming nitrogen is
3 retained within the ecosystem. Temperature, moisture, availability of other nutrients and
4 stand age influence vegetation demand for nitrogen. Uptake rates decline as forests mature,
5 especially after the buildup of nutrient-rich foliar biomass ceases following crown closure.
6 Processes such as fires and harvesting will naturally decrease ecosystem nitrogen saturation or
7 even induce nitrogen deficiency. Intense fire causes a large loss of nitrogen, while frequent
8 fires of low intensity may have little effect.
9 Forest fertilization has proven quite successful in producing growth increases in
10 nitrogen-deficient forests, even though trees typically recover only 5-50% of nitrogen
11 fertilizer. (The functioning of enzyme systems frequently limits metabolic uptake. See
12 Chapter 9.) Fertilization studies, however, differ from pollutant nitrogen deposition in
13 several important respects: (1) pollutant nitrogen enters the ecosystem at the canopy level
14 whereas fertilizer is typically (but not always) applied to the soil, (2) fertilization leads to
15 high concentrations of NH4+ and, in the case of urea, high pH, both of which are conducive
16 to non-biological reactions between soil humus and NH4+, and (3) pollutant nitrogen
17 deposition enters ecosystems at a slow, steady input hi rather low concentrations, whereas
18 fertilizer is typically applied in. 1 to 5 large doses. Both plants and nitrifying bacteria are
19 favored by slow, steady inputs of nitrogen, possibly giving them an competitive advantage
20 over heterotrophic microorganisms.
21 Nitrification and NO3" leaching become significant only after heterotroph and plant
22 demand for nitrogen are substantially satisfied, a condition that has been referred to as
23 "nitrogen-saturated". Additions of nitrogen in any biologically-available form (NH4+, NOg",
24 or organic) to a nitrogen-saturated system will cause equivalent leaching of NO3", except in
25 those very rare systems where nitrification is inhibited by factors other than competition from
26 heterotrophs and plants. Considering the effects of NO3" only will result in a substantial
27 underestimation of the acidification potential of atmospheric deposition in nitrogen-saturated
28 ecosystems. ' ,
29 Since nitrification results in the creation of nitric acid within the soil, there are concerns
30 that elevated nitrogen inputs to nitrogen-saturated systems will result in soil acidification and
31 aluminum mobilization. There are very few proven documented cases in which excessive
August 1991 1-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 atmospheric nitrogen deposition has caused soil acidification (e.g., in forests in The
2 Netherlands subject to very high nitrogen deposition levels), but there is no doubt that the
3 potential exists for many mature forests with low uptake rates, given high enough inputs for a
4 sufficiently long time.
5 Soil acidification is usually thought of as an undesireable effect, but in some cases, the
6 benefits of alleviating nitrogen deficiency clearly outweigh the detriments of soil acidification
7 (e.g.1, the benefits of excessive nitrogen-fixation by red aider always outweigh the detriments
8 of concurrent soil acidification as exhibited by the superior growth of succeeding Douglas fir
9 stands in the Pacific Northwest). , , •
10 Increased concentrations of NO3" or any other mineral acid anion (e.g., SO^', or Cl")
LI in soil solution lead to increases in the concentrations of all cations in order to maintain
12 charge balance in solution. Equations describing cation exchange in soils dictate that as the
13 total anion (and cation) concentrations increase, individual cation concentrations increase as
14 follows: Al3* > Ca2+, Mg2+ > K+, Na+, H+. Thus, soil solution A13+ concentrations
15 increase not only as the soil acidifies (i.e., as the proportion of A13+ on the exchange
L6 complex increases) but also as the total ionic concentration of soil solution increases.
17 Van Breemen et al. (1982, 1987) in The Netherlands and Johnson et al. (in press) in the
18 Smoky Mountains of North Carolina noted NO3" - A13+ pulses in soil solutions from forest
19 sites. Aluminum toxicity is one of several nitrogen-related hypotheses posed to explain forest
$ decline in both countries. Other hypotheses include weather extremes and climate change,
II Mg and K deficiencies which occur in sites naturally low in these nutrients, and foliar
12 damage due to acid mist. Researchers on aquatic effects of acid deposition have long noted
13 springtime pulses of NO3", A13 + , and H+ in acid-affected surface waters of the Northeastern
»4 U.S. (Galloway et al., 1980; Driscoll et al., 1989).
>5 . ' • • •'.'.....-- • -••.-.•;-...•.• =
',6 1.3.2.4 Effects of Nitrogen on Sensitive Terrestrial Vegetation
17 Very little information is available on the direct effects of nitric acid vapor on
!8 vegetation and essentially no information on its effects on ecosystems. The effects of
19 ammonia, a reduced nitrogen gas, have been summarized by van der Eerden (1982),
50 however, ammonia concentrations seldom reach phytotoxic levels in the United States (U.S.
51 Environmental Protection Agency, 1982a).
August 1991 1-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 Because current ambient concentrations of NO, NO2 and NH3 are low across much of
2 the United States except in certain highly populated urban areas, significant direct effects of
3 these nitrogen compounds on ecosystems seems unlikely at the current time. Concentration
4 and effects data are unavailable for making similar conclusions regarding other reactive
5 nitrogen compounds like nitric acid vapor or the gaseous nitrate radical.
6 Serious consideration is currently being given to hypotheses that excess total nitrogen
7 deposition may impact plant productivity directly or through changes in soil chemical
8 properties. Furthermore it has been proposed that excess nitrogen deposition to ecosystems
9 may be modifying interplant competitive balances leading to changes in species composition
10 and/or diversity.
11 Excessive nitrogen inputs to terrestrial ecosystems can cause differential competitive
12 advantage among plants within a heathland (Heil and Bruggink, 1987; Heil et al., 1988). In
13 unmanaged heathlands in The Netherlands, Calluna vulagris is being replaced by grass
14 species, as a consequnce of the eutrophic effect of acidic rainfall and large nitrogen inputs
15 arising from intensive farming practices in the region. Roelofs et al. (1987) also observed
16 that nitrophilous grasses (Molinia and Deschampsid) were displacing slower growing plants
17 (Erica and Calluna) on heathlands in The Netherlands, and suggested that a correlation
18 existed between this change and nitrogen loading. Van Breemen and van Dijk (1988) found a
19 substantial displacement of heathland plants by grasses from 1980 to 1986 and noted increases
20 in nitrophilous plants in forest herb layers. Ellenberg (1988) suggested that ionic inputs
21 (NO3~ and NH4+) influence competition between organisms long before toxic effects appear
22 on individual plants. These changes in The Netherlands have occurred under nitrogen
23 loadings of between 20 and 60 kg N ha"1 yr"1. Liljelund and Torstensson (1988) have shown
24 clear signs of vegetation changes in response to nitrogen deposition rates of 20 kg ha"1 yr"1.
25
26 1.3.2.5 Nitrogen Saturation, Critical Loads, and Current Deposition
27 Ecosystem nitrogen saturation and the definition of the critical nitrogen deposition levels
28 have been the subject of recent conferences (Nilsson and Grennfelt, 1988; Brown et al.,
29 1988; Skeffington and Wilson, 1988). The Workshop held at Skokloster, Sweden in March
30 1988 (Nilsson and Grennfelt, 1988) adopted the following definition for a critical load:
31 "A quantitative estimate of an exposure to one or more pollutants below which significant
August 1991 1-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 harmful effects on specified sensitive elements of the environment do not occur according to
2 present knowledge". The aim of a nitrogen critical load was "to protect soils from long-term
3 chemical changes with respect to base saturation: (p. 17, Nilsson and Grennfelt, 1988;
4 p. 451, Schulze et al., 1989).
5 Brown et al. (1988), however, reported that a recent workshop, based on a model of
6 plant/soil nitrogen saturation put forth by Agren and Bosatta (1988), concluded that nitrogen
7 saturation could best be defined as occurring when nitrogen losses from ecosystems exceeded
8 inputs. Aber et al. (1989) similarly define nitrogen saturation as the availability of
9 ammonium and nitrate in excess of total combined plant and microbial nutritional demands.
10 The concept of nitrogen saturation, it is hoped, will make it possible to define a critical
11 nitrogen load (deposition rate) at which no change or deleterious impacts will occur to an
12 ecosystem (Nilsson, 1986). It is important to recognize that the magnitude of such a "critical
13 load" will be site and species specific because it is highly dependent on initial soil chemistries
14 and biological growth potentials (i.e., nitrogen demands). Skeffmgton and Wilson (1988)
15 point out that intrinsic in all definitions of a "critical load" is the notion that there is a load at
16 which no long-term effects occur. The complexity of the nitrogen cycle and ecosystem
17 diversity make defining a critical load for nitrogen very difficult.
18 Schulze et al. (1989) proposed critical loads for nitrogen deposition based on an
19 ecosystem total anion and cation balance. This approach makes the assumption that processes
20 determining ecosystem stability are related to soil acidification and nitrate leaching (see
21 Section 10.3.6). They concluded that in order to limit the mobilization of aluminum and
22 other heavy metals resulting from acidification and nitrate leaching (a negative result), critical
23 nitrogen deposition rates could not exceed 3-14 Kg nitrogen ha"1 yr"1 for silicate soils or 3 to
24 48 kg nitrogen ha"1 yr"1 for calcareous based soils. Other critical loads have been proposed
25 at rates of nitrogen deposition ranging from as little as 1 to levels near 100 Kg nitrogen ha"1
26 yr"1 depending on the impacts considered acceptable and the criteria used to define the critical
27 load.
28 Using criteria to minimize species change, critical loads less than 30 kg ha"1 yr"1 have
29 been proposed (van Breeman and van Dijk, 1988; Liljelund and Torstensson, 1988). When
30 using the criteria that ecosystem nitrogen inputs should not exceed outputs, critical loads have
31 been proposed as low as 1-5 kg nitrogen ha"1 yr"1 for poor sites with low productivity or in
August 1991 1-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 the range from 5-30 kg nitrogen ha"1 yr"1 for sites having medium quality soils and for
2 common forested systems (Boxman et al., 1988; Rosen, 1988; Skeffington and Wilson, 1988;
3 World Health Organization, 1987).
4 In summarizing the results of a recent conference on critical nitrogen loading, after
5 discussing various options for setting a critical nitrogen load, Skeffington and Wilson (1988)
6 concluded that "we do not understand ecosystems well enough to set a critical load for •
7 nitrogen deposition in a completely objective fashion." Brown et al. (1988) further concluded
8 that there was probably no universal critical load definition that could be applied to all
9 ecosystems, and a combination of scientific, political, and economic considerations would be
10 required for the application of the critical load concept. '
11 The following terrestrial ecosystems have been suggested as being at risk from the
12 deposition of nitrogen-based compounds:
13 • heathlands with a high proportion of lichen cover, •
14 • low meadow vegetation types used for extensive grazing and
15 haymaking, and " ........
16 • coniferous forests, especially those at high altitudes (World Health
17 Organization, 1987).
18 ' - ' ' ' • •
19 The above oligotrophic ecosystems are considered at risk from atmospheric nitrogen
20 deposition because plant species normally restricted by low nutrient concentrations could gain
21 a competitive advantage, and their growth at the expense of existing species would change the
22 "normal" species composition and displace some species entirely (Ellenberg, 1987; Waring,
23 1987). Sensitive natural ecosystems, unlike highly manipulated agricultural systems, may be
24 prone to change from exposure to dry deposited nitrogen compounds because processes of
25 natural selection whereby tolerant individuals survive may not be keeping pace with the
26 current levels of atmospheric nitrogen deposition (World Health Organization, 1987);
27 There is little doubt that nitrogen deposition has a prounced effect on many if not most
28 terrestrial ecosystems. Since most forest ecosystems in North America are nitrogen deficient,
29 one of the most noticeable initial changes in response to increased nitrogen deposition is.
30 likely to be a growth increase. Whether this growth increase is deemed desireable or
31 undesireable in a particular ecosystem is entirely a matter of management objectives (timber
32 production or wilderness preservation), and, ultimately, value judgements by society.
August 1991 1-34 DRAFT-DO NOT QUOTE OR CITE
-------
1 Application of the concept of critical nitrogen loading has not yet been widely adopted
2 in North America (based on amount of published data), but a comparison of total nitrogen
3 deposition data for North America and proposed critical loads just discussed should provide a
4 reasonable comparison of the status of terrestrial systems with respect to changes expected
5 from elevated levels of nitrogen deposition. Much of our information on nitrogen deposition
6 is limited to information on nitrate and ammonium deposition in rainfall. Lindberg et al.
7 (1987) concluded that the lack of data on multiple forms of nitrogen deposition limits our
8 ability to accurately determine current levels of nitrogen loading. •
9 Olsen (1989) summarized nitrate and ammonium concentration and wet deposition data
10 for the United States and southern Canada for the period from 1979 through 1986. For
11 1986, the greatest annual rates of ammonium and nitrate deposition were localized in the
12 northeastern United Sates and southern Canada (Olsen, 1989). Peak values were 5 and
13 25 kg ha"1 yr"1 for ammonium and nitrate, respectively. Similar wet deposition data for 1987
14 showed peak deposition rates of 3.5 and 16 kg ha"1 yr"1 for ammonium and nitrate,
15 respectively (National Atmospheric Deposition Program, 1988). Zemba et al. (1988)
16 summarized wet nitrate deposition data from 77 stations located in Eastern North America ,
17 and found that peak nitrate deposition (>20 kg ha"1 yr"1) occurred between lakes Michigan '
18 and Ontario. They also found the temporal pattern of nitrate deposition was quite even «
19 throughout the year (Schwartz, 1989). Wet deposition of ammonium (NH4+) in Europe ^
20 ranges between 3.5 and 17.3 kg NH4+ ha"1 yr"1 (Buijsman and Erisman, 1987; Heil et al.,
21 1987). Boring et al. (1988) have also published an extensive review of the sources, fatesrand
22 impacts of nitrogen inputs to terrestrial ecosystems. '
23 Based on the current rates of nitrogen deposition (loading) occuring in North America
24 (Sections 10,3.6 and 10.4.3.1) one might conclude that current rates of nitrogen deposition in
25 North America are sufficient to induce minor changes in some ecosystems (i.e., rates of
26 deposition in North America exceed some of the proposed critical load levels). However,
27 because ecosystems have a variable capacity to buffer changes caused by elevated inputs of
28 nitrogen, it is difficult to make general conclusions about the type and extent of change
29 currently resulting from nitrogen deposition in North America. Furthermore, current
30 estimates of total nitrogen deposition to ecosystems and regions of the United States usually
31 do not account for gaseous nitrogen losses from ecosystems (e.g., N2O and NH3); therefore,
August 1991 1-35 DRAFT-DO NOT QUOTE OR CITE
-------
1 the estimates of total nitrogen deposition may be overestimated (Wetselaar and Farquhar,
2 1980; Bowden, 1986; Anderson and Levine, 1987; Schimel et al., 1988).
3 - -
4 1.3.2.6 Effects of Nitrogen on Wetlands and Bogs
5 The anaerobic (oxygen-free) nature of their waterlogged soils is the feature that sets
6 wetlands apart. Anaerobic wetland soils favor the accumulation of organic matter and losses
7 of mineral nitrogen to the atmosphere through denitrification reactions (the conversion of
8 nitrate to gaseous nitrogen by microbes). Nitrogen deposition can impact plant and microbial
9 processes either directly or indirectly by acidifying the environment. An increase in nitrogen
10 supply through atmospheric deposition or other means alters the competitive relationships
11 among plant species such that fast growing nitrophilous species (species that have a high
12 nitrogen requirement) are favored. Microbial rates of decomposition, nitrogen fixation (the
13 conversion of gaseous nitrogen to ammonium), nitrification (the conversion of ammonium to
14 nitrate), and dissimilatory nitrate reduction (conversion to gaseous nitrogen or ammonium)
15 are all affected. Acidification below pH 4.0 to 5.7 blocks the nitrogen cycle by inhibiting
16 nitrification, and the accumulation of NH4+ (ammonium) in the environment represses
17 nitrogen fixation (Roelofs, 1986; Schuurkes et al., 1986, 1987; Rudd et al., 1988). The
18 proportion of ^O (nitrous oxide) produced by denitrification reactions increases with
19 decreasing pH below 7, and the absolute rate of production of N2O increases with increasing
20 eutrophication (nutrient enrichment of the environment) (Focht, 1974). This is potentially
21 important on a global scale because of chemical reactions with N2O in the atmosphere that
22 result in a loss of ozone.
23 The importance of atmospheric nitrogen deposition to the community structure (species
24 composition and interrelationships) of wetlands increases as rainfall increases as a fraction of
25 the total water budget. Primary production (plant growth) in wetlands is commonly limited
26 by nitrogen availability. Primary production is proportional to the rate of internal nitrogen
27 cycling, which is influenced by the quantity of mineralizable soil nitrogen as well as the
28 supply of nitrogen to the ecosystem from the atmosphere or surface flow. Total nitrogen
29 inputs range from about 10 kg N ha"1 yr"1 in ombrotrophic bogs (rain-fed bogs), which
30 receive water only through precipitation, to 750 kg N ha"1 yr"1 or more in intertidal wetlands
31 with large ground and surface hydrologic inputs.
August 1991 1-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 From studies of nine North American wetlands, bulk nitrogen deposition ranges from
2 5.5 to 12 kg nitrogen ha"1 yr"1 and occurs in the form of NO3" (nitrate), NH4+ (ammonium),
3 and dissolved organic nitrogen in roughly equal proportions. More recent studies, however,
4 suggest that these rates are too low and that the wet deposition of NO^" alone is greater than
5 15 kg nitrogen ha"1 yr"1 over much of eastern North America (Zemba et al., 1988). Dry
6 deposition, which probably accounts for greater than 50% of total deposition, adds to the
7 total. Leaf-capture of nitrogen in fog droplets is a third form of deposition that is locally
8 important. Applications of nitrogen fertilizer in the field, ranging from 7 to
9 3,120 kg nitrogen ha"1 yr"1, have increased standing biomass by 6 to 413%. Other nutrients,
10 like phosphorus, become secondarily limiting to primary production after nitrogen inputs
11 reach a threshold. Fertilization and increased atmospheric deposition have increased the
12 dominance of grass species over other plant species in bogs, and extreme eutrophication is
13 associated with a decrease in plant species diversity.
14 Single additions to vegetated wetland soils of 15N-labelled mineral nitrogen at rates of
15 about 100 kg nitrogen ha"1 yr"1 indicate that up to 93% of applied NH4+ is rapidly
16 assimilated into organic matter within a single growing season. The majority of the labelled
17 nitrogen is lost from the system after 3 years by the combined processes of advective
18 transport in water (carried in moving water) of particulate organic matter, advective and
19 diffusive transport of dissolved nitrogen, and denitrifieation. In the absence of plants, the
20 major fate of inorganic nitrogen applied to wetland soils is loss to the atmosphere by
21 denitrifieation.
22 Peat-forming Sphagnum spp. are largely absent from bogs in western Europe where
23 bulk deposition rates are about 20-40 kg nitrogen ha"1 yr"1, and soft water communities once
24 dominated by isoetids in The Netherlands have been converted to later successional stages
25 dominated by Juncus spp. (rush) and Sphagnum spp. or to grasslands. Heathlands dominated
26 by shrubs have also been converted to grasslands. Experimental studies indicate that
27 ombrotrophic bogs can be maintained if nitrogen inputs are less man 20 kg nitrogen ha"1 yr"1.
28 Increased productivity associated with eutrophication is accompanied by increased rates of
29 transpiration (evaporation of water from leaf surfaces) which can alter wetland hydrology and
30 influence the direction of wetland succession. By this mechanism, one modeling study
31 suggests that a succession (change) from open ombrotrophic bog to forested wetland occurs
August 1991 1-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 . when a threshold of 7 kg nitrogen ha"1 yr'1 is exceeded. These estimates are consistent with
2 conclusions from studies of species distributions that place the limit for many species from
3 10 to 20 kg nitrogen ha"1 yr"1 (Liljelund and Torstensson, 1988).
4 Fourteen percent (18 species) of the plant species from the conterminous United States
5 that are formally listed as endangered, and an additional 284 species listed as potentially
6 threatened (Code of Federal Regulations, 1987), are found principally in wetland habitats.
7 Some of the endangered plants, like the green pitcher plant, are known to be adapted to
8 infertile habitats and are threatened by current levels of nitrogen deposition in parts of North
9 America. Plant species that are threatened by high nitrogen deposition are not confined to
10 wetland habitats, however, but are common across many ecosystem types (Ellenberg, 1988).
11
12 1.3.2.7 Effects of Nitrogen on Aquatic Systems
13 Nitrogen deposition has not historically been considered a serious threat to the integrity
14 of aquatic ecosystems.
15 Assessment of the aquatic effects of nitrogen oxides depends on a close examination of
16 the processes by which nitrogen may enter streams, lakes and estuaries. Sources of nitrogen
17 may include: (1) atmospheric deposition directly to the water surface; (2) deposition to the
18 watershed that is subsequently routed to the drainage waters; (3) gaseous uptake by plants that
19 is subsequently routed, by way of litter fall and decomposition, to drainage waters; and
20 (4) nitrogen fixation, either in the water itself, or in watershed soils. In addition, numerous
21 processes act to transform nitrogen species into forms that are only indirectly related to the
22 original deposition or fixation. These transformations include: (1) nitrogen assimilation (the
23 biological uptake of inorganic nitrogen species); (2) nitrification (the oxidation of ammonium
24 to nitrate); (3) denitrification (the biological reduction of nitrate to form gaseous forms of
25 nitrogen, N2, NO, or ^O); and (4) mineralization (the decomposition of organic forms of
26 nitrogen to form ammonium). The multiple sources of nitrogen to aquatic systems, and the
27 complexities of nitrogen transformations in water and watersheds, have the effect of
28 de-coupling nitrogen deposition from nitrogen effects, and reduce our ability to attribute
29 known aquatic effects to known rates of nitrogen deposition. While it is not currently
30 possible to trace the pathway of nitrogen from deposition through any given watershed and
31 into drainage waters, we can, in areas of the United States where non-atmospheric sources of
/
August 1991 1-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 nitrogen are small, begin to infer cases where nitrogen deposition is having an impact on
2 aquatic ecosystems.
3 Any discussion of the aquatic effects of nitrogen oxides must focus on the concept of
4 nitrogen saturation. Under conditions of nitrogen saturation, forested watersheds that
5 previously retained nearly all of nitrogen inputs, due to a high demand for nitrogen by plants
6 and microbes, begin to supply more nitrogen to the surface waters that drain them. Our
7 conceptual understanding of nitrogen saturation suggests that, in aquatic systems, the earliest
8 stages of nitrogen saturation will be observable as increases in the severity and duration of
9 springtime pulses of nitrate. .:
10 The aquatic effects of nitrogen oxides can be divided into three general categories:
11 (1) acidification, both chronic and episodic; (2) eutrophication of both freshwaters and
12 estuaries; and (3) directly toxic effects.
13
14 Acidification
15 Acidification effects are traditionally divided into chronic (long-term) and episodic
16 (short-term effects usually observable only during seasons of high runoff) effects. Nitrate,
17 the dominant form of inorganic nitrogen in almost all aquatic systems, is commonly present
18 in measurable concentrations only during winter and early spring, when terrestrial demand for
19 nitrogen is low because plants in the watershed are dormant. Nitrogen will therefore only be
20 a problem in chronic acidification in rare cases where the process of nitrogen saturation is
21 very much progressed. Chronic acidification by nitrogen can be conclusively demonstrated
22 only in parts of Europe (e.g., Hauhs, 1989; Hauhs et al., 1989; van Breemen and van Dijk,
23 1988).
24 Episodic acidification by nitrate is far more common than chronic acidification, and is
25 well-documented for streams (Driscoll et al., 1987b) and lakes (Galloway et al,, 1980;
26 Driscoll et al., 1991; Schaefer et al., 1990) in the Adirondack Mountains, for streams in the
27 CatsMll Mountains (Stoddard and Murdoch, 1991; Murdoch and Stoddard, in review), and in
28 a small proportion of lakes in Vermont (Stoddard and Kellogg, in press), as well as in many
29 parts of Canada (Jeffries, 1990) and Europe (e.g., Hauhs et al., 1989).
30 Based on intensive monitoring data, it is possible to divide lakes and streams into three
31 groups, based on their seasonal NO3" behavior. In many parts of the country, nitrogen
August 1991 1-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 demand on the part of the terrestrial ecosystem is sufficiently high that no leakage of NO3"
2 from watersheds occurs, even when nitrogen deposition rates are relatively high, and cold
3 temperatures should limit the biological demand for nitrogen. Lakes and streams in these
4 areas show no evidence that nitrogen deposition is producing adverse aquatic effects.
5 In a second group of lakes and streams, NO3" concentrations show strong seasonality,
6 with peak concentrations during snow melt, or following large rain events. In many cases,
7 these episodic increases in NO3", along with already low baseline acid neutralizing capacity
8 (ANC), are sufficient to cause short-term acidification and potential adverse biological
9 effects.
10 The third group of lakes and streams exhibits both the strong seasonality in NO3~
11 concentration described in the previous paragraph, and increasing trends in NO3"
12 concentrations. Because the early stages of nitrogen saturation are expected to produce
13 increases in NO3" concentrations, especially during episodes, long-term increases in NO3"
14 may represent the strongest evidence that nitrogen deposition is responsible for aquatic
15 effects. In all cases where increasing trends in NO3" have been documented in the
16 United States (Smith et al., 1987; Stoddard and Murdoch, 1991; Murdoch and Stoddard, in
17 review; Driscoll et al., in review) they have occurred at a time when nitrogen deposition is
18 relatively constant (e.g., Simpson and Olsen, 1990). Increased leakage of NO3" from
19 watersheds in these areas therefore represents a long-term decrease in the ability of
20 watersheds to retain nitrogen. A likely cause of such long-term changes is a lowering in the
21 demand for nitrogen as a nutrient on the part of the terrestrial ecosystem, which may result
22 from long-term high rates of nitrogen deposition to affected watersheds (e.g., Aber et al.,
23 1989), forest maturation (Elwood et al., 1991), or, more likely, a combination of both
24 factors.
25 The locations of lake and streams sites in each of the three NO3" groups are shown on
26 maps of the Northeast (Figure 1-4) the Southeast (Figure 1-5) and the West (Figure 1-6).
27 The maps illustrate the existence of severe problems in the Northeast (especially the
28 Adirondack and Catskill Mountains), and the Southeast (in the Mid-Appalachians and Great
29 Smoky Mountains), and the potential for future problems in the West.
August 1991 1-40 DRAFT-DO NOT QUOTE OR CITE
-------
o Data Indicate no Influence of NO,
* Data indicate strong influence of NO
* Data indicate strong influence of NO
and Increasing trend In NO3
Figure 1-4, Location of acid-sensitive lakes and streams in the northeastern United
States where the importance of NO%~ to seasonal water chemistry can be
determined. Data from: Kahl et al. (1991); Wigington et al. (1989);
Driscoll et al. (1987a); Driscoll et al. (in review); Kramer et al. (1986);
Murdoch and Stoddard (in review); Eshleman and Hemond (1985); Morgan
and Good (1988); Baird et al. (1987); Likens (1985); Sharpe et al. (1984);
Stoddard and Kellogg (in press); DeWalle et al. (1988); Barker and Witt
(1990); Schofield et al. (1985); Phillips and Stewart (1990).
1 It is also possible to draw correlations between rates of nitrogen deposition, and rates of
2 nitrogen loss from watersheds; while these analyses cannot indicate causal relationships, they
3 can suggest patterns that merit further attention. Two independent attempts have been made
4 to relate deposition and watershed nitrogen export in the United States, and both suggest
5 similar conclusions. The results of the correlation suggest a strong correspondence between
6 median wet deposition of nitrogen in a region, and the median spring base-flow concentration
August 1991
1-41
DRAFT-DO NOT QUOTE OR CITE
-------
O Data indicate no Influence of NOj
* Data indicate strong influence of NO *
•*• Data Indicate strong Influence of NO
and increasing trend in NOj
Virginia
_--—~
North Carolina
-^—_>•—v
South Carolina
Figure 1-5. Location of acid-sensitive lakes and streams in the southeastern United
States where the importance of NO^" to seasonal water chemistry can be
determined. Data from: Elwood et al. (1991); Cosby et al. (1991); Elwood
and Turner (1989); Buell and Peters (1988); Swank and Waide (1988);
Jones et al. (1983); Silsbee and Larson (1982); Katz et al. (1985); Weller
et al. (1986); Wigington et al. (1989); Kramer et al. (1986); Edwards and
Helvey (1991).
1 of nitrogen in a region. In addition, the results suggest a threshold rate of wet nitrogen
2 deposition of approximately 2.8 kg nitrogen ha"1 yr"1 above which significant losses of
3 nitrogen from watersheds can begin to occur. Driseoll et al. (1989a) collected input/output
4 budget data for a large number of undisturbed forested watersheds in the United States and
5 Canada, and summarized the relationship between nitrogen export (of NO3~) and wet nitrogen
August 1991
1-42
DRAFT-DO NOT QUOTE OR CITE
-------
o Data indicate no influence of
* Data indicate strong influence of
* Data indicate strong influence of
and increasing trend In NOg
N
Figure 1-6. Location of acid-sensitive lakes and streams in the western United States
where the importance of NO3" to seasonal water chemistry can be
determined. Data from: Melack and Stoddard (1991); Stoddard (1987a);
Loranger et al. (1986); Wigington et al. (1989); Kramer et al. (1986); Welch
et al. (1986); Eilers et al. (1990); Gilbert et al. (1989).
1
2
3
deposition (of NO3" + NH4+). Like the data of Kaufmann et al. (1991), these budget data
suggest a threshold rate of wet nitrogen deposition of ca. 2.8 kg nitrogen ha"1 yr"1 above
which significant export of NO3" from watersheds may occur.
August 1991 1-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 Eutrophicafion
2 Assigning responsibility for the eutrophieation of lakes and estuaries to nitrogen oxides
3 requires a determination of two key conditions: (1) productivity of the aquatic system is
4 limited by the availability of nitrogen, (2) nitrogen deposition is a significant source of
5 nitrogen to the system. In many cases of eutrophieation, the supply of nitrogen from
6 atmospheric deposition is minor when compared to other anthropogenic sources, such as
7 pollution from either point or non-point sources.
8 It is generally accepted that the productivity of fresh waters is limited by the availability
9 of phosphorus, rather than the availability of nitrogen (reviewed by Hecky and Kilham,
10 1988). Conditions of nitrogen limitation do occur in lakes, but are often either transitory, or
11 the result of high inputs of phosphorus from anthropogenic sources. Often when nitrogen
12 limitation does occur it is a short-term phenomenon, because nitrogen-deficient conditions
13 favor the growth of nitrogen-fixing blue green algae (e.g., Smith, 1982). It appears that
14 nitrogen limitation may occur naturally (i.e., in the absence of anthropogenic phosphorus
15 inputs) in lakes with very low concentrations of both nitrogen and phosphorus, as are
16 common in the western United States, and in the Northeast. In all cases, because the
17 concentrations of both nitrogen and phosphorus are low, additional inputs of nitrogen may
18 have a limited potential to cause eutrophieation, because their input will quickly lead to a
19 switch in the limiting nutrient; additions of nitrogen to these systems would soon lead to
20 nitrogen-sufficient and phosphorus-deficient conditions.
21 Few topics in aquatic biology have received as much attention in the past decade as the
22 debate over whether estuarine and coastal ecosystems are limited by nitrogen, phosphorus, or
23 some other factor (reviewed by Hecky and Kilham, 1988). Numerous geochemical and
24 experimental studies have suggested that nitrogen limitation is much more common in
25 estuarine and coastal waters than in freshwater systems. Taken as a whole, the productivity
26 of estuarine waters of the United States correlates more closely with supply rates of nitrogen
27 than of other nutrients (Nixon and Pilson, 1983).
28 Estimation of the contribution of nitrogen deposition to the eutrophieation of estuarine
29 and coastal waters is made difficult by the multiple direct anthropogenic sources (e.g., from
30 agriculture and sewage) of nitrogen against which the importance of atmospheric sources must
31 be weighed. Estuaries and coastal areas are natural locations for cities and ports, and most of
August 1991 1-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 the watersheds of major estuaries in the United States have been substantially developed. The
2 crux of any assessment of the importance of nitrogen deposition to estuarine eutrophication is
3 establishing the relative importance of direct anthropogenic effects (e.g., sewage and
4 agricultural runoff) and indirect effects (e.g., atmospheric deposition). In the United States,
5 a large effort has been made to establish the relative importance of sources of nitrogen to the
6 Chesapeake Bay (e.g., D'Elia et al., 1982; Smullen et al., 1982; Fisher et al., 1988; Tyler,
7 1988). Estimates of the contribution of nitrogen to the Chesapeake Bay from each individual
8 source are very uncertain; estimating the proportion of nitrogen deposition exported from
9 forested watersheds is especially problematic but critical to the analysis, because ca. 80% of
10 the Chesapeake Bay basin is forested. Nonetheless, three attempts at determining the
11 proportion of the total NO3" load to the Bay,attributable to nitrogen deposition all produce
12 estimates in the range of 18 to 31% (Table 1-2). Supplies of nitrogen from deposition exceed
13 supplies from all other non-point sources to the Bay (e.g., agricultural runoff, pastureland
14 runoff, urban runoff), and only point source inputs represent a greater input than deposition.
15
16 Direct Toxicity
17 Toxic effects of nitrogen on aquatic biota result from un-ionized ammonia (NBlj)* which
18 occurs in equilibrium with ionized ammonium (NH4+) and OH". Ammonia concentrations
19 approach toxic concentrations most commonly at high pH and temperature values, which are
20 most typical of heavily polluted lakes and streams (e.g., Effler et al., 1990). In the well-
21 oxygenated conditions typical of unpolluted lakes and streams (as well as in most watersheds)
22 NH4+ is rapidly oxidized to NO3", which does not have toxic effects on aquatic organisms.
23 Within the typical range of pH and temperature that unpolluted lakes and streams experience,
24 toxic concentrations of NH3 resulting from nitrogen deposition would be extremely unusual.
25 At a pH of 7, and a temperature of 15 °C, for example, concentrations of total NH4+ would
26 have to reach over 750 /imol • L"1 before chronically toxic concentrations of free NH3 would
27 develop. Currently no areas of North America are known to experience rates of nitrogen
28 deposition that are sufficient to produce such high concentrations of total NH4+ in surface
29 waters.
August 1991 1-45 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 1-2. THREE NITROGEN BUDGETS FOR THE CHESAPEAKE BAY
Source of
Nitrogen
Direct Deposition
NO3'
NH4
N Load to Bay (from direct
deposition)1*
Forests
NO3* Deposition
NH4+ Deposition
Watershed Rentention
In-stream Retention
Atmospheric NO3" Load to
Bay (from forests)
N Load to Bay (from
forests)*
Pastureland
NO3" Deposition
NH4* Deposition
Animal Wastes
Watershed Retention
In-stream Retention
Atmospheric NO3~ Load to
Bay (from pastures)
N Load to Bay (from
pastures)b
Cropland
NO3" Deposition
NH4* Deposition
Fertilizers
Watershed Retention
In-stream Retention
Atmospheric NO3' Load to
Bay (from cropland)
N Load to Bay (from
cropland)*
Residential/Urban
NO3* Deposition
NH4+ Deposition
Watershed Retention
In-stream Retention
Atmospheric NO3" Load to
Bay (from urban areas)
N Load to Bay (from urban
areas)b
EOF
Budget
(eq X 10» • yr-1)
0.6
0.3
0.9
6.4
3.5
80%
50%
0.6
1.0
1.7
0.9
13.9
95%°
50%°
0.5
1.1 .
1.8
1.0
11.3
70%
70%
0.6
4.2 '
0.3
0.2
35%
0%
0.2
0.3
Versar
Budget
(eq X 10* • yr4)
0.5
_»
0.5
6.0
a
95%
50%
0.15
0.15
1.2
_a
8.4
94-99%
50%
0.01-0.04
0.05-0.3
2.0
_*
5.9-19.3
76-99%
50%
0.0-0,2
0.04-2.6
,0.5
-•
62-96%
20%
0,01-0.1
0.01-0.1
Refined
Budget
(eq X 109 • yr-1)
0.4
0.2
0.6
4.6
2.5
0.5
0.7
0.9
0.5
13.9
0.09
0.6
1.5
0.8
11.3
0.05
0.4
0.4
0.2
. 0,1
0.2
84.6%
35%
95 %d
35%
95%
35%
50%
35%
August 1991
1-46
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 1-2 (cont'd). THREE NITROGEN BUDGETS FOR THE
CHESAPEAKE BAY
Source of
Nitrogen
Point Sources
NO3' LOAD TO BAY
(FROM DEPOSITION
TOTAL NITROGEN LOAD
TO BAY*
% of N from NO3" depostion
EOF
Budget
(eq X 10* • yr'1)
2,4
2.50
9.95
25%
Versar
Budget
(eq X 10* • yf1)
1.4-2.3
0.67-1.06
2.16-5.90
18-31%
Refined
Budget ,
(eq X 109 • yr1)
2.4
1.09
4.87
22.5%
"The Versar Budget (Tyler, 1988) does not calculate loads of NH/.
bFor the EOF Budget (Fisher et al., 1988a) and refined budget total nitrogen load to the Bay includes both NO3"
and NH4+. The Versar Budget (Tyler, 1988) includes only NO3".
^Watershed and In-stream retention values for pastureland in the EDF Budget apply only to animal wastes. For
atmospheric deposition, the cropland retention value (70%) was used.
d95% retention was used for animal wastes; 85% retention was used for deposition (see text).
"The range of contributions of NO3" deposition to the total budget were calculated by comparing maximum to
maximum estimates, and minimum to minimum estimates. These combinations are more likely to occur during
extreme (e.g., very wet or very dry) years.
1 1.3.3 Effects of Nitrogen Oxides on Visibility
2 Emissions of NOX can contribute significantly to visibility impairment in the form of
3 plumes and hazes. Nitrogen dioxide (NO-j) and ammonium nitrate aerosol (NH4NO3) are the
4 optically active species of NOX. Other species, including nitric oxide (NO) and nitric acid
5 (HNO3), are gases with insignificant optical effects. Nitrogen dioxide is a gas that
6 preferentially absorbs blue light, thus tending to cause yellow-brown atmospheric
7 discoloration (e.g., Hodkinson, 1966). There is agreement among many studies that NO2 is
8 a strong and consistent colorant (e.g., Horvath, 1971; Charlson et al., 1972). Aerosols,
9 however, including nitrate, can cause atmospheric discoloration (e.g., Waggoner et al., 1983;
0 Husar and White, 1976), particularly when bright objects are observed or the sun is behind
.1 the observer (e.g., Ahlquist and Charlson, 1969; Husar and White, 1976; Finlan, 1981).
.2 Nitrogen dioxide has been shown to be the most significant plume colorant for the
,3 yellow-brown power plant plumes that have been observed (Latimer, 1979, 1980; Melo and
.4 Stevens, 1981; Vanderpol and Humbert, 1981) primarily in the western United States and
15 that are of current regulatory concern to EPA and the States (e.g., Blumenthal et al., 1981;
16 BergstrometaL, 1981).
August 1991 1-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 Nitrogen dioxide and nitrate aerosol are significant contributors to urban haze,
2 especially in California (e.g., Appel et al., 1985; Cass, 1979; Trijonis et al., 1982; Pratsinis
3 et al., 1984) and the western United States (GroblicM et al., 1981; Solomon and Moyers,
4 1984, 1986; Dzubay et al., 1982; Stevens et al., 1988). Their combined share of total
5 extinction can be 20 to 40 % of total light extinction in such urban areas. In nonurban areas,
6 NOX appears to be a relatively small contributor to light extinction because NO2, nitrate
7 aerosol, and ammonia concentrations tend to be lower or because moderate or high
8 temperatures tend to prevent nitrate aerosol from condensing (e.g., Trijonis et al., 1988;
9 Malm et al., 1989; Dzubay and Clubb, 1981). Nitrate aerosol does not appear in areas of
10 high concentrations of sulfate, such as the eastern United States, mainly because acidic sulfate
11 compounds consume the available atmospheric ammonia that is needed to condense nitrate
12 aerosol from nitric acid vapor (e.g., Wolff and Korsog, 1989; Vossler et al., 1989; Mathai
13 and Tombach, 1987).
14 Theoretical models have been developed for describing the chemical reactions that result
15 in the formation of optically active NOX species, aerosol dynamics of nitrate aerosol,
16 chemical equilibrium of nitrate-water aerosols, the light scattering and absorption properties
17 as a function of the wavelength of light, and effects on visual range, haze contrasts, and
18 atmospheric color. The available comparison of plume visibility models (White et al., 1985,
19 1986) suggests that the effects of plume NO2 can be accurately predicted (Latimer and
20 Samuelsen, 1975, 1978; Latimer et al., 1978; Johnson et al., 1980; Seigneur et al., 1984;
21 Drivas et al., 1980) but that model predictions of the effects of particles are less adequate.
22 Limited work has been done to develop and test models for urban (Russell and Cass, 1986;
23 Russell et al., 1988), layered (Latimer et al., 1986; Latimer, 1988), and regional (Latimer
24 et al., 1985a,b; 1986) haze; but much more work is clearly needed.
25 Measurement of nitrate aerosol is complicated by its volatility (e.g., Sloane and White,
26 1986). However, newer measurement techniques based on the use of denuders have provided
27 reliable measurements (e.g., Mulawa and Cadle, 1985; Stevens et al., 1988). Because older
28 techniques (such as Teflon filters) can seriously underestimate nitrate aerosol concentrations
29 (e.g., Appel et al., 1981; Cadle, 1985; John, 1986; Stevens, 1987), care must be taken when
30 interpreting data.
August 1991 1-48 DRAFT-DO NOT QUOTE OR CITE
-------
Work is needed to understand the apparently nonlinear effects of NOX emission controls
on nitrate aerosol concentrations and resulting visibility effects. Also, work is needed to
understand the effects of SO2 emission controls on nitrate aerosol production, because the
large-scale reduction of sulfate, which competes with nitrate for available ammonia, may
result in increases in nitrate aerosol.
The current reliance on and continuing methodological uncertainties in the contingent
valuation method (CVM) for obtaining economic estimates related to changes in visibility
means future methodological research related to CVM may provide important information
relevant to interpreting previous and designing new visibility value studies using CVM.
Some evidence is beginning to emerge as to those factors in CVM applications that appear to
have the most effect on visibility value estimates and that, therefore, need to receive more
attention (e.g., Irwin et al., 1990; Carson et al., 1990). Among these factors are:
1. The treatment of related air pollution effects such as health;
2. The geographic location of the impact;
3. Whether values are obtained in a single or multiple good context, and are
appropriately adjusted for non-visibility components;
4. Whether a frequency distribution presentation of visibility is used;
5. Various features of the hypothetical context of the WTP questions; and
6. Survey implementation and data handling procedures.
Available estimates of economic values for plumes are directly applicable for only a
thin, dark plume present on most days in the sky at Grand Canyon National Park (Schulze
et al., 1983; MacFarland et al., 1983; Brookshire et al., 1976), and possibly at a few other
national parks in the Southwest. There is little empirical evidence (from economics studies to
date) about how the values would vary with the frequency, location, or visual characteristics
of a plume. Available results (Schulze et al., 1983; but see also Chestnut and Rowe, 1990)
do suggest the potential importance of considering total preservation values, rather than only
on-site use values, in any assessment of the economic impact of NOX plumes visible from
major recreation areas.
The best economic information available for visibility effects associated with NOX is for
on-site use values related to changes in visual range in urban areas caused by uniform haze
(Loehman et al., 1981; Brookshire et al., 1982; Trijonis et al., 1984; Tolley et al., 1986;
August 1991 1-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 Rae, 1984). These values fall roughly between $10 and $100 per year per local household
2 for a 10% change in visual range in major urban areas in California and throughout the
3 eastern United States. Reasonable extrapolations of on-site use values (with an order-of-
4 magnitude range of uncertainty) could be made from these studies for estimates of changes in
5 visual range that are attributable to changes in NOX levels in these and other major urban
6 areas, where NOX contributes to uniform haze that can be characterized by changes in visual
7 range. Extrapolations to less urbanized areas and/or to other visibility changes would require
8 additional assumptions and might introduce additional uncertainty. Because each of the
9 studies completed to date has some important weaknesses and limitations, it would be
10 desirable to continue to enhance the geographic extent and the technical breadth of issues
11 addressed in these studies to arrive at a broader and more defensible set of estimates.
12 Available results with regard to visual range in urban areas, however, appear to be sufficient
13 to determine the importance of visibility values (on-site use) related to NOx-caused uniform
14 haze in urban areas relative to other potential benefits of NOX controls and to provide order-
15 of-magnitude estimates of such visibility values. To do so would require estimates of the
16 changes in visual range that might be expected as a result of NOX controls.
17 Very little work has been done regarding layered hazes in recreation or residential
18 settings. The work conducted in Denver (Irwin et al., 1990) is the only study in this
19 category. More information is needed about what visual characteristics of such hazes are
20 most important to viewers, as well as on the value people may place on reducing or
21 preventing them. A related question that is relevant for NOX, but that has not been addressed
22 in economic studies to date, is the effect of variations in air pollution color (in plumes or
23 hazes) on visibility values. In order to apply available economic values for changes in
24 layered hazes to changes in NOX levels, a link would need to be made between NOX levels
25 (and changes that might result from control efforts) and visual characteristics of layered hazes
26 that can be tied to the economic values.
27
28 1.3.4 Effects of Nitrogen Oxides on Materials
29 Nitrogen oxides have been shown to cause or accelerate damage to manmade materials
30 exposed to the atmosphere. Strong evidence exists for the negative impact of NOX on dyes
31 and fabrics (e.g., Howe and Chamberlain, 1937; Salvin, 1969; Haynie et al., 1976; Hemphill
August 1991 1-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 1976; Beloin, 1972, 1973). Many varieties of dyes are known to fade, become duller,
2 or acquire a different cast, and white fabrics may "yellow" from exposure to NOX (e.g.,
3 McLendon and Richardson, 1965). Fade-resistant dyes and inhibitors have been developed,
4 but are generally more costly to employ (Salvin and Walker, 1959; Salvin, 1964). Nitrogen
5 oxides also attack textile fibers and result in a loss of strength (Jellinek et al., 1969; Jellinek,
6 1970; Zeronian et al., 1971; Vijayakumar et al., 1989). Plastics and elastomers are subject
7 to NOX reactions that cause discoloration and changes in physical properties, including loss of
8 strength (Jellinek et al., 1969; Jellinek, 1970; Haynie et al., 1976; National Research
9 Council, 1977; Krause et al., 1989). The rate of NOx-induced deterioration to plastics and
.0 elastomers is generally overshadowed, however, by O3-induced damage (Haynie et al.,
.1 1976). Although NOX attacks metals, attack by SO2 is more aggressive. Damage to metals
.2 from NOX can generally be discounted (e.g., Edney et al., 1986, 1987), except perhaps in
.3 indoor exposures, where NOX may react synergistically with SO2 (Svedung et al., 1983;
.4 Kucera, 1986; Johansson, 1986). Also largely indoors, NOX is deposited on electronic
.5 components and magnetic recording equipment and may lead to failures in these systems
.6 (e.g., Svedung et al., 1983; Kucera, 1986; Abbott, 1987; Freitag et al., 1980; Voytko and
.7 Guilinger, 1988). The influence of NOX on paints (e.g., Spence et al., 1975; Haynie and
.8 Spence, 1984) and stone (e.g., Livingston and Baer, 1983; Amoroso and Fassina, 1983;
.9 Johansson et al., 1988) has not been clearly demonstrated. Many authors indicate that NOX
10 plays a role in damaging these materials, but most concede that SO2 and O3 are more directly
II damaging. Nitrogen oxides, along with other "acid pollutants," attack the cellulose fibers in
12 paper (e.g., National Research Council, 1986), leading to discoloration and weakened
13 structure.
14 The presence of NOX will shorten the use-life of susceptible materials and generally the
15 rate of damage is proportional to the pollutant concentration. Adequate NOX damage
16 functions for a wide variety of materials are not available. Consequently, cost/benefit
17 analyses of permissible NOX levels vis-a-vis shortened use-life estimates could be misleading.
18 Cost estimates for NOx-specific damage at existing concentrations are available (National
19 Research Council, 1977) only for dye fading ($97 million annually in 1977 dollars), and
10 these estimates are very out of date.
August 1991 1-51 . DRAFT-DO NOT QUOTE OR CITE
-------
1 The highest NOX levels are to be found indoors where unvented combustion systems
2 (e.g., gas stoves) are used and the widest variety of materials are routinely exposed.
3 Therefore, the principal effects of NOx-indueed damage to materials are probably seen in the
4 indoor environment, but few data are available regarding materials deterioration indoors.
5 This is an area of needed research.
6
7
8 1.4 HEALTH EFFECTS OF OXIDES OF NITROGEN
9 1.4.1 Animal Toxicology
10 A large number of experiments, designed to evaluate the health effects of NOX
11 (primarily NO^ on various animal species, have been conducted; and NO2 appears to be the
12 most toxic. Many of the experiments were performed at very high concentrations of NO2
*Ct
13 (above 9,400 /*g/m [5.0 ppm] for days, weeks, or months) to ensure eliciting the effects
14 being studied and thus provide only limited or no information directly applicable to standard
15 setting. Therefore, this document focuses on studies conducted at less than 9,400 /*g/m3
16 (5.0 ppm). Even so, ambient levels are much lower. For example, except for the Los
17 Angeles Basin, annual averages of NO2 are less than the NAAQS of 100 jug/m3 (0.053 ppm).
18 How then are animal studies at levels considerably above ambient to be interpreted in
19 terms of human health risk? Extrapolating animal data to humans has both qualitative and
20 quantitative components. As has been discussed and will be summarized below, NO2 causes
21 a constellation of effects in several animal species, most notably, effects on host defenses
22 against infectious pulmonary disease, lung metabolism/biochemistry, lung function, and lung
23 structure. Because of basic physiological, metabolic, and structural similarities in all
24 mammals (laboratory animals and humans), the commonality of the observations in several
25 animal species leads to a reasonable conclusion that NO2 could cause similar types of effects
26 in humans. However, because of the differences between mammalian species, the levels of
27 exposures that could actually cause these effects in humans are unknown. That is the topic of
28 quantitative extrapolation.
29 Limited research on the dosimetric aspect (i.e., the dose to the target tissue/cell that
30 actually causes toxicity) of quantitative extrapolation suggests that dose patterns within the
31 respiratory tract of animals and humans are similar, without yet providing adequate values to
August 1991 1-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 use for extrapolation. Unfortunately, very little information is available on the other key
2 aspect of extrapolation, species sensitivity (i.e., the response of the tissues of different species
3 to a given dose). Thus, currently available animal studies indicate what human health effects
4 NO2 could cause, rather than what effects NO2 actually causes. Such knowledge is still quite
5 valuable because it can enable: (1) identification of potential hazards for humans that are
6 either unmeasured or unmeasurable in humans; (2) estimation of the biological plausibility of
7 effects observed in epidemiological studies, which, by their very nature, cannot provide
8 definitive cause-effect relationships; (3) identification of mechanisms of effects that can
9 enable enhanced interpretation of the potential severity of effects, in human studies; and
10 (4) generation of hypotheses for testing in human studies, thereby advancing the human data
11 base.
12 With the above issues in mind, the animal toxicology data base for NO2 will be
13 summarized according to major classes of effects and topics of special interest. These
14 include: animal-to-human dosimetric estimates, biochemical and cellular mechanisms, effects
15 on host defenses, effects of chronic exposure on the development of chronic lung disease,
16 potential carcinogenic or co-carcinogenic effects, extrapulmonary effects, susceptibility of
17 subpopulations, influence of exposure patterns, and interactions with other co-occurring
18 pollutants. Several other topics or issues are not summarized here because of the paucity of
19 the data or lack of current interpretability. For more information on these topics (e.g.,
IQ mortality from NO2 and effects of other NOX compounds, such as NO and HNO3), see
II Chapter 13,
11
13 1.4.1.1 Animal-to-Human Dosimetric Estimates
14 There are few experiments in which NO2 uptake values are reported: two for the upper
15 respiratory tract, one for the lower respiratory tract, and two for the total respiratory tract.
16 The upper respiratory tract uptake efficiencies were 25 to 28% for rats and 45 to 85% for
11 dogs, depending on the mode of breathing and ventilation. Thirty-six percent uptake was
1% obtained in the lower respiratory tract of isolated ventilated rat lungs; however, because the
19 low ventilation used is not representative of living rats, the measured uptake most probably is
50 not representative. The two total respiratory tract uptake experiments involved normal
51 humans and asthmatics and resulted in uptakes of 81 to 92% and 73 to 88%, respectively. In
August 1991 1-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 both experiments percentage uptake increased as ventilation increased. Thus, the available
2 data on in vivo uptake of NO2 by various species indicate that a large percentage of inhaled
3 gases is removed in the respiratory tract, depending on the species, mode of breathing,
4 ventilation, and possibly just as important, the experimental procedures.
5 A comparison of the results from the one NO2 dosimetry modeling study to
6 morphological data that shows the centriacinar region to be most affected by NO2 indicates
7 qualitative agreement between predicted maximum tissue doses and observed effects in the
8 pulmonary region. Comparisons in the tracheobronchial region, however, indicate that dose-
9 effect correlations may be improved by considering other expressions of dose such as total
10 absorption by an airway. Further research is needed to define toxicity mechanisms, to refine
11 knowledge of important physical, chemical, and morphological parameters, to develop NO2
12 dosimetry models using this information, and to perform dosimetry experiments with a
13 variety of species.
14
15 1.4.1.2 Biochemical and Cellular Mechanisms
fj
16 Acute exposure to NO2 concentrations at or below 9,400 ptg/m (5 ppm) can oxidize
17 polyunsaturated fatty acids in cell membranes as well as functional groups of proteins (either
18 soluble proteins in the cell, such as enzymes, or structural proteins, such as components of
19 cell membranes). Such oxidation (free radical-mediated) reactions may well be the
20 mechanism by which NO2 elicits direct toxicity on lung cells. Such a proposed mechanism
21 of action suggests the importance of lung antioxidant defenses, both endogenous (e.g.,
22 maintenance of lung glutathione levels) and exogenous (e.g., dietary vitamin C and E), in
23 identification of potential susceptible populations at risk of NO2 inhalation. The direct
24 cytotoxic effects of NO2 on epithelial cell membranes may be the fundamental mechanism of
25 edemagenesis in response to NO2 exposure, while the direct cytotoxicity of NO2 to
26 membranes of alveolar macrophages could well be the mechanism underlying increased
27 infectivity of bacteria and viruses in lungs of animals and humans exposed to NO2. A large
28 number of studies have suggested that various enzymes in the lung, including glutathione
29 peroxidase, superoxide dismutase, and catalase, may also serve to defend the lung against
30 oxidant attack.
31
August 1991 1-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 1.4.1.3 Effects on Host Defenses
2 Although its primary function is to ensure an efficient exchange of gases, the respiratory
3 tract must also provide the body with a first line protective barrier against inhaled viable and
4 nonviable airborne agents. Thus, any breach in this defense system might increase the risk of
5 disease in the host. During the past several years numerous studies that have contributed to
6 an understanding of these various defense systems have been reported. From this literature it
7 becomes readily evident that exposure to NO2 can result in the dysfunction of these host
8 defenses, increasing susceptibility to infectious respiratory disease. The typical host defense
9 parameters affected by NO2 include the rate of mucociliary clearance, functional and
10 biochemical activity of alveolar macrophages, immunological competence, and susceptibility
11 to experimentally induced respiratory infections. '
12 Cellular injury from NO2 occurs throughout the lower respiratory tract. In the
13 conducting airways, NO2 reacts with the ciliated epithelium, causing a decrease in ciliary
14 beating rates and, as a consequence, a significant reduction in mucus transport. The first of
fj
15 these effects begins to appear after several weeks of exposure to approximately 1,880 ^g/mj
fj
16 (1 ppm) NO2 and becomes clearly evident at 3,760 jig/nr (2 ppm). Shorter periods (several
17 days) of exposure to levels as low as 3,760 jig/m3 (2 ppm) are also effective. Such
18 impairment could render this physiological line of defense less effective in the process of lung
19 clearance.
20 At the alveolar level the host defense system that receives direct exposure to the inhaled
21 pollutant is the alveolar macrophage, which is responsible for maintaining sterility of the
22 pulmonary region, clearing particles from this region, and participating in immunological
23 functions. Morphological appearance of these defense cells changes after a 6-mo exposure to
24 as little as 940 jig/m3 (0.5 ppm) with a 1-h daily spike to 3,760 jig/m3 (ppm 2.0 ppm) of
25 NO2. Measurements of the functional integrity of macrophages isolated from the lungs of
26 NO2-exposed animals indicate that macrophages can be depressed to such a degree that the
27 host can no longer effectively maintain pulmonary sterility. Functional changes that have
28 been reported are: at 560 jig/m3 (0.3 ppm), intermittent for 13-14 days, the suppression in
29 phagocytic ability and stimulation of lung clearance; at 4,320 jig/m3 (2.3 ppm) for 17 h,
30 a decrease in bactericidal activity; and at 3,760 jig/m3 (2.0 ppm) for 6 mo, a decreased
31 response to migration inhibition factor. Macrophages isolated from humans exposed for 3 h
August 1991 1-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 to 1,120 /xg/m3 (0.6 ppm) NO2 were less able to inactivate influenza virus than macrophages
2 from unexposed humans.
3 The importance of these defenses in maintaining pulmonary sterility against invading
4 microorganisms becomes clearly evident when these "impaired" animals have to cope with a
5 laboratory-induced pulmonary infection. Animals exposed to NO2 succumb to the bacterial
6 or viral infection in a concentration-response manner. Mortality also increases with
7 increasing length of exposure to a given concentration of NO2. After acute exposure, effects
8 are seen at levels as low as 3,760 /*g/m3 (2 ppm). Exposure to concentrations as low as
*>
9 940 /zg/mr (0.5 ppm) will cause effects in the infectivity model after 6 mo, but as the
10 concentration increases, the effect becomes evident sooner and is more significant. Thus,
11 evidence clearly indicates that these cells are no longer capable of isolating, transporting, or
12 detoxifying these deposited microbes due to their reduced level of phagocytosis, lytic
13 potential, ability to produce interferon, and other responses. Impacts of different exposure
14 patterns on infectivity are described later (Section 13.6.8).
15 Since NO2 has been shown to cause an increase in susceptibility to both bacterial and
16 viral infections, a few investigators have examined the immune system to explain this change
17 in susceptibility. In the cases in which lung humoral and cell-mediated immunity have been
18 investigated, effects have been observed after acute exposure to concentrations exceeding
19 9,400 /ig/m3 (5 ppm) NO2. These findings illustrate the complexity of pulmonary
20 immunotoxicity, since the direction of the effect (i.e., increase or decrease) was dependent
21 upon NO2 concentration and the length of exposure. Furthermore, systemic and pulmonary
22 immune responses sometimes differed.
23 Subchronic exposure to NO2 has also been found to affect the production of serum
24 neutralizing antibody to viruses and humoral primary antibody responses at concentrations as
25 low as 752 /*g/m3( 0.4 ppm) and 940 /ig/m3 (0.5 ppm) with 1-h daily spikes to 3,760 /ig/m3
26 (2 ppm). Other significant immunological effects attributed to subchronic NO2 exposure
27 include alterations in serum immunoglobulin levels at 1,880 /*g/m3 (1 ppm).
28 Acute exposure to high concentrations of NO2 significantly decreased the number of
29 spleen and thymus cells; however, at a lower concentration, the number of spleen cells
30 actually increased. Others have shown that the reduction in primary antibody response of
31 spleen cells against sheep red blood cells was more influenced by a suppression of B cell
August 1991 1-56 DRAFT-DO NOT QUOTE OR CITE
-------
1 function, while for the secondary antibody response, the T cells were more affected. Levels
2 of NO2 (intermittent exposure, 7 weeks) that affect the percentages of subpopulations of
3 T and B cells have been as low as 470 /xg/m3 (0.25 ppm). Chronic intermittent exposure to
4 high levels of NO2 results in a biphasic response. Antibody liters to sheep red blood cells
5 were increased after 10 weeks of exposure, unchanged at 20 weeks, and decreased at
6 30 weeks. There were no effects when a T-independent antigen was used. A similar
7 biphasic response was observed for a graft versus host reaction. Response of lymphocytes to
8 a T-cell mitogen (phytohemagglutin) was depressed at all exposure times tested.
9
10 1.4.1.4 Effects of Chronic Exposure on the Development of Chronic Lung Disease
11 Since humans can be exposed chronically to NO2, such exposures in animals have been
12 studied rather extensively, typically using morphologic and/or morphometric methods. Less
13 focus has been on pulmonary function effects. Generally, this research has shown that a
14 variety of structural and correlated functional alterations occur, many of which are reversible
L5 when exposure ceases; however, in some cases, emphysema has been observed.
16 Pulmonary function following NO2 exposures in experimental animals showed consistent
17 patterns among different treatment conditions and animal species. Exposures to diurnal
18 spikes of NO2 superimposed on a constant background level, simulating NO2 patterns in the
19 urban environment, produced a decrease in lung distensibility in both mice and rats. These
20 changes were very subtle, occurring in mice after 1 year of exposure to very low NO2
A A
21 concentrations (i.e., 380 /xg/nr [0.2 ppm] NO2 with daily spikes of 1,500 /xg/m [0.8 ppm]
22 NO2). The sensitivity of the lung to irritants was generally increased following NO2
23 exposures, except when the irritant was higher concentrations of NO2. Impaired gas
A
24 exchange occurred following 3 mo of exposure to 7,520 /xg/nr (4 ppm) NO2 and was
25 reflected in decreased arterial O2 tensions, impaired physical performance, and increased
26 anaerobic metabolism. All these studies taken together demonstrate that NO2 produces subtle
27 to dramatic changes in pulmonary function depending on the concentration and duration of
28 exposure. Lung distensibility and gas exchange are the parameters most consistently affected
29 by exposure.
30 Although NO2 produces morphological changes in the respiratory tract, the data base is
31 sometimes confusing due to quantitative and qualitative variability in responsiveness between,
August 1991 1-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 and even within, species. Thus, for example, the rat appears to be relatively resistant to
2 NO2, although this is the most commonly used experimental animal involved in
3 morphological assessments of exposure. In any case, when effective levels are used, the
4 target site is the region that includes the terminal and/or respiratory bronchioles, alveolar
5 ducts, and alveoli. Sensitive cells are the ciliated epithelial cells of the bronchioli and the
6 Type 1 epithelial cells of the alveoli.
7 Acute exposures to NO2 concentrations of 9,400 /-tg/m3 (5 ppm) or less generally
8 produce minimal to no lesions in the rat; however, similar exposures in the guinea pig may
9 result in some epithelial damage. Higher exposure concentrations result in changes typified
10 by hypertrophy and hyperplasia of bronchiolar and alveolar Type 1 epithelium and
11 proliferation of Type 2 alveolar cells.
12 Longer-term NO2 exposures result in lesions in some species with concentrations as low
Q
13 as 560 to 940 ng/m (0.3 to 0.5 ppm). These are characterized by epithelial damage similar
14 to that described above but with the involvement of more proximal airways. Many of these
15 changes, however, will resolve even with continued exposure, and long-term exposures to
16 levels above about 3,760 jtg/m3 (2 ppm) are required for more extensive and permanent
17 changes in the lungs. Some effects are relatively persistent, for example bronchiolitis; while
18 others tend to be reversible and limiting even with continued exposure, for example,
19 epithelial cell hyperplasia. In any case, it seems that for both acute or longer term exposure
20 regimens, the response is more dependent upon concentration than exposure duration.
21 Although ambient exposures are generally to a baseline level with superimposed spikes
22 reaching some higher level, the contribution of these spikes to morphologic damage over that
23 due to baseline exposure is not as yet certain.
24 There is very substantial evidence that long-term exposure of several species of
25 laboratory animals to high concentrations of NO2 results in morphologic lung lesions that
26 meet the current NDH Workshop criteria for an animal model of emphysema. Those criteria
27 are: "An animal model of emphysema is defined as an abnormal state of the lungs in which
28 there is enlargement of the airspaces distal to the terminal bronchiole. Airspace enlargement
29 should be determined qualitatively in appropriate specimens and quantitatively by stereologic
30 methods." Destruction of alveolar walls, an essential additional criterion for human
31 emphysema, has been reliably reported in lungs from animals in a limited number of studies.
August 1991 1-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 While the lowest NO2 concentration for the shortest exposure duration which will result in
I emphysematous lung lesions can not be determined from these published studies, the required
3 NO2 concentrations for such long-exposure durations are far greater than those currently
\ reported in ambient air.
5 • • , ,• • • • ' •. '.. y • - • - . - . .' -.••.-- :•'. :T .•;... ; • • ...
5 1.4.1.5 Potential Carcinogenic or Co-Carcinogenic Effects
7 Literature searches revealed no published reports on NO2 studies using classical whole
3 animal bioassays for carcinogenesis. Research with mice having spontaneously high tumor
? rates was equivocal. Although several co-carcinogenesis investigations have been undertaken,
3 methodological and interpretative problems prevent drawing adequate conclusions. Reports
I on whether NO2 facilitates the metastases of tumors to the lung are also inadequate to form
I conclusions, since the most appropriate models and statistical methods were not used. Other
3 investigations have centered on whether NO2 could produce nitrates and nitrites that, by
\ reacting with amines in the body, could produce animal carcinogens (nitrosamines). A few
5 studies found that in animals treated with high doses of amines and exposed to NO2,
5 nitrosoamines were formed. However, in a related study in two species, a 1- or 2-year
7 exposure to 18,800 jtxg/m3 (10 ppm) did not cause an excess incidence of cancer. In
3 summary, the evidence is inadequate to assess the potential carcinogenic and co-carcinogenic
} effects and suggests that further study may be warranted. ,
D • • . . . . •• •••-..-,
1 1.4.1.6 Extrapulmonary Effects
2 Although it is clear that the effects of NO2 exposure, ex tend beyond the confines of the
3 lung, the interpretation of these effects relative to potential human risk is not clear.
4- After acute exposure, there appears to be a transient rise in white blood cells. Red
5 blood cell count has been shown to increase or decrease, but changes in hematocrit and
6 hemoglobin do not occur. Although nitrite and especially nitrate do appear in the blood,
7 increased methemoglobin is not typically reported at near ambient levels of exposure. The
8 evidence does suggest that there is a change in composition of the red blood cell membrane
9 constituents; however, whether this indicates an increased younger population of erythrocytes
D or an apparent aging has not been resolved. Several of the traditional serum markers of
1 tissue injury are elevated with acute near-ambient exposure to NO2 concentrations as low as
August 1991 1-59 DRAFT-DO NOT QUOTE OR CITE
-------
o
1 470 fig/nr (0.25 ppm), but this effect is also transient. Chronic exposure (at least 3 mo)
2 seems to decrease serum proteins, lipoproteins, and plasma cholinesterase; all of which
3 suggest hepatic damage. Although changes in the microsomal oxidative system are observed
A
4 during acute exposure to concentrations as low as 470 jig/nr (0.25 ppm), little evidence of
5 chronic hepatic damage has been reported. Swollen liver mitochondria, edema, and
6 inflammatory parenchyma! changes after chronic exposure to between 940 and 2,000 ^g/m3
7 (0.5 and 1.05 ppm) NO2 have been reported in two studies. The limited studies of NO2
8 effects on the kidney do not allow any conclusions.
9 Nitrogen dioxide can produce effects on the cardiovascular system, but these effects
10 seem to occur only at concentrations 20 to 100 times ambient. The changes that have been
11 observed in the central nervous system require further study before any kind of interpretation
12 is possible. The behavioral changes seen, especially the changes in forced swim and running
13 wheel performance, are reminiscent of the acute performance decrements seen in humans with
14 ozone exposure.
15
16 1.4.1.7 Susceptibility of Subpopulations
17 Two subpopulations have been examined for susceptibility, young animals and animals
18 with elastase-induced emphysema. Neither group was more susceptible, and in some studies
19 both were found to be more resistant, than their appropriate NO2-exposed controls.
20 Susceptibility to NO2 exposures was not affected in animals treated intratracheally with
21 elastase to produce a condition of experimental emphysema, but sensitivity was increased in
22 mice with a genetic defect in connective-tissue metabolism that results in pulmonary
23 emphysema. Newborn and mature animals are affected differently by NO2 exposures,
24 particularly rats exposed to continuous background NO2 concentrations of 1,880 or
25 3,760 fjLg/m3 (1 or 2 ppm) with diurnal spikes to three times the baseline. A 6-week
26 exposure of the adult rats caused an increase in the thickness of the alveolar basement
27 membrane and a greater increase in the volume of Type 2 cells, compared to the neonatally-
28 exposed rats. Pulmonary function effects were correlated with these changes. Lung
29 distensibility was increased transiently (after 3 but not 6 weeks of exposure) in rats exposed
30 as newborns, but was decreased in rats exposed for 6 weeks as young adults. Hamsters
31 exposed to high NO2 concentrations as newborns had pulmonary function changes one year
August 1991 1-60 DRAFT-DO NOT QUOTE OR CITE
-------
1 later that were indicative of mild pulmonary emphysema, but these changes were not found in
2 hamsters exposed when older.
3
4 1.4.1.8 Influence of Exposure Patterns
5 Several animal lexicological studies have elucidated the relationships between NO2
6 concentration (C) and duration (T, time) of exposure, clearly indicating that the relationship
7 is quite complex, rather than the simple formula, C1 X T1 = Effect. Most of this research
8 used the infectivity model in which air- and NO2-exposed mice are exposed to bacteria, and
9 decreases in host defenses are measured as increases in mortality. Early C X T studies
0 demonstrated that concentration had more impact than exposure duration on mortality. More
1 recent work has evaluated the effects of ambient exposure patterns consisting of a continuous,
2 low baseline of NO2 on which are superimposed brief peaks of higher concentrations of NO2.
3 When components of this pattern were tested and compared to the entire exposure pattern, the
4 decrease in host defenses was dependent upon the specific exposure pattern; that is, the
5 outcome could not be predicted based upon knowledge of the components. Few studies of
6 ambient patterns of NO2 have been investigated for other classes of effects (e.g., pulmonary
7 function, morphology); they generally support the conclusion that the pattern, rather than the
8 simple product of C x T, is responsible for the results. Therefore, this body of work does
9 not provide a mathematical model to relate various exposure patterns to specific effects.
0 However, by illustrating the importance of concentration, duration, and exposure patterns, it
1 does demonstrate the need for rather advanced exposure assessments for more precise
2 estimates of health effects and the need to understand the C x T relationship far better to
3 enable extrapolations of effects from one exposure scenario to another.
4
5 1.4.1.9 Interactions with Other Pollutants
6 Humans are exposed to complex pollutant mixtures, not NO2 alone, making it important
7 to understand the interaction of NO2 with these other pollutants. Unfortunately, animal
8 studies employing pollutant mixtures are rare and most were conducted with binary mixtures
9 containing O3. For information on other interactions, see Section 13.3. Studies of O3 plus
0 NO2, the focus of this summary, are especially pertinent because: (1) they both are
1 photochemical oxidants, (2) NO2, being a primary precursor of O3, is temporally and
August 1991 1-61 DRAFT-DO NOT QUOTE OR CITE
-------
1 spatially related to O3, and (3) O3 and NO2 generally cause most of the same classes of
2 effects, with O3 being significantly more potent. Numerous studies of lung morphology,
3 antioxidant metabolism, and host defenses against infection have found that the effects were
4 primarily due to O3 (typically in cases using low, noneffective levels of NO^, were additive,
5 or were synergistic, depending on the concentrations, exposure durations, exposure patterns,
6 and end point examined. The findings of either additivity or synergism are of concern
7 because of the ubiquity of O3 and NO2 in ambient mixtures and the type of effects observed.
8 For example, if one of these pollutants is causing a decrease in host defenses, even an
9 additive response to the other pollutant would increase the incidence or severity of the effect.
10 Precise interpretation of these findings to ambient scenarios is confounded, however. In the
11 ambient air, the common diurnal pattern is a series of peaks of the photochemical oxidants
12 and their precursors (e.g., NO, NO2, O3); there is some overlapping between the peaks.
13 Such a "real-world" pattern has not been tested under controlled conditions allowing
14 estimations of the relative lexicological role of the compounds. Studies of NO2 and NO2-O3
15 mixtures illustrate the importance of exposure patterns and interaction with other oxidant
16 pollutants. However, precise conclusions concerning mixture study results in relation to
17 ambient patterns are limited.
18
19 1.4.2 Epidemiology Studies of Oxides of Nitrogen
20 The epidemiological evidence for the effects of NOX on human health is discussed in
21 Chapter 14. The major emphasis is on the effects of NO2 because it is the NOX species
22 studied in most epidemiological studies and is the species currently of greatest concern from a
23 public health perspective. The results from various epidemiological studies of NO2 exposure
24 effects on human health outcomes are shown in Appendix 14A.
25 The studies considered in Chapter 14 were evaluated for several key factors, including
26 (1) measurement error in exposure, (2) misclassification of the health outcome, (3) selection
27 bias, (4) adjustment for covariates, (5) publication bias, (6) internal consistency, and
28 (7) plausibility of the effect based on other evidence. The health outcome should be an end
29 point for which there is good reason to suspect that NO2 exposure has an effect. Two health
30 outcome measures are generally considered: lung function measurements and respiratory
31 illness. Each study was reviewed with special attention given to the factors just outlined.
August 1991 1-62 DRAFT-DO NOT QUOTE OR CITE
-------
I The studies which address these factors more appropriately provide a stronger basis for the
I study conclusions. Consistency between studies indicates the level of strength of the total
3 data base. ,
t . Respiratory illness and factors that affect its rate and/or severity are important .public
5 health concerns because of the potential for exposure to NO2 and since childhoQd respiratory
5 illness is common (Samet et al., 1983; Samet and Utell, 1990). This takes on added
7 importance since recurrent childhood respiratory illness may be a risk factor for later
3 susceptibility to damage to the lungs (Glezen, 1989). ,
3 A biologic basis for an increased susceptibility for respiratory illness with exposure to
3 NO2 is found in studies of the respiratory host defense system. The host defense system
1 provides protection against inhaled infectious and chemical agents. The available
2 information, discussed in the animal toxicology and clinical studies chapter, indicates that
3 exposure to NO2 impairs one or more defense mechanisms, leaving the host susceptible to
\ respiratory illness.
5 Several of the epidemiological studies provide some evidence that repeated NO2
6 exposure increases respiratory illness in children, although many were not statistically
7 significant. Melia et al. (1977) first reported on a survey of children in randomly selected
8 areas of England and Scotland using the presence of a gas stove as a measure of NO2
9 exposure. A reanalysis of that data gave an estimated odds ratio of 1.31 for the presence of
0 respiratory symptoms. The cross-sectional study of Melia et al. (1979) also found that the
1 presence of a gas stove was associated with increased risk of respiratory disease. The odds
2 ratio was 1.24 with 95% confidence limits of 1.09 to 1.42. Melia et al. (1980) reported on
3 the results of a third study of respiratory symptoms in children aged six to seven in northern
4 England. Multiple logistic regression analysis of the data presented by Melia et al. (1980)
5 showed a significant increase in symptoms as a function of bedroom NO2 levels. Melia et al.
6 (1982) reported on a fourth study of children in England. Multiple logistic regression
7 analysis of this data was not statistically significant, although the symptoms were positively
8 related to NO2 exposure. An EPA reanalysis suggests that an increase of 30 /ig/m3
9 (0.016 ppm) in bedroom NO2 levels would result in an 11% increase in the odds of .
0 respiratory illness. In 1983, Melia et al. (1983) reported their findings on the effects of NO2
August 1991 1-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure in infants under 1 year of age. No relationship was established between the type of
2 fuel used for cooking and the prevalence of respiratory symptoms,
3 The analysis of the Six City studies by Ware et al. (1984) estimated an unadjusted odds
4 ratio of 1,08 (95% confidence limits of 0.97 to 1.19) for a lower respiratory illness (LRI)
5 index associated with gas stove use. Other indicators such as bronchitis, cough, and wheeze
6 did not show any increased incidence. Neas et al. (1990) analyzed a different cohort enrolled
7 later, used a different symptom questionnaire, and made indoor NO2 measurements for all
8 subjects. They found increased respiratory disease which gave an estimated odds ratio of
9 1.47 (95% confidence limits of 1.17 to 1.86) at an exposure of 31 jag/m3 (0.016 ppm).
10 Ogston et al. (1985) studied respiratory disease in 1-year-olds in the Tayside region of
11 northern Scotland. The presence of a gas stove resulted in an odds ratio of 1.14, with 95%
12 confidence limits of 0.86 to 1.50. Ekwo et al. (1983) studied respiratory symptoms in
13 relation to gas stove use in Iowa City, IA. Gas stove use resulted in an odds ratio of 2.4 for
14 hospitalization for chest illness before age 2 and 1.1 for chest congestion and phlegm with
15 colds. Dijkstra et al. (1990) studied the effect of indoor factors on respiratory health in
16 children in The Netherlands. A logistic regression analysis yielded an odds ratio of 0.94 with
17 95% confidence limits of 0.66 to 1.33, thus showing no evidence of an increase in
18 respiratory disease with increasing NO2 exposure. Keller et al. (1979) did not find any
19 statistically significant changes in respiratory disease associated with gas stove use, but the
20 unadjusted estimated odds ratio for lower respiratory illness was 1.10, with 95% confidence
21 limits of 0.74 to 1.54.
22 Other studies did not provide sufficient information to derive any quantitative estimates
23 of the effect of NO2 or gas stove use on respiratory disease. Several other studies gave
24 information about the effects of NO2 on respiratory illness, but most of the studies either
25 used very different health end points or did not provide quantitative estimates of the effects.
26 Several of the studies suggest an increase in respiratory symptoms in children from
27 exposure to levels seen with gas stove use as compared to electric stove use. These
28 associations, in the majority of the studies, do not reach statistical significance. The
29 consistency of these studies was examined and the evidence synthesized in a quantitative
30 analysis. The studies described used different indicators to study health end points. In order
31 to compare these studies, a standard end point was defined, and then each study was
August 1991 1-64 DRAFT-DO NOT QUOTE OR CITE
-------
1 compared with this end point. For purposes of discussion, the end point chosen was the
2 presence of lower respiratory illness in children aged 12 or under. It was assumed that the
3 relative odds of developing lower respiratory illness is similar across this age range as a
4 function of NO2 exposure, even though the actual rates may not be. (This is a common
5 assumption in many analyses.) The goal was to estimate the odds ratio corresponding to an
6 increase of 30 /ig/m3 (0.016 ppm) in NO2 exposure. This is approximately the increase seen
7 as a result of gas stove use as compared with electric stove use in both the United States and
8 England.
9 An attempt was made to include as many studies as possible. The requirements for
10 inclusion were: (1) the health end point measured must be reasonably close to the standard
11 end point, (2) exposure differences must exist and some estimate of exposure must be
12 available, and (3) an odds ratio for a specified exposure must have been calculated, or data
13 presented so that it can be calculated.
14 The exposure estimates used in these studies are either a surrogate (gas vs. electric) or a
15 2-week integrated NO2 average measured by Palmes tubes. The effects studied may be
16 related to peak exposures, average exposures, or a combination of the two. To the extent
17 that health effects depend on peak exposures rather than average exposures, the above
18 exposure estimates introduce measurement error. These studies can not distinguish between
19 the relative contributions of peak and average exposures and their relationship with the
20 observed health effects.
21 Graphs of the odds ratio from each study are presented in Figure 1-7. Each curve can
22 be given one of three interpretations: (1) the normal approximation to the likelihood of the
23 logarithm of the odds ratio, (2) a distribution such that the 0.025 percentile and the 0.975
24 percentile points of the distribution are the 95 % confidence limits of the estimated odds ratio,
25 and (3) the posterior for the odds ratio for a particular study given a flat prior on the log-
26 odds ratio.
27 Two methods of combining evidence are used here. The first of these is a Bayesian
28 method, and is described by Eddy (1989) and Eddy et al. (1990a,b) as it is used here. The
29 result of the analysis is a distribution about the location of the true value of the odds ratio
30 (see Figure 1-7). The second basic model assumed that the parameter of interest is not fixed,
August 1991 1-65 DRAFT-DO NOT QUOTE OR CITE
-------
COMBINED (fixed)
COMBINED (random)
Mellaetal. (1979)
Meliaetal. (1977)
Ware etal. (1984)
Ekwootal. (1953)
Keller etal. (1979)
Ogstonetal. (198S)
Meliaetal. (1982)
Neas etal. (1990)
Mellaetal. (1980)
Dijkstra etal. (1990)
Mela etal. (1983)
Odds Ratio
Figure 1-7. U.S. Environmental Protection Agency meta-analysis of epidemiologie
studies of N(>2 exposure effects on respiratory disease in children 312 years
old. Each curve can be treated as a likelihood function or posterior
probability distribution. If treated as a likelihood function, the 95%
confidence limits for the odds ratio can be calculated as those two points on
Hie horizontal axis for which 95% of the area under the curve is contained
between the two points. If treated as a posterior probability distribution,
then the area under the curve between any two points is the probability
that the odds ratio lies between those two points.
1 but is itself a random variable from a distribution. These models go by several names,
2 including random effects models or hierarchical models. The purpose of a random effects
3 model is to relax the assumption that each study is estimating exactly the same parameter.
4 DerSimonian and Laird (1986) discussed the random effects model.
5 The conclusion derived from the models was the same, namely that the odds ratio is
6 estimated to be about 1.2 with 95% confidence limits ranging from about 1.1 to 1.3. The
August 1991
1-66
DRAFT-DO NOT QUOTE OR CITE
-------
1 measured NO2 studies gave an estimated odds ratio of 1.27, whereas the others gave an
2 estimate of 1.18, which is consistent with a measurement error effect.
3 The individual evidence of the effect of NO2 on respiratory disease is somewhat mixed.
4 All but one of the studies used in the synthesis showed increased respiratory disease rates
5 with increased NO2 exposure. A few of the individual studies were statistically significant.
6 Combining the studies giving quantitative estimates of effects tends to show that increases of
7 respiratory illness in children are associated with long-term exposure to NO2. When
fj
8 combined, the studies indicated that an increase in NO2 exposure of 30 ^g/m (0.016 ppm)
9 would result in an increase of about 20% in respiratory disease, subject to the assumptions
10 made for the synthesis. Although several assumptions were made to combine the studies, the
11 consistency between the individual studies is demonstrated, giving greater strength to the data
12 base and suggesting that the effect is real. Furthermore, the estimated effect is almost surely
13 an underestimate, given the problems of misclassification of exposures and outcomes. The
14 effect was not dependent on any one or two studies, nor are the results of this analysis
15 sensitive to the inclusion (or exclusion) of any one study. In fact, any two studies can be
16 eliminated, and the 95% confidence limits will exclude the no effect odds ratio of 1.0. Thus,
17 the combined evidence is supportive for the effects of exposure to NO2 on respiratory disease
18 in children under 12 years of age.
19 Only the Harvard Six City study attempted to relate some measure of indoor and
20 outdoor NO2 exposure to long-term changes in pulmonary function. These changes were
21 marginally significant. No short-term studies had indoor exposures. Most studies did not
22 find any effects, which is consistent with the clinical data (see Chapter 15). However, the
23 basic conclusion is that there is insufficient epidemiological evidence to make any conclusion
24 about the long- or short-term effects of NO2 on pulmonary function.
25
26 1.4.3 Controlled Human Exposure Studies of Oxides of Nitrogen
27 Nitrogen dioxide exposure at sufficiently high concentrations produces changes in lung
28 function in healthy subjects. A number of investigators have reported increased airway
29 resistance after exposure to NO2 concentrations exceeding 1,880 ptg/m3 (1 ppm) (Beil and
30 Ulmer, 1976; von Nieding et al., 1979; von Nieding and Wagner, 1977; von Nieding et al.,
31 1980). However, at NO2 concentrations between 3,760 and 7,520 /ig/m3 (2 and 4 ppm),
August 1991 1-67 DRAFT-DO NOT QUOTE OR CITE
-------
1 some investigators have not observed any NO2-induced changes in airway resistance or
2 spirometry (Linn et al., 1985b; Mohsenin, 1987b; Mohsenin, 1988; Sandstrom et al., 1990a).
3 At NO2 exposure concentrations below 1,880 ^tg/m3, there is little if any convincing evidence
4 of change in lung volumes, flow-volume characteristics of the lung, or airways resistance in
5 healthy subjects. Nitrogen dioxide is believed to have its primary effect on small airways.
6 However, routine spirometry and airway resistance measurements are not sensitive indicators
7 of small airways function. Thus, the absence of change in these physiological indicators of
8 large airways function at low NO2 concentrations should not be viewed as evidence that NO2
9 has no effects on lung function. Further developments will be necessary to permit sensitive,
10 reproducible, noninvasive evaluation of small airways, the primary site of NO2 deposition in
11 the lung.
12 Symptoms associated with NO2 exposure in healthy subjects have been limited to
13 detection of odor (reportedly as low as 188 pig/m3 [0.1 ppm]) (Bylin et al., 1985). Few of
14 the studies examined in this review note a significant increase in respiratory symptoms.
15 Sandstr5m et al. (1990a) noted mild nasopharyngeal irritation after exposure to 7,520 /xg/m3
16 (4 ppm) for 20 min.
17 Nitrogen dioxide exposure does result in increased airway responsiveness in normal
18 subjects exposed to concentrations in excess of 1,880 jig/m3 (1.0 ppm). Mohsenin (1987b)
19 and Frampton et al. (1991) reported an increase in airway responsiveness after exposure to
20 3,760 and 2,820 jtg/m3 (2.0 and 1.5 ppm), respectively. Increased airway responsiveness
21 may be associated with airway inflammation. Repeated bouts of airway inflammation could
22 promote deleterious long-term changes in the lung such as the loss of elasticity and
23 acceleration of age-related changes in lung function. However, the development of such
24 responses is only speculative, given the present level of scientific evidence.
25 Potentially sensitive subjects in the population include children, older adults, patients
26 with asthma or chronic obstructive pulmonary disease (COPD), or individuals who may be
27 unusually sensitive to NO2 for other reasons. There are insufficient data on children,
28 adolescents or older adults, either healthy or with asthma, to determine their NO2
29 responsiveness relative to healthy young adults.
A
30 At the concentrations that may fall within the ambient range (e.g., < 1,880 ^tg/m
31 [1.0 ppm]), the effects of NO2 on lung function (i.e., spirometry, airway resistance) in
August 1991 1-68 DRAFT-DO NOT QUOTE OR CFTE
-------
1 asthmatics have tended to be small. For example, Bauer et al. (1986) observed a 4 to 6%
2 decline in FEVj 0 in asthmatics exposed to 564 fAg/m3 (0.3 ppm) NO2 for 30 min. Koenig
3 et al. (1988) reported a 4% decrease in FVC, but no significant change in other spirometry
*3t
4 variables, after exposure of adolescent asthmatics to 564 /*g/m NO2. On the other hand,
5 several other investigators (Avol et al., 1988; Bylin et al., 1985; Hazucha et al., 1982, 1983;
6 Kleinman et al., 1983; Koenig et al., 1985; Linn et al., 1985b, 1986; Mohsenin, 1987a;
7 Roger et al., 1990) have not found any significant changes in spirometry or airway resistance
8 of asthmatics exposed to concentrations < 1,880 /ttg/m3 (1.0 ppm). Again, spirometry and
9 airway resistance are not sensitive measures of small airways function, where NO2 is known
0 to be primarily deposited.
1 A second important category of sensitive subjects includes patients with COPD who
2 have shown increased airway resistance after brief exposures to greater than 3,000 /*g/m3
3 (1.6 ppm) NO2 (von Nieding et al., 1970, 1971, 1973a). In addition, during a longer (4-h)
4 exposure, Morrow and Utell (1989) reported decreased (approx. 5%) FVC in COPD patients
5 exposed to 564 jttg/m3 (0.3 ppm). Other investigators (Linn et al., 1985a; Kerr et al., 1979)
.6 did not find responses in COPD patients even with exposures to levels as high as
o
.7 3,760 /ig/m (2.0 ppm). It appears that brief acute exposure to relatively high concentrations
.8 of NO2 (> 3,760 /ig/m3) will cause bronchoconstriction in some COPD patients and that
[9 these responses may also be observed with longer exposures to lower concentrations.
IQ An unresolved issue with the current data base is the existence of NO2-induced
11 pulmonary responses in asthmatics that have been reported at low but not at high
12 NO2 exposures. Although small functional responses have been observed in studies from
13 various laboratories, effects are not consistently present and demonstrating reproducibility of
14 responses has been difficult, even within the same laboratory. Furthermore, all responses to
25 NO2 that have been observed in asthmatics have occurred at concentrations between 376 and
"3
26 940 /Ltg/m (0.2 and 0.5 ppm). Changes in lung function or airway reactivity have not been
27 seen even at much higher concentrations (i.e., up to 7,520 ^g/m3 [4 ppm]). There is, at
28 present, no plausible explanation for this apparent lack of a concentration-response
29 relationship. There is a possibility that a portion of the variability in response to NO2 may
30 be attributed to differences in the severity of asthma. This is a complex issue and is
31 discussed in Appendix A of Chapter 15. In patients with chronic obstructive lung disease,
August 1991 1-69 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bauer et al. (1987) and Morrow and Utell (1989) have observed decreased lung function
2 (FVC, FEVLO) after exposure to 564 jig/m3 for 4 h but Linn et al. (1985a) and von Nieding
3 and Wagner (1979) found no effects in COPD patients below 3,760 />tg/m3 (2.0 ppm) for
4 short duration exposures. It appears that further work will be necessary to provide enough
5 information to estimate the concentration-response relationships for NO2 exposure of
6 asthmatics and COPD patients, who appear to be the sensitive subpopulations.
7 In several studies of asthmatics exposed to NO2, airway responsiveness to a variety of
8 agents has been demonstrated. However, in many other studies using similar experimental
9 exposures, there was no significant change in airway responsiveness. In order to evaluate
10 this apparent dilemma, a meta-analysis was utilized; the approach is described in
11 Section 15.4. Without regard to the type of airway challenge, NO2 concentration, exposure
12 duration, or other variables, the overall trend was for airway responsiveness to increase (59%
13 of 354 subjects increased). This trend was somewhat more convincing for exposures
14 conducted under nonexercising conditions (69% of 154 subjects increased); indeed the excess
15 positive responses were almost entirely accounted for by exposures conducted under resting
16 conditions. The implications of this overall trend are unclear and will require further
17 investigations to verify if there is an interaction with exercise-induced changes in lung
18 function that may possibly obscure changes in airway responsiveness due to NO2 exposure.
19 Increased airway responsiveness could potentially lead to temporary exacerbation of asthma
20 leading possibly to increased medication usage or even increased hospital admissions. The
21 lowest-observed-effect level for this response appears to be in the 376 to 564 fig/m3 (0.2 to
22 0.3 ppm) range.
23 Several recent studies have examined the possibility that NO2 could induce a pulmonary
24 inflammatory response and/or alter immune system host defenses. These studies typically
25 include collection of cells and airways fluids washing from the lung using BAL. In contrast
26 to O3 exposure, NO2 does not, at the concentrations studied, induce an increase in
27 neutrophils or eosinophils, the typical markers of inflammation following O3 exposure,
28 However, Sandstrom et al. (1990a) have observed an increase in mast cells and lymphocytes
29 in BAL fluid, which they attribute to an unspecific inflammatory response. Boushey et al.
30 (1988) have reported an increase in natural Mller lymphocytes in BAL fluid. Macrophage
31 numbers have not been increased by NO2 exposure nor did their ability to kill virus appear to
August 1991 1-70 DRAFT-DO NOT QUOTE OR CITE
-------
1 have been altered by exposure, although Frampton et al. (1989a) suggested that, in some
2 subjects, macrophage responses may have been impaired. At present there is no evidence of
3 increased pulmonary epithelial permeability, although this possibility has not been examined
4 systematically. Mucociliary clearance was not altered after NO2 exposure in the one study in
5 which it was measured (Rehn et al., 1982). Nitrogen dioxide was found to cause a reduction
6 in alpha-1-antiprotease activity in one study (Mohsenin and Gee, 1987) but not in another
7 (Johnson et al., 1990). Following NO2 exposure, Frampton et al. (1989b) found an increase
8 in alpha-2-macroglobulin, a molecule that has immunoregulatory as well as antiprotease
9 activity. Immunological responses to NO2 exposure are just beginning to be elucidated and
0 additional research will be required to determine whether these responses have any
1 implications for epidemiologically determined associations between NO2 exposure and
2 increased respiratory tract infections.
3 The effects of repeated NO2 exposure have been examined in two studies (Sandstrom
4 et al., 1990b; Boushey et al., 1988). Boushey et al. (1988) reported only a slight increase
5 (12%) in circulating lymphocytes and a possible increase in natural killer lymphocytes after
1J
.6 four 2-h exposures to 1,128 ftg/m (0.6 ppm). There were no detectable changes in
.7 inflammatory mediators. Sandstrom et al. (1990b), on the other hand, found decreased
.8 numbers of mast cells, macrophages, and lymphocytes in the BAL fluid. Despite the
.9 decreased numbers, the phagocytic activity of alveolar macrophages was enhanced. These
IQ observations suggest that host defense responses are different after repeated exposure than
II after a single acute exposure. More research appears to be necessary to confirm and expand
12 these observations because of the important potential connection between altered host defense
13 responses and increased respiratory infectivity. '
14 In healthy adults, a variety of mixtures of other pollutants with NO2 have been
25 examined, primarily using spirometry and airway resistance measurements as end points.
26 In general, NO2 does not significantly alter responses to other pollutants, such as O3, SO2, or
27 paniculate matter. In other words, there is no more than an additive response when NO2 is
28 included in the pollutant mixture. However, further investigation of NO2 mixtures appears
29 warranted using other biological markers, including measures of epithelial permeability,
30 clearance, airway responsiveness, airway inflammation, and measures that are sensitive to
31 changes in small airways function. In asthmatics, there is a tendency for increased
August 1991 1-71 DRAFT-DO NOT QUOTE OR.CJTE
-------
1 responsiveness to cold air, methacholine, carbachol, and histamine after NO2 exposure (see
2 previous discussion). In one study, asthmatics were also more responsive to SO2 after a
3 previous exposure to NO2 (Torres and Magnussen, 1990). In addition to other pollutants,
4 NO2 exposure could potentially alter responses to other substances, particularly airborne
5 antigens. In two studies (Ahmed et al., 1983a; Orehek et al., 1981) the response to grass
6 pollen inhalation was examined in sensitive subjects after exposure to 188 jtg/m3 (0.1 ppm)
7 N02 but no significant difference in the response to air or NO2 exposure was observed.
8 Given the increase in responsiveness to nonantigenic substances such as methacholine,
9 histamine, SO2, or cold air discussed previously, it may be worthwhile to re-examine this
10 hypothesis using higher NO2 concentrations or more prolonged exposures.
11 Responses to other NOX species have also been studied. Nitric oxide does not appear to
12 cause any change in lung function at low concentrations (< 1.0 ppm) either alone (Kagawa,
13 1982) or combined with NO2 (Kagawa, 1990). Also, von Nieding et al. (1973b) reported
14 increased airways resistance in subjects exposed to excessively high concentrations
15 (>20 ppm). Responses to HNO3 "vapor, have been studied in adolescent asthmatics (Koenig
16 et al., 1989a,b). The results are suggestive of small changes in lung function but further
17 investigation is needed to confirm these apparent responses to HNO3 vapor. Nitrates (e.g.,
18 sodium nitrate) have not been found to cause any deleterious effects (Utell et al., 1979, 1980;
19 Kleinman et al., 1980; Stacy et al., 1983) at levels that might be expected in the atmosphere.
20
21 Conclusions:
22 1. Nitrogen dioxide causes decrements in lung function, particularly increased airway
23 resistance in healthy subjects at concentrations exceeding 1,880 /ig/m3 (1.0 ppm).
24
25 2. Nitrogen dioxide exposure results in increased airway responsiveness in healthy subjects
26 exposed to concentrations exceeding 1,880 jig/m3 for exposure durations of 1 h or
27 longer.
28
29 3. Nitrogen dioxide exposure is associated with cellular inflammatory responses in the
30 airways that may include increased levels of mast cells and lymphocytes but not
31 neutrophils and eosinophils. Changes in some biochemical mediators of inflammation
32 or enzymes may be altered by NO2 exposure.
33
34 4. Nitrogen dioxide exposure of asthmatics causes in some subjects increased airway
35 responsiveness to a variety of provocative mediators, including cholinergic and
36 histaminergic chemicals, SO2, and cold air. However, the presence of these responses
August 1991 1-72 DRAFT-DO NOT QUOTE OR CITE
-------
1 appears to be influenced by the exposure protocol, particularly whether or not the
2 exposure includes exercise. However, NO2 concentration-response relationships are not
3 evident.
4
5 5. Modest decrements in spirometric measures of lung function (3 to 8%) may occur in
6 asthmatics and COPD patients exposed to NO2 concentrations (564 /*g/m3 [0.3 ppm]).
7
8 6. Nitric acid exposure may cause some pulmonary function responses in asthmatics '.,.•£
9 other commonly occurring NOX species do not appear to cause any function responses at
10 concentrations expected or at higher levels than in worst-case scenarios in the ambient
11 environment.
12
13
14
15 1,4.4 Discussion of Health Effects Associated With Exposure to Nitrogen
16 Oxide
17
18 This section concisely summarizes and integrates key information and conclusions from
19 preceding discussions into a coherent framework or perspective upon which interpretations
20 can be based concerning human health risks posed by ambient or near-ambient levels of NO2
21 in the United States. This section discusses qualitative and quantitative characterization of
22 key health effects of NO2 and their biological bases and identification of population groups
23 potentially at enhanced risk for health effects associated with NO2 exposure.
24
25 1.4.4.1 Airway Reactivity in Asthmatics and Exposure to NO2
26 Subjects with asthma who have airway hyperresponsiveness to a variety of chemical and
27 physical stimuli are considered one of the potentially most NO2-responsive groups in the
28 population. The physiological end point, which to date appears to be the most sensitive
29 indicator of response in asthmatics, is a change in airway responsiveness. An extensive
30 discussion of these responses is presented in Chapter 15. Asthmatics as a group are
31 significantly more responsive than healthy normal subjects to a variety of airway challenges.
32 The differences in airway responsiveness may span several orders of magnitude (at least
33 100 fold) between normal and asthmatic individuals (O'Connor et al., 1987).
34 There is some evidence that NO2 may cause an increase in airway responsiveness in
35 asthmatics. This response has been observed only at relatively low NO2 concentrations,
36 mostly within the range of 376 to 564 /*g/m3 (0.2 to 0.3 ppm) NO2, the concentration range
37 of concern within the ambient environment. A meta-analysis of the over 300 asthmatics
„ •*• ~~ N
August 1991 1-73 DRAFT-DO NOT QUOTEi OR CITE
-------
1 exposed to NO2 indicates a slight excess increase in airway responsiveness following NO2
2 exposure (see Chapter 15). In the concentration range between 376 and 564 peg/m3, the
3 excess increase in airway responsiveness was attributable to subjects exposed to NO2 at rest.
4 The change in nonspecific (i.e., not specific allergens) airway responsiveness may reflect
5 increased permeability of the airway epithelium leading to increased access of the provocative
6 agent to airway receptors, release of local mediators of inflammation, or alterations in airway
7 smooth muscle tone. Since NO2 does not appear to cause airway inflammation at these levels
8 and the increase in airway responsiveness appears to be fully reversible, the implications of
9 the observed increase in responsiveness are unclear. Although it is conceivable that increased
10 nonspecific airway responsiveness caused by NO2 could lead to increased responses to a
11 specific antigen, there is presently no evidence to support this hypothesis. It is also possible
12 that persistence of airway hyperresponsiveness may be associated with an accelerated rate of
13 decline in pulmonary function (O'Connor et al., 1987).
14 An unresolved issue with the current data base is the existence of NO2-indueed
15 pulmonary responses in asthmatics that have been reported at low but not at high NO2
16 exposure concentrations. Although small functional responses have been observed in studies
17 from various laboratories, effects are not consistent and reproducibility of responses has been
18 difficult, even within the same laboratory. Furthermore, all responses to NO2 that have been
19 observed in asthmatics have occurred at exposure concentrations between 376 and 940 jtg/m3
20 (0.2 and 0.5 ppm). Changes in lung function or airway responsiveness have not been seen
21 even at much higher concentrations (i.e., up to 7,520 ^g/m3 [4 ppm]). There is, at present,
22 no plausible explanation for this apparent lack of a concentration-response relationship for
23 both airway responsiveness and pulmonary function changes.
24 Controlled human exposure studies are limited to acute fully reversible functional and/or
25 symptomatic responses. This may in many cases limit the magnitude of expected responses
26 and, hence, the statistical significance of responses in studies with small numbers of subjects.
27
28 1.4.4.2 Respiratory Morbidity in Children Associated with Exposure to NO2
29 The effects of NO2 on respiratory illness and the factors determining occurrence
30 and severity are important public health concerns because the potential for exposure to
31 NO2 and childhood respiratory illness is common (Samet et al., 1983; Samet and Utell,
August 1991 1-74 DRAFT-DO NOT QUOTE OR CITE
-------
1 1990). This takes on added importance since recurrent childhood respiratory illness may be a
2 risk factor for later susceptibility to lung damage (Glezen, 1989; Samet et al., 1983; Gold
3 • et al., 1989).
4 The discussion on epidemiological findings in Chapter 14 indicates that the combined
5 evidence is supportive for effects of exposure to NO2 on respiratory disease.in children under
6 12 years of age. Combining the studies giving quantitative estimates of effects tend to show
7 increases in respiratory illness among children associated with exposure to NO2. Combining
8 the studies as if the end points were similar gives an estimated odds ratio of 1.20 (95%
9 confidence limits of 1.1 and 1.3) for gas stove use and increased respiratory illness. The
10 EPA meta-analysis indicates that, when combined, the studies collectively provide evidence
11 that an increase in NO2 exposure of 30 /ig/m3 (0.016 ppm) is associated with about a 20%
12 increase in the odds of respiratory illness, subject to the assumptions made for the synthesis.
13 These studies can not distinguish between the relative contribution of short-term peak
14 exposures and longer-term average exposures to the observed health effect.. Although several
15 assumptions were made, the consistency between the individual studies adds greater strength
16 to the data base. That NO2 is a major causative factor in the increased incidence of
17 respiratory illness reported in the epidemiological studies is also implied by the animal
18 lexicological literature (see Chapter 13).
19
20 1.4.4.3 Biological Basis Relating NO2 Exposure to Respiratory Morbidity: Effects of
21 NO2 on the Respiratory Host Defense System
22 The biological basis relating NO2 exposure to respiratory symptoms and infection shown
23 in epidemiological studies are presented in the extensive discussion of effects on host defense
24 system in Chapter 13 (animal toxicology). Nitrogen dioxide exposure can impair one or
25 more components of this important defense system, resulting in the host being more
26 susceptible to respiratory infection. Epidemiological studies have reported an association
27 between an increase in symptoms of respiratory disease of infectious origin and NO2
28 exposure (Chapter 14). Animal studies provide important evidence indicating that several
29 defense systems are a target organ for inhaled NO2. Nitrogen dioxide affects all of the host
30 defenses studied (e.g., mucociliary clearance, macrophages, and the humoral and cell-
31 mediated immune system).
August 1991 1-75 DRAFT-DO NOT QUOTE OR CITE
-------
1 Studies on the effects of NO2 exposure on mucociliary transport in the conducting
2 airways show a reduction in the number and activity of the cilia, morphological changes in
3 the cilia, and decreased mucociliary velocity at concentrations as low as 940 /*g/m3 (0.5 ppm)
4 after 7 months of exposure (Yamamoto and Takahashi, 1984). Others have observed similar
5 effects at higher levels for shorter exposure durations. Several short-term (days to weeks)
6 studies with concentrations at or above 3,760 /*g/m3 (2.0 ppm) demonstrated structural
7 changes in cilia and ciliated cells, decreases in numbers of ciliated cells, and decreases in
8 ciliary beating. As a foreign agent deposits below the mucociliary region, in the gaseous
9 exchange region of the lung, host defenses are provided by the alveolar macrophage which
10 acts to remove or kill viable particles, to remove nonviable particles, and to process and
11 present antigens to lymphocytes for antibody production. Exposure to NO2 has produced a
12 variety of effects on alveolar macrophages in several animal species after several weeks of
13 exposure to levels as low as 564 uglm3 (0.3 ppm); however, most effects were observed at
14 higher levels. These effects included decreased phagocytosis and bactericidal activity, altered
15 metabolism, increases in numbers of macrophages, and morphological changes (Rombout
16 et al., 1986; Aranyi et al., 1976; Goldstein et al., 1974; Suzuki et al., 1986; Chang et al,
17 1986; SuzuM et al., 1986; Schlesinger et al., 1987; Mochitate et al., 1986). Decreases in the
18 ability of alveolar macrophages to engulf foreign particles (phagocytosis) and bactericidal
19 activity are likely highly related to increased susceptibility to pulmonary infections.
20 Controlled human exposure studies have also examined macrophage function and show that
21 these cells, when exposed to NO2, tended to inactivate influenza virus in vitro less effectively
22 than cells collected after air exposure (Frampton et al., 1989a).
23 An example of the alteration of host defenses against tumor cells following exposure to
24 NO2 is seen in studies examining the colonization of the lung by B16 melanoma tumor cells.
25 Richters and Kuraitis (1981, 1983) and Richters et al. (1985) reported that when mice were
26 exposed to 752 or 1,504 jttg/m3 (0.4 or 0.8 ppm) NO2 for 10 to 12 weeks and then injected
27 with transplantable tumor cells, colonization of these cells increased in the lung, indicating
28 decreased resistance to such challenge. Questions have been raised regarding the
29 appropriateness of the statistical analysis used in evaluating these data and the interpretation
30 of the model used. Although the authors attribute effects to increased metastases, an
31 alternative interpretation is an NO2-induced change in lung permeability to cells. Another
August 1991 1-76 DRAFT-DO NOT QUOTE OR CITE
-------
1 study reported that NO2 inhibits formation of B16F10 tumors in the lung (Weinbaum et al.,
2 1987).
3 The humoral and cell-mediated immune systems together are essential for antibody
4 production and the secretion of cellular products that are lethal to certain invading organisms
5 and also regulates the normal host defense response. There is some indication that exposure
6 to NO2 suppresses some of these specific immune responses and that the effect is both
7 concentration-and time-dependent. For example, a significant suppression of antibody
8 production by spleen cells has been reported in experimental animals exposed for 1 week to
9 1 month to NO2 concentrations as low as 752 to 940 jig/m3 (0.4 to 0.5 ppm) (Lefkowitz et
10 al., 1986; Fujimaki et al., 1982). Subchronic exposure to NO2 also resulted in decreased
11 numbers of circulating T lymphocytes, T helper/inducer lymphocytes, and
12 T cytotoxic/suppressor lymphocytes in mice (Damji and Richters, 1989). The cause of this
13 suppression is not clear.
14 Animal infectivity studies present key data relating exposure to NO2 and effects on the
15 overall functioning of host defense mechanisms. In these studies, animals were exposed to
16 varying concentrations and durations of NO2 followed by exposure to an aerosol of an
17 infectious agent. Microbial-induced mortality was used as the end point. Exposure to NO2
18 increased both bacterial- and influenza-induced mortality after subchronic exposures to levels
19 as low as 940 to 1,880 /ug/m3 (0.5 to 1.0 ppm) NO2 (Ehrlich and Henry, 1968; Ito, 1971;
20 Ehrlich et al., 1977). After acute (2 h) exposure, 3,760 ng/m3 (2.0 ppm) NO2 is the lowest
21 effective concentration measured (Ehrlich et al., 1977). Nitrogen dioxide increases
22 microbial-induced mortality by impairing the host's ability to defend the respiratory tract
23 from infectious agents, thereby increasing susceptibility to viral, mycoplasma, and bacterial
24 infections (Ehrlich and Henry, 1968; Ito, 1971; Ehrlich et al., 1977; Parker et al., 1989;
25 Gardner et al., 1977a,b, 1979, 1980, 1982; Graham et al., 1987; Jakab, 1987; Motomiya et
26 al., 1973; Miller et al., 1987). Using an animal model designed to evaluate the effects of
27 NO2 on non-fatal respiratory infection, NO2 decreased the intrapulmonary bactericidal
28 activity in mice in a concentration-related manner, without a decrease in the number of
29 alveolar macrophages (Goldstein et al., 1973). Exposure to NO2 was found to increase the
30 severity of mycoplasma-induced lesions within the lung, but did not increase the susceptibility
31 of the mice to the infection (Parker et al., 1989). Animal studies have also shown that
August 1991 1-77 DRAFT-DO NOT QUOTE OR CITE
-------
1 influenza infection is exacerbated with NO2 exposure (Ito, 1971). In studies with
2 cytomegalovirus and paramyxovirus (Jacob, 1987; Rose et al., 1988), the pathogenesis of
3 these infections are enhanced.
4 The strength of the animal lexicological, human clinical, and epidemiological studies on
5 host defenses provides the rationale and plausible biological basis to explain the relationship
6 seen in population studies showing increased frequency and severity of respiratory symptoms
7 and/or infections in humans exposed to NO2. The human body must be able to defend itself
8 against a wide variety of inhaled foreign substances. When these defenses are overcome,
9 serious consequences can occur. The significance of the observations from relevant animal
10 models is clear. With minor variations, the mammalian species, including hurnans, share in
11 common an array of defensive mechanisms that are anatomically, functionally, and
12 physiologically integrated in the respiratory tract to prevent and control infectious disease.
13 All information available to date would indicate that qualitative extrapolation of data observed
14 in animals to humans is valid. The accumulated evidence that NO2 causes dysfunction in
15 several defenses provides the plausible biological mechanism necessary to link NO2 exposure
16 to increased morbidity and respiratory symptomatology in children indicative of respiratory
17 infection.
18
19 1.4.4.4 Emphysema and Exposure to NO2
20 Studies in several species of animals have shown that chronic exposure to high levels of
21 NO2 (relative to ambient) can cause emphysema. Since emphysema is an irreversible disease,
22 representing important public health concerns, whether NO2 creates a risk for this disease in
23 humans is a major question. Although this question cannot be definitely answered, the
24 potential for risk requires discussion. The definition of emphysema as used in the United
25 States is an anatomic one best characterized by National Institutes of Health (NIH) (1985)
26 criteria. These criteria are: "An animal model of emphysema is defined as an abnormal state
27 of the lungs in which there is enlargement of the airspace distal to the terminal bronchiole.
28 Airspace enlargement should be determined qualitatively in appropriate specimens and
29 quantitatively by stereologic methods." An additional essential criterion for human
30 emphysema is the destruction of alveolar walls.
August 1991 1-78 DRAFT-DO NOT QUOTE OR CITE
-------
1 Several studies (Haydon et al., 1967; Freeman et al., 1977; Port et al., 1977) relate
2 long-term (1 to over 30 months) exposure of rats and rabbits to high concentrations of
3 NO2 (> 15,040 /ig/m3 [8 ppm], much greater than ambient levels) with morphologic lung
4 lesions which meet the 1985 NIH workshop criteria for a human model of emphysema (i.e.,
5 alveolar wall destruction occurred in addition to other characteristic changes). One study
5 (Hyde et al., 1978) reported on dogs exposed to a mixture of 1,203 /zg/m3 (0.64 ppm) NO2
7 and 0.25 ppm NO for 68 months. Upon examination 32 to 36 months after exposure ceased,
3 the dogs had morphologic lesions that meet the 1985 NIH workshop criteria for human
•) emphysema. In the same dogs, pulmonary function was also measured. Pulmonary function
3 decrements observed at the end of exposure progressed postexposure. This suggests that the
1 morphological effects may also have been progressive. Another group of dogs in the same
I study was exposed to a mixture of "low" NO2 (263 ^g/m3 [0.14 ppm]) and "high" NO
3 (1.1 ppm), but emphysema was not observed. Since the study did not include an NO2-only
4 group, it is not possible to discern the effects of NO2 in the mixture. However, the presence
5 of emphysema in the "high" NO2-"low" NO group and its absence in the "low" NO2 - "high"
5 NO group implies that NO2 was a significant etiologic factor.
7 Emphysema was reportedly observed in numerous other NO2 studies with several
3 species of animals, but either the reports lacked sufficient detail for independent conclusions
•) or only the criteria for animal (not human) emphysema were met. Several other studies
3 discussed in Chapter 13 were negative for emphysema. Various factors such as the exposure
1 protocol may also play a role in the outcome of studies. Potential differences may relate to
2 age of the animals during exposure, concentration and duration of exposure, and the duration
3 after exposure ceases before the animals are evaluated for emphysematous pathology.
4- While there is a fairly extensive toxicologic data base concerning morphologic effects of
5 NO2, it is still not possible to establish a reasonably accurate "no-observed-effect" level for
5 emphysema. This is likely due to a combination of factors: the complexity of changes
7 occurring with NO2 exposure, the lack of published papers utilizing highly sensitive
3 morphometric techniques, interspecies difference in response, and inadequate description of
2 methods and findings in some published reports. Qualitatively, it is clear that NO2 can
3 cause emphysema in animals. While the lowest NO2 concentration for the shortest
1 exposure duration that will result in emphysematous lung lesions can not be reliably
August 1991 1-79 DRAFT-DO NOT QUOTE OR CITE
-------
1 determined from the published studies, the exposures that did cause emphysema, according to
2 the NIH criteria, are far higher than those currently reported in ambient air.
3
4 1.4.4.5 Subpopulations Potentially Susceptible to NO2 Exposure
5 Certain groups within the population may be more susceptible to the effects of NO2
6 exposure, including persons with preexisting respiratory disease, children, and the elderly.
7 The reasons for paying special attention to these groups is that they may be affected by lower
8 levels of NO2 than other subpopulations or the impact of a given response may be greater.
9 Some causes of heightened susceptibility are better understood than others. Subpopulations
10 that already have reduced lung functions and reserves (e.g., the elderly, asthmatics,
11 emphysemics, chronic bronchitics) will be more impacted than other groups by decrements in
12 pulmonary function. For example, a young healthy person may not even notice a small
13 percentage change in pulmonary function, but a person whose activities are already limited by
14 reduced lung function may not have the reserve to compensate for the same percentage
15 change.
16 The airways of asthmatics may be hyperresponsive to a variety of inhaled materials
17 including pollens, cold-dry air, allergens, and air pollutants. Asthmatics have the potential to
18 be among the most susceptible members of the population with regard to respiratory
19 responses to NO2 (Section 15.3.1). On the average, asthmatics are much more sensitive to
20 inhaled bronchoconstrictors such as histamine, methacholine, or carbachol. The potential
21 addition of an NO2-induced increase in airway response to the already heightened
22 responsiveness to other substances raises the possibility of exacerbation of this pulmonary
23 disease by NO2. This is discussed in Section 15-.4.
24 Other potentially susceptible groups include patients with COPD, such as emphysema
25 and chronic bronchitis. Many of these patients have airway hyperresponsiveness to physical
26 and chemical stimuli. A major concern with COPD patients is the absence of an adequate
27 pulmonary reserve, a susceptibility factor described above. The poor distribution of
28 ventilation in COPD may lead to a greater delivery of NO2 to the segment of the lung that is
29 well ventilated, thus resulting in a greater regional tissue dose.
30 Since over 2 million Americans have emphysema, it would be important to know
31 whether NO2 has the potential to excaberate the disease. Lafuma et al. (1987) exposed both
August 1991 1-80 DRAFT-DO NOT QUOTE OR CITE
-------
I normal hamsters and hamsters with laboratory-induced (with elastase) emphysema to
I 3,760 Mg/m3 (2 ppm) NO2 for 8 h/day, 5 days/week for 8 weeks. The emphysematous
} lesions produced by elastase and NO2 showed increases in mean linear intercepts and
1 pulmonary volumes and a decrease in internal alveolar surface areas, compared to those
5 treated with elastase and exposed to clean air. The NO2-exposed animals developed a
j different type of emphysematous lesions than those caused by elastase. The investigators
1 suggested that the results may imply a role for NO2 in enhancing preexisting emphysema.
\ Stavert et al. (1986) state that whereas NO2 inhalation alone may not cause emphysema, it is
) conceivable that this oxidant gas may act as a co-determinant to allow the progression or
) amplification of emphysema caused by other means. However, it is not clear what the
I potential would be for exacaberation of emphysema in humans at ambient concentrations.
I Based upon epidemiology studies, children 12 years or younger constitute a
^ subpopulation potentially susceptible to an increase in respiratory morbidity associated with
I NO2 exposure (Section 14.5). Children may be more susceptible due to an immature host
)" defense system. Data on the resident population of the United States provides information on
) the numbers of children in various age ranges. Approximately 37 million children are in the
7 age group 9 years and younger, while around 54 million children are in the age group 14
? years and younger. It is also possible that the increase in respiratory morbidity may be more
) detectible in this age group compared to adults due to the greater frequency of lower
) respiratory illness in this age group of children (Glezen and Denny, 1973).
August 1991 1-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2
3 Abbott, W. H. (1987) Corrosion of porous gold plating in field and laboratory environments. Plat. Surf. Finish.
4 74: 72-75.
5
6 Aber, J. D.; Nadelhoffer, K. J.; Steudler, P.; Melillo, J. M. (1989) Nitrogen saturation in northern forest
7 ecosystems: excess nitrogen from fossil fuel combustion may stress the biosphere. Bioscience
8 39: 378-386.
9
10 Abraham, W. M.; Welker, M.; Oliver, W., Jr.; Mingle, M.; Januszkiewicz, A. J.; Wanner, A.; Sackner, M. A.
11 (1980) Cardiopulmonary effects of short-term nitrogen dioxide exposure in conscious sheep. Environ.
12 Res. 22: 61-72.
13
14 Acton, J. D.; Myrvik, Q. N. (1972) Nitrogen dioxide effects on alveolar macrophages. Arch. Environ. Health
15 24: 48-52.
16
17 Agren, G. L; Bosatta, E. (1988) Nitrogen saturation of terrestrial ecosystems. Environ. Pollut. 54: 185-197.
18
19 Ahlquist, N. C.; Charlson, R. J. (1969) Measurement of the wavelength dependence of atmospheric extinction
20 due to scatter. Atmos. Environ. 3: 551-564.
21
22 Ahmed, T.; Dougherty, R.; Sackner, M. A. (1983) Effect of NO2 exposure on specific bronchial reactivity in
23 subjects with allergic bronchial asthma [final report]. Warren, MI: General Motors Research
24 Laboratories; contract report no. CR-83/07/BI.
25
26 Amoroso, G. G.; Fassina, V. (1983) Stone decay and conservation: atmospheric pollution, cleaning,
27 consolidation and protection. Amsterdam, The Netherlands: Elsevier Science Publishers B.V. (Materials
28 science monographs 11).
29
30 Anderson, I. C.; Levine, J. S. (1987) Simultaneous field measurements of biogenic emissions of nitric oxide and
31 nitrous oxide. J. Geophys. Res. [Atmos.] 92: 965-976.
32
33 Andreae, M.; Delany, A. C.; Liu, S.; Logan, S.; Steele, L. P.; Westberg, H.; Zika, R. (1989) Key aspects of
34 species related to global biogeochemical cycles. In: Lenschow, D. H.; Hicks, B. B., eds. Global
35 tropospheric chemistry: chemical fluxes in the global atmosphere. Boulder, CO: NCAR.
36
37 Appel, B. R.; Tokiwa, Y.; Haik, M. (1981) Sampling of nitrates in ambient air. Atmos. Environ. 15: 283-289.
38
39 Appcl, B. R.; Tokiwa, Y.; Hsu, J.; Kothny, E. L.; Hahn, E. (1985) Visibility as related to atmospheric aerosol
40 constituents. Atmos. Environ. 19: 1525-1534.
41
42 Aranyi, C.; Fenters, J.; Erhlich, R.; Gardner, D. (1976) Scanning electron microscopy of alveolar macrophages
43 after exposure to oxygen, nitrogen dioxide, and ozone. Environ. Health Perspect. 16: 180.
44
45 Avol, E. L.; Linn, W. S.; Peng, R. C.; Valencia, G.; Little, D.; Hackney, J. D. (1988) Laboratory study of
46 asthmatic volunteers exposed to nitrogen dioxide and to ambient air pollution. Am. Ind. Hyg. Assoc. J.
47 49: 143-149.
48
49 Azoulay, E.; Soler, P.; Blayo, M. C. (1978) The absence of lung damage in rats after chronic exposure to 2 ppm
50 nitrogen dioxide. Bull. Eur. Physiopathol. Respir. 14: 311-325.
51
52 Baird, S. F.; Buso, D. C.; Hombeck, J. W. (1987) Acid pulses from snowmelt at acidic Cone Pond, New
53 Hampshire. Water Air Soil Pollut. 34: 325-338.
54
August 1991 1-82 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Baldocchi, D. (1988) A multi-layer model for estimating sulfur dioxide deposition to a deciduous oak forest
canopy, Atmos. Environ. 22; 869-884.
Baldocchi, D. D.; Hicks, B. B.; Camara, P. (1987) A canopy stomatal resistance model for gaseous deposition to
vegetated surfaces. Atmos. Environ. 21: 91-101.
Barker, J. L.; Witt, E. C. (1990) Effects of acidic precipitation on the water quality of streams in the Laurel Hill
area, Somerset County, Pennsylvania, 1983-1986. Harrisburg, PA: U. S. Geological Survey, Water
Resources Investigations; report 89-4113.
Barrie, L. A.; Sirois, A. (1986) Wet and dry deposition of sulphates and nitrates in eastern Canada: 1979-1982.
Water Air Soil. Pollut. 30: 303-310.
Bauer, M. A.; Utell, M. J.; Morrow, P. E.; Speers, D. M.; Gibb, F. R. (1986) Inhalation of 0.30 ppm nitrogen
dioxide potentiates exercise-induced bronchospasm in asthmatics. Am. Rev. Respir. Dis. 134: 1203-1208.
Bauer, M. A.; Utell, M. J.; Speers, D. M.; Gibb, F. R.; Morrow, P. E. (1987) Effects of 0.30 ppm nitrogen
dioxide on lung function and breathing patterns in subjects with chronic obstructive lung disease (COPD).
Am. Rev. Respir. Dis. 135(suppl.): ASS.
Beil, M.; Ulmer, W. T. (1976) Wirkung von NO2 im MAK-Bereich auf Atemmechanik und bronchiale
AcetylchoUnempfindlichkeit bei Normalpersonen [Effect of NO2 in workroom concentrations on
respiratory mechanics and bronchial susceptibility to acetylcholine in normal persons]. Int. Arch. Occup.
Environ. Health 38: 31-44.
Beloin, N. J. (1972) Fading of dyed fabrics by air pollution. Text. Chem. Color. 4: 77-82. ,
Beloin, N. J. (1973) Fading of dyed fabrics exposed to air pollutants. Text. Chem. Color. 5: 128-133.
Bergstrom, R. W,; Seigneur, C.; Babson, B. L,; Holman, H.-Y.; Wojcik, M. A. (1981) Comparison of the
observed and predicted visual effects caused by power plant plumes. Atmos. Environ. 15: 2135-2150.
Blumenthal, D. L.; Richards, L. W.; Macias, E. S.; Bergstrom, R. W.; Wilson, W. E.; Bhardwaja, P. S. (1981)
Effects of a coal-fired power plant and other sources on southwestern visibility (interim summary of
EPA's project VISTTA). In: White, W. H.; Moore, D. J.; Lodge, J. P., Jr., eds. Plumes and visibility:
measurements and model components, proceedings of the symposium; November 1980; Grand Canyon
National Park, AZ. Atmos. Environ. 15: 1955-1969.
Boring, L. R.; Swank, W. T.; Waide, J. B.; Henderson, G. S. (1988) Sources, fates, and impacts of nitrogen
inputs to terrestrial ecosystems: review and synthesis. Biogeochemistry 6: 119-159.
Boushey, H. A., Jr.; Rubinstein, L; Bigby, B. G.; Stites, D. P.; Locksley, R. M. (1988) Studies on air
pollution: effects of nitrogen dioxide on airway caliber and reactivity in asthmatic subjects; effects of
nitrogen dioxide on lung lymphocytes and macrophage products in healthy subjects; nasal and bronchial
effects of sulfur dioxide in asthmatic subjects. Sacramento, CA: California Air Resources Board; report
no. ARB/R-89/384. Available from: NTIS, Springfield, VA; PB89-183057.
Bowden, W. B. (1986) Gaseous nitrogen emissions from undisturbed terrestrial ecosystems: an assessment of
their impacts on local and global nitrogen budgets. Biogeochemistry 2: 249-279. ,
Boxman, D.; van Dijk, H.; Houdijk, A.; Roelofs, J. (1988) Critical loads for nitrogen with special emphasis on
ammonium. In: Nilsson, J.; Grennfelt, P., eds. Critical loads for sulphur and nitrogen: report from a
workshop; March; Skokloster, Sweden. Copenhagen, Denmark: Nordic Council of Ministers;
pp. 295-322.
August 1991
1-83
DRAFT-DO NOT QUOTE OR CITE
-------
1 Brookshire, D. S,; Ives, B. C,; Schulze, W. D. (1976) The valuation of aesthetic preferences. J. Environ. Econ.
2 Manage. 3: 325-346.
3
4 Brookshire, D, S,; Thayer, M. A,; Schulze, W. D.; d'Arge, R, C, (1982) Valuing public goods: a comparison
5 of survey and hedonic approaches. Am. Econ. Rev. 72: 65-83.
6
7 Brown, K. A.; Freer-Smith, P. H.; Howells, G. D.; Skeffington, R. A.; Wilson, R. B. (1988) Rapporteurs'
8 report on discussions at the workshop on excess nitrogen deposition, Leatherhead, September 1987. In:
9 Dempster, J. P.; Manning, W. J.; Skeffington, R. A., eds. Excess nitrogen deposition: [papers from the
10 workshop]; September 1987; Leatherhead, Surrey, United Kingdom. Environ. Pollut. 54: 285-295.
11
12 Buell, G, R,; Peters, N. E. (1988) Atmospheric deposition effects on the chemistry of a stream in northeastern
13 Georgia, Water Air Soil Pollut. 39: 275-291.
14
15 Buijsman, E.; Erisman, J.-W. (1987) Wet deposition of ammonium in Europe. In: Asman, W. A. H.; Diederen,
16 H. S. M. A., eds. Ammonia and acidification: proceedings of a symposium of the European Association
17 for the Science of Air Pollution (EURASAP); April; Bilthoven, The Netherlands. European Association
18 for the Science of Air Pollution; pp. 202-210.
19
20 Butler, D. A,; Ozkaynak, H.; Billick, I. H.; Spengler, J. D. (1990) Predicting indoor NO2 concentrations as a
21 function of home characteristics and ambient NO2 levels. In: Indoor air '90: precedings of the 5th
22 international conference on indoor air quality and climate, volume 2, characteristics of indoor air;
23 July-August; Toronto, ON, Canada. Ottawa, ON, Canada: International Conference on Indoor Air
24 Quality and Climate, Inc.; pp. 519-524. »
25
26 Bylin, G.; Lindvall, T.; Rehn, T.; Sundin, B. (1985) Effects of short-term exposure to ambient nitrogen dioxide
27 concentrations on human bronchial reactivity and lung function. Bur. J. Respir. Dis. 66: 205-217.
28
29 Cadle, S. H. (1985) Seasonal variations in nitric acid, nitrate, strong aerosol acidity, and ammonia in an urban
30 area. Atmos. Environ. 19: 181-188.
31
32 Carson, R. T.; Mitchell, R. C.; Ruud, P, A. (1990) Valuing air quality improvements: simulating a hedonic
33 equation in the context of a contingent valuation scenario. In: Ma thai, C. V., ed. Visibility and fine
34 particles: an AWMA/EPA international specialty conference; October 1989; Estes Park, CO. Pittsburgh,
35 PA; Air and Waste Management Association; pp. 639-646, (AWMA transactions series no. TR-17).
36
37 Cass, G, R. (1979) On the relationship between sulfate air quality and visibility with examples in Los Angeles.
38 Atmos. Environ. 13: 1069-1084.
39
40 Chang, T. Y.; Norbeck, J. M.; Weinstock, B. (1979) An estimate of the NOX removal rate in an urban
41 atmosphere. Environ. Sci. Technol. 13: 1534-1537.
42
43 Chang, L.-Y.; Graham, J. A.; Miller, F. J.; Ospital, J. J.; Crapo, J. D. (1986) Effects of subchronic inhalation
44 of low concentrations of nitrogen dioxide. I. The proximal alveolar region of juvenile and adult rats.
45 Toxicol. Appl. Pnarmacol. 83: 46-61.
46
47 Charlson, R. J.j Covert, D. S.; Tokiwa, Y.; Mueller, P. K. (1972) Multiwavelength nephelometer measurements
48 in Los Angeles smog aerosol: M, comparison to light extinction by NO2. In: Hidy, G. M., ed. Aerosols
49 and atmospheric chemistry: the Kendall award symposium at the proceedings of the American Chemical
50 Society; March-April 1971; Los Angeles, CA. New York, NY: Academic Press, Inc.; pp. 333-338.
51
52 Chestnut, L. G.; Rowe, R. D. (1990) Preservation values for visibility in the national parks. Washington, DC:
53 U. S, Environmental Protection Agency.
54
August 1991 1-84 DRAFT-DO NOT QUOTE OR CITE
-------
1 Code of Federal Regulations. (1987) Endangered and threatened wildlife and plants. C. F. R. 50: 17.11 & 17.12.
2 '
3 Cosby, B. J.; Ryan, P. F.; Webb, R.; Hornberger, G. M.; Galloway, J. N. (1991) Mountains of West Virginia.
4 In: Charles, D, F., ed. Acidic deposition and aquatic ecosystems: regional case studies. New York, NY:
5 Springer-Verlag,
6 .
7 Cowling, D. W.; Lockyer, D. R. (1981) Increased growth of ryegrass exposed to ammonia. Nature (London)
8 292: 337-338.
9
10 D'Elia, C. F.; Taft, J.; Smullen, J. T.; Madams, J. (1982) Nutrient enrichment. In: Chesapeake Bay Program
11 technical studies: a synthesis. Annapolis, MD: U. S. Environmental Protection Agency; pp. 36-102.
12 Available from: NTIS, Springfield, VA; PB84-111202.
13
14 Damji, K. S.; Richters, A. (1989) Reduction in T lymphocyte subpopulations following acute exposure to 4 ppm
15 nitrogen dioxide. Environ. Res. 49: 217-224.
16
17 DerSimonian, R.; Laird, N. (1986) Meta-analysis in clinical trials. Controlled Clin. Trials 7: 177-188.
18
19 DeWalle, D. R.; Sharpe, W. E.; Edwards, P. J. (1988) Biogeochemistry of two Appalachian deciduous forest
£0 sites in relation to episodic stream acidification. Water Air Soil Pollut. 40: 143-156.
Jl ' - '
12 Dijkstra, L.; Houthuijs, D.; Brunekreef, B.; Akkerman, I.; Boleij, J. S. M. (1990) Respiratory health effects of
13 the indoor environment in a population of Dutch children. Am. Rev. Respir. Dis. 142: 1172-1178.
>4
15 Driscoll, C. T. et al. (n.d.) Processes regulating seasonal variations in the water, chemistry of Adirondack lakes.
t6 Water Air Soil Pollut.: in review.
17
18 Driscoll, C. T.; Wyskowski, B. J.; Cosentini, C. C.; Smith, M. E. (1987a) Processes regulating temporal and
59 longitudinal variations in the chemistry of a low-order woodland stream in the Adirondack region of New
SO York. Biogeochemistry 3: 225-241.
\l
S2 Driscoll, C. T.; Yatsko, C. P.; Unangst, F. J. (1987b) Longitudinal and temporal trends hi the water chemistry
S3 of the north branch of the Moose River. Biogeochemistry 3: 37-61.
S4
(5 Driscoll, C. T.; Schaefer, D. A.; Molot, L, A.; Dillon, P. J. (1989) Summary of North American data. In:
S6 Malanchuk, J. L.; Nilsson, J., eds. The role of nitrogen in the acidification of soils and surface waters,
S7 ' Gotab, Sweden: Nordic Council of Ministers; pp. 6-1 - 6-45,
18
>9 Driscoll, C. T.; Wyskowski, B. J.; DeStaffan, P. (1989) Chemistry and transfer of aluminum in a forested
1-0 watershed in the Adirondack region of New York, USA. In: Lewis, T., ed. The environmental chemistry
H of aluminum. Chelsea, MA: Lewis Publishers; pp. 1-26.
\2
$ Driscoll, C. T.; Newton, R. M.; Gubala, C. P.; Baker, J. P.; Christensen, S. W. (1991) Adirondack mountains.
14 In: Charles, D. F., ed. Acidic deposition and aquatic ecosystems: regional case studies. New York, NY:
f5 Springer-Verlag; pp. 133-202.
^6
17 Drivas, P. J.; Machiraju, S.; Heinold, D. W. (1980) ERT visibility model: version 3, technical description and
r8 user's guide. Concord, MA: Environmental Research & Technology.
:9
>0 Dzubay, T. G.; Clubb, K. W. (1981) Comparison of telephotometer measurements of extinction coefficients with
»1 scattering and absorption coefficients. In: White, W. H.; Moore, D. J.; Lodge, J. P., Jr., eds. Plumes
i2 and visibility: measurements and model components, proceedings of the symposium; November 1980;
i3 Grand Canyon National Park, AZ. Atmos. Environ. 15: 2617-2624.
4
August 1991 1-85 DRAFT-DO NOT QUOTE OR CITE
-------
1 Dzubay, T. G.; Stevens, R. K.; Lewis, C. W.; Hern, D. H.; Courtney, W. J.; Tesch, J. W.; Mason, M. A.
2 (1982) Visibility and aerosol composition in Houston, Texas. Environ. Sci. Technol. 16: 514-525.>
3
4 Eddy, D. M. (1989) The confidence profile method: a Bayesian method for assessing health technologies. Oper.
5 Res. 37: 210-228.
6
7 Eddy, D. M.; Hasselblad, V.; Shachter, R. (1990a) An introduction to a Bayesian method for meta-analysis: the
8 confidence profile method. Med. Decision Making 10: 15-23.
9
10 Eddy, D. M.; Hasselblad, V.; Shachter, R. (1990b) A Bayesian method for synthesizing evidence: the confidence
11 profile method. Int. J. Technol. Assess. Health Care 6: 31-55.
12
13 Edney, E. O.; Stiles, D. C.; Spence, J. W.; Haynie, F. H.; Wilson, W. E. (1986) A laboratory study to evaluate
14 the impact of NOX, SOX, and oxidants on atmospheric corrosion of galvanized steel. In: Baboian, R., ed.
15 Materials degradation caused by acid rain: developed from the 20th state-of-the-art symposium of the
16 American Chemical Society; June 1985; Arlington, VA. Washington, DC: American Chemical Society;
17 pp. 172-193. (ACS symposium series 318).
18
19 Edney, E. O.; Stiles, D. C.; Corse, E. W.; Wheeler, M. L.; Spence, J. W.; Haynie, F. H.; Wilson, W. E., Jr.
20 (1987) Field study to determine the impact of air pollutants on the corrosion of galvanized steel.
21 Presented at: Corrosion/87; San Francisco, CA. NACE; paper no. 410.
22
23 Edwards, P. J.; Helvey, J. D. (1991) Long-term ionic increases from a central Appalachian forested watershed.
24 J. Environ. Qual. 20: 250-255.
25 ......
26 Effler, S. W.; Brooks, C. M.; Auer, M. T.; Doerr, S. M. (1990) Free ammonia and toxicity criteria in a
27 polluted urban lake. Res. J. Water Pollut. Control Fed. 62: 771-779.
28
29 Ehrlich, R.; Henry, M. C. (1968) Chronic toxicity of nitrogen dioxide: I. effect on resistance to bacterial
30 pneumonia. Arch. Environ. Health 17: 860-865.
31
32 Ehrlich, R.; Findlay, J. C.; Fenters, J. D.; Gardner, D. E. (1977) Health effects of short-term inhalation of
33 nitrogen dioxide and ozone mixtures. Environ. Res. 14: 223-231.
34
35 Eilers, J. M.; Sullivan, T. J.; Hurley, K. C. (1990) The most dilute lake in the world? Hydrobiologia 199: 1-6.
36
37 Ekwo, E. E.; Weinberger, M. M.; Lachenbruch, P. A.; Huntley, W. H. (1983) Relationship of parental smoking
38 and gas cooking to respiratory disease in children. Chest 84: 662-668.
39
40 Ellenberg, H. (1987) Floristic changes due to eutrophication. In: Asman, W. A. H.; Diederen, S. M. A., eds.
41 Ammonia and acidification: proceedings of a symposium of the European Association for the Science of
42 Air Pollution (EURASAP); April; Bilthoven, The Netherlands. European Association for the Science of
43 Air Pollution; pp. 301-308.
44
45 Ellenberg, H. (1988) Floristic changes due to nitrogen deposition in central Europe. In: Nilsson, J.; Grennfelt,
46 P., eds. Critical loads for sulphur and nitrogen: report from a workshop; March; Skokloster, Sweden.
47 Copenhagen, Denmark: Nordic Council of Ministers; pp. 375-383.
48
49 Elwood, J. W.; Turner, R. R. (1989) Streams: water chemistry and ecology. In: Johnson, D. W.; Van Hook,
50 R. I., eds. Analysis of biogeochemical cycling processes in Walker Branch Watershed. New York, NY:
51 Springer-Verlag; pp. 301-350.
52
August 1991 1-86 DRAFT-DO NOT QUOTE OR CITE
-------
1 Elwood, J. W.; Sale, M. J.; Kauftnann, P. R.; Cada, G. F. (1991) The Southern Blue Ridge province. In:
2 Charles, D, F., ed. Acidic deposition and aquatic ecosystems: regional case studies. New York, NY:
3 Springer-Verlag.
4
5 Eshleman, K. N.; Hemond, H. F. (1985) The role of organic acids in the acid-base status of surface waters at
6 Biekford Watershed, Massachusetts. Water Resour. Res. 21: 1503-1510.
7 - . • : • • •• •-.-'.--•
8 Fahey, D. W.; Murphy, D. M.; Kelly, K. K.; Ko, M. K. W.; Proffitt, M. H.; Eubank, C. S.; Ferry, G. V.;
9 Loewenstein, M.; Chan, K. R. (1989) Measurements of nitric oxide and total reactive nitrogen in the
,0 Antarctic stratosphere: observations and chemical implications. J. Geophys. Res. [Atmos.] 94:
.1 16665-16681.
.2
,3 Finlan, J, M. (1981) Scattering of light and the color of atmospheric hazes. In: White, W. H.; Moore, D. J.;
.4 Lodge, J. P., Jr., eds. Plumes and visibility: measurements and model components, proceedings of the
.5 symposium; November 1980; Grand Canyon National Park, AZ. Atmos. Environ. 15: 2599-2616.
.6 -• • •'••.•'"' .''"''
.7 Fisher, D.; Ceraso, J.; Mathew, T.; Oppenheimer, M. (1988) Polluted coastal waters: the role of acid rain. New
.8 York, NY: Environmental Defense Fund.
.9
tO Fisher, F. M.; Zak, J. C.; Cunningham, G. L.; Whitford, W. G. (1988) Water and nitrogen effects on growth
' 1 and allocation patterns of creosotebush in the northern Chihuahuan Desert. J. Range Manage.
',2 41: 387-391.
13 - - '•'••-•"-..•
'4 Fisher, T. R.; Harding, L. W., Jr.; Stanley, D. W.; Ward, L. G. (1988) Phytoplankton, nutrients, and turbidity
'5 in the Chesapeake, Delaware, and Hudson estuaries. Estuarine Coastal Shelf Sci. 27: 61-93.
:6 • • • * ' . ' ' '
'.7 Focht, D. D. (1974) The effect "of temperature, pH, and aeration on the production of nitrous oxide and gaseous
18 nitrogen - zero-order kinetic model. Soil Sci. 118: 173-179.
:9 • • ...-••. •-•
iO Foster, J. R.; Cottrell, R. C.; Herod, I. A.; Atkinson, H. A. C.; Miller, K. (19S5) A comparative study of the
• 1 pulmonary effects of NO2 in the rat and hamster. Br. J. Exp. Pathol. 66: 193-204.
,2 -•.'. ; -: •
•3 Frampton, M. W.; Smeglin, A. M.; Roberts, N. J., Jr.; Finkelstein, J. N.; Morrow, P. E.; Utell, M. J. (1989a)
4 Nitrogen dioxide exposure in vivo and human alveolar macrophage inactivation of influenza virus in
5 vitro. Environ. Res. 48: 179-192.
6
•7 Frampton, M. W.; Finkelstein, J. N.; Roberts, N. J., Jr.; Smeglin, A. M.; Morrow, P, E.; Ufell, M. J. (1989b)
'8 Effects of nitrogen dioxide exposure on bronchoalveolar lavage proteins in humans. Am. J. Respir. Cell
9 Mol. Biol. 1: 499-505.
0 •.-'•;'
1 Frampton, M. W.; Morrow, P. E.; Cox, C.; Gibb, F. R.; Speers, D. M,; Utell, M. T. (1991) Effects of
2 nitrogen dioxide exposure on pulmonary function and airway reactivity in normal humans. Am. Rev.
3 Respir. Dis. 143: 522-527.
4
5 Freeman, G.; Crane, S. C.; Furiosi, N. J.; Stephens, R. J.; Evans, M. J.; Moore, W. D. (1972) Covert
6 reduction in ventilatory surface in rats during prolonged exposure to subacute nitrogen dioxide. Am. Rev.
7 Respir. Dis. 106: 563-579.
8
9 Freitag, W. O.; Mee, P.; Petersen, R. (1980) Glass/ferrite interactions and corrosion of gap glasses in recording
0 heads. IEEE Trans. Magn. 16: 876-878.
1
2 Fujimaki, H.; Shimizu, F. (1981) Effects of acute exposure to nitrogen dioxide on primary antibody response.
3 Arch. Environ. Health 36: 114-119.
4
August 1991 1-87 DRAFT-DO NOT QUOTE OR CITE
-------
1 Galloway, J. N.; Whelpdale, D. M. (1987) WATOX-86 overview and western North Atlantic Ocean S and
2 N atmospheric budgets. Global Biogeochem. Cycles 1: 261-281.
3
4 Galloway, J. N.; Schofield, C. L.; Hendrey, G. R.; Peters, N. E.; Johannes, A. H. (1980) Sources of acidity in
5 three lakes acidified during snowmelt. In: Drablos, D.; Tollan, A., eds. Ecological effects of acid
6 precipitation: proceedings of an international conference. Mysen, Norway: Johs. Grefslie Trykkeri A/S;
7 pp. 264-265.
8
9 Galloway, J. N.; Whelpdale, D. M.; Wolff, G. T. (1984) The flux of S and N eastward from North America.
10 Atmos. Environ. 18: 2595-2607.
11
12 Gardner, D. E.; Holzman, R. S.; Coffin, D. L. (1969) Effects of nitrogen dioxide on pulmonary cell population.
13 J. Bacteriol. 98: 1041-1043.
14
15 Gardner, D. E.; Coffin, D. L.; Pinigin, M. A.; Sidorenko, G. I. (1977a) Role of time as a factor in the toxicity
16 of chemical compounds in intermittent and continuous exposures. Part I. Effects of continuous exposure.
17 J. Toxicol. Environ. Health 3: 811-820.
18
19 Gardner, D. E.; Miller, F. J.; Blommer, E. J.; Coffin, D. L. (1977b) Relationships between nitrogen dioxide
20 concentration, time, and level of effect Using an animal infectivity model. In: Dimitriades, B., ed.
21 International conierence on photochemical oxidant pollution and its control: proceedings, v. I; September
22 1976; Raleigh, NC. Research Triangle Park, NC: U. S. Environmental Protection Agency,
23 Environmental Sciences Research Laboratory; pp. 513-525; EPA report no. EPA-600/3-77-001a.
24 Available from: NTIS, Springfield, VA; PB-264232. (Ecological research series).
25
26 Gardner, D. E.; Miller, F. J.; Blommer, E. J.; Coffin, D. L. (1979) Influence of exposure mode on the toxicity
27 of NO2. Environ. Health Perspect. 30: 23-29. ^
28
29 Gardner, D. E.; Graham, J. A.; Illing, J. W.; Blommer, E. J.; Miller, F. J. (1980) Impact of exposure patterns
30 on the lexicological response to NO2 and modifications by added stressors. In: Proceedings of the
31 US-USSR third joint symposium on problems of environmental health; October 1979; Suzdal, USSR.
32 Research Triangle Park, NC: National Institute of Environmental Health Sciences; pp. 17-40.
33
34 Gardner, D. E.; Miller, F. J.; Illing, J. W.; Graham, J. A. (1982) Non-respiratory function of the lungs: host
35 defenses against infection. In: Schneider, T.; Grant, L., eds. Air pollution by nitrogen oxides:
36 proceedings of the US-Dutch international symposium; May; Maastricht, The Netherlands. Amsterdam,
37 The Netherlands: Elsevier Scientific Publishing Company; pp. 401-415. (Studies in environmental science
38 21).
39
40 Garland, J. A.; Penkett, S. A. (1976) Absorption of peroxy acetyl nitrate and ozone by natural surfaces. Atmos.
41 Environ. 10: 1127-1131.
42
43 Gilbert, D. A.; Sagraves, T. H.; Lang, M. M.; Munson, R. K.; Gherini, S. A. (1989) R&D Lake Acidification
44 Assessment Project: Blue Lake acidification study. San Ramon, CA: Pacific Gas and Electric; report
45 009.5-89.2.
46
47 Glezen, W. P. (1989) Antecedents of chronic and recurrent lung disease: childhood respiratory trouble. Am. Rev.
48 Respir. Dis. 140: 873-874.
49
50 Glezen, W. P.; Denny, F. W. (1973) Epidemiology of acute lower respiratory disease in children. N. Engl.
51 J. Med. 288: 498-505.
52
August 1991 1-88 DRAFT-DO NOT QUOTE OR CITE
-------
1 Gold, D. R.; Tager, I. B.; Weiss, S. T.; Tosteson, T. D.; Speizer, F. E. (1989) Acute lower respiratory illness
2 , in childhood as a predictor of lung function and chronic respiratory symptoms. Am. Rev. Respir. Dis.
3 140: 877-884.
4
5 Graham, J, A.; Gardner, D. E.; Blommer, E. J.; House, D. E.; Menache, M, G.; Miller, F. J. (1987) Influence
6 of exposure patterns of nitrogen dioxide and modifications by ozone on susceptibility to bacterial
7 infectious disease in mice. J. Toxicol. Environ. Health 21: 113-125.
8
9 Gravenhorst, G.; Hoeflcen, K. D.; Georgii, H. W. (1983) Acidic input to a beech and spruce forest. In:
.0 Beilke, S.; Elshout, A. J., eds. Acid deposition: proceedings of the CEC workshop organized as part of
.1 the concerted action "Physico-chemical behaviour of atmospheric pollutants"; September 1982; Berlin,
.2 Federal Republic of Germany. Dordrecht, The Netherlands: D. Reidel Publishing Company;
.3 pp. 155-171.
,4
,5 Grennfelt, P.; Hultberg, H. (1986) Effects of nitrogen deposition on the acidification of terrestrial and aquatic
.6 ecosystems. Water Air Soil Pollut. 30: 945-963.
.7
.8 Grennfelt, P.; Bengtson, C.; Skarby, L. (1983) Dry deposition of nitrogen dioxide to Scots pine needles. In:
.9 Pruppacher, H. R.; Semonin, R. G.; Slinn, W. G. N., eds. Precipitation scavenging, dry deposition, and
»0 resuspension. Volume 2. Dry deposition and resuspension: proceedings of the fourth international
11 conference; November-December 1982; Santa Monica, CA. New York, NY: Elsevier; pp. 753-762.
12
13 Groblicki, P. J.; Wolff, G. T.; Countess, R. J. (1981) Visibility-reducing species k the Denver "brown cloud"—
14 I. relationships between extinction and chemical composition. In: White, W. H.; Moore, D. J.; Lodge,
15 J. P., Jr., eds. Plumes and visibility: measurements and model components, proceedings of the
16 symposium; November 1980; Grand Canyon National Park, AZ. Atmos. Environ. 15: 2473-2484.
17
18 Guderian, R.; Tingey, D. T.; Rabe, R. (1985) Effects of photochemical oxidants on plants. In: Guderian, R., ed.
19 Air pollution by photochemical oxidants. Berlin, Federal Republic of Germany: Springer-Verlag;
JO pp. 127-333.
»
12 Hansen, J.; Lacis, A.; Prather, M. (1989) Greenhouse effect of chlorofluorocarbons and other trace gases.
13 J. Gedphys. Res. [Atmos.] 94: 16417-16421.
54
55 Hanson, P. J.; Rott, K.; Taylor, G. E., Jr.; Gunderson, C. A.; Lindberg, S. E.; Ross-Todd, B. M. (1989) NO2
56 deposition to elements representative of a forest landscape, Atmos. Environ. 23: 1783-1794.
M
58 Hauhs, M. (1989) Lange Bramke: an ecosystem study of a forested watershed. In: Adriano, D. C.; Salmons, W.,
59 eds. Acidic precipitation. New York, NY: Springer-Verlag; pp. 275-305.
40
\\ Hauhs, M.; Rost-Siebert, K.; Raben, G.; Paces, T.; Vigerust, B. (1989) Summary of European data. In:
42 Malanchuk, J. L.; Nilsson, J., eds. The role of nitrogen in the acidification of soils and surface waters.
43 Nordic Council of Ministers, Miljorapport 1989:10; pp. 5-1 - 5-37.
44
45 Haydon, G. B.; Davidson, J. T.; Lillington, G. A.; Wasserman, K. (1967) Nitrogen dioxide-induced emphysema
46 in rabbits. Am. Rev. Respir. Dis. 95: 797-805.
47
4-8 Haynie, F. H.; Spence, J. W. (1984) Air pollution damage to exterior household paints. J. Air Pollut. Control
49 Assoc. 34: 941-944.
50
51 Haynie, F. H.; Spence, J. W.; Upham, J. B. (1976) Effects of gaseous pollutants on materials—a chamber study.
52 Research Triangle Park, NC: U.S. Environmental Protection Agency, Environmental Sciences Research
53 Laboratory; EPA report no. EPA-600/3-76/015. Available from: NTIS, Springfield, VA; PB-251580.
54
August 1991 1-89 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hazucha, M. J.; Ginsberg, J. F.; McDonnell, W. F.; Haak, E. D., Jr.; Pimmel, R. L.; House, D. E.;
2 Bromberg, P. A. (1982) Changes in bronchial reactivity of asthmatics and normals following exposures to
3 0,1 ppmNO2, In: Schneider, T.; Grant, L., eds. Air pollution by nitrogen oxides: proceedings of the
4 US-Dutch international symposium; May; Maastricht,' The Netherlands. Amsterdam, The Netherlands:
5 Elsevier Scientific Publishing Company; pp. 387-400. (Studies in environmental science 21).
6
7 Hazucfaa, M. J.; Ginsberg, J, F.; McDonnell, W. F.; Haak, E. D., Jr.; Pimmel, R. L.; Salaam, S. A.; House,
8 D. E.; Bromberg, P. A. (1983) Effects of 0.1 ppm nitrogen dioxide on airways of normal and asthmatic
9 subjects. J. Appl, Physiol.: Respir. Environ. Exercise Physiol. 54: 730-739.
10
11 Hecky, R. B,| Kilham, P. (1988) Nutrient limitation of phytoplankton in freshwater and marine environments:
12 a review of recent evidence on the effects of enrichment. Limnol. Oceanogr. 33: 796-822.
13
14 Hegg, D. A.; Hobbs, P. V. (1979) Some observations of particulate nitrate concentrations in coal-fired power
15 plant plumes. Atmos. Environ. 13: 1715-1716.
16
17 Hegg, D.; Hobbs, P. V.; Radke, L. F.; Harrison, H. (1977) Ozone and nitrogen oxides in power plant plumes.
18 In: Dimitriades, B., ed. International conference on photochemical oxidant pollution and its control:
19 proceedings, v. I; September 1976; Raleigh, NC. Research Triangle Park, NC: U. S. Environmental
20 Protection Agency, Environmental Sciences Research Laboratory; pp. 173-183; EPA report no.
21 EPA-600/3-77-001a. Available from: NTIS, Springfield, VA; PB-264232.
22
23 Heil, G. W.; Bruggink, M. (1987) Competition for nutrients between Calluna vulgaris (L.) Hull and Molinia
24 caerulea (L.) Moench. Oecologia 73: 105-108.
25
26 Heil, G. W.; van Dam, D.; Heijne, B. (1987) Catch of atmospheric deposition in relation to vegetation structures
27 of heathland. In: Asman, W. A. H.; Diederen, H. S. M. A., eds. Ammonia and acidification:
28 proceedings of a symposium of the European Association for the Science of Air Pollution (EURASAP);
29 April; Bilthoven, The Netherlands. European Association for the Science of Air Pollution; pp. 107-123.
30
31 Heil, G. W.; Werger, M. J. A.; de Mol, W.; van Dam, D.; Heijne, B. (1988) Capture of atmospheric
32 ammonium by grassland canopies. Science (Washington, DC) 239: 764-765.
33
34 Hemphill, J. E.; Norton, J. E.; Ofjord, O. A.; Stone, R. L. (1976) Colorfastness to light and atmospheric
35 contaminants. Text. Chem. Color. 8: 60-62.
36
37 Hicks, B. B.; Baldocchi, D. D.; Hosker, R. P., Jr.; Hutchison, B. A.; Matt, D. R.; McMillen, R. T.;
38 Satterfield, L. C. (1985) On the use of monitored air concentrations to infer dry deposition (1985). Silver
39 Spring, MD: National Oceanic and Atmospheric Administration, Air Resources Laboratory; NOAA
40 technical memorandum no. ERL-ARL-141. Available from: NTIS, Springfield, VA; PB86-158409.
41 ,
42 Hicks, B. B.; Draxler, R. R.; Albritton, D. L.; Fehsenfeld, F. C.; Dodge, M.; Schwartz, S. E.; Tanner, R. L.;
43 Hales, J. M.; Meyers, T. P.; Vong, R. J.; Davidson, C. L; Wesely, M. L.; Bates, T. S.; Lindberg,
44 S. E.; Liu, S. C.; Gillette, D. A.; Westberg, H. H.; Lamb, B.; Sigmon, J. T.; Mueller, S. F. (1990)
45 NAPAP report 2: atmospheric processes research and process model development. In: Acidic deposition:
46 state of science and technology, v. I, emissions, atmospheric processes and deposition. Washington, DC:
47 National Acid Precipitation Assessment Program; state of science and technology report no. 2.
48 . • . •
49 Hill, A. C. (1971) Vegetation: a sink for atmospheric pollutants. J. Air Pollut. Control Assoc. 21: 341-346.
50
51 Hodkinson, J. R. (1966) Calculations of colour and visibility in urban atmospheres polluted by gaseous NO2. Int.
52 J. Air Water Pollut. 10: 137-144.
53 . -.: •
54 Horvath, H. (1971) On the brown colour of atmospheric haze. Atmos. Environ. 5: 333-344.
August 1991 1-90 DRAFT-DO NOT QUOTE OR CITE
-------
1 Huebert, B. J.; Robert, C. H. (1985) The dry deposition of nitric acid to grass. J. Geophys. Res. [Atmos.]
2 90: 2085-2090.
3
4 Husar, R. B.; White, W. H. (1976) On the color of the Los Angeles smog. Atmos. Environ. 10: 199-204.
5 -:
6 Hutchinson, G. L.; Millington, R. J.; Peters, D. B. (1972) Atmospheric ammonia: absorption by plant leaves.
7 Science (Washington, DC) 175: 771-772. . . .
8
9 Hyde, D.; Orthoefer, J.; Dungworth, D.; Tyler, W.; Carter, R.; Lum, H. (1978) Morphometric and
10 morphologic evaluation of pulmonary lesions in beagle dogs chronically exposed to high ambient levels
11 of air pollutants. Lab. Invest. 38: 455-469.
12
13 Irwin, J.; Schenk, D.; McClelland, G. H.; Schulze, W. D.; Stewart, T.; Thayer, M. (1990) Urban visibility:
14 some experiments on the contingent valuation method. In: Mathai, C. V., ed. Visibility and fine
15 particles: an AWMA/EPA international specialty conference; October 1989; Estes Park, CO. Pittsburgh,
16 PA: Air and Waste Management Association; pp. 647-658. (AWMA transactions series no. TR-17).
17
18 Ito, K. (1971) [Effect of nitrogen dioxide inhalation on influenza virus infection in mice]. Nippon Eiseigaku
19 Zasshi 26: 304-314.
20
21 Jacob, D. J.; Wofsy, S. C. (1988) Photochemistry of biogenic emissions over the Amazon forest. J. Geophys.
22 Res. [Atmos.] 93: 1477-1486.
23
24 Jacobson, J. S.; McManus, J. M. (1985) Pattern of atmospheric sulphur dioxide occurrence: an important
25 criterion in vegetation effects assessment. Atmos. Environ. 19: 501-506.
26
27 Jakab, G. J. (1987) Modulation of pulmonary defense mechanisms by acute exposures to nitrogen dioxide.
28 Environ. Res. 42: 215-228.
29
30 Jakab, G. J. (1987) Modulation of pulmonary defense mechanisms by acute exposures to nitrogen dioxide.
31 Experientia Suppl. 51: 235-242.
32
33 Jeffries, D. S. (1990) Snowpack storage of pollutants, release during melting, and impact on receiving waters. In:
34 Norton, S. A.; Lindberg, S. E.; Page, A. L., eds. Acidic precipitation: v. 4, soils, aquatic processes,
35 and lake acidification. New York, NY: Springer-Verlag; pp. 107-132.
36
37 Jellinek, H. H. G. (1970) Chain scission of polymers by small concentrations (1 to 5 ppm) of sulfur dioxide and
38 nitrogen dioxide, respectively, in presence of air and near ultraviolet radiation. J. Air Pollut. Control
39 Assoc. 20: 672-674.
40
41 Jellinek, H. H. G.; Flajsman, F.; Kryman, F. J. (1969) Reaction of SO2 and NO2 with polymers. J. Appl.
42 Polym. Sci. 13: 107-116.
43
44 Joerres, R.; Magnussen, H. (1990) Airways response of asthmatics after a 30 min exposure, at resting
45 ventilation, to 0.25 ppm NO2 or 0.5 ppm SO2. Eur. Respir. J. 3: 132-137.
46
47 Johansson, L.-G. (1986) A laboratory study of the influence of NO2 and SO2 on the atmospheric corrosion of
48 steel, copper, zinc and aluminium. Proc. Electrochem. Soc. 86-6: 267-279.
49
50 Johansson, C. (1987) Pine forest: a negligible sink for atmospheric NO^ in rural Sweden. Tellus Ser. B
51 39: 426-438.
52
53 Johansson, L.-G.; Lindqvist, O.; Mangio, R. E. (1988) Corrosion of calcareous stones in humid air containing
54 SO2 and NO2. Durability Build. Mater. 5: 439-449.
August 1991 1-91 DRAFT-DO NOT QUOTE OR CITE
-------
1 John, W. (1986) A new method for nitric acid and nitrate aerosol measurement using the dichotomous sampler.
2 Sacramento, CA: California Air Resources Board; CARS document no. CA/DOH/AIHL/R-304.
3
4 Johnson, C. D.; Latimer, D. A.; Bergstrom, R. W,; Hogo, H. (1980) User's manual for the plume visibility
5 model (PLUVUE). Research Triangle Park, NC: U. S. Environmental Protection Agency, Office of Air
6 Quality Planning and Standards; EPA report no. EPA-450/4-80-032. Available from: NTIS, Springfield,
7 VA; PB81-163297.
8
9 Johnson, D. A.; Frampton, M. W.; Winters, R. S.; Morrow, P. E.; Utell, M. J. (1990) Inhalation of nitrogen
10 dioxide fails to reduce the activity of human lung alpha-1-proteinase inhibitor. Am. Rev. Respir. Dis.
11 142: 758-762,
12
13 Johnson, D. W.; van Miegroet, H.; Lindberg, S. E,; Harrison, R. B.; Todd, D. E. (n.d.) Nutrient cycling in red
14 spruce forests of the Great Smoky Mountains. Can. J. For. Res.: in press.
15
16 Jones, H. C.; Noggle, J. C.; Young, R. C.; Kelly, J. M.; Olem, H.; Ruane, R. J.; Pasch, R. W.; Hyfantis,
17 G. J.; Parkhurst, W. J. (1983) Investigations of the cause of fishkills in fish-rearing facilities in Raven
18 Fork watershed. Chattanooga, TN: Tennessee Valley Authority, Office of Natural Resources; report no.
19 TVA/ONR/WR-83-9.
20
21 Kagawa, J. (1982) Respiratory effects of 2-hr exposure to 1.0 ppm nitric oxide in normal subjects. Environ. Res.
22 27:485^*90,
23
24 Kagawa, J. (1990) Health effects of exposure to mixtures of nitric oxide and nitrogen dioxide in healthy young
25 women. In: Indoor air '90: precedings of the 5th international conference on indoor air quality and
26 climate, volume 1, human health, comfort and performance; July-August; Toronto, ON, Canada. Ottawa,
27 ON, Canada: International Conference on Indoor Air Quality and Climate, Inc.; pp. 307-312.
28
29 Kahl, J. S.; Norton, S. A.; Cronan, C. A.; Fernandez, I. J.; Bacon, L. C.; Haines, T. A. (1991) Maine. In:
30 Charles, D. F., ed. Acidic deposition and aquatic ecosystems: regional case studies. New York, NY:
31 Springer Verlag.
32
33 Katz, B. G.; Bricker, O. P.; Kennedy, M. M. (1985) Geochemical mass-balance relationships for selected ions in
34 precipitation and stream water, Catoctin Mountains, Maryland. Am. J. Sci. 285: 931-962.
35
36 Kaufmann, P. R.; Herlihy, A. T.; Mitch, M. E.; Messer, J. J.; Overton, W. S. (1991) Stream chemistry in the
37 eastern United States: 1. synoptic survey design, acid-base status, and regional patterns. Water Resour.
38 Res. 27: 611-627.
39
40 Keller, M. D.; Lanese, R. R.; Mitchell, R. L; Cote, R. W. (1979) Respiratory illness in households using gas
41 and electricity for cooking: n. symptoms and objective findings. Environ. Res. 19: 504-515.
42
43 Kelly, N. A. (1987) The photochemical formation and fate of nitric acid in the metropolitan Detroit area:
44 ambient, captive-air irradiation and modeling results. Atmos. Environ. 21: 2163-2177.
45
46 Ken-, H. D.; Kulle, T. J.; McHhany, M. L.; Swidersky, P. (1979) Effects of nitrogen dioxide on pulmonary
47 function in human subjects: an environmental chamber study. Environ. Res. 19: 392-404,
48
49 Kleinman, M. T.; Linn, W. S.; Bailey, R. M.; Jones, M. P.; Hackney, J. D. (1980) Effect of ammonium nitrate
50 aerosol on human respiratory function and symptoms. Environ. Res. 21: 317-326.
51
52 Kleinman, M. T.; Bailey, R. M.; Linn, W. S.; Anderson, K. R.; Whynot, L D,; Shamoo, D. A.; Hackney,
53 J. D. (1983) Effects of 0.2 ppm nitrogen dioxide on pulmonary function and response to
54 bronchoprovocation in asthmatics. J, Toxicol. Environ. Health 12: 815-826.
August 1991 1-92 DRAFT-DO NOT QUOTE OR CITE
-------
1 Koenig, J. Q.; Covert, D. S.; Morgan, M. S.; Horike, M.; Horike, N.; Marshall, S. G.; Pierson, W, E. (1985)
2 Acute effects of 0.12 ppm ozone or 0,12 ppm nitrogen dioxide on pulmonary function in healthy and
3 asthmatic adolescents. Am. Rev. Respir. Dis. 132: 648-651.
4
5 Koenig, J. Q.; Covert, D. S.; Pierson, W. E.; McManus, M. S. (1988) The effects of inhaled nitric acid on
6 pulmonary function in adolescent asthmatics. Am. Rev. Respir. Dis. 137(suppl.): 169.
7
8 Koenig, J. Q.; Covert, D. S.; Pierson, W. E. (1989a) Effects of inhalation of acidic compounds on pulmonary
9 function in allergic adolescent subjects. In: Symposium on the health effects of acid aerosols; October
10 1987; Research Triangle Park, NC. Environ. Health Perspect. 79: 173-178.
11
12 Koenig, J. Q.; Hanley, Q. S.; Anderson, T. L.; Rebolledo, V.; Pierson, W. E. (1989b) An assessment of
13 pulmonary function changes and oral ammonia levels after exposure of adolescent asthmatic subjects to
14 sulfuric or nitric acid. Presented at: 82nd annual meeting and exhibition of the Air and Waste
15 Management Association; June; Anaheim, CA. Pittsburgh, PA: Air and Waste Management Association;
16 paper no. 89-92.4.
17
18 Kramer, J. R.; Andren, A. W.; Smith, R. A.; Johnson, A. H.; Alexander, R. B.; Oehlert, G. (1986) Streams
19 and lakes. In: Acid deposition: long-term trends. Washington, DC: National Academy Press;
20 pp. 231-299.
21
22 Krause, H. H.; Bohn, M. A.; Pfeil, A.; Schmitt, D.; Ziegahn, K.-F. (1989) Polymer failures induced by air
23 pollutants. In: Brasser, L. J.; Mulder, W. C., eds. Man and his ecosystem: proceedings of the 8th world
24 clean air congress, v. 2; September; The Hague, The Netherlands. Amsterdam, The Netherlands: Elsevier
25 Science Publishers B.V.; pp. 337-342.
26
27 Kucera, V. (1986) Influence of acid deposition on atmospheric corrosion of metals: a review. In: Baboian, R.,
28 ed. Materials degradation caused by acid rain: developed from the 20th state-of-the-art symposium of the
29 American Chemical Society; June 1985; Arlington, VA. Washington, DC: American Chemical Society;
30 pp. 104-118. (ACS symposium series 318).
31
32 Lafuma, C.; Harf, A.; Lange, F.; Bozzi, L.; Poncy, J. L.; Bignon, J. (1987) Effect of low-level NO2 chronic
33 exposure on elastase-induced emphysema. Environ. Res. 43: 75-84.
34
35 Lane, P. L; Bell, J. N. B. (1984) The effects of simulated urban air pollution on grass yield: part 2—
36 performance of Lolium perenne, Phleum pratense and Dactylis glomerata fumigated with SO2, NO2
37 and/or NO. Environ. Pollut. Ser. A 35: 97-124.
38
39 Lane, P. L; Bell, J. N. B. (1984) The effects of simulated urban air pollution on grass yield: part I - description
40 and simulation of ambient pollution. Environ. Pollut. Ser. B 8: 245-263.
41
42 Latimer, D. A. (1979) Power plant impacts on air quality and visibility: siting and emission control implications.
43 Washington, DC: U. S. Environmental Protection Agency, Office of Planning and Evaluation; EPA
44 publication no. EPA-230/11-79-001. Available from: NTIS, Springfield, VA; PB80-123516.
45
46 Latimer, D. A. (1980) Power plant impacts on visibility in the West: siting and emissions control implications.
47 J. Air Pollut. Control Assoc. 30: 142-146.
48
49 Latimer, D. A. (1988) Application of disperion models for layered haze source apportionment. Presented at: 81st
50 annual meeting of the Air Pollution Control Association; June; Dallas, TX. Pittsburgh, PA: Air Pollution
51 Control Association; paper no. 88-52.6.
52 '
53 Latimer, D. A.; Samuelsen, G. S. (1975) Visual impact of plumes from power plants. Irvine, CA: University of
54 California, School of Engineering, Air Quality Laboratory; publication no. UCI-ARTR-75-3.
August 1991 1-93 DRAFT-DO NOT QUOTE OR CITE
-------
1 Latimer, D. A.; Samuelsen, G. S. (1978) Visual impact of plumes from power plants: a theoretical model.
2 Atmos. Environ. 12: 1455-1465,
3
4 Latimer, D, A.; Bergstrom, R. W.; Hayes, S. R.; Liu, M.-K.; Seinfeld, J. FL; Whitten, G. Z.; Wojcik, M. A.;
5 Hillyer, M. J. (1978) The development of mathematical models for the prediction of anthropogenic
6 visibility impairment: volumes I-in, Research Triangle Park, NC: U. S. Environmental Protection
7 Ageacy, Office of Air Quality Planning and Standards; EPA report no. EPA-450/3-78-110a-c. Available
8 from: NTIS, Springfield, VA; PB-293118-SET.
9
10 Latimer, D. A,; Hogo, H.; ChiaMn, L. R.; Dudik, M, C.; Ireson, R. G.; Irpan, P.; Jacobson, H.; Killus, J. P.;
11 Lundberg, G. W.; Saxena, P.; Whitten, G. Z.; Yocke, M. A. (1985a) Modeling regional haze in the
12 Southwest: a preliminary assessment of source contributions, volume 1 [revised draft report]. San Rafaelj
13 CA: Systems Applications, Inc.; chps. 4 and 5; SAI report no. SYSAPP/85-038.
14
15 Latimer, D. A.; Chinkin, L. R.; Dudik, M. C.; Hogo, H.; Ireson, R. G. (1985b) Uncertainties associated with
16 modeling regional haze in the Southwest. Washington, DC: American Petroleum Institute, Health and
17 Environmental Sciences Department, API publication no. 4403.
18
19 Latimer, D. A.; Gery, M. W.; Hogo, H. (1986) A theoretical evaluation of the role of nighttime nitrate
20 formation in the formation of layered haze. San Rafael, CA: Systems Applications, Inc.; SAI document
21 no. SYSAPP-86-167.
22
23 Leflcowitz, S. S.; McGrath, J. J.; Lefkowitz, D. L. (1986) Effects of NO2 on immune responses. J. Toxicol.
24 Envkon. Health 17: 241-248.
25
26 Lefohn, A. S.; Tingey, D. T. (1984) The co-occurrence of potentially phytotoxic concentrations of various
27 gaseous air pollutants. Atmos. Environ. 18: 2521-2526.
28
29 Lefohn, A. S.; Davis, C. E.; Jones, C. K.; Tingey, D. T.; Hogsett, W. E. (1987) Co-occurrence patterns of
30 gaseous air pollutant pairs at different minimum concentrations in the United States. Atmos. Environ.
31 21:2435-2444.
32
33 Levander. (1990)
34
35 Likens, G. E,» ed. (1985) An ecosystem approach to aquatic ecology: Mirror Lake and its environment. New •
36 York, NY: Springer-Verlag.
37
38 LUjelund, L.-E.; Torstensson, P. (1988) Critical load of nitrogen with regards to effects on plant composition.
39 In: NUsson, J.; Grennfelt, P., eds. Critical loads for sulphur and nitrogen: report from a workshop;
40 March; Skokloster, Sweden. Copenhagen, Denmark: Nordic Council of Ministers; pp. 363-373.
41
42 Lindberg, S. E,; Lovett, G. M.; Meiwes, K.-J. (1987) Deposition and forest canopy interactions of airborne
43 nitrate. In: Hutchinson, T. C.; Meema, K. M., eds. Effects of atmospheric pollutants on forests,
44 wetlands and agricultural ecosystems. Berlin, Federal Republic of Germany: Springer-Verlag;
45 pp. 117-130. (NATO as 1 series: v. G16).
46
47 Linn, W. S.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Anzar, U. T.; Venet, T. G.; Avol, E. L.;
48 Hackney, J. D. (1985a) Controlled exposure of volunteers with chronic obstructive pulmonary disease to
49 nitrogen dioxide. Arch. Environ. Health 40: 313-317.
50
51 Linn, W. S.; Solomon, J. C.; Trim, S. C.; Spier, C. E.; Shamoo, D. A.; Venet, T. G,; Avol, E. L.; Hackney,
52 J. D. (1985b) Effects of exposure to 4 ppm nitrogen dioxide in healthy and asthmatic volunteers. Arch.
53 Environ, Health 40: 234-239.
54
August 1991 1-94 DRAFT-DO NOT QUOTE OR CITE
-------
1 Linn, W. S.; Shamoo, D. A.; Avol, E. L.; Whynot, J. D.; Anderson, K. R.; Venet, T. G.; Hackney, J. D.
2 (1986) Dose-response study of asthmatic volunteers exposed to nitrogen dioxide during intermittent
3 exercise. Arch. Environ. Healfli 41: 292-296.
4
5 Livingston, R. A,; Baer, N. S. (1983) Mechanisms of air pollution-induced damage to stone. In: Proceedings of
6 the Vlth world congress on air quality, v. 3, atmospheric pollution and materials; May; Paris, France.
7 Paris, France: International Union of Air Pollution Prevention Associations; pp. 33-40.
8
9 Loehman, E.; Boldt, D.; Chaikin, K. (1981) Measuring the benefits of air quality improvements in the San
.0 Francisco Bay area. Washington, DC: U. S. Environmental Protection Agency.
.1
.2 Logan, J. A. (1983) Nitrogen oxides in the troposphere: global and regional budgets. J. Geophys. Res.
3 C; Oceans Atmos. 88: 10785-10807.
A
.5 Loranger, T. J.; Brakke, D. F.; Bonoff, M. B.; Gall, B. F. (1986) Temporal variability of lake waters in the
.6 North Cascades Mountains (Washington, U.S.A.). Water Air Soil Pollut. 31: 123-129.
17
18 Lovett, G. (1991) Atmospheric deposition and canopy interactions of nitrogen. In: Johnson, D. W.; Lindberg,
19 S. E., eds. Atmospheric deposition and nutrient cycling in forest ecosystems. New York, NY:
iO Springer-Verlag; in press.
II
12, Lovett, G. M.; Lindberg, S. E. (1984) Dry deposition and canopy exchange in a mixed oak forest as determined
13 by analysis of throughfall. J, Appl. Ecol. 21: 1013-1027.
84
15 Lovett, G. M.; Lindberg, S. E. (1986) Dry deposition of nitrate to a deciduous forest. Biogeochemistry
16 2: 137-148.
n
18 Luke, W. T.; Dickerson, R. R. (1987) The flux of reactive nitrogen compounds from eastern North America to
19 the western Atlantic Ocean. Global Biogeochem. Cycles 1: 329-343.
JO
II MacFariand, K. K.; Malm, W.; Molenar, J. (1983) An examination of methodologies for assessing the value of
52 visibility. In: Rowe, R. D.; Chestnut, L. G., eds. Managing air quality and scenic resources at national
J3 parks and wilderness areas. Boulder, CO: Westview Press; pp. 151-172.
W
15 Malm, W.; Gebhard, K.; Cahill, T.; Eldred, R.; Pielke, R.; Watson, J.; Latimer, D. (1989) Winter haze
56 intensive experiment: draft final report. Fort Collins, CO: National Park Service.
J7
$8 Marshall, J. D.; Cadle, S. H. (1989) Evidence for trans-cuticular uptake of HNO3 vapor by foliage of eastern
19 white pine (Pinus strobus L.). Environ. Pollut. 60: 15-28.
to
11 Mathai, C. V.; Tombach, I. H. (1987) A critical assessment of atmospheric visibility and aerosol measurements
VI in the eastern United States. J. Air Pollut. Control Assoc. 37: 700-707.
13
14 McElroy, M. B.; Salawitch, R. J. (1989) Stratospheric ozone: impact of human activity. Planet. Space Sci.
15 37: 1653-1672.
t6
17 McElroy, M. B.; Salawitch, R. J.; Wofsy, S. C.| Logan, J. A. (1986) Reductions of Antarctic ozone due to
18 synergistic interactions of chlorine and bromine. Nature (London) 321: 759-762.
!9
50 McLendon, V.; Richardson, F. (1965) Oxides of nitrogen as a factor in color changes of used and laundered
51 cotton articles. Am. Dyest. Rep. 54: 15-21.
52
53 Melack, J. M.; Stoddard, J. L. (1991) Sierra Nevada. In: Charles, D. F., ed. Acidic deposition and aquatic
54 ecosystems: regional case studies. New York, NY: Springer Verlag; In press.
August 1991 1-95 DRAFT-DO NOT QUOTE OR CITE
-------
1 Melia, R. J. W.; Florey, C. du V.; Altman, D, G.; Swan, A. V, (1977) Association between gas cooking and
2 respiratory disease in children. Br. Med, J, 2: 149-152.
3
4 Melia, R. J. W.; Florey, C. du V.; Chinn, S. (1979) The relation between respiratory illness in primary
5 schoolchildren and the use of gas for cooking: I - results from a national survey. Int. J. Epidemiol.
6 8: 333-338,
7
8 Melia, R. J. W.; Florey, C. du V.; Chinn, S.; Goldstein, B. D.; Brooks, A. G. F.; John, H. H.; Clark, D.;
9 Craighead, I. B.; Webster, X. (1980) The relation between indoor air pollution from nitrogen dioxide
10 and respiratory illness in primary schoolchildren. Clin. Respir. Physiol. 16: 7P-8P.
11
12 Melia, R. J. W.; Florey, C. du V.; Morris, R. W.; Goldstein, B. D.; John, H. H.; Clark, D.; Craighead, I. B.;
13 Mackinlay, J. C. (1982) Childhood respiratory illness and the home environment: II. association between
14 respiratory illness and nitrogen dioxide, temperature and relative humidity. Int. J. Epidemiol.
15 11: 164-169.
16
17 Melia, J,; Florey, C.", Sittampalam, Y.; Watkins, C. (1983) The relation between respiratory illness in infants
18 and gas cooking in the UK: a preliminary report. In: Air quality Vlth world congress: [proceedings of
19 the International Union of Air Pollution Prevention Associations]; May; Paris, France. Paris, France:
20 SEPIC (APPA); pp. 263-269.
21
22 Melo, O. T.; Stevens, R. D. S. (1981) The occurrence and nature of brown plumes in Ontario. Atmos. Environ.
23 15: 2521-2529.
24
25 Miller, F. J.; Graham, J. A.; Raub, J. A.; nMng, J. W.; Menache, M. G.; House, D. E.; Gardner, D. E. (1987)
26 Evaluating the toxicity of urban patterns of oxidant gases, n. Effects in mice from chronic exposure to
27 nitrogen dioxide. J. Toxicol. Environ. Health 21: 99-112.
28
29 Mochitate, K.; Takahashi, Y.; Ohsumi, T.; Miura, T. (1986) Activation and increment of alveolar macrophages
30 induced by nitrogen dioxide. J. Toxicol. Environ. Health 17: 229-239.
31
32 Mohsenin, V. (1987a) Airway responses to nitrogen dioxide in asthmatic subjects. J. Toxicol. Environ. Health
33 22: 371-380.
34
35 Mohsenin, V. (1987b) Effect of vitamin C on NO2-induced airway hyperresponsiveness in normal subjects:
36 a randomized double-blind experiment. Am. Rev. Respir. Dis. 136: 1408-1411.
37
38 Mohsenin, V. (1988) Airway responses to 2.0 ppm nitrogen dioxide in normal subjects. Arch. Environ. Health
39 43: 242-246.
40
41 Mohsenin, V.; Gea, J. B. L. (1987) Acute effect of nitrogen dioxide exposure on the functional activity of
42 alpha-1-protease inhibitor in bronchoalveolar lavage fluid of normal subjects. Am. Rev. Respir. Dis.
43 136: 646-650.
44
45 Molina and Molina. (1987)
46
47 Molina, M. J.; Tso, T.-L.; Molina, L. T.; Wang, F. C.-Y. (1987) Antarctic stratospheric chemistry of chlorine
48 nitrate, hydrogen chloride, and ice: release of active chlorine. Science (Washington, DC)
49 238: 1253-1257.
50
51 Morgan, M. D.; Good, R. E. (1988) Stream chemistry hi the New Jersey Pinelands: the influence of
52 precipitation and watershed disturbance. Water Resour. Res. 24: 1091-1100.
53
August 1991 1-96 DRAFT-DO NOT QUOTE OR CITE
-------
1 Morrow, P. E,; Utell, M. J. (1989) Responses of susceptible subpopulations to nitrogen dioxide. Cambridge,
2 MA: Health Effects Institute; research report no, 23.
3
4 Motomiya, K.; Ito, K.; Yoshida, A.; Idewara, S.; Otsu, Y.; Nakajima, Y. (1973) (The effects of exposure to
5 NO2 gas on the infection of influenza virus of mouse - long term experiment in low concentration],
6 Kankyo Kagaku Kenkyu Hokoku (Chiba Daigaku) 1: 27-33.
7
8 Mulawa, P. A.; Cadle, S. H. (1985) A comparison of nitric acid and particulate nitrate measurements by the
9 penetration and denuder difference methods. Atmos. Environ. 19: 1317-1324.
10
11 Murdoch, P. S.; Stoddard, J. L. (n.d.) The importance of nitrate in the acidification of Catskill Mountain
12 streams. Water Resour. Res.: Submitted.
1.3
14 National Atmospheric Deposition Program. (1988) NADP/NTN annual data summary. Fort Collins, CO:
15 National Atmospheric Deposition Program/National Trends Network Coordination Office,
16
17 National Institutes of Health. (1985) The definition of emphysema: report of a National Heart, Lung, and Blood
18 Institute, Division of Lung Diseases workshop. Am. Rev. Respir. Dis. 132: 182-185.
19
20 National Research Council. (1977) Nitrogen oxides. Washington, DC: National Academy of Sciences.
21
22 National Research Council. (1978) Nitrates: an environmental assessment. Washington, DC: National Academy
23 of Sciences.
24
25 National Research Council. (1983) Acid deposition: atmospheric processes in eastern North America, a review of
26 current scientific understanding. Washington, DC: National Academy Press.
27
28 National Research Council. (1986) Preservation of historical records. Washington, DC: National Academy Press.
29
30 Neas, L. M.; Ware, J. H.; Dockery, D. W.; Spengler, J. D.; Ferris, B. G., Jr.; Speizer, F. E. (1990) The
31 association of indoor nitrogen dioxide levels with, respiratory symptoms and-pulmonary function in
32 children. In: Indoor air '90: precedings of the 5th international conference on indoor air quality and
33 climate, volume 1, human health, comfort and performance; July-August; Toronto, ON, Canada. Ottawa,
34 ON, Canada: International Conference on Indoor Air Quality and Climate, Inc.; pp. 381-386.
35
36 Nilsson, J., ed. (1986) Critical loads for nitrogen and sulfur. Copenhagen, Denmark: Nordic Council of
37 Ministers.
38
39 Nilsson, J.; Grennfelt, P., eds. (1988) Critical loads for sulphur and nitrogen: report from a workshop; March;
40 Skokloster, Sweden. Copenhagen, Denmark: Nordic Council of Ministers.
41
42 Nixon, S. W.; Pilson, M. E. Q. (1983) Nitrogen in estuarine and coastal marine ecosystems. In: Carpenter,
43 E. J.; Capone, D. G., eds. Nitrogen in the marine environment. New York, NY: Academic Press;
44 pp. 565-648.
45
46 O'Connor, G.; Sparrow, D.; Taylor, D,; Segal, M.; Weiss, S. (1987) Analysis of dose-response curves to
47 methacholine: an approach suitable for population studies. Am. Rev. Respir. Dis. 136: 1412-1417,
48
49 Ogren, J, A.; Blumenthal, D. L.; Vanderpol, A. H. (1977) Oxidant measurements in western power plant
50 plumes. Volume I: technical analysis. Volume II: data. Palo Alto, CA: Electric Power Research Institute;
51 report no. EPRI EA-421. Volume II available from: NTIS, Springfield, VA; DE82-901024.
52
53 Ogston, S. A.; Florey, C. du V.; Walker, C. H. M. (1985) The Tayside infant morbidity and mortality study:
54 effect on health of using gas for cooking. Br. Med. J. 290: 957-960.
August 1991 1-97 DRAFT-DO NOT QUOTE OR CITE
-------
1 Olsen, A, R. (1989) 1986 Wet deposition temporal and spatial patterns in North America, Richland, WA: U. S,
2 Department of Energy, Pacific Northwest Laboratory; report no. PNL-6933. Available from: NTIS,
3 Springfield, VA; DE89-014372.
4
5 Orefaek, J.; Grimaldi, F.; Muls, E.; Durand, J. P.; Viala, A.; Charpin, J. (1981) Reponse bronchique aux
6 allergenes apres exposition controlee au dioxyde d'azote [Bronchial response to allergens after controlled
7 NO2 exposure]. Bull. lur. Physiopathol. Respir. 17: 911-915.
8
9 Parker, R, F.; Davis, J. K.; Cassell, G. H.; White, H,; D2iedzic, D.; Blalock, D. K.; Thorp, R. B.; Simecka,
10 J. W. (1989) Short-term exposure to nitrogen dioxide enhances susceptibility to murine respiratory
11 mycoplasmosis and decreases intrapulmonary killing of Mycoplasma pulmonis. Am. Rev. Respir. Dis.
12 140: 502-512.
13
14 Phillips, R. A.; Stewart, K. M. (1990) Longitudinal and seasonal water chemistry variations in a northern
15 Appalachian stream. Water Resour. Bull. 26: 489-498.
16
17 Port, C. D,; Ketels, K. V.; Coffin, D. L.; Kane, P. (1977) A comparative study of experimental and
18 spontaneous emphysema. J. Toxicol. Environ. Health 2: 589-604.
19
20 Pratsinis, S.; Ellis, E. C.; Novakov, T.; Friedlander, S. K. (1984) The carbon containing component of the Los
21 Angeles aerosol: source apportionment and contributions to the visibility budget. J. Air Pollut. Control
22 Assoc. 34: 643-650.
23
24 Rae, D. A. (1984) Benefits of visual air quality in Cincinnati - results of a contingent ranking survey, final
25 report. Palo Alto, CA: Electric Power Research Institute; report no. RP-1742.
26
27 Rehn, T.; Svartengren, M.; Philipson, K.; Camner, P. (1982) Mukociliar transport i lunga och nasa samt
28 luftvagsmotstand efter exponering for kvavedioxid [Mucociliary transport in the lung and nose after
29 exposure to nitrogen dioxide]. Vallingby, Sweden: Swedish State Power Board; project KHM technical
30 report no. 40.
31
32 Richters, A.; Damji, K. S. (1988) Changes in T-lymphocyte subpopulations and natural killer cells following
33 exposure to ambient levels of nitrogen dioxide. J. Toxicol. Environ. Health 25: 247-256.
34
35 Rodhe, H. (1990) A comparison of the contribution of various gases to the greenhouse effect. Science
36 (Washington, DC) 248: 1217-1219.
37
38 Roelofs, J. G. M. (1986) The effect of airborne sulphur and nitrogen deposition on aquatic and terrestrial
39 heathland vegetation. Experientia 42: 372-377.
40
41 Roelofs, J. G. M.; Boxman, A. W.; van Dijk, H. F. G. (1987) Effects of airborne ammonium on natural
42 vegetation and forests. In; Asman, W. A. H.; Diederen, H. S. M. A., eds. Ammonia and acidification:
43 proceedings [of a] symposium of the European Association for the Science of Air Pollution (EURASAP);
44 April; Bilthoven, The Netherlands. European Association for the Science of Air Pollution; pp. 266-276.
45
46 Roger, L. J.; Horstman, D. H.; McDonnell, W.; Kehrl, H.; Ives, P. J.; Seal, E.; Chapman, R.; Massaro, E.
47 (1990) Pulmonary function, airway responsiveness, and respiratory symptoms in asthmatics following
48 exercise in NO2. Toxicol. Ind. Health 6: 155-171.
49
50 Rombout, P, J. A.; Dormans, J. A. M. A.; Marra, M.; van Esch, G. J. (1986) Influence of exposure regimen
51 on nitrogen dioxide-induced morphological changes in the rat lung. Environ. Res. 41: 466-480.
52
August 1991 1-98 DRAFT-DO NOT QUOTE OR CITE
-------
1 Rosen, K. (1988) Effects of biomass accumulation and forestry on nitrogen in forest ecosystems. In: Nilsson, J.;
2 Grennfelt, P., eds. Critical loads for sulphur and nitrogen: report from a workshop; March; Skokloster,
3 Sweden. Copenhagen, Denmark: Nordic Council of Ministers; pp. 269-293.
4
5 Rowe, F. M.; Chamberlain, K. A. J. (1937) The "fading" of dyeings on cellulose acetate rayon: the action of
6 "burnt gas fumes" (oxides of nitrogen, etc. in the atmosphere) on cellulose acetate rayon dyes. J. Soc.
7 Dyers Colour. 53: 268-278.
8
9 Rudd, J. W. M.; Kelly, C. A.; Schindler, D. W.; Turner, M. A. (1988) Disruption of the nitrogen cycle in
10 acidified lakes. Science (Washington, DC) 240: 1515-1517.
11
12 Russell, A. G.; Cass, G. R. (1986) Verification of a mathematical model for aerosol nitrate and nitric acid
13 formation and its use for control measure evaluation. Atmos. Environ. 20: 2011-2025.
14
15 Russell, A. G.; McCue, K. F.; Cass, G. R. (1988) Mathematical modeling of the formation of
16 nitrogen-containing air pollutants. 1. Evaluation of an Eulerian photochemical model. Environ. Sci.
17 Technol. 22: 263-271.
18
19 Salvin, V. S. (1964) Relation of atmospheric contaminants and ozone to lightfastness. Am. Dyest. Rep.
20 53: 33-41. *
21
22 Salvin, V. S. (1969) Testing atmospheric fading of dyed cotton and rayon. Am. Dyest. Rep. 58: 28-29.
23
24 Salvin, V. S.; Walker, R. A. (1959) Relation of dye structure to properties of disperse dyes: part I—
25 anthraquinone blues. Am. Dyest. Rep. 48: 35-43.
26
27 Samet, J. M.; Utell, M. J. (1990) The risk of nitrogen dioxide: what have we learned from epidemiological and
28 clinical studies? Toxicol. Ind. Health 6: 247-262.
29
30 Samet, J. M.; Tager, I. B.; Speizer, F. E. (1983) The relationship between respiratory illness in childhood and
31 chronic air-flow obstruction in adulthood. Am. Rev. Respir. Dis. 127: 508-523.
32
33 Sandstrom, T.; Andersson, M. C.; Kolmodin-Hedman, B.; Stjernberg, N.; Angstrom, T. (1990a)
34 Bronchoalveolar mastocytosis and lymphocytosis after nitrogen dioxide exposure in man: a time-kinetic
35 study. Eur. Respir. J. 3: 138-143.
36
37 Sandstrom, T.; Bjenner, L.; Kolmodin-Hedman, B.; Stjernberg, N. (1990b) Nitrogen dioxide (NOj) induced
38 inflammation in the lung; attenuated response after repeated exposures. Am. Rev. Respir. Dis.
39 141(suppl.): A73.
40
41 Saxe, H. (1986) Stomatal-dependent and stomatal-independent uptake of NOX. New Phytol. 103: 199-205.
42
43 Schaefer, D. A.; Driscoll, C. T., Jr.; Van Dreason, R.; Yatsko, C. P. (1990) The episodic acidification of
44 Adirondack lakes during snowmelt. Water Resour. Res. 26: 1639-1647.
45
46 Schiff, L. J. (1977) Effect of nitrogen dioxide on influenza virus infection in hamster trachea organ culture. Proc.
47 Soc. Exp. Biol. Med. 156: 546-549.
48
49 Schimel, D. S.; Simians, S.; Rosswall, T.; Mosier, A. R.; Parton, W. J. (1988) Scale and the measurement of
50 nitrogen-gas fluxes from terrestrial ecosystems. In: Rosswall, T.; Woodmansee, R. G.; Risser, P. G.,
51 eds. Scales and global change. New York, NY: John Wiley & Sons; pp. 179-193.
52
August 1991 1-99 DRAFT-DO NOT QUOTE OR CITE
-------
1 Schlesiager, R. 1.; Driscoll, K. E.; Vollmuth, T. A. (1987) Effect of repeated exposures to nitrogen dioxide and
2 sulfuric acid mist alone or in combination on mucociliary clearance from the lungs of rabbits. Environ.
3 Res. 44: 294-301.
4
5 Sehofield, C. L.; Galloway, J. N.; Hendry, G. R. (1985) Surface water chemistry in the DLWAS basins. Water
6 Air Soil Pollut. 26: 403-423.
7
8 Schulze, W. D.; Brookshire, D. S.; Walther, E. G.; MacFarland, K. K.; Thayer, M. A.; Whitworth, R. L.;
9 Ben-David, S.; Malm, W.; Molenar, J. (1983) The economic benefits of preserving visibility in the
10 national parMaads of the southwest. Nat. Resour. J. 23: 149-173.
11
12 Schulze, E.-D.; de Vries, W.; Hauhs, M.; Rosen, K.; Rasmussen, L.; Tamm, C.-O.; Nilsson, J. (1989) Critical
13 loads for nitrogen deposition on forest ecosystems. Water Air Soil Pollut. 48: 451-456.
14
15 Schuurkes, J. A. A. R.; Kok, C. J.; den Hartog, C. (1986) Ammonium and nitrate uptake by aquatic plants from
16 poorly buffered and acidified waters. Aquat. Bot. 24: 131-146.
17
18 Schuurkes, J. A. A. R.; Elbers, M. A.; Gudden, J. J. F.; Roelofs, J. G. M. (1987) Effects of simulated
19 ammonium sulphate and sulphuric acid rain on acidification, water quality and flora of small-scale soft
20 water systems. Aquat *Bot. 28: 199-226.
21
22 Schwartz, S. E. (1989) Acid deposition: unraveling a regional phenomenon. Science (Washington, DC)
23 243: 753-763.
24
25 Seigneur, C.; Johnson, C. D.; Latimer, D. A.; Bergstrom, R. W.; Hogo, H. (1984) User's manual for the plume
26 visibility model (PLUVUE II). Research Triangle Park, NC: U. S, Environmental Protection Agency,
27 Environmental Sciences Research Laboratory; EPA report no. EPA-600/8-84-005. Available from: NTIS,
28 Springfield, VA; PB84-158302.
29
30 Sharps, W. E.; DeWalle, D. R.; Leibfried, R. T.; Dinicola, R. S.; Kimmel, W. G.; Sherwin, L. S. (1984)
31 Causes of acidification of four streams on Laurel Hill in southwestern Pennsylvania. J. Environ. Qual.
32 13: 619-631.
33
34 Silsbee, D. G.; Larson, G. L. (1982) Water quality of streams in the Great Smoky Mountains National Park.
35 Hydrobiologia89:97-115.
36
37 Simpson, J. C.; Olsen, A. R. (1990) Wet deposition temporal and spatial patterns in North America, 1987.
38 Research Triangle Park, NC: U. S. Environmental Protection Agency, Atmospheric Research and
39 Exposure Assessment Laboratory; EPA report no. EPA-600/4-90-019. Available from: NTIS,
40 Springfield, VA; PB90-251836.
41 .'•'".
42 Sinclair, T. R.; van Houtte, R. F. (1982) Simulative analysis of ammonia exchange between the atmosphere and
43 plant communities. Agric. Environ. 7: 237-242.
44
45 Sinn, J. P.; Pell, E. J. (1984) Impact of repeated nitrogen dioxide exposures on composition and yield of potato
46 foliage and tubers. J. Am. Soc. Hortic. Sci. 109: 481-484.
47
48 Skarby, L.; Bengtson, C.; Bostrom, C.-A.; Grennfelt, P.; Troeng, E. (1981) Uptake of NOX in Scots pine. Silva
49 Fenn. 15: 396-398.
50
51 Skeffington, R. A.; Wilson, E. J. (1988) Excess nitrogen deposition: issues for consideration. Environ. Pollut.
52 54: 159-184.
53
54 Sloane, C. S.; White, W. H. (1986) Visibility: an evolving issue. Environ. Sci. Technol. 20: 760-766.
August 1991 1-100 DRAFT-DO NOT QUOTE OR CITE
-------
1 Smith, V. H. (1982) The nitrogen and phosphorus dependence of algal biomass in lakes: an empirical and
2 theoretical analysis. Limnol. Oceanogr. 27: 1101-1112.
3
4 Smith, R. A.; Alexander, R. B.; Wolman, M. G, (1987) Water-quality trends in the nation's rivers. Science
5 (Washington, DC) 235: 1607-1615.
6
7 Smith, R. A.; Alexander, R. B.; Wolman, M. G. (1987) Analysis and interpretation of water quality trends in
8 major U.S. rivers, 1974-1981. U. S. Geol. Surv. Water-Supply Pap. 2307.
9
10 Smullen, J, T.; Taft, J. L.; Macknis, J. (1982) Nutrient and sediment loads to the tidal Chesapeake Bay system.
11 In: Chesapeake Bay Program technical studies: a synthesis. Annapolis, MD: U. S. Environmental
12 Protection Agency; pp. 150-251. Available from: NTIS, Springfield, VA; PB84-111202.
13
14 Solomon, P. A.; Movers, J. L. (1984) Use of a high volume dicnotomous virtual impactor to estimate light
15 extinction due to carbon and related species in the Phoenix haze. Sci. Total Environ. 36: 169-175.
16
17 Solomon, P. A.; Moyers, J. L. (1986) A chemical characterization of wintertime haze in Phoenix, Arizona.
18 Atmos. Environ. 20: 207-213.
19
20 Spence, J. W.; Haynie, F.; Upham, J. B. (1975) Effects of gaseous pollutants on paints: a chamber study.
21 J. Paint Technol. 47: 57-63.
22
23 Spicer, C. W. (1982) The distribution of oxidized nitrogen in urban air, Sci. Total Environ. 24: 183-192,
24
25 Spicer, C. W.; Sverdrup, G. M.; Kuhlman, M. R. (1981) Smog chamber studies of NOX chemistry in power
26 plant plumes. Atmos. Environ. 15: 2353-2365.
27
28 Stacy, R. W.; Seal, E., Jr.; House, D. E.; Green, J.; Roger, L. J.; Raggio, L. (1983) A survey of effects of
29 gaseous and aerosol pollutants on pulmonary function of normal males. Arch. Environ. Health
30 38: 104-115.
31
32 Stavert, D, M.; Archuleta, D. C.; Holland, L. M.; Lehnert, B. E. (1986) Nitrogen dioxide exposure and
33 development of pulmonary emphysema. J. Toxicol. Environ. Health 17: 249-267.
34
35 Stephens, R. J.; Freeman, G.; Evans, M. J. (1972) Early response of lungs to low levels of nitrogen dioxide:
36 light and electron microscopy. Arch. Environ. Health 24: 160-179.
37
38 Stevens, R. K. (1987) [Personal communication to J. C. Trijonis]. Bloomington, MN: Sante Fe Research
39 Corporation [as cited in Trijonis (1987)].
40
41 Stevens, R. K.; King, F.; Bell, J.; Wbitfield, J. (1988) Measurement of the chemical species that contribute to
42 urban haze. Presented at: 81st annual meeting of the Air Pollution Control Association; June; Dallas,
43 TX. Pittsburgh, PA: Air Pollution Control Association; paper no. 88-57.3.
44
45 Stoddard, J. L. (1987) Alkalinity dynamics in an unacidified alpine lake, Sierra Nevada, California, Limnol,
46 Oceanogr. 32: 825-839.
47
48 Stoddard, J. L.; Kellogg, J. H. (n.d.) Trends and patterns in lake acidification in the state of Vermont: evidence
49 from the Long-Term Monitoring Project. Water Air Soil Pollut.: in press.
50
51 Stoddard, J. L.; Murdoch, P. S. (1991) Catskill Mountains: an overview of chronic and episodic acidity in dilute
52 Catskill Mountain streams. In: Charles, D. F., ed. Acid deposition and aquatic ecosystems: regional case
53 studies of acid deposition. Springer-Verlag.
54
August 1991 1-101 DRAFT-DO NOT QUOTE OR CITE
-------
1 Suzuki, T.J Dceda, S.; Kanoh, T.; Mizoguchi, I. (1986) Decreased phagocytosis and superoxide anion production
2 in alveolar macrophages of rats exposed to nitrogen dioxide. Arch. Environ. Contam. Toxicol.
3 15: 733-739.
4
5 Svedung, O. A.; Johansson, L.-G.; Vannerberg, N.-G. (1983) Corrosion of gold-coated contact materials
6 exposed to humid atmospheres containing low concentrations of SO2 and NO2. IEEE Trans, Compon.
7 Hybrids Manuf. Technol. CHMT-6: 349-355.
8
9 Swank, W. T.; Waide, J. B. (1988) Characterization of baseline precipitation and stream chemistry and nutrient
10 budgets for control watersheds. In: Swank, W. T.; Crossiey, D. A., Jr., eds. Forest hydrology and
11 ecology at Coweeta. New York, NY: Springer-Verlag; pp. 57-79. (Billings, W. D.; Golley, F.; Lange,
12 O. L.; Olson, J. S.; Remmert, H., eds. Ecological studies: analysis and synthesis, v. 66).
13
14 Tolbert, M. A.; Rossi, M. J.; Malhotra, R.; Golden, D. M. (1987) Reaction of chlorine nitrate with hydrogen
15 chloride and water at Antarctic stratospheric temperatures. Science (Washington, DC) 238: 1258-1260.
16
17 Tolbert, M. A.; Rossi, M. J.; Golden, D. M. (1988) Antarctic ozone depletion chemistry: reactions of N2OS with
18 H,O and HC1 on ice surfaces. Science (Washington, DC) 240: 1018-1021.
19
20 Tolley, G. A.; Randall, A.; Blomquist, G.; Fabian, R.; Fishelson, G.; Frankel, A.; Hoehn, J.; Krumm, R.;
21 Mensah, E.; Smith, T. (1986) Establishing and valuing the effects of improved visibility in eastern
22 United States. Washington, DC: U. S. Environmental Protection Agency.
23
24 Trijonis, J.; Cass, G.; McRae, G.; Hone, Y.; Lim, W.-Y.; Chang, N.; Cahill, T. (1982) Analysis of
25 visibility/aerosol relationships and visibility modeling/monitoring alternatives for California. Sacramento,
26 CA: State of California, Air Resources Board; report no, ARB-R-83-197. Available from: NTIS;
27 Springfield, VA; PB83-245563.
28
29 Trijonis, J.; Thayer, M.; Murdoch, J.; Hageman, R. (1984) Air quality benefits analysis for Los Angeles and
30 San Francisco based on housing values and visibility. Sacramento, CA: California Air Resources Board.
31
32 Trijonis, J.; McGown, M.; Pitchford, M.; Blumenthal, D.; Roberts, P.; White, W.; Macias, E.; Weiss, R.;
33 Waggoner, A.; Watson, J.; Chow, J.; Flocchini, R. (1988) Visibility conditions and causes of visibility
34 degradation in the Mojave Desert of California: executive summary, RESOLVE project final report.
35 China Lake, CA: Department of the Navy, Naval Weapons Center; document no. NWC TP 6869.
36 Available from: NTIS, Springfield, VA; AD-A206 322.
37
38 Tyler, M. (1988) Contribution of atmospheric nitrate deposition to nitrate loading in the Chesapeake Bay.
39 Annapolis, MD: Department of Natural Resources, Chesapeake Bay Research & Monitoring Division;
40 report no. AD-88-7.
41 - .
42 U. S, Cods. (1991) Clean Air Act, §108, air quality criteria and control techniques, §109, national ambient air
43 quality standards. U. S. C. 42: §§7408-7409. •
44
45 U. S. Environmental Protection Agency. (1982a) Air quality criteria for oxides of nitrogen. Research Triangle
46 Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
47 Office; EPA report no. EPA-600/8-82-026. Available from: NTIS, Springfield, VA; PB83-131011.
48
49 U. S, Environmental Protection Agency. (1982b) Review of the national ambient air quality standards for
50 nitrogen oxides: assessment of scientific and technical information; OAQPS staff paper. Research
51 Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/5-82-002.
52 Available from: NTIS, Springfield, VA; PB83-132829.
53
August 1991 M02 DRAFT-DO NOT QUOTE OR CITE
-------
1 Utell, M, J.; Swinburne, A. J.; Hyde, R. W«; Speers, D. M.; Gibb, F. R.; Morrow, P. E. (1979) Airway
2 reactivity to nitrates in normal and mild asthmatic subjects. J. Appl, Physiol.: Respir. Environ. Exercise
3 Physiol, 46: 189-196.
4
5 Utell, M. J.; Aquilina, A. T.; Hall, W. J.; Speers, D. M.; Douglas, R. G., Jr.; Gibb, F. R.; Morrow, P. E.;
6 Hyde, R. W. (1980) Development of airway reactivity to nitrates in subjects with influenza. Am. Rev.
7 Respir. Dis. 121: 233-241.
8
9 van Breemen, N.; van Dijk, H. F. G. (1988) Ecosystem effects of atmospheric deposition of nitrogen in The
10 Netherlands. In: Dempster, J. P.; Manning, W. J.; Skeffington, R. A., eds. Excess nitrogen deposition:
11 [papers from the workshop]; September 1987; Leatherhead, Surrey, United Kingdom. Environ. Pollut.
12 54: 249-274.
13
14 van Breemen, N.; Burrough, P. A.; Velthorst, E. J.; van Dobben, H. F.; de Wit, T.; Ridder, T. B.; Reijnders,
15 H. F. R. (1982) Soil acidification from atmospheric ammonium sulphate in forest canopy throughfall.
16 Nature (London) 299: 548-550.
17
18 van Breemen, N.; Mulder, J.; van Grinsven, J. J. M. (1987) Impacts of acid atmospheric deposition on
19 woodland soils in the Netherlands: n. nitrogen transformations. Soil Sci. Soc. Am. J. 51: 1634-1640.
20
21 Van der Eerden, L. J. M. (1982) Toxicity of ammonia to plants. Agric. Environ. 7: 223-235.
22
23 Vanderpol, A. H.; Humbert, M. E. (1981) Coloration of power plant plumes—NO2 or aerosols? Atmos. Environ.
24 15:2105-2110.
25
26 Vijayakumar, R.; Mansfeld, F.; Henry, R. (1989) Investigation of the effects of acid deposition on materials
27 [final report]. Camarillo, CA: Combustion Engineering, Environmental Monitoring and Services, Inc.;
28 contract nos. A4-110-32 and A5-137-32 for California Air Resources Board, Sacramento, CA.
29
30 von Nieding, G.; Wagner, H. M. (1977) Experimental studies on the short-term effect of air pollutants on
31 pulmonary function in man: two-hour exposure to NO2, O3 and SO2 alone and in combination. In:
32 Kasuga, S,; Suzuki, N.; Yamada, T.; Kimura, G.; Inagaki, K.; Onoe, K., eds. Proceedings of the fourth
33 international clean air congress; May; Tokyo, Japan. Tokyo, Japan: Japanese Union of Air Pollution
34 Prevention Associations; pp. 5-8.
35
36 von Nieding, G.; Wagner, H. M. (1979) Effects of NO2 on chronic bronchitics. Environ. Health Perspect.
37 29: 137-142.
38
39 von Nieding, G.; Wagner, H. M.; Krekeler, H.; Smidt, U.; Muysers, K. (1970) Absorption of NO2 in low
40 concentrations in the respiratory tract and its acute effects on lung function and circulation. Presented at:
41 the second international clean air congress; December; Washington, DC; paper no. MB-15G.
42
43 von Nieding, G.; Wagner, M.; Krekeler, H.; Smidt, U.; Muysers, K. (1971) Grenzwertrjestimmung der akuten
44 NO2-Wirkung auf den respiratorisehen Gasaustausch und die Atemwegswiderstaende des chronisch
45 lungenkranken Mensehen [Minimum concentrations of NO2 causing acute effects on the respiratory gas
46 exchange and airway resistance in patients with chronic bronchitis]. Int. Arch. Arbeitsmed. 27: 338-348.
47
48 von Nieding, G.; Krekeler, H.; Fuchs, R.; Wagner, M.; Koppenhagen, K. (1973a) Studies of the acute effects of
49 NO2 on lung function: influence on diffusion, perfusion and ventilation in the lungs. Int. Arch.
50 Arbeitsmed. 31: 61-72.
51
52
August 1991 1-103 DRAFT-DO NOT QUOTE OR CITE
-------
1 von Nieding, G.; Wagner, H. M.; Krekeler, H. (1973b) Investigation of the acute effects of nitrogen monoxide
2 on lung function in man. In: Proceedings of the third international clean air congress; October;
3 Duesseldorf, Federal Republic of Germany. Duesseldorf, Federal Republic of Germany: Verein Deutscher
4 Ingenieure; pp. A14-A16.
5
6 von Nieding, G.; Wagner, H. M.; Krekeler, H.; Loellgen, H.; Fries, W.; Beuthan, A. (1979) Controlled studies
7 of human exposure to single and combined action of NO2, O3, and SO2. Int. Arch. Occup, Environ.
8 Health 43: 195-210.
9
10 von Nieding, G,; Wagner, H. M.; Casper, H.; Beuthan, A.; Smidt, U. (1980) Effect of experimental and
11 occupational exposure to NO2 in sensitive and normal subjects. In: Lee, S. D., ed. Nitrogen oxides and
12 their effects on health. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.; pp. 315-331.
13
14 Vossler, T. L.; Lewis, C. W.; Stevens, R. K.; Dzubay, T. G.; Gordon, G. E.; Tuncel, S. G.; Russwurm,
15 G. M.; Keeler, G. J. (1989) Composition and origin of summertime air pollutants at Deep Creek Lake,
16 Maryland. Atmos. Environ. 23: 1535-1547.
17
18 Voytko, J. E.; Guilinger, T. R. (1988) Effectiveness of electrodeposited coatings in reducing atmospheric
19 corrosion of copper substrates. Proc. AESF Annu, Tech. Conf. 75: C-3, llpp.
20
21 Waggoner, A. P.; Weiss, R. E.; Ahlquist, N. C. (1983) The color of Denver haze. Atmos. Environ.
22 17:2081-2086,
23
24 Ware, J. H.; Dockery, D. W.; Spiro, A., IH; Speizer, F. E.; Ferris, B. G., Jr. (1984) Passive smoking, gas
25 cooking, and respiratory health of children living in six cities. Am. Rev. Respir. Dis. 129: 366-374.
26
27 Waring, R. H. (1987) Nitrate pollution: a particular danger to boreal and subalpine coniferous forests. In:
28 Fujimori, T.; Kunura, M., eds. Human impacts and management of mountain forests: [proceedings of a
29 symposium]. Ibaraki, Japan: Forestry and Forest Products Research Institute; pp. 93-105.
30
31 Welch, E. B.; Spyridakis, D. E.; Smayda, T. (1986) Temporal chemical variability La acid sensitive high
32 elevation lakes. Water Air Soil Pollut. 31: 35-44.
33
34 Weller, D. E.; Peterjohn, W. T.; Goff, N. M.; Correll, D. L. (1986) Ion and acid budgets for a forested
35 Atlantic Coastal Plain watershed and their implications for the impacts of acid deposition. In: Correll,
36 D. L., ed. Watershed research perspectives. Washington, DC: Smithsonian Institution Press;
37 pp. 392-421.
38
39 Wetselaar, R.; Farquhar, G. D. (1980) Nitrogen losses from tops of plants. Adv. Agron. 33: 263-302.
40
41 White, W, H. (1977) NOX-O3 photochemistry in power plant plumes: comparison of theory with observation.
42 Environ. Sci. Technol. 11: 995-1000.
43
44 White, W. H.; Seigneur, C.; Heinold, D. W.; Eltgroth, M. W.; Richards, L. W.; Roberts, P. T.; Bhardwaja,
45 P. S.; Conner, W. D.; Wilson, W. E., Jr. (1985) Predicting the visibility of chimney plumes: an
46 intercomparison of four models with observations at a well-controlled power plant. Atmos. Environ.
47 19: 515-528.
48
49 White, W. H.; Seigneur, C.; Heinold, D. W.; Richards, L. W.; Wilson, W. E.; Roberts, P. T. (1986) Radiative
50 transfer budgets for scattering and absorbing plumes: measurements and model predictions. Atmos.
51 Environ. 20: 2243-2257.
52
August 1991 1-104 DRAFT-DO NOT QUOTE OR CITE
-------
1 Wigington, P. J.; Davies, T, D,; Tranter, M.; EsMeman, K. (1989) Episodic acidification of surface waters due
2 to acidic deposition. Washington, DC: National Acid Precipitation Assessment Program; State-of-scienee
3 technology report 12.
4
5 Wolff, G. T.; Korsog, P. E. (1989) Atmospheric concentrations and regional source apportionments of sulfate,
6 nitrate and sulfur dioxide in the Berkshire Mountains in western Massachusetts. Atmos. Environ.
7 23: 55-65. •-.....••- . •.'-. •:..;....y.w ,..•..
8
9 Woodwell, G. M. (1970) Effects of pollution on the structure and physiology of ecosystems: changes in natural
10 ecosystems caused by many different types of disturbances are similar and predictable. Science
11 (Washington, DC) 168: 429-433.
12
13 World Health Organization. (1987) The effects of nitrogen on vegetation. In: Air quality guidelines for Europe.
14 Copenhagen, Denmark: Regional Office for Europe; pp., 373-385. (WHO regional publications, European
15 series no. 23).
16
17 Wuebbles, D. J. (1989) On the mitigation of non CO2 greenhouse gases. UCRL-101523. Available from: NTIS,
18 Springfield, VA; DE89-016659.
19
20 Yamamoto, L; Takahashi, M. (1984) Ultrastructural observations of rat lung exposed to nitrogen dioxide for
21 7 months. Kitasato Arch. Exp. Med. 57; 57-65.
22
23 Zemba, S. G.; Golomb, D.; Fay, J. A. (1988) Wet sulfate and nitrate deposition patterns in eastern North
24 America. Atmos. Environ. 22: 2751-2761,
25
26 Zeronian, S. H.; Alger, K. W.; Onraye, S. T. (1971) Reaction of fabrics made from synthetic fibers to air
27 contaminated with nitrogen dioxide, ozone, or sulfur dioxide. In: Englund, H. M.; Beery, W. T., eds.
28 Proceedings of the second international clean air congress; December 1970; Washington, DC. New York,
29 NY: Academic Press; pp. 468-476.
30
August 1991 1-105 DRAFT-DO NOT QUOTE OR CITE
-------
-------
i 2. INTRODUCTION
2
3
4 This revised Air Quality Criteria for Oxides of Nitrogen reviews and evaluates the
5 scientific information on the health and welfare effects associated with exposure to the
6 concentrations of nitrogen dioxide (NO2) found in ambient air. The purpose of this
7 document is to present air quality criteria for oxides of nitrogen (NOX) in accordance with
8 Sections 108 and 109 of the Clean Air Act (CAA).
9 Section 108 (U.S. Code, 1991) directs the Administrator of the U. S. Environmental
0 Protection Agency (EPA) to list pollutants that may reasonably be anticipated to endanger
1 public health and welfare, and to issue air quality criteria for them. These air quality criteria
2 are to reflect the latest scientific information useful in indicating the kind and extent of all
3 identifiable effects on public health and welfare that may be expected from the presence of
4 the pollutant in the ambient air.
5 Section 109(a,b) (U.S. Code, 1991) directs the EPA Administrator to propose and
6 promulgate "primary" and "secondary" National Ambient Air Quality Standards (NAAQS)
7 for pollutants identified under Section 108. Section 109(b)(l) defines a primary standard as a
8 level of air quality, the attainment and maintenance of which in the judgment of the
3 Administrator, based on the criteria and allowing for an adequate margin of safety, is
3 requisite to protect the public health. Section 109(d) of the Act (U.S. Code, 1991) requires
1 periodic review and, if appropriate, revision of existing criteria and standards. In addition,
2 Section 109(c) specifically requires the Administrator to promulgate a primary standard for
3 NO2 with an averaging time of not more than 3 h, unless no significant evidence is found
4- that such a short-term standard is required to protect health. Under Section 109(b) of the
5 Clean Air Act, the Administrator must consider available information to set secondary
S NAAQS that are based on the criteria and are requisite to protect the public welfare from any
7 known or anticipated adverse effects associated with the presence of such pollutants. The
3 welfare effects included in the criteria are effects on vegetation, crops, soils, water, animals,
? man-made materials, weather, visibility, and climate, as weE as damage to and deterioration
3 of property, hazards to transportation, and effects on economic values, personal comfort, and
I well-being.
August 1991 2-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 A variety of NOX compounds and their transformation products occur naturally in the
2 environment and also result from human activities. In addition to NO2, nitric oxide (NO),
3 nitrous oxide (N2O), gaseous nitrous acid (HONO), gaseous nitric acid (HNO3)» and both
4 nitrite and nitrate particles have all been found in the ambient air. The formation of
5 nitrosamines in the atmosphere by reaction of NOX with amines has also been suggested, but
6 not yet convincingly demonstrated. Available scientific research on the potential health and
7 welfare effects of NOX compounds provides the strongest evidence linking specific adverse
8 effects to near-ambient concentrations of NO2. Therefore, EPA has focused its criteria
9 reviews primarily on health and welfare effects reported to be associated with exposure to
10 NO2. Nitrogen dioxide is an air pollutant generated mainly by the photochemical oxidation
11 of NO, which is emitted from a variety of mobile and stationary sources. At elevated
12 concentrations, NO2 can adversely affect human health, vegetation, materials, and visibility.
13 Nitrogen oxide compounds can also contribute to increased rates of acidic deposition and high
14 ozone concentrations.
15
16 Regulatory and Scientific Background
17 On April 30, 1971, EPA first promulgated the NAAQS for NO2 under Section 109 of
18 the Clean Air Act (Federal Register, 1971). Identical primary and secondary standards for
*3
19 NO2 were set at 0.053 ppm (100 /tg/m ), averaged over 1 year. The scientific bases for
20 these standards are contained in the original criteria document, Air Quality Criteria for
21 Nitrogen Oxides (U.S. Environmental Protection Agency, 1971). The primary standard set
22 in 1971 was based largely on community epidemiology studies (Shy et al., 1970a,b; Pearlman
23 et al., 1971) conducted in Chattanooga, TN, which reported respiratory effects in children
24 exposed to low-level NO2 concentrations over a long-term period.
25 In accordance with Sections 108 and 109 of the Clean Air Act, in 1985 EPA completed
26 the review of criteria upon which the existing primary and secondary NO2 NAAQS were
27 based. Reevaluation of the Chattanooga studies in view of later information (especially
28 regarding the accuracy of the air quality monitoring method for NO2 used in the studies)
29 indicated that these studies provide only limited qualitative evidence for an association
30 between health effects and ambient exposures to NO2. In reviewing the scientific basis for an
31 annual standard, EPA found that evidence showing the most serious health effects associated
August 1991 2-2 DRAFT-DO NOT QUOTE OR CITE
-------
I with chronic NO2 exposures (e.g., emphysematous-like alterations in the lung and increased
I susceptibility to infection) came from animal studies conducted at concentrations well above
5 those permitted in the ambient air by the current annual standard. Major factors impeding
!• use of these studies for standard-setting purposes included limitations of methods for
5 quantitatively extrapolating exposure-response relationships from these animal studies directly
3 to humans. However, the seriousness of these effects, coupled with biological similarities
? between humans and test animals, suggested that there was some risk to human health from
I long-term exposure to elevated NO2 levels. Other evidence from community epidemiology
) and gas stove epidemiology studies provided some qualitative support for concluding that
) there is a relationship between adverse health effects and repeated, acute exposures to
elevated (>0.2-1.0 ppm) NO2 concentrations or longer-term human exposure to near-ambient
E levels of NO2. However, concern at that time for limitations associated with these studies
1 (e.g., unreliable or insufficient monitoring data and inadequate treatment of potential
confounding factors such as humidity and pollutants other than NO^ then precluded
> derivation of quantitative exposure-response relationships.
i While it is not possible to quantify the margin of safety provided by the existing annual
' standard, two observations are relevant: (1) a 0.053-ppm standard was consistent with the
! Clean Air Science Advisory Committee's recommendation (Friedlander, 1982) to set the
1 annual standard at the lower end of the range (0.05 to 0.08 ppm) cited in the Office of Air
i Quality Planning and Standards (OAQPS) Staff Paper (U.S. Environmental Protection
Agency, 1982b) to ensure an adequate margin of safety against long-term effects and to
provide some measure of protection against possible short-term health effects; and
(2) a 0.053 ppm standard would keep annual NO2 concentrations considerably below the
long-term exposure levels for which serious chronic effects have been observed in animals.
Maintaining the current annual primary standard thus represented at that time a prudent public
health policy choice aimed at preventing any increased chronic health risk in large, populated
urban areas that attain the annual standard. On July 19, 1985, EPA announced the final
decision that retained the existing annual primary and secondary standards. The decision on
the need, if any, for a separate short-term standard (less than 3 h) was deferred, pending
results from additional research focused on reducing uncertainties associated with evaluating
short-term health effects of NO2.
August 1991 2-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 Critical Issues
2 Based on the available scientific evidence several critical questions or issues are
3 addressed in this document. Following are some key issues.
4
5 1. Is the apparent relationship between short-term exposure (1 to 3 h) in the
6 range 0.2 to 0.5 ppm NO2 and increased bronchial reactivity in asthmatics
7 an adequate basis to form criteria for a short-term NO2 ambient standard?
8 How frequently do such short-term ambient levels occur in or above this
9 concentration range?
10
11 2. Is the strength and consistency of the epidemiologic data base and its
12 analysis relating NO2 exposure and an increased rate and/or severity of
13 respiratory disease and symptoms adequate to quantitatively assess whether
14 ambient or near ambient levels pose an increased level of risk?
15
16 3. Is the host defense data base adequate to provide a biological basis for the
17 relationship seen in epidemiologic studies between respiratory disease and
18 exposure to NO2? »
19
20 4. What is the emphysematous potential in humans from exposure to long-
21 term chronic ambient levels of NO2?
22
23 5. What subpopulation groups might be more susceptible to effects from
24 ambient NO2 exposure and potentially at heightened risk?
25 • ' -.-•••'
26 6. Is the data base concerning the ecological effects of critical nitrogen loading
27 adequate to allow conclusions to be determined?
28
29
30 Organization of the Document
31 The document consists of 16 chapters. The Executive Summary for the entire document
32 is contained in Chapter 1, followed by this general introduction in Chapter 2. Chapters 3
33 through 8 provide background information on: physical and chemical properties of NO2 and
34 related compounds; sources and emissions; atmospheric transport, transformation and fate of
35 NO2; methods for the collection and measurement of NO2; and ambient air concentrations
36 and factors affecting exposure of the general population. Chapter 9 evaluates NO2 effects on
37 crops and natural vegetation, while Chapter 10 discusses effects on terrestrial and aquatic
38 ecosystems. Chapter 11 describes effects on visibility, and Chapter 12 describes damage to
39 materials attributable to NO2. Chapters 13 through 16 evaluate information concerning the
40 health effects of NO2. More specifically, Chapter 13 discusses respiratory tract deposition of
August 1991 2-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 NO2 and information derived from experimental lexicological studies of animals. Chapter 14
2 discusses epidemiological studies, and Chapter 15 discusses clinical studies. Chapter 16
3 integrates information on critical health issues arising from all three experimental approaches.
4 Neither control techniques nor control strategies for the abatement of NOX are discussed
5 in this document, although some of the topics included are relevant to abatement strategies.
6 Technology for controlling the emissions of NOX and of volatile organic compounds is
7 discussed in documents issued by OAQPS (e.g., U.S. Environmental Protection Agency,
8 1978, 1983). Likewise, issues germane to the scientific basis for control strategies, but not
9 pertinent to the development of criteria, are addressed in numerous documents issued by
10 OAQPS.
11 In addition, certain issues of direct relevance to standard-setting are not explicitly
12 addressed in this document, but are contained instead in documentation prepared by OAQPS
13 as part of its regulatory analyses. Such analyses include (1) discussion of what constitutes
14 an"adverse effect," that is, the effects that the NAAQS is intended to protect against;
15 (2) assessment of risk; and (3) discussion of factors to be considered in determining an
16 adequate margin of safety. While scientific data contribute significantly to decisions
17 regarding these three issues, their resolution cannot be achieved solely on the basis of
L8 experimentally acquired information. Final decisions on items (1) and (3) are made by the
19 Administrator, as mandated by the CAA.
IQ A fourth issue directly pertinent to standard-setting is identification of the population at
II risk, which is basically a selection by EPA of the population to be protected by the
12 promulgation of% a given standard. This issue is addressed only partially in this document.
13 For example, information is presented on factors, such as pre-existing disease, that
*
14 biologically may predispose individuals and subpopulations to adverse effects from exposures
15 to NOX. The identification of a population at risk, however, requires information above and
16 beyond data on biological predisposition, such as information on levels of exposure, activity
17 patterns, and personal habits. Such information is included in a Staff Paper developed by
18 OAQPS.
\9 This document consists of the review and evaluation of relevant literature on NO.,
A
»0 through early 1991. The material selected for review and comment in the text generally
il comes from the more recent literature, with emphasis on studies conducted at or near
August 1991 2-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 pollutant concentrations found in ambient air. Older literature that was cited in the previous
2 criteria document for NOX (U.S. Environmental Protection Agency, 1982a) is generally not
3 discussed. However, as appropriate, there is some discussion of older studies judged to be
4 significant because of their potential usefulness in deriving a NAAQS. The newer
5 information on NOX available may in some instances make possible a better understanding of
6 earlier studies, such that a more detailed and comprehensive picture of health and welfare
7 effects is emerging. An attempt has been made to discuss key literature in the text and
8 present it in tables as well. Reports of lesser importance to the purposes of this document
9 may appear in tables only.
10 Generally, only published material that has undergone scientific peer review is included.
11 In the interest of admitting new and important information, however, some material not yet
12 published in the open literature but meeting other standards of scientific reporting may be
13 included. Emphasis has been placed on studies in which exposure concentrations were
14 ^5 ppm. On this basis, studies in which the lowest concentration employed exceeded this
15 level have been included if they contain unique data, such as documentation of a previously
16 unreported effect or of mechanisms of effects; or if they were multiple-concentration studies
17 designed to provide information on concentration-response relationships. In the areas of
18 emphysema, mutagenesis, teratogenesis, and reproductive effects, results of studies conducted
19 at higher levels have been included because of the potential importance of these effects to
20 public health.
21 In reviewing and summarizing the literature, an attempt has been made to present
22 alternative points of view where scientific controversy exists. As warranted, considerations
23 bearing on the quality of studies have been included.
• .
24 The general policy of EPA is to express concentrations of air pollutants in metric units
25 (e.g., in micrograms per cubic meter [jug/m3]); as well as in the more widely used units,
26 parts per million (ppm) or parts per billion (ppb), which are neither metric nor English units.
27 That policy has been followed in those chapters in which most of the data have been obtained
28 from laboratory studies conducted at room temperature (e.g., Chapters 13 and 16). Data
29 reported in parts per million for studies conducted outdoors, such as field and open-top
30 chamber vegetation studies, ambient air monitoring, and research on atmospheric chemistry,
31 have not been converted. Conversion of reported parts per million and parts per billion units
August 1991 2-6 DRAFT-DO NOT QUOTE OR CITE
-------
is questionable in these cases because it assumes standard or uniform temperatures and
pressures. Deposition studies are reported in micrograms per hectare (/*g/ha).
August 1991 2-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Federal Register. (1971) National primary and secondary ambient air quality standards, F. R. (April 30) 36:
3 8186-8201.
4
5 Friedlander, S. K, (1982) CASAC review and closure of the OAQPS staff paper for nitrogen oxides
6 [memorandum to Anne M. Gorsuch]. July 6.
7
8 Pearlman, M. E.; Finklea, J. F.; Creason, J. P.; Shy, C. M.; Young, M. M.; Horton, R. J, M. (1971) Nitrogen
9 dioxide and lower respiratory illness. Pediatrics 47: 391-398.
10
11 Shy, C, M.; Creason, J. P.; Pearlman, M. E.; McClain, K. E.; Benson, F. B.; Young, M. M. (1970a) The
12 Chattanooga school children study; effects of community exposure' to nitrogen dioxide. 1. Methods,
13 description of pollutant exposure, and results of ventilatory function testing. J. Air PoIIut. Control Assoc.
14 20:539-545.
15 .
16 Shy, C. M.; Creason, J. P.; Pearlman, M. E.; McClain, K. E.; Benson, F. B.; Young, M. M. (1970b) The
17 Chattanooga school children study: effects of community exposure to nitrogen dioxide. II. Incidence of
18 acute respiratory illness. J. Air Pollut. Control Assoc. 20: 582-588.
19
20 U. S. Code. (1991) Clean Air Act, § 108, air quality criteria and control techniques, § 109, national ambient air
21 quality standards. U. S. C. 42: §§ 7408-7409.
22
23 U. S. Environmental Protection Agency. (1971) Air quality criteria for nitrogen oxides. Washington, DC: U. S.
24 Environmental Protection Agency, Air Pollution Control Office; EPA report no, AP-84. Available from:
25 NTIS, Springfield, VA; PB-197333/BE.
26
27 U. S. Environmental Protection Agency. (1978) Control techniques for volatile organic emissions from stationary
28 sources. Research Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no.
29 EPA-450/2-78-022. Available from: NTIS, Springfield, VA; PB-284804.
30
31 U. S. Environmental Protection Agency. (1982a) Air quality criteria for oxides of nitrogen. Research Triangle
32 Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
33 Office; EPA report no. EPA-600/8-82-026. Available from: NTIS, Springfield, VA; PB83-131011.
34
35 U. S, Environmental Protection Agency, (1982b) Review of the national ambient air quality standards for
36 nitrogen oxides: assessment of scientific and technical information; OAQPS staff paper. Research
37 Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no, EPA-450/5-82-002.
38 Available from: NTIS, Springfield, VA; PB83-132829.
39
40 U. S. Environmental Protection Agency. (1983) Control techniques for nitrogen oxides emissions from stationary
41 sources - revised second edition. Research Triangle Park, NC: Office of Air Quality Planning and
42 Standards; EPA report no. EPA-450/3-83-002. Available from: NTIS, Springfield, VA;
43 PB84-118330/REB.
44
August 1991 2-8 DRAFT-DO NOT QUOTE OR CITE
-------
i 3. GENERAL CHEMICAL AND PHYSICAL
2 PROPERTIES OF NOX AND NOX-DERIVED
3 POLLUTANTS
4
5
6 3.1 INTRODUCTION AND OVERVIEW
7 In this chapter some general chemical and physical properties of NOX* and
8 NOx-derived pollutants are discussed by way of introduction to the complex chemical and
9 physical interactions which may occur in the atmosphere and other media. The discussion
0 will be significantly augmented throughout the document as particular topics are discussed in
1 depth.
2 There are eight oxides of nitrogen that may be present in the ambient air: nitric oxide
3 (NO), nitrogen dioxide (NO2), nitrous oxide (N2O), nitrogen trioxide (NO3), dinitrogen
4 trioxide (N2O3), dinitrogen tetroxide (N2O£, and dinitrogen pentoxide (N2O5). Of these,
5 NO and NO2 are generally considered the most important in the lower troposphere because
5 they may be present in significant concentrations (Chapter 7). Their interconvertibility in
7 photochemical smog reactions (Chapter 5) has frequently resulted in their being grouped
3 together under the designation NOX, although analytic techniques can distinguish clearly
J between them (Chapter 6). Of the two, NO2 has the greater impact on human health
) (Chapter 16).
I Nitrous oxide is ubiquitous even in the absence of anthropogenic sources, since it is a
I product of natural biologic processes in soil. It is not known, however, to be'involved in any
\ photochemical smog reactions. Although N2O is not generally considered to be an air
!• pollutant, it participates in upper atmospheric reactions involving the ozone layer (Chapter 5).
j While NO3, N2O3, N2O4, and N2O5 may play a role in atmospheric chemical reactions
> leading to the transformation, transport, and ultimate removal of nitrogen compounds from
r ambient air, they are present only in very low concentrations, even in polluted environments.
"In this document, NOX is the sum of nitrogen dioxide (NO^ and nitric oxide (NO). NOj, refers to the sum of
NOX and other oxidized nitrogen compounds except N2O. These include HNO3, NO3, N2O3, N2O4 and N2O5. PAN
(peroxyaeetylnitrate), while not discussed at great length in this document, is normally included with the NOy group
of compounds.
August 1991 3-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ammonia (NH3) is generated, on a global scale, during the decomposition of
2 nitrogenous matter in natural ecosystems and it may also be produced locally in larger
3 concentrations by human activities such as the maintenance of dense animal populations. It is
4 discussed briefly in this document to facilitate understanding of the nitrogen cycle and, also,
5 because some researchers have suggested that NH3 is converted to NOX in the atmosphere.
6 Other NOx-derived compounds which may be found in polluted air include nitrites,
7 nitrates, nitrogen acids, N-nitroso compounds, and organic compounds such as the peroxyacyl
8 nitrates [RC(O)OONO2], where R represents any one of a large variety of possible organic
9 groups) (Chapter 6).
10 The peroxyacyl nitrates, of which peroxyacetyl nitrate [CH3C(O)OONO2], or PAN is of
11 most concern in terms of atmospheric concentrations, have been thoroughly reviewed in the
12 recent EPA document, Air Quality Criteria for Ozone and Other Photochemical Oxidants
13 (U.S. Environmental Protection Agency, 1986), and will be given only the briefest discussion
14 in this chapter and elsewhere in this document.
15 The discovery of N-nitroso compounds in air, water, and food has led to concern about
16 possible human exposure to this family of compounds, some of which have been shown to be
17 carcinogenic in animals. Health concerns also have been expressed about nitrates, which
18 occur as a component of paniculate matter in the respirable size range, suspended in ambient
19 air (Chapter 13). Some of this particulate nitrate is produced in atmospheric reactions
20 (Chapter 5). Nitrates may also occur in significant concentrations in drinking water supplies
21 but this occurrence is not believed to be the result of atmospheric production.
22 Photochemical models predict that up to one-half of the original nitrogen oxides emitted
23 may be converted on a daily basis to nitrates and nitric acid (HNO3). This atmospheric
24 production of nitric acid is an important component of acid rain.
25 Table 3-1 summarizes theoretical estimates of the concentrations of the various nitrogen
26 oxides and acids that would be present in an equilibrium state assuming initially only
27 molecules of nitrogen and oxygen at 1 atm pressure, 25 °C and 50% relative humidity. The
28 low concentrations of many of the oxides and acids preclude direct measurement of most of
29 them in the ambient air. Consequently, most studies leading to predictions of concentrations
30 rely on theoretical estimates derived from small-scale laboratory studies.
31
August 1991 3-2 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3-1. THEORETICAL CONCENTRATIONS OF NITROGEN OXIDES AND
NITROGEN ACIDS WHICH WOULD BE PRESENT AT EQUILIBRIUM WITH
MOLECULAR NITROGEN, MOLECULAR OXYGEN, AND WATER IN AIR
AT 25 °C, 1 ATM, 50 PERCENT RELATIVE HUMIDITY
Concentrations in Hypothetical Atmosphere, ppm
Compound
03
N2
H2O
NO
NO2
NO3
N203
N204
NA
HONO (cis)
MONO (trans)
HONO2
At Equilibrium*
2.06 X Ws
7.69 X 10s
1.56 X IO4
2.69 X ID'10
1.91 x IO-4
3.88 X lO'16
2.96 X ID'20
2.48 X IO-13
3.16 X 104T
7.02 X 10'9
1.60 X ID"8
1.33 X 10'3
In Typical Sunlight-Irradiated,
Smoggy Atmosphere8'1'
2.06
7.69
1.56
io-1
ID'1
lO'8 -
lO"8-
io-7 -
io-3 -
io-3
io-3
10-2-
X 10s
X IO5
x IO4
10'9
10-*
io-8
10'5
io-3
"Assumes initially only molecules of nitrogen and oxygen at 1 atm pressure, 25 °C, and 50% R.H.
"Theoretical estimates made using computer simulations of the chemical reactions rates in a synthetic smog
mixture with hydrocarbons present.
Source: Demerjian et al. (1974).
1 In fact, the thermodynamic equilibrium state is not achieved in polluted, sunlight-
2 irradiated atmospheres. Rather, expected concentrations of pollutants are influenced by
3 emissions and subsequent reactions and tend to be much greater than those at equilibrium.
4 Table 3-1 lists one set of estimated concentrations of nitrogen oxides and acids expected
5 under more realistic conditions, derived from computer simulations of photochemical smog
5 reactions that might occur in more or less typical urban environments. White and Dietz
7 (1984) have postulated that more than one steady state is possible at certain rates of NOX
3 emissions. The calculated steady states of the free troposphere are shown on Figure 3-1 for a
? range of NOX concentrations.
August 1991 3-3 DRAFT-DO NOT QUOTE OR CITE
-------
Steady State
8 10
10°
NOX Concentration (ppb)
Figure 3-1. Calculated steady states of the free troposphere as a function of NOX
concentration.
Source: White and Dietz (1984).
1 3.2 NITROGEN OXIDES
2 Table 3-2 summarizes some important physical properties of nitrogen oxides under
3 standard temperature and pressure (STP) conditions of 25 °C and 1 atm, respectively. The
4 remainder of this section describes chemical and physical properties of individual nitrogen
5 oxide species.
August 1991
3-4
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 3-2. SOME PHYSICAL AND THERMODYNAMIC PROPERTIES OF THE NITROGEN OXIDES
c
fa
r*
1— «
vo
t— »
£
DRAFT-DO 1
O
H
O
3
o
Thermodynamic Functions
(Ideal Gas, 1 atm, 25 °Q
Molecular Melting
Weight, Point
Oxide g/mol '
-------
1 3.2.1 Nitric Oxide (NO)
2 Nitric oxide is an odorless gas. It is also colorless since its absorption bands are all at
3 wavelengths less than 230 nm, well below the visible wavelengths. Nitric oxide is only
4 slightly soluble in water (0.006 g/100 g of water at 24 °C and 1 atm pressure). It has an
5 uneven number of valence electrons, but, unlike NO2, it does not dimerize in the gas phase.
6 Nitric oxide is a principal by-product of combustion processes, arising from the
7 oxidation of molecular nitrogen in combustion air and of organically bound nitrogen present
8 in certain fuels such as coal and heavy oil. The oxidation of nitrogen in combustion air
9 occurs primarily through a set of reactions known as the extended Zeldovitch mechanism
10 (Zeldovitch, 1946):
11
12 N2 + O -> NO + N
13 N + O2 -> NO + O
14
15 with the additional equation (extended mechanism)
16
17 N + OH H. NO + H
18
19 The high activation energy of the first reaction above (75 kcal/mol) coupled with its essential
20. function of breaking the strong N2 triple bond make this the rate limiting step of the
21 Zeldovitch mechanism. Due to the high activation energy, this mechanism for NO
22 production proceeds at a somewhat slower rate than the reactions of fuel constituents and is
23 extremely temperature sensitive (Bowman, 1973). Moreover, the production of atomic
24 oxygen required for the first step is also highly temperature sensitive. NO formed via this
25 mechanism is often referred to as "thermal NOX."
26 In addition to the strong temperature dependence of the rate of the first step of the
27 Zeldovich mechanism, the temperature also influences the amount of atomic oxygen (O)
28 available for the reaction. In the immediate vicinity of a flame, the high temperatures
29 coupled with the kinetics of the hydrocarbons in the fuel can drive the oxygen concentration
30 to several times its equilibrium level. The local ratio of fuel to air also has a first order
31 effect on the concentration of atomic oxygen (Bowman, 1973).
August 1991 3-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 The reaction kinetics of thermal NO formation is further complicated by the fact that
2 certain hydrocarbon radicals can be effective in splitting the N2 bond through reactions such
3 as (Fenimore, 1971):
4
5 CH + N2 •* CHN + N
5
7 The rate of oxidation of the fuel (and intermediate hydrocarbon radical fragments) is usually
3 sufficiently rapid that only negligible quantities of the fuel radicals are available to attack the
? molecular nitrogen. However, under fuel-rich conditions, this can become the dominant
) mode of breaking the N2 bond and, in turn, can be responsible for significant NO formation
L (Engleman et al., 1976). Such reactions appear to have a relatively low activation energy
I and can proceed at a rate comparable to oxidation of the fuel. Because of the early formation
} of NO by this mechanism, relative to that formed by the Zeldovitch mechanism, NO thus
I formed is often referred to as "prompt NO." The importance of this mechanism has not been
> quantified for practical systems.
j In fossil fuels such as coal and residual fuel oil, nitrogen compounds are bound within
1 the fuel matrix. Typically, Number 6 residual oil contains 0.2 to 0.8% by weight bound
1 nitrogen and coal typically contains 1 to 2%. If this 1% nitrogen were converted
> quantitatively to NOX, it would account for about 2,000 ppm NOX in the exhaust of a coal-
) fired unit. In practice, only a portion of these nitrogen compounds is converted to NOX, with
the remainder being converted to molecular nitrogen (N2). Tests designed to determine the
t percent of the NOX emissions due to oxidation of bound nitrogen (Pershing and Wendt, 1976)
» show that upward of 80% of the NOX from a coal-fired boiler originate from fuel bound
nitrogen. Details of the Mnetic mechanisms involved in fuel nitrogen oxidation are uncertain
• due in part to the variability of molecular composition among the many types (and sources) of
> coal and heavy oils and to the complex nature of the heterogeneous processes occurring.
' Experimental evidence does, however, lend some insight into the processes involved.
i A number of fuel-bound nitrogen compounds have been cited (Axworthy and Schuman, 1973;
1 Martin et al., 1971; Turner and Siegmund, 1972), but the degree of conversion to NOX does
i not seem to be significantly affected by the compound type. NOX conversions arising from
fuel sources seem also to be relatively insensitive to temperature in diffusion flames. The
August 1991 3-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 most important parameters in determining fuel-bound nitrogen conversion appear to be the
2 local conditions prevailing when the nitrogen is evolved from the fuel. Under fuel-rich
3 conditions this nitrogen tends to form N2, whereas under fuel-lean conditions significant
4 amounts of NCX, are formed.
A
5 Nitric oxide formation kinetics in typical furnaces are not fast enough to reach
6 equilibrium levels in the high temperature flame zone, while the NO destruction mechanisms
7 are far too slow to allow the NO, once formed, to reach equilibrium at typical stack
8 temperatures. This is to say that the NO formation process is Mnetically controlled.
9 Nitric oxide and nitrogen dioxide produced in relatively large concentrations at high
10 temperatures in combustion processes would revert to lower concentrations characteristic
11 approximately of the equilibrium values shown in Table 3-3 were it not for the fact that
12 combustion equipment rapidly converts a large fraction of the thermal energy available to
13 useful work. This results in a rapid cooling of the combustion gases and a "freezing-in" of
14 the produced NO and NO2 near concentrations characteristic of the high temperature phase of
15 the process. '
16
TABLE 3-3. THEORETICAL EQUILIBRIUM CONCENTRATIONS OF
NITRIC OXIDE AND NITROGEN DIOXIDE IN AIR (50%
RELATIVE HUMIDITY) AT VARIOUS TEMPERATURES
Temperature, K (°C)
298 (24.85)
500 (226.85)
1,000 (726.85)
1,500 (1,226.85)
2,000 (1,726.85)
Concentration, ftg/m3
NO
3.29 x 10"10
(2.63 x 10'10)
8.18 x 10"4
(6.54 x 10"4)
43
(34.4)
. 1,620
(1,296)
9,946.25
(7,957)
(ppm)
NO2
3.53 X 10'4
(1.88 x 10"4)
7.26 x 10'2
(3.86 x ID'2)
3,38
(1.80)
12.35
(6,57)
23.88
(12.70)
Source; National Research Council (1977).
August 1991
3-8
DRAFT-DO NOT QUOTE OR CITE
-------
1 A major implication of the fact that NOX emissions are defined by the kinetics of the
2 process rather than being an equilibrium phenomenon is that NOX emissions can be
3 effectively modified by changes in the details of the combustion process. For clean fuels
4 such as natural gas or Number 2 distillate oil with no bound nitrogen, the NO formation is
5 dominated by the Zeldovitch mechanism. Thus, combustion modifications which reduce peak
6 flame temperature, limit the gas residence time at peak temperatures and/or reduce the
7 amount of atomic oxygen available at high temperatures will reduce the NOX emissions.
8 Examples of such modifications are flue gas reeireulation, reduced load, reduced combustion
9 air preheat temperature, water injection and reduced excess air (Bowen and Hall, 1976a,
.0 1976b, 1976c; Bowen and Hall, 1977a, 1977b, 1977c, 1977d, 1977e).
1 In furnaces fired with coal or heavy oil, the major portion of the NOX emissions is from
2 fuel-bound nitrogen conversion. Thus, combustion modifications which reduce the
3 availability of oxygen when the nitrogen compounds are evolved will reduce the NOX
4 produced. Examples of such modifications are reduction of the amount of excess air during
5 firing, establishing fuel-rich conditions during the early stages of combustion (staged
6 combustion), or new burner designs that tailor the rate of mixing between the fuel and air
7 streams (Bowen and Hall, 1976a, 1976b, 1976c).
8
9 3.2.2 Nitrogen Dioxide (NO2)
:0 Nitrogen dioxide is a feddish-orange-brown gas with a characteristic pungent odor,
1 Although its boiling point is 21.1 °C, the low partial pressure of NO2 in the atmosphere
',2 prevents condensation. Nitrogen dioxide is corrosive and highly oxidizing. It has an uneven
3 number of valence electrons and forms the dimer N2O4 at higher concentrations and lower
'A temperatures, but the dimer is not important at ambient concentrations. In the atmosphere
:5 NO can be oxidized to NO2 by the thermal reactions:
:6 NO + O3 •* NO2 + O2
:7 HO2 + NO •* NO2 + HO
:8 R02 + NO •+ RO + N02
;9 However, these reactions are of minor importance in most urban ambient situations, since
'0 other chemical processes are faster. The above reactions are mainly responsible for the NO2
>1 present in combustion exhaust gases. About 5 to 10% by volume of the total emissions of
August 1991 3-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 NOX from combustion sources is in the form of NO2, although substantial variations from
2 one source to another have been observed. Under more dilute ambient conditions,
3 photochemical smog reactions involving hydrocarbons convert NO to NO2 (Chapter 5).
4 Nitrogen dioxide's principal involvement in photochemical smog stems from its
5 absorption of sunlight and subsequent decomposition (photolysis) to NO and atomic oxygen
6 (O). Nitrogen dioxide is an efficient absorber of light over a broad range of ultraviolet and
7 visible wavelengths. Only quanta with wavelengths less than about 430 nm, however, have
8 sufficient energy to cause photolysis. It should also be noted that photons having
9 wavelengths less than about 290 nm are largely absorbed in the upper atmosphere. The
10 effective range of wavelengths responsible for photolysis of NO2 at ground level is,
11 therefore, 290 nm to 430 nm. Because of its absorption properties, NO2 produces
12 discoloration and reduces visibility in the polluted lower troposphere.
13 NO2 + hi; -> O + NO
14 O + O2 •* O3
15
16 3.2.3 Nitrous Oxide (N2O)
17 Nitrous oxide is a colorless gas with a slight odor at high concentrations. Nitrous oxide
18 in the atmosphere arises as one product of the reduction of nitrate by a ubiquitous group of
19 bacteria that use nitrate as their terminal electron acceptor in the absence of oxygen
20 (denitrification) (Brezonik, 1972; Delwiche, 1970; Focht and Verstraete, 1977; Keeney,
21 1973).
22 Although N2O does not play a significant role in atmospheric reactions in the lower
23 troposphere, it participates in a mechanism for ozone decomposition in the stratosphere.
24 Nitrous oxide transported to the stratosphere undergoes photolysis by absorbing
25 ultraviolet (UV) radiation at wavelengths below 300 nm to produce N2 and singlet oxygen
26 (Johnston and Selwyn, 1975):
27
28 N2O + hv -»• N2 + O('D) where hv is a unit of radiant energy
29
30 Singlet oxygen, which also is produced by ozone photolysis, reacts with more nitrous
31 oxide to produce two sets of products:
August 1991 3-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 N2O + O('D) •* N2 + O2
2
3 and
4
5 ^ NO + NO
6
7
8 The NO produced enters a catalytic cycle, the net result of which is the regeneration of
9 NOX and the destruction of ozone: ,
.0
1 NO + O3 •* NO2 + O2
.2
3 O3 + hv •* O2 + O
.4
5 NO2 + O -•• NO + O2
6
.7
8 These reactions are of concern because of the possibility that increased N2O resulting
9 from denitrification of excess fertilizer may lead to a decrease of stratospheric ozone (Council
',0 for Agricultural Science and Technology, 1976; Crutzen, 1976) with consequent potential for
11 adverse human health effects.
'2 '
3 3.2.4 Nitrogen Trioxide (NO3)
!4 Nitrogen trioxide has been identified in laboratory systems containing NO2/O3, NO2/O,
15 and N2O5 as an important reactive transient (Johnston, 1966). It is likely to be present in
',6 photochemical smog. This compound can be formed as follows:
:7
:8 O3 + NO2 •* NO3 + O2
:9
10 O + NO2 (+ M) -» NO3 (+ M)
il
12 N2O5 (+ M) -> NO3 + NO2 (+ M)
J3
4.
'5 (where M represents any third molecule available to remove a fraction of the energy involved
•6 in the reaction.)
17
August 1991 3-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 In the presence of sunlight, NO3 is rapidly converted to either NO or NO2 by these
2 reactions (Wayne etal., 1991):
3 NO3 + hi/ -* NO + O2
4
5 NO3 +hv -* NO2 + O
6
7
8 Nitrogen trioxide is highly reactive towards both nitric oxide and nitrogen dioxide:
9 .
10 NO3 + NO -* 2NO2
11
12 NO3 + NO2 (+ M) -* N205 (+ M)
13
14
15 Its expected concentration in polluted air is very low (about 10"6 ^g/m3 or 10"9 ppm).
16 However, NO3 may play an important role in the atmospheric chemistry of nitrogen,
17 especially at night when it may serve as a reservoir for NOX (Wayne et al., 1991). •
18
19 3.2.5 Dinitrogen Trioxide (N2O3) (Also Known as Nitrogen Sesquioxide)
20 In the atmosphere, N2O3 is in equilibrium with NO and NO2 according to the following
21 equation:
22
23 NO + NO2 » N2O3
24
25 The equilibrium concentrations at typical urban levels of NO and NO2 range from about
26 10"4 jig/m3 (~ 10"7 ppm) to 10"6 jtg/m3 (~ 10'9 ppm) (Table 3-4). N2O3 is the anhydride of
27 nitrous acid and reacts with liquid water to form the acid:
28
29 N2O3 + H2O -* 2HONO
30
31 3.2.6 Dinitrogen Tetroxide (N2O4) (Also Known as Nitrogen Tetroxide)
32 Dinitrogen tetroxide is the dimer of NO2 formed by the association of NO2 molecules.
33 It also readily dissociates to establish the equilibrium:
34
35 2NO2 « N2O4
August 1991 3-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 Table 3-4 presents theoretical predictions of concentrations of N2O3 and N2O4 in equilibrium
2 with various NO and NO2 concentrations.
3
TABLE 3-4. THEORETICAL CONCENTRATIONS OF DINTTROGEN TRIOXIDE
AND DINITROGEN TETROXTOE IN EQUILIBRIUM WITH VARIOUS LEVELS
OF GASEOUS NITRIC OXIDE AND NITROGEN DIOXIDE IN AIR AT 25 °C
NO
0.05
0.10
0.50
1.00
NO9
£j
0.05
0.10
0.50
1.00
Concentration, ppm
N203
1.3 x 10-9
5.2 x 10-9
1.3 x 10-7
5.2 x 10-7
N204
1.7 x 10-8
6.8 x 10-8
1.7 x 10-6
6.8 x 10-6
Source: National Research Council (1977).
1 3.3 NITRATES, NITRITES, AND NITROGEN ACIDS
2 Nitric acid in the gaseous state is colorless and photochemically stable. The major
3 pathway for atmospheric formation of nitric acid is given by:
4
5 »OH + NO2 -*• HNO3 where »OH is a hydroxyl free radical
6
7 It is a volatile acid, so that at ambient concentrations in the atmosphere, the vapor would not
8 be expected to coalesce into aerosol and be retained unless the aerosol contains reactants such
9 as sodium chloride (NaCl) or ammonia (NH3) to neutralize the acid, producing paniculate
0 nitrates (Chapter 6).
1 The nitrate ion (NO3~) is the most oxidized form of nitrogen. Since nitrate is
2 chemically unreactive in dilute aqueous solution, nearly all of the transformations involving
3 nitrate in natural waters result from biochemical pathways. The nitrate salts of all common
4 metals are quite soluble.
August 1991 3-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 Nitrates can be reduced to nitrites by microbial action. Many of the deleterious effects
2 of nitrate result from its conversion to nitrite. The nitrite ion represents an intermediate and
3 relatively unstable oxidation state (+3) for nitrogen. Both chemical and biological processes
4 can result in its further reduction to various products, or its oxidation to nitrate. Nitrite salts
5 are also quite soluble.
6 The nitrite ion is the Lewis base of the weak acid, nitrous acid (HNO^. When NO and
7 NO2 are present in the atmosphere, HNO2 will be formed as a result of the reaction:
8
9 NO + NO2 + H2O -*"2HNO2
10
11 However, in sunlight-irradiated atmospheres, the dominant pathway for nitrous acid formation
12 is:
13
14 •OH + NO •» HONO
15
16 During the daytime, atmospheric concentrations of HONO are limited by the reverse
17 reaction:
18
19 HONO + hv -» -OH + NO
20
21 Nitrous acid is a weak reducing agent and is oxidized to nitrate only by strong chemical
22 oxidants and by nitrifying bacteria. Nitrous acid reacts with amino acids (the Van Slyke
23 reactions) to yield N2. The reaction of nitrous acid with secondary amines to form
24 N-nitrosamines is discussed in Section 3.5.
25
26
27 3.4 AMMONIA (NH3)
28 Ammonia is a colorless gas with a pungent odor. It is extremely soluble in water,
29 forming the ammonium (NH4+) and hydroxy (OH") ions. In the atmosphere, ammonia has
30 been reported (Soderlund and Svensson, 1976) to be converted into oxides of nitrogen when
August 1991 3-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 it reacts with hydroxyl free radicals (»OH). Burns and Hardy (1975) report that ammonia is
2 oxidized into nitrates and nitrites in the atmosphere, and in geothermal wells. In the
3 stratosphere, ammonia can be dissociated by irradiation with sunlight at wavelengths below
4 230 nm (McConnell, 1973).
5
6
7 3.5 N-NITROSO COMPOUNDS
8 Organic nitrqso compounds contain a nitroso group (-N=O) attached to a nitrogen or
9 carbon atom. According to Magee (1971), N-nitroso compounds generally can be divided
10 into two groups—one group includes the dialkyl, alkylaryl, and diaryl nitrosamines, and the
11 other, alkyl and aryl nitrosamides.
12 The principal chemical reaction involved in the formation of N-nitrosamines is that of
13 the secondary amines with nitrous acid. Nitrosation is effected by agents having the structure
14 ONX, where X = O-alkoxyl, NO2", NO3", halogen, tetrafluoroborate, hydrogen sulfate or
15 OH2+. The equilibrium reaction of nitrosonium ion (ON+), nitrous acid and nitrite ion:
16
17 ON+ + OH' « HNO2 - H+ + NOf
18
19 is shifted to the right at pH >7. The simplest form of nitrosation of amines involves
20 electrophilic attack by the nitrosonium ion and subsequent deprotonation.
21 Mirvish (1970) studied the kinetics of dimethymitrosamine (DMN) nitrosation and
22 pointed out that the chief nitrosating agent at pH 1 is dinitrogen trioxide, the anhydride of
23 nitrous acid, which forms reversibly from two HNO2 molecules. The formation of
24 nitrosamines is dependent on the pK of the amine.
25 Nitroso compounds are characteristically photosensitive and the nitroso group is split by
26 UV radiation. Gaseous nitrosamines may be denitrosated by visible light. The electron
27 absorption spectra of several nitrosamines are given in the literature (Rao and Bhaskar, 1969);
28 the characteristic spectra show a low intensity absorption maximum around 360 nm and an
29 intense band around 235 nm. Nitrosamines show three relatively intense bands in the
August 1991 3-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 infrared region of 7.1-7.4 j«m, 7.6-8.6 /*m, and 9.15-9.55 pm. Nuclear magnetic resonance,
2 infrared, ultraviolet, and mass spectrometry spectra have been reviewed by Magee et al.
3 (1976).
4 Atmospheric reactions involving nitrosamines are discussed in Chapter 5.
5
6
7 3.6 SUMMARY
8 There are eight nitrogen oxides that may be present in the ambient air: nitric oxide
9 (NO), nitrogen dioxide (NO^, nitrous oxide (N2O), unsymmetrical nitrogen trioxide
10 (OONO), symmetrical nitrogen trioxide (O-N(O)-O), dinitrogen trioxide (N2O3), dinitrogen
11 tetroxide (N2O,,|), and dinitrogen pentoxide (N2O5).
12 Of these, NO and NO2 are generally considered the most important in the lower
13 troposphere because they may be present in significant concentrations in polluted
14 atmospheres. Their interconvertibility in photochemical smog reactions has frequently
15 resulted in their being grouped together under the designation NOX, although analytic
16 techniques can distinguish clearly between them. Of the two, NO2 is the more toxic and
17 irritating compound.
18 Nitrous oxide is ubiquitous even in the absence of anthropogenic sources since it is a
19 product of natural biologic processes in soil. It is not known, however, to be involved in any
20 photochemical smog reactions. Although N2O is not generally considered to be an air
21 pollutant, it is a principal reactant in upper atmospheric reactions involving the ozone layer.
22 While OONO, ON(O)O, N2O3, N2O4, and N2O5 may play a role in atmospheric
23 chemical reactions leading to the transformation, transport, and ultimate removal of nitrogen
24 compounds from ambient air, they are present only in very low concentrations, even in
25 polluted environments.
26 Ammonia (NH3) originates on a global scale during the decomposition of nitrogenous
27 matter in natural ecosystems but it may also be produced locally by human activities such as
28 the maintenance of dense animal populations. Some researchers have suggested conversion of
29 NH3 to NOX in the atmosphere.
30 Compounds derived from NOX including nitrites, nitrates, nitrogen acids, N-nitroso
31 compounds, and organic compounds such as the peroxyacyl nitrates [RC(O)OONO2], where
August 1991 3-16 DRAFT-DO NOT QUOTE OR CITE
-------
I R represents any one of a large variety of possible organic groups, may also be found in
I polluted air.
5 The peroxyacyl nitrates, of which peroxyacetyl nitrate [CH3C(O)OONO2] or PAN is of
!• most concern in terms of atmospheric concentrations, have been thoroughly reviewed in the
5 recent EPA document, Air Quality Criteria for Ozone and Other Photochemical Oxidants
5 (U.S. Environmental Protection Agency, 1986).
7 The discovery of N-nitroso compounds (some of which have been shown to be
? carcinogenic in animals) in air, water, and food has led to concern about possible human
) exposure to this family of compounds. Health concerns also have been expressed about nitric
) acid vapor and other nitrates, occurring as a component of particulate matter in the respirable
I size range, suspended in ambient air. Some of these nitrates are produced in atmospheric
I reactions. Nitrates may also occur in significant concentrations in public and private drinking
3 water, but this occurrence is not believed to be the result of atmospheric production.
1 Photochemical models predict that up to one-half of the original nitrogen oxides emitted
5 may be converted on a daily basis to nitrates and nitric acid. This atmospheric production of
5 nitric acid is an important component of acidic rain.
7
3 3.6.1 Nitrogen Oxides
? Nitric oxide (NO) is an odorless and colorless gas. It is a major by-product of the,
3 combustion process, arising both from the oxidation of molecular nitrogen in the combustion
1 air and of nitrogen compounds bound in the fuel molecule. The amount of NO formed from
2 the oxidation of molecular nitrogen is dependent upon such parameters as peak flame
3 temperature, quantity of combustion air, and gas residence time in the combustion chamber.
4 The amount of NO arising from oxidation of fuel-bound nitrogen does not seem to depend
5 significantly on either the type of nitrogen compound involved or the flame temperature, but
6 instead upon the specific air-to-fuel ratio at various stages in combustion.
7 Nitrogen dioxide (NO2) is produced in minor quantities in the combustion process
8 (5 to 10% of the total oxides of nitrogen). In terms of significant atmospheric loading in
9 populated areas, NO2 arises mainly from the conversion of NO to NO2 by a variety of
0 chemical processes in the atmosphere. Nitrogen dioxide is corrosive and highly oxidizing.
1 Its reddish-orange-brown color arises from its absorption of light over a broad range of
August 1991 . 3-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 visible wavelengths. Because of its strong absorption in this range (and also in the ultraviolet
2 spectrum), NO2 can cause visibility reduction and affect the spectral distribution of solar
3 radiation in the polluted, lower atmosphere.
4
5 3.6.2 Nitrates, Nitrites, and Nitrogen Acids
6 Other compounds derived from oxides of nitrogen (NOX) by means of atmospheric
7 chemical processes include nitrites, nitrates, nitrogen acids, organic compounds such as the
8 peroxyacyl nitrates, and, possibly, the N-nitroso compounds.
9 Nitric acid, a strong acid and powerful oxidizing agent, is colorless and photochemically
10 stable in the gaseous state. Its high volatility prevents condensation into droplets in the
11 atmosphere unless the droplets contain reactants such as sodium chloride or ammonia, which
12 neutralize the acid. Atmospheric reactions such as this may result in the formation of
13 paniculate nitrates suspended in ambient air.
14
15 3.6.3 N-Nitroso Compounds
16 The N-nitroso family comprises a wide variety of compounds all containing a nitroso
17 group (-N=O) attached to a nitrogen or carbon atom. Their formation in the atmosphere has
18 been postulated to proceed through chemical reaction of amines with NOX and
19 NOx-derivatives in gas phase reactions and/or through atmospheric reactions involving
20 aerosols. Nitroso compounds are characteristically photosensitive and the nitroso group is
21 split by the ultraviolet radiation in sunlight. Gaseous nitrosamines may also be denitrosated
22 by visible light.
23
24
25
August 1991 3-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Axworthy, A. E.; Schuman, M. (1973) Investigation of the mechanism and chemistry of fuel nitrogen conversion
3 to nitrogen oxides in combustion. In: Proceedings, coal combustion seminar; June; Research Triangle
4 Park, NC. Research Triangle Park, NC: U. S, Environmental Protection Agency, Office of Research and
5 Development; pp. 9^1; EPA report no. EPA-650/2-73-021, Available from: NTIS, Springfield, VA;
6 PB-224210.
7
8 Bowen, J. S.; Hall, R. E., eds. (1976a) Proceedings of the stationary source combustion symposium. Volume I.
9 Fundamental research; September 1975; Atlanta, GA. Washington, DC: U. S. Environmental Protection
0 Agency, Office of Research and Development; EPA report no, EPA-600/2-76-152a. Available from:
1 NTIS, Springfield, VA; PB-256320.
2 ......
3 Bowen, J. S.; Hall, R. E., eds. (1976b) Proceedings of the stationary source combustion symposium. Volume II.
4 Fuels and process research and development; September 1975; Atlanta, GA. Washington, DC: U. S.
5 Environmental Protection Agency, Office of Research and Development; EPA report no.
6 EPA-600/2-76-152b. Available from: NTIS, Springfield, VA; PB-256321.
7
8 Bowen, J. S.; Hall, R. E., eds. (1976c) Proceedings of the stationary source combustion symposium. Volume HI.
9 Field testing and surveys; September 1975; Atlanta, GA. Washington, DC: U. S. Environmental
0 Protection Agency, Office of Research and Development; EPA report no. EPA-600/2-76-152c. Available
1 from: NTIS, Springfield, VA; PB-257146.
2
3 Bowen, J. S.; Hall, R. E., eds. (1977a) Proceedings of the second stationary source combustion symposium. •
4 Volume I. Small industrial, commercial, and residential systems; August-September; New Orleans, LA.
5 Research Triangle Park, NC: U. S. Environmental Protection Agency, Office of Research and
6 Development; EPA report no. EPA-600/7-77-073a. Available from: NTIS, Springfield, VA; PB-270923.
7 - .•-..'..••
8 Bowen, J. S.; Hall, R. E., eds. (1977b) Proceedings of the second stationary source combustion symposium.
9 Volume II. Utility and large industrial boilers; August-September; New Orleans, LA. Research Triangle
D Park, NC: U. S. Environmental Protection Agency, Office of Research and Development; EPA report
1 no. EPA-600/7-77-073b. Available from: NTIS, Springfield, VA; PB-271756/9.
2
3 Bowen, J. S.; Hall, R. E., eds. (1977c) Proceedings of the second stationary source combustion symposium.
4 Volume in. Stationary engine, industrial process combustion systems, and advanced processes;
5 August-September; New Orleans, LA. Research Triangle Park, NC: U. S. Environmental Protection
5 Agency, Office of Research and Development; EPA report no. EPA-600/7-77-Q73e. Available from:
7 NTIS, Springfield, VA; PB-271757/7.
8
9 Bowen, J. S.; Hall, R. E., eds. (1977d) Proceedings of the second stationary source combustion symposium.
D Volume IV. Fundamental combustion research; August-September; New Orleans, LA. Research Triangle
1 Park, NC: U. S. Environmental Protection Agency, Office of Research and Development; EPA report
I no. EPA-600/7-77-073d. Available from: NTIS, Springfield, VA; PB-274029.
3
4 Bowen, J. S.; Hall, R. E., eds. (1977e) Proceedings of the second stationary source combustion symposium.
5 Volume V. Addendum; August-September; New Orleans, LA. Research Triangle Park, NC:
5 U.S. Environmental Protection Agency, Office of Research and Development; EPA report no.
7 EPA-600/7-77-073e, Available,from: NTIS, Springfield, VA; PB-274897,
3
3 Bowman, C. T. (1973) Kinetics of nitric oxide formation in combustion processes. In: Fourteenth symposium
3 (international) on combustion; August 1972; University Park, PA. Pittsburgh, PA: The Combustion
1 Institute; pp. 729-738.
I
August 1991 3-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 Brezonik, P. L. (1972) Nitrogen: sources and transformations in natural waters. In: Allen, H. E.; Kramer, J. R.,
2 eds. Nutrients in natural waters. New York, NY: John Wiley & Sons, Inc.; pp. 1-50.
3
4 Burns, R. C.; Hardy, R. W. F. (1975) Nitrogen fixation in bacteria and higher plants. New York, NY:
5 Springer-Verlag. (Molecular biology biochemistry and biophysics: 21).
6
7 Council for Agricultural Science and Technology. (1976) Effect of increased nitrogen fixation on stratospheric
8 ozone. Ames, IA: Iowa State University, Department of Agronomy; report no. 53; January 19.
9
10 Crutzen, P. J. (1976) Upper limits on atmospheric ozone reductions following increased application of fixed
11 nitrogen to the soil. Geophys. Res. Lett. 3: 169-172.
12
13 Delwiche, C. C. (1970) The nitrogen cycle. Sci. Am. 223: 137-147.
14
15 Demerjian, K. L.; Kerr, J. A.; Calvert, J. G. (1974) The mechanism of photochemical smog formation. In: Pitts,
16 J. N., Jr.; Metcalf, R. L.; Lloyd, A. C., eds. Advances in environmental science and technology,
17 v. 4. New York, NY: John Wiley & Sons; pp. 1-262.
18
19 Engleman, V. S.; Siminski, V. J.; Bartok, W. (1976) Mechanism and kinetics of the formation of NOX and other
20 combustion pollutants. Phase n. Modified combustion. Research Triangle Park, NC: U. S.
21 Environmental Protection Agency, Industrial Environmental Research Laboratory; EPA report no.
22 EPA-600/7-76-009b. Available from: NTIS, Springfield, VA; PB-258875.
23
24 Fenimore, C. P. (1971) Formation of nitric oxide in premixed hydrocarbon flames. In: Thirteenth symposium
25 (international) on combustion; August 1970; Salt Lake City, UT. Pittsburgh, PA: The Combustion
26 Institute; pp. 373-380.
27
28 Focht, D. D.; Verstraete, W. (1977) Biochemical ecology of nitrification and denitrification. Adv. Microb.
29 Ecol. 1: 135-214.
30
31 Johnston, H. S. (1966) Experimental chemical kinetics. In: Gas phase reaction rate theory. New York, NY:
32 The Ronald Press Company; pp. 14-34.
33
34 Johnston, H. S.; Selwyn, G. S. (1975) New cross sections for the absorption of near ultraviolet radiation by
35 nitrous oxide (N2O). Geophys. Res. Lett. 2: 549-551.
36
37 Keeney, D. R. (1973) The nitrogen cycle in sediment-water systems. J. Environ. Qual. 2: 15-29.
38
39 Magee, P. N. (1971) Toxicity of nitrosamines: their possible human health hazards. Food Cosmet. Toxicol.
40 9: 207-218.
41
42 Magee, P. N.; Montesano, R.; Preussmann, R. (1976) N-nitroso compounds and related carcinogens. In: Searle,
43 C. E., ed. Chemical carcinogens. Washington, DC: American Chemical Society; pp. 491-625. (ACS
44 monograph 173).
45
46 Martin, G. B.; Pershing, D. W.; Berkau, E. E. (1971) Effects of fuel additives on air pollutant emissions from
47 distillate-oil-fired furnaces. Research Triangle Park, NC: U. S. Environmental Protection Agency, Office
48 of Air Programs; report no. AP-87. Available from: NTIS, Springfield, VA; PB-213630.
49
50 Matheson Company, Inc. (1966) Matheson gas data book. 4th ed. East Rutherford, NJ: The Matheson Company,
51 Inc.
52
53 McConnell, J. C. (1973) Atmospheric ammonia. J. Geophys. Res. 78: 7812-7821.
54
August 1991 3-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 Mirvish, S, S. (1970) Kinetics of dimethylamine nitrosation in relation to nitrosamine carcinogenesis. J. Natl.
2 Cancer Inst. 44: 633-639.
3
4 National Research Council, (1977) Nitrogen oxides, Washington, DC: National Academy of Sciences.
*/
6 Pershing, D. W.; Wendt, J. O. L. (1976) Pulverized coal combustion: the influence of flame temperature and
7 coal composition on thermal and fuel NOX. In: Sixteenth symposium (international) on combustion;
8 August; Cambridge, MA. Pittsburgh, PA: The Combustion Institute; pp. 389-399.
9
10 Rao, C. N. R.; Bhaskar, K. R. (1969) Spectroscopy of the nitroso group. In: Feuer, H,, ed. The chemistry of
11 the nitro and nitroso groups: part 1. New York, NY: Interscience Publishers; pp. 137-163.
12
13 Schwartz, S. E.; White, W. H. (1981) Solubility equilibria of the nitrogen oxides and oxyacids in dilute aqueous
14 solution. Adv. Environ. Sci. Eng, 4: 1-45.
15
16 Soderlund, R.; Svensson, B. H. (1976) The global nitrogen cycle. Ecol. Bull. 22: 23-73.
17
18 Turner, D. W.; Siegmund, C. W. (1972) Staged combustion and flue gas recycle: potential for minimizing NOX
19 from fuel oil combustion. Presented at: The American Flame Research Committee Flame Days;
20 September; Chicago, IL.
21
22 U. S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants
23 [draft final]. Research Triangle Park, NC: Office of Health and Environmental Assessment,
24 Environmental Criteria and Assessment Office; EPA report no. EPA-600/8-84-020aF-ef. 5v. Available
25 from: NTIS, Springfield, VA; PB87-142949.
26
27 Wayne, R. P.; Barnes, L; Biggs, P.; Burrows, J. P.; Canosa-Mas, C. E.; Hjorth, J.; Le Bras, G.; Moortgat,
28 G. K.; Perner, D.; Poulet, G.; Restelli, G.; Sidebottom, H. (1991) The nitrate radical: physics,
29 chemistry, and the atmosphere. Atmos. Environ. Part A 25A: 1-203.
30
31 Weast, R. C.; Astle, M, J.; Beyer, W. H., eds. (1986) CRC handbook of chemistry and physics:
32 a ready-reference book of chemical and physical data. 67th ed, Boca Raton, FL: CRC Press, Inc.;
33 pp. B-lll-B-112.
34
35 White, W. H.; Dietz, D. (1984) Does the photochemistry of the troposphere admit more than one steady state?
36 Nature (London) 309: 242-244.
37
38 Zeldovich, J, (1946) The oxidation of nitrogen in combustion and explosions. Acta Physicochim, URSS
39 21: 577-628.
M)
August 1991 3-21 DRAFT-DO NOT QUOTE OR CITE
-------
-------
i 4. SOURCES OF NITROGEN OXIDES INFLUENCING
2 AMBIENT AND INDOOR AIR QUALITY
3
4
5 4.1 INTRODUCTION
6 Nitrogen oxides are important air pollutants, and this chapter serves to document the
7 emissions of nitrogenous compounds that directly effect human health or which may
8 participate in atmospheric chemical pathways leading to effects on human health and welfare.
9 Nitrogen containing compounds are also of particular interest because of their secondary
0 impacts. For example, the photochemical production of ozone and highly reactive radicals
1 like hydroxyl (OH) depends on the concentration of nitrogen oxides. In turn, chemical
2 species that affect OH are also relevant, as reactions with the OH radical provide the
3 dominant pathway for removal of a variety of other atmospheric species, including CO, CH4
4 and halocarbons. Nitric acid (HNO3), produced by the reaction of OH with nitrogen dioxide
5 (NO^, is one of the major components of acid precipitation. Agricultural usage of
5 nitrogenous compounds is important because of the possible perturbation of the stratospheric
7 ozone layer by nitrous oxide (N2O). As a result a knowledge of the emissions patterns for
8 nitrogen oxides is critically important for air quality planning. The chapter presents an
? update of the material presented in Chapter 5 of the previous Air Quality Criteria Document
3 for Oxides of Nitrogen (U.S. Environmental Protection Agency, 1982). Particular attention
1 in this revision is given to the global budget for NOX, the emissions distribution over the
I continental United States and to indoor sources of NOX.
3
t
5 4.2 AMBIENT SOURCES OF NITROGEN OXIDES
5 Nitrogen compounds are present in the atmosphere in the aerosol as well as in the gas
7 phase. While Gradel et al. (1986) identified more than 22 nitrogen containing species
3 observed in the atmosphere, two of the most important are nitric oxide (NO) and nitrogen
) dioxide (NO^, both of which are produced primarily in combustion processes. Because NO
) is converted to NO2 in the atmosphere, emissions of both species frequently are lumped
I together with the designation NOX (= NO + NO^. When NOX emissions are presented in
August 1991 4-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 mass units, the mass of NOX is calculated as if all the NO had been converted to NO2.
2 Another convention adopted in some of the following sections is to report the emissions on a
3 mass basis in terms of the nitrogen composition.
4 The principal routes to the production of NOX are combustion processes, nitrification
5 and denitrification in soils, and lightning discharges. The major removal mechanism is
6 oxidation to HNO3 followed by wet and dry deposition. Recent efforts by Boettger et al.
7 (1978), Ehhalt and Drummond (1982), Galbally (1985) and Warneck (1988) to quantify the
8 sources and sinks have led to an improved understanding of the global budget of NOX in
9 which the flux of NOX into the troposphere and the rate of nitrate deposition are
10 approximately balanced. A summary of the global budget for emissions and removal of
11 nitrogen oxides is presented in Section 4.2.3.
12
13 4.2.1 Combustion Generated NOX Emissions
14 The annual production of NOX from the combustion of fossil fuels is typically estimated
15 from emission factors for various combustion processes, combined with the worldwide
16 consumption data for coal, oil, and natural gas. Logan (1983) provided a convenient tabular
17 summary of emission factors, which has been updated by the National Acid Precipitation
18 Assessment Program (Placet et al., 1991). Due to variations in process operating conditions,
19 the emission factors must be considered to be uncertain by about 30%. Table 4-1 provides a
20 breakdown of global emissions for NOX according to fuel consumption rates. The estimates
21 of Logan (1983) are slightly higher than those of Ehhalt and Drummond (1982), with the
22 largest discrepancies showing up in the emissions from the transportation sector. The
23 differences arise because Logan (1983) used fuel usage as a way to estimate the emissions,
24 and Ehhalt and Drummond (1982) scaled the totals by using the world automobile population.
25
26 4.2.1.1 Generation of NOX from Biomass Burning
27 Table 4-1 also includes a breakdown of estimates for the release of nitrogen oxides from
28 the burning of biomass. In natural fires and the burning of wood, the temperatures are rarely
29 high enough to cause the oxidation of the nitrogen in the air. The emissions are thereby
30 more closely related to the fixed nitrogen content of the fuel. Logan (1983) reviewed a
31 number of experimental determinations of emissions factors that indicate yields are highest for
August 1991 4-2 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-1. GLOBAL NO. EMISSIONS FROM THE BURNING OF FOSSIL FUELS AND BIOMASS
C
o
§>
H
6
O
• ' » "* J* J ~* JL.% X^ A-^-^JLPi*-*-* AT^-r-^ Jfc-UL».«^v^»w».JLX^A T»w» JL. Jfc»V^i1
Annual Consumption
(10 metric tons unless, otherwise noted)
Source Type (E&D) (L) (Cetal.)
Fossil fuelsa'b
Hard coal 2,150 2,696
Lignite 810
Light fuel oil 300 1.39
Heavy fuel oil 470
Natural gas 1.04 1.2m3xl09
Industrial sources - - .
Automobiles . (4,1-5.4) X 1012 km 1.0 m3 x 109
Total '
Biomass burning0
Savanna (6-14) X 103 . 2,000 1,200
Forest clearings (2.7-6.7) x 103 . 4,100 2,700
Fuel wood - 850 1,100
Agricultural waste - 15 1,900
Total
rjk -M. ,m...m..M~* JUT **i jhmjk IJLJ. ^ ^i™« -*~r jk, M. •v^fc-'fc-'J
JL-f JL^.a,^_^A.> MJL, »»«*»«*
Global Source Strength
Emission Factors
(E&D) (L)
1.0-2.8 g/kg 2.7 g/kg
0.9-2.7 g/kg
1.5-3.0 g/kg 2.2 g/m3
1.5-3.1 g/kg
0.6-3.0 g/kg 1.9 g/m3
0.9-1. 2 g/km 8.0 g/m3
1.0 g/kg 1.7 g/kg
L0-1.6g/kg 2.0 g/kg
0.5 g/kg
• - 1.6 g/kg
aEstimates according to Ehhalt and Drummond (1982) (E &D) and Logan (1983) (L). Emission factors refer to grams nitrogen per unit
(10°
(E&D)
3.9 (1.9-5.8)
1.6(0.8-2.3)
0.7 (0.5-0.9)
1.1 (0.7-1.5)
1.9 (0.6-3.1)
-
4.3 (3.7-6.4)
13.5 (8.2-18.5)
3.1 (1.8-4.3)
2.1 (0.8-3.4)
2.0 (1-3)
4.0 (2-6)
11.2(5.6-16.4)
of fuel consumed.
metric tons N/yr)
(L) (C
6.4
3.1
2.3
1.2
8.0
19.9
3.4
8.2
0.4
0.02
12.0
et al.)
-
-
-
-
-
-
2.1
4.7
0.5
3.3
10.6
^, Petroleum refining and manufacture of nitric acid and cement; global emissions were obtained by scaling U.S. emissions for each industrial process.
o
H
O
1
w
o
cFor biomass-buming, Crutzen et al, (1979) (C et al.) have given annual consumption rate
production rates are included for comparison.
differing somewhat from those of the other authors. The data of C
et al. and the resulting
NOX
n
-------
1 grass and agricultural refuse fires (1.3 g N/kg fuel), less for prescribed forest fires
2 (0.6 g N/kg fuel), and still lower for the burning of fuel wood in stoves and fireplaces
3 (0.4 g N/kg fuel). The values roughly reflect the differences in nitrogen content of the
4 materials burned. Biomass burning is mainly associated with agricultural practices in the
5 tropics, which include plant, slash, and shift practices as well as natural or intentional
6 burning of savanna vegetation at the end of the dry season. Forest wildfires and the use of
7 wood as fuel make a lesser contribution.
8
9 4.2.1.2 Generation of NOX From Lightning
10 Thunderstorm activity has been considered to be a major source of NOX ever since
11 Leibig proposed it as a natural mechanism for the fixation of atmospheric nitrogen in 1827.
12 Electrical discharges in air generate nitrogen oxides by the thermal dissociation of N2 due to
13 ohmic heating inside the discharge channel and Shockwave heating of the surroundings.
14 Laboratory studies by Chameides et al. (1977) and Levine et al. (1981) indicate a NOX yield
15 of 6 x 1016 molecule per joule of spent energy. Great uncertainties exist, however, about the
16 total energy deposited by lightning in the atmosphere. Noxon (1976, 1978) first studied the
17 increase of NOX in the air during a thunderstorm. His results provide the basis for many of
18 the estimates shown in Table 4-2. Thorough reviews by Bordcki; and Chameides (1984);
19 Albritton (1984) and Kowalczyk and Bauer (1981, 1982) have provided a best estimate of
20 1 X 106 metric ton/year of NOX in North America and 13 X 106 metric tons/year globally
21 (Placet etal., 1991).
22
23 4.2.1.3 Generation of NOX From Soils
24 The biochemical release of NOX from soils is poorly understood, and the flux estimates
25 must be viewed with caution. Both rely on the observations by Galbally and Roy (1978),
26 who used the flux box method in conjunction with chemiluminescence detection of NOX.
27 They found average fluxes of 5.7 and 12.6 /ng N/m2«h on ungrazed and grazed pastures,
28 respectively, where NO was the main product. More recent measurements of Slemr and
29 Seiler (1984) indicate that the release of NOX from soils depends critically on the temperature
30 and moisture content of the soil, which in turn complicates the estimate of the global
31 emissions. Slemr and Seiler also found an average release rate of 20 /ug N/m2»h for
August 1991 4-4 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-2. GLOBAL AND NORTH AMERICA NATURAL EMISSIONS OF NOX
FROM LIGHTNING, SOILS AND OCEANS
Lightening
Soils
Oceans
Global NOX
(106 metric tons/year)
8.6 (2.6-26)
18
13 7-26
50 (as NO?)
30 (as NO)
36
0.35
North America
(106 metric tons/year)
1.7
1 0.3-2
2
Reference
(1)
(2)
(3) (4)
(5)
(6) (7)
(8)
(4)
(9) (10)
( 1) Borucki and Chameides (1984)
( 2) Attrition et al. (1984)
( 3) Kowalczuk and Bauer (1981, 1982)
( 4) Placet et al. (1991)
( 5) Lipschultz et al. (1981)
( 6) Levine et al. (1984)
( 7) Galbally and Roy (1978)
( 8) Slemr and Seiler (1984)
( 9) Zafiriou and McFarland (1981)
(10) Logan (1983)
uncovered natural soils, evenly divided between NO and NO2. Grass coverage reduced the
escape flux, whereas fertilization enhanced it. Ammonium fertilizers were about five times
more effective than nitrate fertilizers. This suggests that nitrification as a source of NOX is
more important than denitrification. According to Slemr and Seiler (1984), a global flux of
10 x 106 metric tons N/yr represents an upper limit to the release of NOX from soils.
Galbally (1985) presents some more detailed estimates for arid lands, and Table 4-2 provides
a compilation of current literature used to develop the global budgets. Global emissions of
nitrous oxide and ammonia are summarized in Table 4-3.
4.2.1.4 Generation of NOX From the Ocean
There are few if any measurements of NOX, N2O or NH3 fluxes over the ocean, and
current literature suggests that the sea is a negligible source of NO. Zafiriou and McFarland
(1981) observed a supersaturation of seawater with regard to NO in regions of relatively high
August 1991 4-5 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-3. ESTIMATES OF NITROUS OXIDE (N2O) AND AMMONIA (NH4+)
EMISSIONS TO THE TROPOSPHERE (106 METRIC TONS N/YR)
Source
Soils
Ocean
Biomass burning
Fossil fuels
Fertilizer
Domestic animals
(1) Dawson (1977)
(2) Boettger et al. (1978)
(3) Hahn(1981)
(4) Crutzen et al. (1979,
N2O
10 (2-20)
26 (12-38)
2
1.6
0.1
1983)
NH4
15
2-8
0.2
3
22
(5)
(6)
(7)
+ Reference
(1)(2)
(3)
(4)
(5) (2)
(2) (4) (6)
(2) (4) (6) (7)
Weiss and Craig (1976)
Stedman and Shelter (1983)
Soderlund and Swensson (1976)
1 concentrations of nitrite due to upwelling conditions. Here, the excess NO must arise from
2 the photochemical decomposition of nitrite by sunlight. Logan (1983) estimated a local
O ' '
3 source strength of 1.3 x 1,012 molecules/m s under these conditions. Linear extrapolation
4 results in a global flux of 0.35 x 106 metric tons N/yr.
5
6 4.2.2 Removal of NOX From the Atmosphere
7 Wet precipitation and dry deposition provide two of the major mechanisms for removal
8 of nitrogen oxides from the atmosphere. The addition of nitrate (and ammonium) by
9 rainwater to the plant soil ecosystem constitutes an important source of fixed nitrogen to the
10 terrestrial biosphere, and until 1930 practically all studies of nitrate in rainwater were
11 concerned with the input of fixed nitrogen into agricultural soils. Eriksson (1952) and
12 Boettger et al. (1978) have compiled many of the available data. Despite the wealth of
13 information it remains difficult to derive a global average for the deposition of nitrate,
14 because of an uneven global coverage of the data, unfavorably short measurement periods at
15 many locations, and inadequate collection and handling techniques for rainwater samples. In
16 addition, the concentration of nitrate in rainwater has increased in those parts of
August 1991 4-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 the world where the utilization of fossil fuels has led to a rise in the emissions of NOX,
2 primarily Western Europe and the United States.
3 Dry deposition is important as a sink for those gases that are readily absorbed by
t materials covering the earth surface. In the budget of NOX, the gases affected most by dry
5 deposition are NO2 and HNO3. The deposition velocity of NO is too small and the
5 concentration of PAN is not high enough for a significant contribution. In developing
7 Table 4-4, the dry deposition velocities for NO2 over bare soil grass, and agricultural crops
3 were assumed to fall in the range of 3-8 mm/s, but over water the velocities are significant
) smaller, so that the losses of NO2 by deposition onto the ocean surface can be ignored. The
3 absorption of nitric acid by soil, grass and water is rapid, and dry deposition correspondingly
I important, but the global flux is difficult to estimate because information on HNO3 mixing
I ratios is still sparse. Logan (1983) adopted NO mixing ratios of 50 pptv over the oceans and
3 100 pptv over the continents. The mixing ratios assumed for NO2 were 100 and 400 pptv,
!• respectively. Allowance was made for higher mixing ratios in industrialized areas affected by
5 pollution. Logan included the deposition of particulate nitrate over the oceans, using a
5 settling velocity of 3 mm/s. This process contributes 2 X 106 metric tons N/year to a total
1 dry deposition rate of 12-22 X 106 metric tons N/yr.
I On the continents one should also consider the interception of aerosol particulates by
) high growing vegetation. The interception of nitrate is expected to be particularly effective,
) as a large fraction of aerosol nitrate is associated with giant particles, which are most
L efficiently removed by impaction. Hofken et al. (1981) have studied the enrichment of
I nitrate in rainwater collected underneath forest canopies over that collected in open areas
$ outside the forests. The effect is caused by the washoff of dry-deposited material from the
I- foliage. Hofken et al. (1981) found that, in a beech forest, nitrate was enhanced by a factor
) of 1.4, whereas in a spruce forest the enhancement was by a factor of 4.1. Unfortunately,
5 they were unable to differentiate between the contributions of particulate and gaseous nitrate
1 to the total dry deposition.
\ If losses of NO2 and HNO3 by dry deposition are included in the total budget of NOX,
) one obtains a reasonable balance between the sources and sinks, as Table 4-4 shows. Ehhalt
) and Drummond (1982) felt that an appreciable part of dry deposition is already included in
August 1991 4-7 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-4. GLOBAL BUDGET OF NITROGEN OXIDES IN THE TROPOSPHERE
Type of Source or Sink
Global Flux (10* metric tons N/year)
Ehhalt and Dnunmond
Logan
Production
Fossil-fuel combustion
Biomass burning
Release from soils
Lightning discharges
NH3 oxidation
Ocean surface (biologic)
High-flying aircraft
Stratosphere
Total production
13.5 (8.2-18.5)
11.5 (5.6-16.4)
5.5 (1-10)
5.0 (2-8)
3.1 (1.2-4.9)
-
0.3 (0.2-0.4)
0.6 (0.3-0.9)
39 (19-59)
19.9 (14-28)
12.0 (4-24)
8.0 (4-16)
8.0 (2-20)
Uncertain (0-10)
1
-
0.5
48.4 (25-99)
Losses
Wet deposition of NO3-, land
Wet deposition of NO3-, oceans
Wet deposition, combined
Dry deposition of NOX
Total loss
17 (10-24)
8 (2-14)
24 (15-33)
-
24 (15-40)
19 (8-30)
8 (4-12)
27 (12-42)
16 (11-22)
43 (23-64)
Derived from estimates according to Ehhalt and Drummond (1982) and Logan (1983).
1
2
3
4
5
6
7
8
9
10
their wet deposition rate, because rain gauges frequently are left open continuously, so that
the collection of nitrate occurs all the time during both wet and dry periods. For NO2, they
estimated a dry deposition rate of 7 x 106 metric tons N/year. Because of the uncertainty,
they chose to include it in the error bounds and not in the mean value of total NOX derived
nitrogen deposition. Clearly, the total budget of NOX is far from being well defined.
Moreover, in view of the short residence times of the species involved in the NOX cycle, it is
questionable whether a global budget gives an adequate description of the tropospheric
behavior of NOX and its reaction products. Supplemental regional budgets could be more
appropriate.
August 1991
4-8
DRAFT-DO NOT QUOTE OR CITE
-------
4.2.3 Global Budgets for NOX
In view of these difficulties, it is somewhat surprising to find that the two independent
estimates given in Table 4-4 for the wet deposition of nitrate are essentially identical. Ehhalt
and Drummond (1982) relied on the detailed evaluation of data by Boettger et al. (1978).
Their analysis emphasized measurements from the period 1950-1977. They prepared a world
map for NO3 deposition rates, which was then integrated along 5° latitude belts. Logan
(1983) considered recent network data from North America and Europe in addition to the
new measurements of nitrate in precipitation at remote locations (Galloway et al., 1982).
Both estimates gave wet NO3 deposition rates in the range of 2-14 x 106 metric tons N/year
for the marine environment, and 8-30 x 106 metric tons N/year on the continents. It is
interesting to note than an earlier appraisal by Soderlund and Svensson (1976) had led to
rather similar values, namely, 5-16 and 13-30 X 106 metric tons N/year, respectively,
although it was primarily based on Eriksson's (1952) compilation of data from the period
1880-1930. This remarkable consistency contrasts with the much large scatter of individual
estimates.
4.2.4 Major Sources of NOX Emissions in the United States
Given the global budgets for the nitrogen oxides species the next step is to estimate the
emission fluxes for the United States. Since publication of the previous Air Quality Criteria
Document for Oxides of Nitrogen (U.S. Environmental Protection Agency, 1982) a major
study of emissions has been made as part of the National Acid Precipitation Assessment
Program (Placet et al., 1991). Table 4-5 presents the summary results from the detailed
inventory of man made NOX emissions in the United States for the 1985 inventory period.
Although national totals from all sources combined are close, estimates from individual
source categories vary somewhat. The left side of Table 4-5 presents a best estimate of 1985
nitrogen oxides emissions and Figure 4-1 presents some historical trends. Areas of high
emissions density are shown in Figure 4-2 for the United States and Canada. A comparison
between anthropogenic and natural sources for both the United Stated and Canada is shown in
Table 4-6, where it can be seen that anthropogenic sources are estimated to contribute close
to 90% of total emissions of nitrogen oxides.
August 1991 4-9 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-5. ESTIMATES OF ANTHROPOGENIC NOX EMISSIONS IN THE UNITED STATES (1985)
EXPRESSED IN MILLIONS OF METRIC TONS
VO
1— "•
f"
I-*
o
1985
Emission Source
Power plants
Industrial combustion
Transportation
Other sources
Totald
"Placet et al. (1991).
bKohout et al. (1991).
TJ.S. Environmental Protection Agency (1987).
^Values may not sum to totals due to independent r<
NAPAP Emissions
Inventory8
6.0
3.5
8.0
1.0
18.6
junding.
Monthly/State Current
Emissions Trendb
6.2
3.6
7.6
0.8
18.2
EPA Trends Report0
6.8
3.6
7.6
0.8
18.2
3
O
1
X3
a
I
o
I
-------
LONG-RANGE TRENDS: 1900-1980 AND 19BO-1989
FOR ANTHROPOGENIC EMISSIONS, Itf METRIC TONS/YR
30
1900 1910 1920 1930 1940 1950 1960 1970 1980
1980
Source Category j||ndU8Wa,
12 Fuel Combustion Q Solid Waste
H Transportation Hi Other
1989
Figure 4-1. Historical estimates of NOX emissions from man-made and natural sources
in the United States: (A) 1900 to 1980 and (B) 1980 to 1989.
Source: Gsehwandtaer et al. (1985, 1986); U.S. EPA (1991).
1
2
3
4
5
6
7
4.2.5 Conclusions
Estimates of nitrogen oxides emissions, derived from a detailed literature review, have
been presented for global budgets and for continental North America. These estimates show
that ambient anthropogenic NOX emissions in the United States are about 18-19 X 106 metric
tons per year, representing nearly 90% of the total emissions. Historically, these ambient
emissions have not changed significantly during the past decade. In Chapter 5, the
contribution of these emissions to atmospheric chemical processes is discussed.
August 1991
4-11
DRAFT-DO NOT QUOTE OR CITE
-------
Density of Annual Anthropogenic NOx Emissions
i
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
Mgure 4-2. Density of NOX emission from man-made sources in 1980.
Source: Wagner et al. (1986).
4.3 SOURCES OF NITROGEN OXIDES INFLUENCING INDOOR AIR
QUALITY
4.3.1 Introduction
This section summarizes emissions of nitrogen oxides from combustion sources
commonly found in the indoor residential environment which affect indoor air quality. There
are several reasons for considering these emissions. First, such information is needed to
understand the fundamental physical and chemical processes influencing emissions. This
understanding can be used to help develop strategies for reducing emissions. Second,
examining emissions from several types of sources and source categories can help identify the
August 1991
4-12
DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
5
7
3
1
2
3
TABLE 4-6. ESTIMATES OF NOX EMISSIONS FROM ANTHROPOGENIC AND
NATURAL SOURCES IN THE UNITED STATES AND CANADA (MILLIONS OF
METRIC TONS/YEAR OF NO2 EQUIVALENT EMISSIONS)
Emissions
Emission Source Best Estimate Rangea
Anthropogenic sources0
Highway vehicles 8.1 6.2-10
Power plants 6.0 5.0 -7.2
% Of Total b
.5 32
23
Industrial combustion 4.1 3.4-4.9 16
Other 4.3 3.4 - 5.4 17
Total Anthropogenic*1 22.6 18.0-28
Natural sources®
Lightning 1 0.3-2
Soil 2 0.3-5
Stratospheric injection < 0. 1
Total natural 3 0.6-7
Totald 26 19 - 35
"Based on uncertainties in Table 4-5 for man-made source categories.
bPercentages are based on "best estimate" values.
Includes Canada and the United States. Values for the United States are the
Table 4-5.
dValues may not sum to totals due to independent rounding.
"Values are for North America from Table 4-6.
fNot available.
relative importance of each source in affecting indoor air quality,
decisions by house occupants on combustion appliance purchases
appliances. Finally, studying emissions from indoor sources can
.0 88
4
8
~0
12
100
"best estimate" for 1985 shown in
, This information can guide
and methods of using such
provide source strength
4- input data needed for indoor air quality modeling. Predicting indoor airborne concentrations
5
5
7
3
is important for estimating the total exposure of individuals to nitrogen oxides.
This section begins with a brief discussion of the physics and chemistry of nitrogen
oxides formation in flames. Several major source categories are
include gas stoves used for cooking, unvented gas space heaters,
then considered. These
unvented kerosene heaters,
August 1991
4-13
DRAFT-DO NOT QUOTE OR CITE
-------
1 wood stoves primarily used for heating, and tobacco products. For each category, several
2 studies will be discussed, including the measurement methods used and the resulting data.
3 Following these presentations, the emissions of nitrogen oxides for all categories will be
4 compared.
5 Note that several types of vented appliances commonly found indoors usually emit
6 nitrogen oxides to the outdoors. Examples include furnaces, water heaters, and clothes
7 dryers using gas, as well as stoves and furnaces using wood, coal, and other fuels. Under
8 some circumstances, these sources may contribute to elevated NOX levels indoors; for
9 example, Hollowell et al. (1977) reported high NO and NO2 concentrations in a house where
10 a vented forced-air gas-fired heating system was used. Elevated concentrations may also be a
11 problem with malfunctioning vented appliances. Other data (e.g. Fortmann et al., 1984),
12 however, suggested that fugitive emissions of NOX from vented appliances are small.
13 Because the importance of unvented appliances to indoor NOX levels is well-documented, this
14 chapter focuses on emissions from such appliances.
15
16 4.3.2 Formation of Nitrogen Oxides in Combustion in Gas-fueled
17 Household Appliances
18 Many household appliances incorporate laminar flames in which the input fuel is
19 premixed with a known amount of air before being subjected to combustion. As an example,
20 Figure 4-3 shows a diagram of a single-port flame fueled with natural gas, taken from a
21 review by the Institute of Gas Technology (ZawacM et al., 1986). The figure is somewhat
22 simplified in that most appliances include burners with several flame ports; the combustion
23 product distributions are affected by interference from adjacent flames. Nevertheless, the
24 single flame model has been used successfully to approximate the behavior of more complex
25 systems.
26 The primary mixture consists of gas and 40-70% of the air required for complete
27 combustion. Excess air needed for complete combustion is provided through an annular
28 region surrounding the central port. The combustion takes place in a thin layer surrounding
29 the inner cone, termed the combustion zone. A number of intermediate products are formed
30 in this zone, including H2, CO, and radical species such as OH, O, and N. Significant
31 amounts of the major products CO2 and H2O are also formed here.
August 1991 .4-14 DRAFT-DO NOT QUOTE OR CITE
-------
§
III
CQ
DC
LU
Z
DC
CQ
LU
s
GO
<
LU
o
I
^
Q
LU
16
14
12
10
8
FUEL FLOW:
OUTER
CONE
SECONDARY
AIR
PRIMARY
AIR: 70%
EXCESS
AIR: 80%
RADIAL
PROFILE
POST-COMBUSTION
REACTION ZONE
COMBUSTION ZONE
(Bright blue)
'SECONDARY
AIR
10 t
BURNER BiASE, cm
Figure 4-3. Laminar blue-flame. Reproduced from Zawacki et al. (1986). Permission
to be obtained.
August 1991
4-15
DRAFT-DO NOT QUOTE OR CITE
-------
As the combustion products flow outward, other reactions occur in a luminous
secondary reaction zone termed the post-combustion reaction zone. The maximum
temperatures in the flame occur at the boundary of the outer cone defined by these reactions.
Beyond this boundary, the combustion products mix with secondary air resulting in a rapid
temperature decrease.
Many investigators have considered NO and NO2 production in this type of combustion
process, as reviewed by ZawacM et al. (1986). Examples of the distributions of temperature,
NO, and NO2 taken from this reference are given in Figures 4-4 and 4-5. The flame
conditions for these figures correspond to those of Figure 4-3. The NO concentration peaks
at the outer cone boundary where production is assisted by high temperatures. The low O2
content there prevents production of NO2. The latter species is produced at a slightly greater
radial distance from the flame center, where oxygen provided by secondary air reacts with the
high NO concentration. Overall, the production of NO is highest in regions of maximum
temperature. Thus, NO production may be particularly high at "hot spots" in the flame
(Hayhurst and Vince, 1980),
It is of interest to consider the chemical reactions involved in the formation of NO and
NO2 in this system. At high temperatures (> 1,600 K), NO is produced primarily by the
reactions first proposed by Zeldovich (1946):
O + N2 •* NO + N
N + O2 •* NO + O
Coutant et al. (1982) observe that NO also appears to form around the base of the flame
where the temperatures are too low for the Zeldovich mechanism to occur (about 1,350 K).
They hypothesize reactions of the form:
CN, NH + O2 -* NO + products
which may be important at high as well as low temperatures in the flame. These authors use
the work of Fenimore (1971) to further hypothesize a number of specific reactions:
CH4 + O -* CH3 + OH
CH3 + OH - CH2 + H2O
August 1991 4-16 DRAFT-DO NOT QUOTE OR CITE
-------
3500
- 3000
- 2500
2000
- 1500
1000
500
0
0 0.5 1.0 1.5 2.0 2.5
DISTANCE FROM FLAME AXIS, cm
figure 4-4. Radial profiles of NO and temperature at 7.19 cm above base. (Zawacki
et al.» 1986).
August 1991
4-17
DRAFT-DO NOT QUOTE OR CITE
-------
15r
D.
Q.
•%
1 10
SE
oc
LU
O
O B
O ^
0 0.5 1.0 1.5 2.0 2.5
DISTANCE FROM FLAME AXIS, cm
Figure 4-5. Radial profile of NO2 at 7.19 cm above burner base. (Zawacki et al.,
1986).
August 1991
4-18
DRAFT-DO NOT QUOTE OR CITE
-------
1 CH3 + N2 -» HCN + NH2
2 CH2 + N2 •* HCN + NH .
3
4 After NO is produced, oxidation to NO2 occurs via the reaction:
5
6 NO + HO2 -* NO2 + OH
7
8 The species HO2 is produced from the reaction of oxygen with methane:
9
0 CH4 + O2 -* H02 + CH3
1 '
2 or from reactions with CH2O or CHO as proposed by Peelers and Mahnen (1973):
3
4 CH2O, CHO + O2 -> HO2 + CHO, CO
5
6 Knowledge of these reactions has been used to design burners which may have smaller
7 NOX emissions. For example, reducing the flame temperature decreases the production of
8 NO; use of infrared tiles to lower the temperature has been shown to be an effective strategy
9 in decreasing NO emissions (Zawacki et al., 1986). The emissions of CO, however, tend to
0 increase under these conditions because of reduced oxidation of CO to CO2. Fuel type may
1 affect temperature and hence NOX emissions: propane combustion is much hotter than that of
2 natural gas, resulting in greater NO emissions. Other factors such as the level of primary
3 aeration, use of recirculation of combustion products, and control of air currents near the
4 burner have been shown to affect NOX emissions; the reader is referred to Zawacki et al.
5 (1986) for more/detail, ' ',. . -
6 It is important to note that our understanding of NOX emissions even from simple
7 combustion systems is far from complete. Reuther et al. (1988) summarize data from
8 12 investigations of premixed stoichiometric air/methane combustion, and conclude that wide
9 variations in reported emissions are probably due to differences .in measurement protocol.
0 Reuther et al. recommend that standardized measurement techniques be established for further
1 investigation of NOX emissions.
August 1991 4-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 Besides studies of simple well-controlled combustion, a number of investigators have
2 measured NOX emissions from common household appliances in the laboratory and in the
3 field. Data resulting from these measurements are now briefly reviewed.
4
5 4.3.3 Gas Stoves Used for Cooking
6 Several research programs have investigated nitrogen oxides emissions from stoves
7 fueled with natural gas and propane. Most of these studies have included a number of other
8 pollutants as well, such as carbon monoxide, aldehydes, and unburned hydrocarbons. This
9 summary will address only the nitrogen oxides emission data. Furthermore, only studies
10 using fuels of composition commonly found in the U.S. will be included. For additional
11 information on emissions from other types of gas, the reader is referred to Yamanaka et al.
12 (1979) and Caceres et al. (1983).
13
14 4.3.3.1 Study of Himmel and Dewerth (1974)
15 This represents one of the earliest studies of gas stove emissions, conducted at the
16 American Gas Association Laboratories. A total of 18 commercially available residential
17 stoves were examined. The authors estimated that the population of stoves tested was
18 representative of at least 90% of the total population of gas stoves in use within the U.S. at
19 that time.
20 The emissions were sampled using a quartz hood 28 cm in diameter and 25 cm high.
21 The combustion exhaust gases were drawn through the top of the hood, passed through a cold
22 trap to reduce moisture content, and directed into pollutant analyzers. These investigators
23 used the standard American National Standard Institute (ANSI) pot filled with water as a
24 load. The ANSI pot is chrome-plated brass, having dimensions 19 cm in diameter by
25 15 cm high. The quartz hood was positioned over the burner and pot such that the CO2
26 concentration in the emissions was approximately 2%. This value was selected to provide a
27 reasonably concentrated sample that did not have an excessive correction factor when
28 calculating the air-free pollutant concentration:
29
30 Air-Free Pollutant Concentration = Ultimate CO2 x Sample Pollutant Concentration (4-1)
gi Sample CO2
August 1991 4-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 where the air-free and sample pollutant concentrations are given on a dry weight basis,
2 Ultimate CO2 refers to the concentration expected based on stoichiometric considerations,
3 The combustion process was allowed to reach steady-state before recording data.
4 This procedure was applied to each of four top burners on 16 stoves, and to three top
5 burners on each of two additional stoves. Oven and broiler burners were also sampled, as
6 were the pilot lights for all types of burners. Tests were conducted using well adjusted blue
7 flames, and also poorly adjusted yellow flames. In limited additional testing, emissions from
8 the burners were sampled with water-filled cooking pots made of different materials: quartz,
9 pyrex, aluminum, copper, steel, iron, and Pyroceram.
10 An overall summary of the data for burner operation is presented in Table 4-7. The top
11 burners were operated at maximum heat input rate, which was close to the rated 158 kJ/min
12 (9,000 BTU/h) for some of the stoves and 211 kJ/min (12,000 BTU/h) for others. The oven
13 and broiler burners were operated at heat input rates ranging from 193 kJ/min
14 (11,000 BTU/h) to 422 kJ/min (24,000 BTU/h), The data show that emissions of NO are
15 generally in the range 16-24 /ig/kJ (0.037-0.056 lb/106 BTU), while emissions of NO2 are in
16 the range 5-14 /tg/kJ (0.012-0.033 lb/106 BTU). Exceptions include an infrared burner with
17 a very low NO emission factor and a pyrolytic self-cleaning oven with a relatively high NO
18 emission factor.
19 Table 4-7 also shows emission factors as a function of heat input rate for a top burner
IQ on one of the stoves. As the heat input rate increases, the data show that the emission factor
II of NO increases but the emission factor of NO2 decreases.
12 Table 4-8 shows emission factors for pilot lights associated with the top burners and
13 with the ovens and broilers. Three types of top burner pilot lights and two types of oven and
14 broiler pilot lights were tested. The data show that emission factors for NO and NO2 average
IS roughly 12.9 and 7.3 j«g/kJ (0.030 and 0.016 lb/106 BTU), respectively, for the top burner
16 pilots. For the oven and broiler pilots, the NO emission factor is very small; the original
17 data show high variability, with individual measurements ranging from near zero to 2.2 jtig/kJ
18 (0.005 lb/106 BTU). The NO2 emission factor is not as variable.
19 Himmel and Dewerth (1974) also conducted statistical tests with the emission factor data
50 to determine the influence of various parameters on the total nitrogen oxides (NOX)
Jl emissions. Results showed that the heat input rate and the number of burners in operation
August 1991 4-21 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-7. EMISSION FACTORS FOR NO AND NO2 FROM BURNERS
ON GAS STOVES, AFTER HBVCMEL AND DEWERTH (1974)
Top Burners
Top Burners
Ovens and Broilers
Ovens and Broilers
Top Burners with
Thermostat
Top Burners 142 kJ/min
Top Burners 190 kJ/min
Infrared Burners
Ovens and Broilers with
Catalytic Clean
Ovens and Broilers with
Catalytic Clean
Pyrolytic Self Clean
Oven
Top Burner of One
Stove, 150 kJ/min
Top Burner of One
Stove, 79 kJ/min
Top Burner of One
Stove, 39 kJ/min
Number
of
Burners
70
70
27
27
6
35
35
2
8
8
1
1
1
1
Flame
Type
Blue
Yellow
Blue
Yellow
Blue
Blue
Blue
Blue
Blue
Yellow
750 F*
Yellow
Yellow
Yellow
Emission Factor
for NO
/ig/kj
20.4 + 4.4
16.0 ± 5.0
22.2 ± 5.5
16.6 ± 7.8
20.7 ± 3.0
20.5 ± 4.0
20.6 ± 4.4
3.5
24.0 + 7.1
18.5 ± 10.8
38.1
22.5
9.42
1.45
Emission Factor
for NO2
/*g/kj
8.4
13.5
5.5
11.5
10.5
7.5
8.6
5.2
5.1
7.44
14.4
10.2
14.0
15.3
± 2.22
± 5.4
± 1.91
±,8.1 ,
± 3.2
+ 1.30
± 2.35
± 2.74
± 2.48
"Temperature setting as given in the original reference.
1 each had a significant effect on emissions at the 99% confidence level. Cooking utensil
2 material had a significant effect at the 95% but not at the 99% level. The supporting grate
3 material, grate height, and whether or not the oven was in operation had no significant effect
4 on the top burner NOX emissions. There was no significant difference in NO2 emitted from
5 stove to stove, although differences in total NOX were observed. The front burners had 13%
August 1991
4-22
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-8. EMISSION FACTORS FOE NO AND NO2 FROM PILOT LIGHTS ON
GAS STOVES, AFTER fflMMEL AND DEWERTH (1974)
Top Burners
Top Burners
Top Burners
Ovens/Broilers
Ovens/Broilers
yilot Type
I
n
m
i
n
Emission Factor
for NO
Mg/kj
13.7
13.2
11.9
0.265
1.01
Emission Factor
for NO2
^g/k)
7.83
6.83
7.35
14.4
10.7
Types of pilot lights on top burners:
LA free standing single flame which is surrounded by a flashtube assembly,
II. Same as above with a shield or baffle around the flame. The baffle typically comes up to about 1/2 the
flame height, protects the flame from drafts and tends to channel the combustion air supply. Combustion
air supply openings may be in the baffle, or inside or outside the baffle on the range top.
HI. Same as II with the addition of a shield above the flame. This shield can be a flat baffle or an arch type
baffle. The baffle appears to be for the purpose of keeping the range top cool. Type III seems to be the
most popular of the three pilot types.
Types of pilot lights on ovens and broilers:
I. A constant input pilot which typically is a horizontally oriented flame positioned directly below a flame
sensing element.
II. Similar to I, but operated in two fuel input modes: (a) a standby pilot mode and (b) ignition mode where a
secondary fuel supply ignites the burner. The standby pilot mode is not directed onto a flame sensing
element, while the ignition flame is. Generally this type of pilot is primary aerated.
1 higher emission factors for NO2 than the rear, attributed to differences in entrainment of
2 secondary air to feed the flame. The authors also discussed experimental burners that may be
3 effective in reducing NOX emissions from top burners. They concluded that controlling the
4- ingress of secondary air to the flame could reduce NOX emissions by about 50%. Reducing
5 flame temperature with a screen or other heat absorber in the flame could reduce NOX
6 emissions by as much as 68%, while operating a burner at 100% plus primary aeration could
7 reduce NOX emissions by as much as 66%.
8
August 1991 4-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 4.3.3.2 Study of Cote et al. (1974)
2 These investigators measured emission factors from one top burner on each of two gas
3 stoves. Each stove was placed in a 7.9 m3 (280 ft3) chamber having an air exchange rate of
4 64 air changes per hour. The stoves were operated at maximum heat input rate. By
5 measuring the pollutant concentrations in the chamber, the investigators were able to calculate
6 the source strength using a one-compartment mass balance model.
7 Results showed emission factors averaging 27 /xg/kJ (0.062 lb/106 BTU) for NO and
8 15 /ig/kJ (0.035 lb/106 BTU) for NO2. These values are similar to those reported by
9 Himmel and Dewerth (1974).
10
11 4.3.3.3 Study by the Massachusetts Institute of Technology (1976)
12 Tests conducted at MIT involved one top burner on a single stove. The experiments
13 were performed at maximum heat input using a sampling hood as done by Himmel and '
14 Dewerth (1974). Results showed an emission factor of 25 /-tg/kJ (0.057 lb/106 BTU) for NO
15 and 18 /-tg/kJ (0.042 lb/106 BTU) for NO2. Because the MIT tests were performed without a
16 cooking utensil to serve as a load, the results may not be representative of emissions found in
17 indoor residential environments.
18
19 4.3.3.4 Study of Traynor et al. (1982b)
20 These investigators operated two burners on a single stove using a 27 m3 (950 ft3)
21 chamber. Air exchange rates in the chamber varied from 0.24 to 0.52 air changes per hour.
22 Results for maximum heat input of 162 KJ/min per burner (9,200 BTU/h) yielded an average
23 emission factor of 10 /-tg/kJ (0.024 lb/106 BTU) for NO and 16 /xg/kJ (0.037 lb/106 BTU)
24 for NO2. Traynor et al. (1982b) also tested an oven burner, with emissions of 6.5 /-tg/kJ
25 (0.015 lb/106 BTU) for NO and 10 /-tg/kJ (0.024 lb/106 BTU) for NO2.
26
27 4.3.3.5 Studies of Cole et al. (1983) and Moschandreas et al. (1985)
28 This research involved a variety of emissions tests with top burners, ovens, broilers,
29 and pilot lights. The tests for the top burners were conducted using two methods. The first
30 method was similar to that of Himmel and Dewerth (1974) involving a quartz hood placed
31 directly over the burner. The second method involved placing a stove in a 33 m3 (1,150 ft3)
August 1991 4-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 all-aluminum chamber with controlled air exchange (1 to 5 air changes per hour) and
2 temperature characteristics; emissions were determined by measuring the ambient
3 concentrations in the chamber and using a one-compartment mass balance model to calculate
4 the source strength, as in the experiments of Cote et al. (1974).
5 Four top burners on each of three ranges were tested using the quartz hood. Both lean
6 gas with a heating content of 36,600 kJ/m3 (983 BTU/ft3) and rich gas with a heating content
7 of 38,100 kJ/m3 (1,022 BTU/ft3) were used. Some of the tests were conducted with a
8 properly adjusted blue flame, while others used a poorly adjusted yellow flame. The tests
9 were conducted at maximum heat inputs, which were in the range 140-175 kJ/min
10 (8,000-10,000 BTU/h).
11 The results of these tests are shown in Table 4-9. The data show average values for both
12 lean and rich gas. Note that both NO and NO2 emission factors agree well with those of the
13 Himmel and Dewerth study obtained using similar sampling procedures. All of these values
14 refer to steady-state conditions. A limited amount of data were also reported showing
15 variations in emission factor with time before reaching steady-state; results showed that
16 emissions of NO increase during the approach to steady-state, while NO2 emissions decrease
17 until a steady condition is achieved.
18 The chamber tests involved operating one burner on each of the same three stoves,
19 using blue and yellow flames. Results are also shown in Table 4-9. The data from these
20 tests agree reasonably well with the results of direct sampling using the quartz hood,
11 suggesting the viability of either method.
12 Moschandreas et al. (1985) also examined the influence of heat input rate on emissions.
13 Measurements were made with one stove operated in the chamber at four heat input rates.
14 The results, shown in Table 4-9, are in qualitative agreement with Himmel and Dewerth:
25 increasing the heat input rate increases NO but decreases NO2 emission factors. Additional
26 tests conducted by Moschandreas et al. (1985) showed that NO and NO2 emissions decreased
27 with increasing relative humidity in their chamber.
28 As with the burner emissions, NO and NO2 emissions from ovens were measured using
29 two different techniques. The first method involved use of a sampling probe positioned at the
30 oven flue outlet on the back of the range. The second method involved placing the stove in
August 1991 4-25 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-9. EMISSION FACTORS FOR NO, NO2, AND NOX FOR TOP
BURNERS ON GAS STOVES MEASURED WITH A SAMPLING HOOD AND
WITH A CHAMBER, AFTER COLE ET AL. (1983) AND
MOSCHANDREAS ET AL. (1985)
Sampling Hood
Sampling Hood
Chamber, Mass Balance
Chamber, Mass Balance
Top Burner of One
Stove, 145 kJ/min
Top Burner of One
Stove, 121 kl/min
Top Burner of One
Stove, 24 kJ/min
Top Burner of One
Stove, 13 kJ/rnin
Flame
Type
Blue
Yellow
Blue
Yellow
Blue
Blue
Blue
Blue
Emission
Factor
for NO
/ag/kJ
18 ± 1.1
16 ± 0.7
16 ±1.1
11 + 1.8
15 ± 0.4
13 ± 0.4
20 ± 0.9
0.86 ± 0.4
Emission
Factor
for NO2
/*g/kJ
9.9 ± 1.1
14 ± 1.1
10 ± 2.0
16 ± 3.5
7.3 ± 0.4
8.6 ± 0.4
8.6 ± 0.4
14 ± 2.6
Emission
Factor
for NOX
Mg/kJ
36 ± 2.0
39 ± 1.0
35 ± 2.6
33 ± 3.0
29 ± 0.9
29 ± 0.4
23 ± 1.3
16 ± 2.6
The first four rows refer to tests conducted at maximum heat input rate; values of average and standard
deviation in these tests are computed based on three values, each representing the average of 9-24 measurements
oa a single stove. Values given for the chamber mass balance tests are for sampling 12-29 minutes after turning
on the stove.
1 the chamber, as before. Average heat input rates ranged from 90 to 340 kJ/min (5,100 to
2 19,500 BTU/h). Table 4-10 shows the results of the oven emissions tests using the first
3 method. The original datm show that results for the second method of testing agree well with
4 those in Table 4-10 and therefore are not shown.
5 Finally, Moschandreas et al. (1985) determined emissions from three pilot lights in one
6 of the ranges using the chamber method. Results are shown in Table 4-11. The two top
7 pilots had a combined heat input rate of 4.4 kJ/min (250 BTU/h), compared with the single
8 bottom pilot heat input rate of 15 kJ/min (850 BTU/h). Despite the lower heat input rate of
August 1991
4-26
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-10. EMISSION FACTORS FOR NO, NO2, and NOX FOR OVENS,
AFTER COLE ET AL. (1983) AND MOSCHANDREAS ET AL. (1985)
Range 1, Bake
Range 1, Broil
Range 2, Broil
Range 2, Broil
Range 3, Bake
Range 3, Broil
Range 3, Broil
Range 3, Self-clean
. Type of
Gas
Rich
Lean
Lean
Rich
Lean
Lean
Rich
Rich
Emission
Factor :
for NO
14 + 3.2
29 ± 0.3
22 + 0.1
22 + 0.3
14 ± 0.1
23 + 0.3
23 + 0.2
27 + 1.0
Emission
Factor
for NO2
11 ± 1.5
7.1 + 0.8
9.7 + 0.3
9.9 ± 0.1
13 ± 0.7
12 + 0.1
11 ± 0.4
12 + 0.2
Emission
Factor
for NOX
41 + 5.4
53 + 0.9
44 ± 0.3
41 + 0.5
35 ± 1.5
47 ± 0.5
46 ± 0.6
54 + 1.8
Bake test: thermostat at 500 °F, burner cycles normally.
Broil test: thermostat at broil, burner on continuously.
Self-clean test: thermostat at clean, programmed burner sequential operation.
TABLE 4-11. EMISSION FACTORS FOR NO AND NO2 FROM PILOT LIGHTS
ON GAS STOVES, AFTER MOSCHANDREAS ET AL. (1985)
All three pilots
All three pilots
All three pilots
Two top pilots
Bottom pilot
Air Exchange
Rate
1.0
2.5
5.0
1.0
1.0
Emission Factor
for NO jwg/kJ
7.3 ± 0.9
7.3 + 1.3
9.0 + 2.6
17 + 1.7
3.9 + 1.3
Emission Factor
for NO2 ^g/kJ
8.6 + 1.3
9.0 + 0.9
12 ± 1.7
11 ± 3.0
8.2 ± 1.3
Air exchange rates refer to chamber operating conditions.
August 1991 . 4-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 the top pilots, both NO and NO2 emissions were substantially greater from the top pilots.
2 Overall, the emissions in Table 4-11 are similar to those of the pilot lights tested by Himmel
3 and Dewerth (1974).
4
5 4.3.3.6 Studies of Fortmann et al. (1984) and Borrazzo et al. (1987)
6 Tests conducted by Fortmann et al. (1984) involved two top burners on each of two
7 stoves, operated with a blue flame at maximum heat input rate. A sheet metal sampling hood
8 covering all four burners was used, following the procedures of Yamanaka et al. (1979).
9 A water-filled stainless steel pot 22 cm in diameter by 17 cm high was used as a load. These
10 tests gave average steady-state emission factors of 17 /ig/kJ (0.040 lb/106 BTU) for NO and
11 12 /itg/kJ (0.028 lb/106 BTU) for NO2. Later tests by the same research group involved one
12 of these stoves sampled with a Teflon-coated hood (Borrazzo et al., 1987). Experiments
13 were performed during initial start-up and at steady-state. Results for steady-state operation
14 showed average emission factors of 17 /ig/kJ (0.039 lb/106 BTU) for NO and 12 ^g/kJ
15 0.028 lb/106 BTU) for NO2. For initial start-up, emissions of NO increased while those of
16 NO2 decreased until steady-state was achieved, in qualitative agreement with the data of Cole
17 et al. (1983). Both Fortmann et al. (1984) and Borrazzo et al. (1987) also sampled at several
18 heat input rates. Results showed that although there is considerable variability, NO emissions
19 generally increase and NO2 emissions decrease with increasing heat input rate.
20
21 4.3.3.7 Study of Cole and Zawacki (1985)
22 These investigators prepared a literature survey of emissions from gas-fired appliances,
23 including gas stoves. Within the summary, they report emissions of nitrogen oxides from
24 two Gas Research Institute studies of advanced design burners. The American Gas
25 Association (AGA) design involves stainless steel inserts applied to conventional burners.
26 The Shukla and Hurley (1983) design incorporates a new infrared jet burner. Preliminary
27 results are presented in Table 4-12. Reductions are seen in the improved burners.
28
29 4.3.3.8 Study of Tikalsky et al. (1987)
30 This research involved measurements of emissions from ten gas stoves currently in use
31 in residences. The emissions were measured using the hood method of Himmel and DeWerth
August 1991 4-28 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-12. EMISSION FACTORS FOR NO AND NO2 FROM
RANGE-TOP BURNERS OF IMPROVED DESIGN,
AFTER COLE AND ZAWACKI (1985)
AGA Standard Burner
AGA 2-Ring Insert
AGA 3-Ring Insert
Shukla and Hurley Infrared
Shukla and Hurley Infrared
Shukla and Hurley Infrared
Heat Input
Rate kJ/min
158
158
158
118
72.1
45.7.
Emission Factor
for NO ftg/mkJ
20.2
9.04
.8.61
2.6
2.2
1.7 .
Emission Factor
for NO2 /u-g/mkJ
14.6
12.5
10.8
1.3
1.3
0.86
1 (1974). Data from each stove were obtained independently by two research groups in order
2 to obtain comparative data. The results are shown in Table 4-13.
3 Overall, the emission factors are similar to those reported in the literature from other
4 studies. Although the overall mean values reported by the two groups are in agreement,
5 results of the individual tests for each burner showed significant differences between the two
6 groups. The original data also showed greater variability in the results of these field tests
7 compared with results of laboratory tests reported in the literature. The emissions did not
8 appear to vary with gas flow rate for the conditions of this study.
9
0 4.3.3.9 Summary of Emissions From Gas Stoves
1 Table 4-14 lists average emission factors for range top burners and for oven and broiler
2 burners operated at maximum heat input rate. Data are shown for both well adjusted blue
3 flames and for poorly adjusted yellow flames. Each of the averages is based on the total
4 number of stoves tested for that category using data from the above studies. For top burners
5 with blue flames, a total of 27 values are represented; for yellow flames, there are, a total of
6 23 values (24 for NOX). Averages for the oven and broiler burners represent 20 blue flame
7 and 16 yellow flame values. Where data are reported for both oven and broiler burners for a
August 1991 4-29 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
TABLE 4-13. EMISSION FACTORS FOR NO2 FROM TEN
GAS STOVES
IN USE IN RESIDENCE, MEASURED INDEPENDENTLY BY RESEARCH
GROUPS (TTKALSKY ET AL., 1987). VALUES SHOWN ARE ARITHMETIC
AVERAGES AND
Gas Flow Rate
Right Front Burner High
Right Front Burner Medium
Right Front Burner Low
Left Rear Burner High
Left Rear Burner Medium
Left Rear Burner Low
Oven - Bake
Oven - Broil
STANDARD DEVIATIONS
Emission Factors
Measured by Group 1
A*g/kJ
15+3
15+5
14+3
16+3
16+3
18+4
13+9
19+14
Emission Factors
Measured by Group 2
A*g/kJ
14+6
15+8
26+35a
15+6
15+6
17+6
12+6
16+7
"Average influenced significantly by one extreme value.
TABLE 4-14. AVERAGE EMISSION FACTORS FOR NO, NO2, AND NOX FROM
BURNERS ON GAS STOVES BASED
Flame Type
Top Burners Blue
Top Burners Yellow
Ovens and Broilers Blue
Ovens and Broilers Yellow
ON DATA REPORTED IN
Factor for Factor for
NO £tg/kJ NO2 A*g/kJ
20.0 + 4.5 10.2 + 3.1
16.9 + 4.5 15.0 + 4.8
21.9 + 6.3 7.23 + 3.
THE LITERATURE
Factor for
NOxA*g/kJ
41.0 + 8.2
42.0 + 9.1
01 40.9 + 8.6
19.8 + 9.6 11.4 + 5.7 39.0 + 10.8
August 1991
4-30
DRAFT-DO NOT QUOTE OR CITE
-------
1 single stove, the values have been averaged to produce one emission factor for the oven and
2 broiler category for that stove. Values are generally very similar for emissions from these
3 two types of burners on the same stove. Overall, the results show that well adjusted blue
4 flames emit more NO but less NO2 than poorly adjusted yellow flames. Emission factors
5 from range top burners are comparable to those from oven and broiler burners. Note that the
6 emission factors from Tikalsky et al. (1987) have not been included in these averages because
7 flame type (blue or yellow) was not specified.
8 A recent literature survey, which includes most of the studies cited in this section, has
9 examined data from range-top burners for the purpose of identifying factors that significantly
10 influence emissions of NO, NO2, and total NOX (Davidson et al., 1987). The data were used
11 with several statistical tests. First, analysis of variance was used to investigate the
12 importance of three binary factors in explaining the observed variations in emissions. The
13 factors considered were (1) type of combustion (poorly adjusted or well adjusted), (2) burner
14 position (front or rear), and (3) method of sampling (direct with a hood or indirect with a
15 chamber). The results showed that roughly one-third of the variance in log EF (base ten
L6 logarithm of the emission factor) for NO2 can be explained by noting whether the combustion
17 is poorly adjusted or well adjusted. For NO and total NOX, the fraction of total variance in
18 log EF explained by this factor depends on the subset of the data chosen. Values of the
19 fraction range from 0.088 to 0.56, depending on whether the subset of data involved front or
10 rear burners, and whether the measurements were conducted with a hood or in a chamber.
II Burner position and method of sampling were both relatively unimportant in explaining the
12 observed variance for NO, NO2, or NOX.
23 The emission factor data were then used to estimate coefficients in various multivariate
24 regression models. The first regression model incorporated several factors: type of
25 combustion, burner position, method of sampling, the three two-way interactions among these
26 . factors, and (M-l) binary factors corresponding to the M stoves for which data were
17 available. Subsequent multivariate regression models were constructed by sequentially
28 eliminating a factor or factors from the previous model. Results of these tests showed that
29 stove differences were significant at the 95 % level in explaining the variance in NO2 and
30 NOX emission factor. Type of combustion was significant for NO and NO2. Burner position
31 had a smaller but still statistically significant effect in explaining variance in NO2 emissions.
August 1991 4-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 Similarly, the method of sampling had a small but statistically significant effect for NOX
2 emissions.
3 The influence of gas flow rate (heat input rate) on emission factors was investigated
4 separately. Statistical tests were not run for this factor due to lack of data and due to the
5 presence of detailed data for only one study. Results of plotting all of the data for NO and
6 NO2 are shown in Figures 4-6 and 4-7, respectively. The graphs illustrate wide variations in
7 emission factors, apparently due to varying stove characteristics, testing conditions, and other
8 nonuniformities among the datasets. However, there are general trends toward increasing
9 emissions for NO and decreasing emissions for NO2 as flow rate increases. Note that the
10 detailed data of Borrazzo et al. (1987) in these figures suggest that the emission/gas flow rate
11 relationships are complex, despite the general trends.
12
1
O
24.0
22,0
20.0
18.0
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
*
*
*
*
5
-
X
\
x
x*"**
xK
X*
'-
A
XXX
'S?
X*
A
O
X
XX ^'
x
A
A
X X *
v x"^
A
A
O
X
&XxX
xx X^
^x
+A
A
>^
A
A
O
f™
A
o Hlmmel and Dewerth (1974)
A Fortmann etal. (19841
+ Moschandreas et al. (1985)
x Borrazzo et al. (1987)
0.0
20.0
40,0 80.0 80.0 100.0 120X3 140.0 180.0
Figure 4-6. Emission factors for NO as a function of gas flow rate.
August 1991
4-32
DRAFT-DO NOT QUOTE OR CITE
-------
17.5
15.0
12.5
1 10.0
iu
7.5
5.0
X
.
+
-
x
x
X
X
A *5fc
XX
xx >
y
X
+
0 x
< XX
X
X
xxa
x
xx
X
o Himmel and Dewerth (1974}
A Fortmann etal. (1984)
+ Moschandreas et al. (1885)
x Sorrazzo etal. (1987)
xx
X
A
A
x^x x
o
\
^
A
X
3%*>>
X
x
x x
^x^^
+
£
A
X
x
f*
x
^
A A
O
&
0.0 20.0 40.0 60.0 80.0
lgas, kJ/min
100.0 120.0 140.0 160.0
Figure 4-7. Emission factors for NO2 as a function of gas flow rate.
4.3.4 Unvented Space Heaters Fueled with Natural Gas and Propane
A number of studies have considered emissions of nitrogen oxides from unvented space
heaters. As with the stove emissions studies, several different types of heaters and a variety
of measurement techniques were used. This section summarizes these studies. Much of the
information has been taken from the literature review of Cole and ZawacM (1985). It is
important to note that the emission factors for heaters must be interpreted differently from
those for stoves, due to differences in use profiles.
4.3.4.1 Study of Thrasher and Dewerth (1979)
These investigators studied emissions from five heaters by measuring NO and NO2
concentrations in the flue products during well adjusted blue flame operation. All of the
heaters had cast iron Bunsen burners with drilled ports. Three of the heaters had suspended
August 1991
4-33
DRAFT-DO NOT QUOTE OR CITE
-------
radiant tiles above the flame, while two did not. The heat input rates varied from 86.1 to
661 kJ/min (4,900 to 37,600 BTU/h). Results of these tests are given in Table 4-15A. The
data show that the emission factors are not monotonically related to heat input rate; the
influence of the radiant tiles is unclear based on these tests. These investigators also
examined emissions during poorly adjusted yellow flame operation. No definite conclusions
could be reached when results of these tests were compared with blue flame emissions.
In other tests, Thrasher and Dewerth (1979) examined the influence of placing a small
metal screen in the flame to decrease the temperature of combustion. Results showed that
NO emissions were reduced by 60%, while NO2 emissions were reduced by 23%. However,
emissions of CO increased substantially.
4.3.4.2 Study of Traynor et al. (1983a)
This research group examined eight heaters using a 27 m3 (953 ft3) environmental
chamber operated at 0.5 air changes per hour. Emissions were computed based on airborne
concentrations in the chamber using the mass balance method. Heaters with cast iron Bunsen
burners of both drilled port and slotted port designs were used. Heat input rates varied from
188 to 830 kJ/min. Emission factors for NO, NO2, and NOX are shown in Table 4-15B.
In addition to these chamber tests, Traynor et al. (1983a) operated four of the heaters in
an experimental research house with a volume of 240 m3. The tests in the house showed
greater total NOX emissions than were measured in the chamber with the same heaters.
However, the ratio NO2 emissions/NOx emissions measured in the house with each heater
were much smaller than the corresponding ratios measured in the chamber. These differences
were attributed to the much longer time periods of heater operation in the house as compared
with the chamber tests.
4.3.4.3 Study of Traynor et al. (1984)
The previous work by this group involved heaters without oxygen depletion sensors
(ODS) which were all fueled by natural gas. This subsequent set of tests involved ODS-
equipped heaters that were fueled by both natural gas and propane. Infrared as well as
August 1991 4-34 DRAFT-DO NOT QUOTE OR CITE
-------
-I-JLtSA*
V^JL' A JJLUtmiJJLLlUX% nilJL* J-»JUi f f .UU.V Jl JUL \M.J Uf A'VJJLX
WITH DRILLED PORTS USING NATURAL GAS
f Ji\^ J.J. f to JUULVfX ± IVI.VU
CfQ
e-
V0
1— «
Heater
ID
RH6
RH7
RH8
RH9
RH17
Distinguishing
Features
With/radiant tiles
No radiant tiles
No radiant tiles
With radiant tiles
With radiant tiles
Heat Input
Rate kJ/min
661
86.1
176
311
156
Emission Factor
for NO /ig/kJ
47
35
24
27
34
Emission Factor
for NO2 /xg/kJ
5.2
7.3
2.2
6.0
6.0
Emission Factor
for NOX /xg/kJ
77
60
39
47
56
n
TABLE 445B. DATA OF TRAYNOR ET AL. (1983a) FOR CONVECTIVE
WITH RADIANT TILES USING NATURAL GAS
•P--
Ul
o
15
H
1
0
o
2!
O
H
O
a
o
d
to
o
Heater
ID
12A
20A
30A
16B
40B
20C
30C
40C
Distinguishing
Features
Drilled Port
Drilled Port
Drilled Port
Drilled Port
Drilled Port
Slotted Port
Slotted Port
Slotted Port
Heat Input
Rate kJ/min
188
424
656
332
830
424
592
798
Emission Factor
for NO ;tg/kJ
9.5
22
22
14
16
16
19
19
Emission Factor
for NO2 ^g/kJ
20
13
12
18
20
11
9.5
9.9
Emission Factor
for NOX /ig/kJ
34
47
43
39
47
36
39
39
-------
TABLE 4-15C. DATA OF TRAYNOR ET AL. (1984), FOR CONNECTIVE HEATERS WITH RADIANT TELES,
AND FOR INFRARED HEATERS, FUELED WITH NATURAL GAS AND PROPANE
(re
c
ON
O
O
1
o
c!
S
w
o
n
Emission
Heater
ID
11
12
13
14
15
Cl
C2
C3 ,
C4
" N= Natural gas, P
TABLE 4-15D.
Distinguishing
Features*
Infrared, P
Infrared, N
Infrared, N
Infrared, N
Infrared, N
Convective, P, Drilled Port
Convective, N, Ribbon Port
Convective, P
Convective, P, Slotted Port
= Propane
DATA OF BELLICK ET AL.
Heat Input Factor
Rate kJ/min for NO /*g/kJ
245
263
280
317
352
335
486
419
626
(1984), MOSCHANDREAS
FOR HEATERS WITH BUNSEN, CATALYTIC, AND CERAMIC
0.1
0.1
0.39
0.1
1.02
28.7
17.8
25.5
28.2
ETAL.
Emission
Emission Factor Factor
for N02 /ig/kJ for NOX /ig/kJ
5.9
5.2
4.3
6.2
4.1
12.4
12.9
18.3
10
(1985), AND ZAWACKI ET
TELE BURNERS USING NATURAL
Emission
Heater
ID
1-Direct
1 -Chamber
2-Direct
3-Direct
3-Chamber
4-Direct
4-Chamber
Distinguishing
Features
Bunsen, no radiant tiles
Bunsen, no radiant tiles
Bunsen, with radiant tiles
Catalytic, no radiant tiles
Catalytic, no radiant tiles
Radiant (ceramic), no radiant tiles
Radiant (ceramic), no radiant tiles
Heat Input
Rate kJ/min for
186
186
255
207
207
260
260
Factor
NO/^g/kJ
18
15
22
0.09
0
0.39
0
5.9
5.2
4.9
6.2
5.6
56.5
40.1
57.5
53.2
AL. (1984)
GAS
Emission
Emission Factor Factor
for NO2 /^g/kJ for NOX /^g/kJ
9.9
15
9.0
0.1
1.3
3.8
4.7
35
37
42
0.3
1.3
4.3
4.7
-------
convective heaters were used. The measurements were conducted in the same manner as
employed previously. All tests were run under well-tuned conditions at full heat input.
Results are shown in Table 4-15C. NO2 emissions from the infrared heaters average
about one-third of those from convective heaters, while NOX emissions from the infrared
heaters are an order of magnitude smaller. NO emissions from the infrared heaters are very
small, below the limits of the measurement methods in some cases. No significant
differences are observed between propane and natural gas heaters.
In other tests, Traynor et al. (1984) considered emissions during short-term use before
the heaters were able to warm up completely. Compared with the longer-term emissions in
Table 4-15C, the tests showed slightly lower NO2 and NOX emissions for the infrared
heaters. Similarly, the short-term emissions of NOX for the convective heaters were slightly
lower than the longer-term emissions.
4.3.4.4 Studies of Billick et al. (1984), Moschandreas et al. (1985), and Zawacki et al.
(1984)
These investigators examined four heaters of different burner designs. Two of the
heaters had ribbon port Bunsen burners, one had a catalytic burner, and one had a ceramic
tile burner. The Bunsen burners were used in the well adjusted blue flame mode. The units
were operated at their respective maximum heat input rates, which varied from 186 to
260 kJ/min (10,600 to 14,800 BTU/h). The tests involved direct measurements of the
exhaust gases for all heaters; three of the heaters were also tested in a 33 m3 (1,150 ft3)
chamber using the mass balance method. The results, shown in Table 4-15D, indicate that
the NO2 emission factors measured in the chamber is much greater than these determined
using the direct testing for each of the three heaters. The opposite is true for the NO
emission factor.
These researchers also investigated ways of reducing emissions by using various inserts
in the flame of one of the Bunsen burner heaters. They found that ceramic rod inserts used
to reduce the temperature of the flame reduced NO emissions by 44%, although NO2
emissions were unchanged and CO emissions increased.
August 1991 4-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 4.3.4.5 Study of Zawacki et al. (1986)
2 In more recent work, this group considered ten natural gas and propane heaters. These
3 included three convective and four infrared heaters fueled with natural gas, two propane . .
4 convective heaters, and one propane infrared heater. Tests were conducted using three
5 measurement methods: a probe, a sampling hood, and a chamber. Of the ten heaters, eight
6 were the same ones used previously by Traynor et al. (1984) listed in Table 4-15C.
7 Results are shown in Table 4-16. Statistical analyses were performed on these data by
8 the investigators; the differences were generally insignificant between emissions determined
9 with the probe and with the hood for NO, NO2 and NOX. However, NO emissions obtained
10 with the chamber method were smaller than those obtained with the other two methods. This
11 was attributed to the vitiated atmosphere maintained within the chamber at the low air
12 exchange rates used. Overall, the results were in general agreement with the data of Traynor
13 et al. (1984) for the same eight heaters. An exception is seen for heater G, where the
14 disagreement is presumably due to damage during shipment.
15
16 4.3.4.6 GATC Task on Environmental Control
17 In their literature review, Cole and Zawacki (1985) cite data for two additional heaters
18 whose emission factors are given in Table 4-17. Heater 5 had a flat design radiant burner,
19 while heater 6 had a round Bunsen burned and included a circulating fan. Both burners were
20 steel and incorporated retention screens. The data were obtained from questionnaires
21 circulated as part of the GATC Task on Environmental Control (Cole and Zawacki, 1985).
22 At the time the data were collected, neither of these heaters were marketed in the U.S.,
23 although Cole and Zawacki reported that the manufacturer intended to market them in this
24 country as soon as possible.
25 • ,
26 4.3.4.7 Summary of Emissions from Unvented Gas Space Heaters
27 This section has summarized the findings of five separate investigations, with the data
28 given in Tables 4-15 through 4-17. The tables show that, on the average, convective space
29 heaters have emissions of NO roughly three times the emissions of NO2. The influence of
August 1991 4-38 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-16. DATA OF ZAWACKI ET AL. (1986) FOR CONVECTTVE AND
INFRARED HEATERS OF VARIOUS DESIGNS,
USING NATURAL GAS AND PROPANE
Heater ID Heater ID from Table 4-15C Distinguishing Features*
A
B
C 12
D 14
E C4
F Cl
G C2
H 11
I 15
J 13
Convective, ribbon burner, no radiant tiles, N
Convective, ribbon burner, with radiant tiles, N
Infrared, N
Infrared, N
Convective, slotted port, with radiant tiles, P
Convective, drilled port, with radiant tiles, P
Convective, ribbon burner, with radiant tiles, N
Infrared, P
Infrared, P
Infrared, P
N = natural gas, P = propane
Heater ID Test Method
A Probe
Hood
Chamber
B Probe
Hood
Chamber
C Probe
Hood
Chamber
D Probe
Hood
Chamber
E Probe
Hood
Chamber
Emission
Emission Emission Factor for
Rated Input .Factor for Factor for NOX (as NO^
Rate kJ/min NO2 /*g/kJ NO2 ME/^J MS^^J
176 17.7 9.8 35.3
16 10.9 35.4
9.7 8.4 23.3
264 21.7 8.4 41.9
20.7 9.1 40.8
11 10.4 27.2
264 0.6 1.8 2.7
0.005 2.8 2.8
0.005 3.5 3.5
304 1.7 2,6 5.1
0.5 4.1 4.8
0.3 5.3 5.7
703 39.8 7.6 68.6
39.3 7.6 67.9
28.3 8 51.2
August 1991
4-39
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-16 (cont'd). DATA OF ZAWACKI ET AL. (1986) FOR CONVECTIVE AND
INFRARED BEATERS OF VARIOUS DESIGNS,
USING NATURAL GAS AND PROPANE
Heater ID Test Method
F Probe
Hood
Chamber
G Probe
Hood
Chamber
H Probe
Hood
Chamber
I Probe
Hood
Chamber
J Probe
Hood
Chamber
Rated Input
Rate kJ/min
352
527
264
334
316
Emission
Factor for
NO2 ^g/kJ
44.4
43.9
31.8
13
14.5
5.3
0.5
0
0.01
1.5
1.2
0.3
0.9
0.5
0.4
Emission
Factor for
NO2 ^tg/kJ
8.4
8.1
10.5
23.3
20
19
2.6
5.2
5.8
1.7
2.3
4.8
1.6
3
3.3
Emission
Factor for
NOX (as NO^
jtg/kJ
76.4
75.4
59.3
43.3
43.1
27.1
3.3
5.2
5.9
4
4.2
5.3
3
3.9
4
TABLE 4-17. DATA FROM COLE AND ZAWACKI (1985) OBTAINED FROM
THE GATC TASK ON ENVIRONMENTAL CONTROL SURVEY.
BOTH HEATERS HAVE FLAME RETENTION SCREEN-TYPE
STEEL BURNERS AND USE NATURAL GAS
Emission
Emission Emission Factor for
Heat Input Factor for NO Factor for NO2 NOX (as NO2)
RatekJ/min /xg/kJ ptg/kJ /zg/kJ
Heater ID
Distinguishing
Features
5 Radiant burner, rad. No data
tiles
6 Bunsen burner, no
rad. tiles
211
1.9
0.52
5.6
0.6
8.6
1.4
August 1991
4-40
DRAFT-DO NOT QUOTE OR CITE
-------
1 radiant tiles on emissions is not clear. Heaters with catalytic burners, radiant ceramic tile
2 burners, and improved design steel burners (radiant and Bunsen) have much smaller NO and
3 NO2 emissions than heaters with conventional cast iron Bunsen burners.
4 These data also indicate that measurement of emissions with a probe, sampling hood, or
5 chamber can all yield equivalent results under certain conditions. ZawacM et al. (1986)
5 suggest the use of a sampling hood as the preferred method, because of its versatility and
7 ease in use.
5
) 4.3.5 Kerosene Heaters
) In this section, the results of three studies reporting nitrogen oxides emissions for
I portable kerosene heaters are examined.
S 4.3.5.1 Study of Yamanaka et al. (1979)
\ These researchers examined emissions from six radiant and five eonveetive kerosene
i heaters. Emissions were sampled using a hood positioned over the heater. Results were
> presented only for total NOX (as NO2). The emission factors for the radiant heaters averaged
' 13 ± 1.8 ng/yj, while the emission factors for the eonveetive heaters averaged
1 70 ± 6.8 /ig/kJ. Overall heat input rates for the 11 heaters were in the range
» 110-200 kJ/min.
4.3.5.2 Study of Leaderer (1982)
This investigator measured emissions of NO, NO2, and other pollutants from one
radiant and one eonveetive kerosene heater. The heaters were rated at 169 kJ/min
(9,600 BTU/hr) and 153 kJ/min (8,700 BTU/h), respectively. Measurements were performed
*2 ^3
in a 34 nr (1,200 ft ) chamber operated at 100 air changes per hour. Emission factors were
determined by mass balance. The data were obtained at three different heat input rates for
each heater; there were three sets of runs for each heat input rate.
Results are shown in Table 4-18. Emission factors for NO from the radiant heater are
very small, while those from the eonveetive heater are more than an order of magnitude
greater. For NO2, the emission factors are also greater from the eonveetive heater than from
the radiant heater, but only by factors of 1.5-3.
August 1991 4-41 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-18. AVERAGE EMISSION FACTORS FOR NO AND NO2 FROM
KEROSENE HEATERS, AFTER LEADERER (1982) AND TRAYNOR ET AL. (1983b)
Leaderer
Leaderer
Leaderer
Leaderer
Leaderer
Leaderer
Traynor
etal.
Traynor
etal.
Traynor
etal.
Traynor
etal.
Type of
Heater
Radiant, new
Radiant, new
Radiant, new
Convective,
new
Convective,
new
Convective,
new
Radiant, new
Radiant,
1 year old
Convective,
new
Convective,
5 years old
Heat Input
Rate kJ/min
144
113
84.4
158
97.9
37.3
137
111
131
94.8
Emission Factor
for NO jtg/kJ
0,45 ± 0.05
0.08 ± 0.05
0
17 + 0.3
12 ± 0.6
11 ± 0.9
1.3 ± 0.7
2.1
25 ± 0.7
11 ± 0.1
Emission Factor
for NO2 jtg/kJ
4.4 + 0.2
5.0 ± 0.2
5.9 :"± 0.3
7.0 + 0.4
15 + 0.3 '
17 ±1.0
4.6 ± 0.8
5.1
13 ± 0.8
32 ± 2.8
Emission Factor
for NOX ^g/kJ
5.1 ± 0.2
5.1 ± 0.2
5.9 +03
33 ± 0.6
33 + 1.0
34 ±1.7
6.6 ± 1.3
8.3
51 >± 1.3
49 ± 2.8
1 4.3.5.3 Study of Traynor et al. (1983b)
2 These investigators tested two radiant and two convective kerosene heaters. Emission
3 factors were determined by the mass balance method, using airborne concentrations measured
*2
4 in a 27 m chamber operated at 0.4 air changes per hour. Two types of tests were
5 conducted. In the first type, the heater was fired in the chamber and allowed to run for one
6 hour. In the second type, the heater was fired outside the chamber and allowed to warm up
7 for ten minutes. The heater was then brought into the chamber for a one hour run.
8 Results are shown in Table 4-18. The data are the averages of both types of tests for
9 the two convective heaters and for the new radiant heater. The one year old radiant heater
10 was studied using only the second type of test. The original data show little difference in
August 1991
4-42
DRAFT-DO NOT QUOTE OR CITE
-------
1 emission factor between the two types of tests. Overall, the results are similar to those of
2 Leaderer (1982), showing smaller emission factors for radiant as compared with convective
3 heaters.
I The data reported by Traynor et al. (1983b) in Table 4-18 refer to operation at the
5 maximum wick length. In additional tests, these investigators reduced the length of the wick
5 to half of full setting of the wick control knob (radiant heater) or until the flame was half the
7 length as it was previously (convective heater). Results showed slightly smaller NO
3 emissions for both heater types; for NO2, the emissions were comparable to those of the full
3 wick for the radiant heater but about double those of the full wick for the convective heater.
3
( 4.3.5.4 Study of Apte and Traynor (1986)
I This review paper includes data obtained by a Lawrence Berkeley Laboratory group on
5 emissions from two-stage kerosene heaters. The two-stage burners resemble those of radiant
!• kerosene heaters, except that there is a second chamber above the radiant element where
5 additional combustion air is introduced. In this region, the flame temperature is allowed to
5 rise and the combustion process is more complete. Emissions of NO from the two-stage
1 heater are slightly greater than those from radiant heaters due to the higher flame
I temperature, although emissions of NO2 as well as of CO and unburned hydrocarbons are
J lower.
)
_ 4.3.5.5 Summary of Emissions from Kerosene Heaters
I The data presented in this section show that emission factors of NO and NO2 for radiant
5 kerosene heaters are generally much smaller than those for convective kerosene heaters.
!• Emissions of NO from two-stage heaters are only slightly greater than those from radiant
> heaters, while emissions of NO2 are the lowest of the three heater types. Most of the NOX
) emissions from radiant heaters are in the form of NO2; for convective heaters that are two-
1 stage heaters, the emissions of NO and NO2 are of comparable magnitude. There are
> insufficient data to evaluate changes in emissions as kerosene heaters age.
August 1991 4-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 4.3.6 Wood Stoves
2 The use of wood stoves for residential space heating has become increasingly popular in
3 recent years. Because of this, a number of studies have examined pollutant emissions from
4 wood stoves. Some of these studies have measured emission factors based on concentrations
5 in the flue gases; such information would be useful for assessing the contribution of wood
6 stove emissions to ambient air quality. Very little information is available, however, on
7 fugitive emissions from wood stoves into the indoor living space.
8 In a detailed literature survey, Smith (1987) reports that emissions of pollutants from
9 wood stoves are highly variable, depending on the type of wood used, stove design, the way
10 the stove is used, and other factors. He reports emission factors for nitrogen oxides and
11 other pollutants for wood stoves used in developing countries. Many of these stoves are
12 unvented, resulting in excessive indoor concentrations as the combustion products are
13 exhausted into the room. This information is not applicable to the U.S., where virtually all
14 wood stoves are vented to the outdoors.
15 Traynor et al. (1984) have studied wood stoves used in a house (three airtight and one
16 non-airtight). For each experiment, airborne concentrations of several pollutants were
17 measured inside and outside the house during operation of one of the stoves. The results
18 showed that all indoor and outdoor concentrations of NO and NO2 were below 0.02 ppm.
19 Indoor airborne concentrations of some of the other pollutants were high during use of the
20 non-airtight stove, however. The airtight stoves had little influence on indoor concentrations
21 of any pollutants. In another study, Traynor et al. (1982a) found elevated airborne
22 concentrations of NO and NO2 in three occupied houses during operation of wood stoves and
23 a wood furnace. The concentrations were highly variable, however, and the authors caution
24 that additional tests would be needed to determine the influence of wood stoves on indoor
25 concentrations of nitrogen oxides and other pollutants.
26 Because of the paucity of data, it is difficult to reach quantitative conclusions regarding
27 the importance of wood stoves. However, the limited information available suggests that
28 wood stoves are not a major contributor to nitrogen oxides exposures indoors. This is
29 consistent with the small NO emission rates expected from the low temperature combustion
30 processes characteristic of wood stoves.
31
August 1991 4-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 4.3.7 Tobacco Products
2 A number of studies have compared concentrations of nitrogen oxides and other
3 pollutants in houses with smokers and houses without smokers. In general, these studies have
4 shown that concentrations are greater in the homes of smokers. There are few data available,
5 however, on NOX emission factors of tobacco products.
6 A few studies have reported emissions of nitrogen oxides from cigarettes while sampling
7 both sidestream and mainstream smoke together. Woods et al. (1983) report
8 0.079 mg/cigarette for NO2, while the University of Kentucky (undated) reports
9 1.56 mg/cigarette for NOX (as NO). Moschandreas et al. (1985) lists emissions of
10 2.78 mg/cigarette for NO and 0.73 mg/cigarette for NO2. The National Research Council
11 (1986) reports total nitrogen oxides emissions of 100-600 ^eg/cigarette for mainstream smoke,
12 with values four to ten times greater for sidestream smoke. According to this reference,
13 virtually all of the emitted NOX is in the form of NO; once emitted, the NO is gradually
14 oxidized to NO2. Thus environments containing cigarette smoke may have higher
15 concentrations of both NO and NO2 than environments without such smoke.
16 ' '."''•
17 4.3.8 Comparison of Emissions from Sources Influencing Indoor Air
8 Quality
.9 This section has considered emissions of nitrogen oxides from gas stoves, heaters using
10 natural gas or propane, kerosene heaters, wood stoves, and tobacco products. A significant
tl number of appliances in the first three categories have been tested; emission factors in these
t2 categories have been averaged and are shown in Table 4-19. Note that all data for a single
'3 appliance have been averaged before being used as an input to compute the grand averages
14 shown in the table. This procedure has been followed even when a single appliance has been
',5 tested by more than one group (e.g., Tables 4-15C and 4-18). Some of the reported data are
16 given as values below a specified limit of detection; these values have been taken as zero in
'.7 this analysis. Only dafa for the hood measurement technique have been used in Table 4-19.
:8 Table 4-19 data have been excluded from the averaging because at last report these heaters
:9 were not marketed in this country. For kerosene heaters, the NOX values include the data of
>0 Yamanaka et al. (1979), while the NO and NO2 values include only the data of Leaderer
1 (1982) and Traynor et al. (1983b).
August 1991 4-45 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 4-19. AVERAGE EMISSION FACTORS FOR NO, NO2, AND NOX FROM
VARIOUS SOURCES BASED ON DATA REPORTED IN THE LITERATURE
Number of Emission Factor Emission Factor Emission Factor
Appliances for NO ^tg/kJ for NO2
for NO
Range Top Burners
Ovens and Broilers
Unvented Gas Heaters
(Cast Iron Bunsen
Burners)
Kerosene Heaters
(Convectivfc)
Kerosene Heaters
(Radiant)
27
20
15
20.0 ± 4.5 10,2 ± 3.1 41.0 ± 8.2
21.9 ± 6.3 7.23 ± 3.01 40.9 ± 8.6
22.9 ± 9.6 10.8 ± 5.3 45.4 ± 11.4
15.2 ± 6.02 16.8 ± 9.3 55.1 ± 17.7*
0.79 + 0.90 5.00 + 0.58 9.69 ± 3.68
**
*TotaI number of convective kerosene heaters tested for NOX — 10.
**Total number of radiant kerosene heaters tested for NO, = 11.
1 These data show that emissions of NOX are about 65-75% NO and 25-35% NO2 for
2 range top burners and for ovens and broilers fueled with natural gas, and for convective
3 heaters fueled with natural gas and propane. In contrast, convective kerosene heaters have
4 emissions of NO and NO2 that are roughly comparable. Radiant heaters using natural gas,
5 propane, or kerosene all have emissions of NO that are negligible compared with those of
6 NO2.
7 For gas stoves, emission factors of NO and NO2 from range top burners operating at
8 maximum heat input rate are comparable to those from the oven and broiler burners. Well
9 adjusted blue flames emit slightly more NO and slightly less NO2 than poorly adjusted yellow
10 flames. The emission factor of NO increases as the heat input rate increases on top burners;
11 the emission factor of NO2 decreases. When first starting a range top burner with a cold
12 load (e.g., a water-filled pot), the emission factor of NO is initially small but steadily
13 increases as the load warms up. The emission factor of NO2, on the other hand, is initially
14 high but steadily decreases. Pilot lights associated with range top burners have emission
15 " factors comparable to or slightly smaller than emission factors for the burners. Pilot lights
August 1991
4-46
DRAFT-DO NOT QUOTE OR CITE
-------
1 associated with oven and broiler burners have NO emission factors much smaller than those
2 of the burners themselves, although NO2 emission factors are comparable.
3 Emission factors of NO and NO2 from convective natural gas and propane heaters with
4 Bunsen burners are similar to emission factors from natural gas stoves. Emission factors of
5 NO from convective kerosene heaters are slightly smaller than those from gas stoves and
6 convective gas heaters, while emission factors of NO2 are slightly greater. Emission factors
7 of NO from radiant heaters using natural gas, propane, or kerosene are very small, while
8 those for NO2 are about one-third of the respective convective heater emissions. There
9 appears to be little difference in emission factor from a warmed kerosene heater compared
0 with emissions from a cold start. The length of the wick can affect emissions: a smaller
1 wick yields smaller NO emissions but possibly greater NO2 emissions. Two-stage kerosene
2 heaters have smaller emissions of NO2 than those from either convective or radiant kerosene
3 heaters.
4 Emissions of nitrogen oxides to the indoor environment from wood stoves are not
5 expected to be significant given the low combustion temperatures involved. No data are
6 available to allow quantification of such emissions, however.
7 Only limited data are available on nitrogen oxides emissions from tobacco. Nearly all
8 of the emitted nitrogen oxides from cigarettes is in the form of NO, although oxidation to
9 NO2 occurs over time periods of several minutes following emission. Sidestream smoke
0 from a cigarette contains up to an order of magnitude more NO than mainstream smoke.
1 The emission factors given in Table 4-19 can be used with indoor air quality models to
2 predict indoor airborne concentrations of nitrogen oxides, provided input data for other
3 parameters included in the model are available. Examples of such parameters include air
4- exchange characteristics of the house, outdoor airborne concentrations of nitrogen oxides, and
5 appliance usage patterns. The last parameter is especially important: very little information
S is available on the frequency of use of stoves and other combustion sources. Information on
7 the way occupants influence air exchange, such as by opening windows and doors, is also
3 very limited. For many situations, our ability to predict indoor airborne concentrations is
? limited by our lack of understanding of occupant behavior, rather than by lack of data on
3 emission factors. Nevertheless, additional information on emissions from the variety of
August 1991 4-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 combustion sources under actual use conditions is needed to enable more accurate predictions
2 of indoor air quality exposure.
3
4
5 4.4 SUMMARY OF EMISSIONS OF NO, FROM AMBIENT AND
X
6 INDOOR SOURCES
7 Quantitative estimates of the total amount of NOX emitted to the ambient global
8 atmosphere are available. These estimates suggest that 40-50 x 106 metric tons of NOX are
9 emitted annually, with about 18-19 x 106 metric tons in the United States alone. These
10 emissions have a direct impact on visibility, human health, and natural ecosystem processes,
11 as well as an indirect impact on atmospheric processes that may contribute to acidic
12 deposition, global warming and stratospheric ozone depletion.
13 The impact of ambient emissions of NOX on human health compounds the effect of
14 indoor NOX emissions. These two factors determine air concentrations and exposure in the
15 human environment. The important indoor sources of NOX are gas stoves, unvented space
16 heaters, kerosene heaters, wood stoves, and tobacco products. Total emissions and the ratio
17 of NO/NO2 from gas stoves and space heaters differ according to fuel flow rate and flame
18 adjustment. Additional factors, such as the load (e.g. cold pot of water), heater type
19 (convective vs. radiant) and fuel type (natural gas, propane or kerosene) may also be
20 important. Only limited information is available for wood stoves and tobacco products.
21
August 1991 4-48 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Albritton, D. L.; Liu, S. C.; Kley, D. (1984) Global nitrate deposition from lightning. In: Proceedings of the
3 conference on the environmental impact of natrual emissions; March; Research Triangle Park, NC.
4 Pittsburgh, PA: Air Pollution Control Association.
5
6 Apte, M, G.; Traynor, G. W. (1986) Comparison of pollutant emission rates from unvented kerosene and gas
7 space heaters. In: Proceedings of the ASHRAE conference IAQ '86: managing indoor air for health and
8 energy conservation; April; Atlanta, GA. Atlanta, GA: American Society of Heating, Refrigerating and
9 Air-Conditioning Engineers, Inc.; pp. 405-416.
0
1 Billick, L; Johnson, D.; Moschandreas, D.; Relwani, S. (1984) An investigation of operational factors that
2 influence emission rates from gas appliances. In: Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor air:
3 proceedings of the 3rd international conference on indoor air quality and climate, v. 4, chemical
4 characterization and personal exposure; August; Stockholm, Sweden. Stockholm, Sweden: Swedish
5 Council for Building Research; pp. 181-188. Available from: NTIS, Springfield, VA; PB85-104214.
6
7 Boettger, A.; Ehhalt, D. H,; Gravenhorst, G. (1978) Atmosphaerische Kreislaeufe von Stickoxiden und
8 Ammoniak [Atmospheric cycles of nitrogen oxides and ammonia], Juelich, Federal Republic of Germany:
9 Kernforsehungsanlage Juelich, Institut filer Chemie, Institut 3: Atmosphaerische Chemie; report no.
0 JUEL-1558. Available from: NTIS, Springfield, VA; DE7992244.
1
2 Borrazzo, J. E,; Osbora, J. F.; Fortmann, R. C.; Keefer, R. L.; Davidson, C. I. (1987) Modeling and
3 • monitoring of CO, NO and NO2 in a modern townhouse. Atmos. Environ. 21: 299-311.
4 - •- '"
5 Borucki, W. J.; Chameides, W. L. (1984) Lightning: estimates of the rates of energy dissipation and nitrogen
5 fixation. Rev. Geophys. Space Phys. 22: 363-372.
7
3 Caceres, T.; Soto, H.; Lissi, E.; Cisternas, R. (1983) Indoor house pollution: appliance emissions and indoor
? ambient concentrations. Atmos. Environ. 17: 1009-1013.
D
1 Chameides, W. L.; Stedman, D. H.; Dickerson, R. R.; Rusch, D. W.; Cicerone, R. J. (1977) NOX production in
2 lightning. J. Atmos. Sci. 34: 143-149.
3
\ Cole, J. T.; Zawacki, T. S. (1985) Emissions from residential gas-fired appliances. Chicago, IL: Institute of Gas
5 Technology; final report for IGT project no. 30570, February.
5
7 Cole, J. T.; Zawacki, T. S.; Maoris, R. A.; Moschandreas, D. J. (1983) Constituent source emission rate
3 characterization of three- gas fired domestic ranges. Presented at: 76th annual meeting of the Air
} Pollution Control Association; June; Atlanta, GA. Pittsburgh, PA: Air Pollution Control Association;
) paper no. 83-64.3.
I
I Cote, W. A.; Wade, W. A., IH; Yocom, J. E. (1974) A study of indoor air quality. Washington, DC: U. S.
3 Environmental Protection Agency, Office of Research and Development; EPA report no.
J EPA-650/4-74-042. Available from: NTIS, Springfield, VA; PB-238556.
5
S Coutant, R. W.; Merryman, E. L.; Levy, A. (1982) Formation of NO2 in range-top burners. Environ. Int.
7 185-192.
?
) Davidson, D. L; Borrazzo, J. E.; Hendrickson, C. T. (1987) Pollutant emission factors for gas stoves: a
) literature survey. Research Triangle Park, NC: U. S. Environmental Protection Agency, Air and Energy
I Engineering Research Laboratory; EPA report no. EPA-600/9-87-005. Available from: NTIS,
> Springfield, VA; PB87-171328.
August 1991 4-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ehhalt, D. H.; Drummond, J. W. (1982) The tropospherie cycle of NO;,. In; Georgii, H. W,; Jaeschke, W., eds.
2 Chemistry of the unpolluted and polluted troposphere: proceedings of the NATO Advanced Study
3 Institute; September-October 1981; Corfu, Greece. Boston, MA: D. Reidel Publishing Company;
4 pp. 219-251. (NATO Advanced Study Institutes series C - mathematical and physical sciences: v. 96).
5
6 Eriksson, B. (1952) Composition of atmospheric precipitation, n. Sulfur, chloride, iodine compounds.
7 Bibliography. Tellus 4: 280-303.
8
9 Fenimore, C. P. (1971) Formation of nitric oxide in premixed hydrocarbon flames. In: Thirteenth symposium
10 (international) on combustion; August 1970; Salt Lake City, UT. Pittsburgh, PA: The Combustion
11 Institute; pp. 373-380.
12
13 Fortmann, R. C.; Borrazzo, J. E.; Davidson, C. I. (1984) Characterization of parameters influencing indoor
14 pollutant concentrations. In: Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of the
15 3rd international conference on indoor air quality and climate, v. 4, chemical characterization and
16 personal exposure; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building
17 Research; pp. 259-264. Available from: NTIS, Springfield, VA; PB85-104214.
18
19 Galbally (1985).
20
21 Galbally, I. E.; Roy, C. R. (1978) Loss of fixed nitrogen from soils by nitric oxide exhalation. Nature (London)
22 275: 734-735.
23
24 Galloway, J. N.; Likens, G. E.; Keene, W. C.;'Miller, J, M. (1982) The composition of precipitation in remote
25 areas of the world. J. Geophys. Res. C: Oceans Atmos. 87: 8771-8786.
26
27 Gschwandmer, G,; Gschwandtner, K. C.; Eldridge, K. (1985) Historic emissions of sulfur and nitrogen oxides in
28 the United States from 1900 to 1980. Volumes 1 and 2. Research Triangle Park, NC: U.S.
29 Environmental Protection Agency, Air and Energy Engineering Research Laboratory; EPA report nos.
30 EPA-600/7-85-009A.B. Available from: NTIS, Springfield, VA; PB85-191195 and PB85-191203.
31
32 Gschwandtner, G.; Gschwandtner, K.; Eldridge, K.; Mann, C.; Mobley, D. (1986) Historic emissions of sulfur
33 and nitrogen oxides in the United States from 1900-1980. J. Air Pollut. Control Assoc. 36: 139-149.
34
35 Grnedel et al. (1986).
36
37 Himmel, R. L.; DeWerth, D. W. (1974) Evaluation of the pollutant emissions from gas-fired ranges. American
38 Gas Association Laboratories; report no. 1392.
39
40 Hofken et al. (1981).
41
42 Hollowell, C. D.; Budnitz, R. J.; Traynor, G. W. (1977) Combustion-generated .indoor air pollution. In:
43 Proceedings of the fourth international clean air congress; May; Tokyo, Japan. Tokyo, Japan: The
44 Japanese Union of Air Pollution Prevention Associations; pp. 684-687.
45
46 Kohout et al. (1991) Month and state current emission trends for NOX, SOx, and VOC: methodology and results.
47 Argonne, IL: Argonne National Laboratory; in preparation.
48
49 Kowalczyk, M. L.; Bauer, E. (1981) Lightning as a source of NOX in the troposphere. IDA-P-1590; December.
50
51
August 1991 4-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 Kowalczyk, M. L.; Bauer, E. (1982) Lightning as a source of nitrogen oxides in the troposphere, Washington,
2 DC: U. S. Department of Transportation; report no. FAA-EE-82-4.
3
4 Leaderer, B. P. (1982) Air pollutant emissions from kerosene space heaters. Science (Washington, DC) 218:
5 1113-1115.
6
7 Levine, J. S.; Rogowski, R. S.; Gregory, G. L,; Howell, W. E.; Fishman, J. (1981) Simultaneous measurements
8 of NO,, NO, and O3 production in a laboratory discharge: atmospheric implications. Geophys. Res. Lett.
9 8: 357-360.
10
11 Logan, J. A. (1983) Nitrogen oxides in the troposphere: global and regional budgets. J. Geophys. Res.
12 C: Oceans Atmos. 88: 10785-10807.
13
14 Massachusetts Institute of Technology. (1976) Experimental evaluation of range-top burner modification to reduce
15 NO, formation. American Gas Association; no. M40677.
16
17 Moschandreas, D. J.; Relwani, S. M.; O'Neill, H. J.; Cole, J. T.; EIMns, R. H.; Macriss, R. A. (1985)
18 Characterization of emission rates from indoor combustion sources. Chicago, 1L: Gas Research Institute;
19 report no. GRI 85/0075. Available from: NTIS; Springfield, VA; PB86-103900.
20
21 National Research Council. (1986) Environmental tobacco smoke: measuring exposures and assessing health
22 effects. Washington, DC: National Academy Press.
23
24 Noxon, J. F. (1976) Atmospheric nitrogen fixation by lightning. Geophys. Res. Lett. 3: 463-465.
25 .
26 Noxon, J. F. (1978) Tropospheric NO2. J. Geophys. Res. C: Oceans Atmos. 83: 3051-3057.
27
28 Peelers, J.; Mahnen, G. (1973) Structure of ethylene-oxygen flames. Reaction mechanism and rate constants of
29 elementary reactions. In: Weinberg, F. J., ed. Combustion Institute European Symposium 1973: papers
30 for presentation at the symposium; September; University of Sheffield, United Kingdom. New York, NY:
31 Academic Press; pp. 53-58.
32
33 Placet, M.; Battye, R. E.; Fehsenfeld, F. C,; Bassett, G. W. (1991) Emissions involved in acidic deposition
34 processes: state-of-science/technology report 1. National Acid Precipitation Assessment Program.
35
36 Shukla, K. C.; Hurley, J. R. (1983) Development of an efficient, low NO, domestic range cooktop. Thermo
37 Electron Corporation research report; GRI contract no. 5081-241-0544.
38
39 Slemr, F.; Seller, W. (1984) Field measurements of NO and NO2 emissions from fertilized and unfertilized soils.
40 J. Atmos. Chem. 2: 1-24.
41
42 Smith, K. R. (1987) Air pollution and rural biomass fuels: a global review. Honolulu, HI: East-West Center: in
43 press.
44
45 Soderlund, R.; Svensson, B. H. (1976) The global nitrogen cycle. In: Svensson, B. H.; Soderlund, R., eds.
46 Nitrogen, phosphorus and sulphur - global cycles. Ecol. Bull. 22: 23-73. (SCOPE report 7).
47
48 Thrasher, W. H.; Dewerth, D. W. (1979) Evaluation of the pollutant emissions from gas-fired room heaters.
49 Cleveland, OH: American Gas Association Laboratories; research report no. 1515.
50
51
August 1991 4-51 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tikalsky, S,; Reisdorf, K.; Flickinger, J.; Totzke, D.; Haywood, J. (1987) Gas range/oven emisssions impact
2 analysis. Final report July 1985-December 1987. Chicago, IL: Gas Research Institute; report ao.
3 D/M-7683-062. Available from: NTIS, Springfield, VA; PB88-232756/XAB.
4
5 Traynor, G. W.; Allen, J. R.; Apte, M. G.; Dillworth, J. F.; Girman, J. R.; Hollowell, C. D.; Koonee,
6 J. F., Jr. (1982a) Indoor air pollution from portable kerosene-fired space heaters, wood-burning stoves,
7 and wood-burning furnaces. In: Proceedings of the Air Pollution Control Association specialty conference
8 on residential wood and coal combustion; March; Louisville, KY. Pittsburgh, PA: Air Pollution Control
9 Association; pp. 253-263.
10
11 Traynor, G, W.; Anthon, D. W.; Hollowell, C. D. (1982b) Technique for determining pollutant emissions from
12 a gas-fired range. Atmos. Environ. 16: 2979-2987.
13
14 Traynor, G. W.; Girman, J. R.; Allen, J. R.; Apte, M. G.; Carruthers, A. R.; Dillworth, J. F.; Martin, V. M.
15 (1983a) Indoor air pollution due to emissions from unvented gas-fired space heaters. Presented at: 76th
16 annual meeting of the Air Pollution Control Association; June; Atlanta, GA. Pittsburgh, PA: Air
17 Pollution Control Association; paper no. 83-9.6.
18
19 Traynor, G. W.; Allen, J. R.; Apte, M. G.; Girman, J. R.; Hollowell, C. D. (1983b) Pollutant emissions from
20 portable kerosene-fired space heaters. Environ. Sci. Technol. 17: 369-371.
21
22 Traynor, G. W.; Apte, M. G.; Carruthers, A. R.; Dillworth, J. F.; Grimsrud, D. T. (1984) Pollutant emission
23 rates from unvented infrared and convective gas-fired space heaters. Berkeley, CA: U. S. Department of
24 Energy, Lawrence Berkeley Laboratory; report no. LBL-18258. Available from: NTIS, Springfield, VA;
25 DE8501G647/XAB.
26
27 U. S. Environmental Protection Agency. (1982) Air quality criteria for oxides of nitrogen. Research Triangle
28 Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
29 Office; EPA report no. EPA-600/8-82-026. Available from: NTIS, Springfield, VA; PB83-131011.
30
31 U. S, Environmental Protection Agency. (1987) National air quality and emissions trends report, 1985. Research
32 Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/4-87-001.
33 Available from: NTIS, Springfield, VA; PB87-180352.
34
35 University of Kentucky. (n.d.) Physical and analytical data on IR4F reference cigarettes from Tobacco and Health
36 Research Institute. Lexington, KY.
37
38 Wameck, P. (1988) Chemistry of the natural atmosphere. New York, NY: Academic Press, Inc.
39
40 Woods, J. et al. (1983) Calculation methods to assess indoor air quality. Presented at the ASHRAE annual
41 meeting; June.
42
43 Yamanaka, S.; ffirose, H.; Takada, S. (1979) Nitrogen oxide emissions from domestic kerosene-fired and
44 gas-fired appliances. Atmos. Environ. 13: 407-412.
45
46 Zafiriou, O. C.; McFarland, M. (1981) Nitric oxide from nitrite photolysis in the central equatorial Pacific.
47 J. Geophys. Res. C: Oceans Atmos. 86: 3173-3182.
48
49 ZawacM, T. S.; Cole, J. T.; Huang, V. M. S.; Banasiuk, H.; Macriss, R. A. (1984) Efficiency and emissions
50 improvement of gas-fired space heaters. Task 2. Unvented space heater emission reduction [final report].
51 Chicago, IL: Gas Research Institute; report no. GRI-84/0021. Available from: NTIS, Springfield, VA;
52 PB84-237734.
53
August 1991 4-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 Zawacki, T. S.; Cole, J. T.; Jasionowski, W. J.; Macriss, R. A. (1986) Measurement of emission rates from
2 gas-fired space heaters. Chicago, IL: Institute of Gas Technology; final report for IGT project no.
3 30570-13; October.
4
5 Zeldovich, J. (1946) The oxidation of nitrogen in combustion and explosions. Acta Physicochim. URSS 21:
6 577-628.
7
August 1991 4-53 DRAFT-DO NOT QUOTE OR CITE
-------
-------
5. TRANSPORT AND TRANSFORMATION
OF NITROGEN OXIDES
5.1 BACKGROUND
i Even in the clean, unpolluted troposphere, oxides of nitrogen (NOX) play an important
role in natural atmospheric processes. They regulate the oxidizing power of the free
1 troposphere by controlling the build up and fate of free radical species such as HO2», »OH,
1 and RO2» (where R is an organic moiety). Consequently, the concentration of NOX is a
i critical factor in determining ozone (O3) production and hydrocarbon (HC) chemistry under
natural circumstances, and even more so under circumstances of added anthropogenic
; emissions.
Much of the tropospheric chemistry discussed in this chapter is based on the cyclic
reactions of nitrogen (Figure 5-1), oxygen (Figure 5-2), and hydrogen (Figure 5-3). These
: are schematically combined in Figure 5-4 to show the complexity of interactions among these
i three groups of gaseous compounds. Collectively, these four diagrams, along with
Figure 5-6 in Section 5.2.1, provide a visual reference for the series of individual reactions
! discussed throughout the chapter. White and Deitz (1984) demonstrated the complexity of
1 these interactions and the role of NOX in regulating the tropospheric steady state as illustrated
i in Figure 5-5. They showed that multiple steady states are possible for levels of NOX that are
comparable to those found in urban and industrial locations.
: In urban environments, NOX react with hydrocarbons in the presence of sunlight to
' produce oxidants such as O3 and peroxyacetylnitrate (PAN). The involvement of NOX in this
photochemical process must be critically examined when considering strategies to reduce the
; impact of O3 on human health. On a global basis, increasing O3 levels in the troposphere are
i also of concern because of the ability of O3 to absorb outgoing radiation and thus contribute
to the greenhouse warming of the earth's atmosphere.
i In recent years, acid deposition in the eastern United States has caused a great deal of
1 concern. Nitrogen oxides contribute to atmospheric acidity through their conversion to nitric
I acid (HNO3), as well as contributing to the chemistry of H2O2, which is an important
oxidant involved in the transformation of SO2 to sulfuric acid (H2SC«4).
August 1991 5-1 DRAFT-DO NOT QUOTE OR CITE
-------
/A'.M
Heterogeneous
loss
Figure 5-1. Summary of the gas phase chemistry of NOX in the clean troposphere.
Source: Finalyson-Pitts and Pitts (1986).
1
2
3
4
In summary, the atmospheric chemistry of NOX has an important impact on:
* concentration of free radical species in the clean troposphere,
* production of O3, and
• acidic deposition from the atmosphere.
August 1991
5-2 DRAFT-DO NOT QUOTE OR CITE
-------
OH
I
H2O
kv S
OOD>
OH
P)
M
02
a^lf
OH
03
/tv
O2+M
NO
Odd oxygen
Figure 5-2. Major chemical reactions affecting oxygen species in the troposphere.
Molecules acting as third bodies are denoted as 'M' (e.g., N2» ©2, Ar,
H2O).
Source: Finalyson-Pitts and Pitts (1986).
1
2
3
4
5
6
7
8
9
10
11
12
13
The chemical mechanisms that control each of these processes will be described in the
following sections. Complementing this chemistry is the dispersion and geographical
movement of NOX and their oxidation products. This chapter will address the factors that
control the transport of these species within the atmosphere.
5.2 THE ROLE OF NOX IN OZONE PRODUCTION
Solar radiation triggers a series of reactions in the atmosphere between gaseous organic
molecules and NOX. This chemistry involves a variety of unstable excited molecules and
molecular fragments that lead to the production of secondary pollutants. In urban plumes, O3
is the predominant product of these reactions. In the free troposphere where pollutant levels
are very low, the amount of NOX present can determine whether the photochemical reactions
lead to the production or consumption of O3. In both types of environment there are similar
August 1991 5-3 DRAFT-DO NOT QUOTE OR CITE
-------
H2O2
jf n»~»2
' OH
H2O
Heterogeneous
removal
Heterogeneous
removal
Figure 5-3. Major chemical reactions affecting hydrogen species (OH, H, H2O,
in the troposphere.
Source; Finalyson-Pitts and Pitts (1986).
1 reaction sequences with the amount of ozone (O3) produced, dependent on the concentration
2 of the various species involved.
3 Most combustion processes emit a variety of nitrogen compounds, with nitric oxide
4 (NO) being the major constituent. Upon entering the ambient atmosphere the NO is oxidized
5 quite rapidly to nitrogen dioxide (NO^. The conversion of NO to NO2 occurs via reactions
6 (5-1) and (5-2):
August 1991
5-4
DRAFT-DO NOT QUOTE OR CITE
-------
Figure 5-4. Schematic diagram of the combined reactions of nitrogen, oxygen, and
hydrogen.
Source: Peters and Carmichael (1983),
1
2
3
4
5
6
7
8
9
0
2NO + O2 •* 2NO2
NO + O3 -*• NO2 + O2
NO2 + hv •+ NO + O(3P)
O(3P) + O2 + M -*• O3 + M
(5-1)
(5-2)
(5-3)
(5-4)
Reaction (5-1), with molecular oxygen (O^ as the oxidizing agent, is relatively slow
and is important only where NO concentrations are elevated, usually in the immediate vicinity
August 1991
5-5
DRAFT-DO NOT QUOTE OR CITE
-------
Steady State
_Q
a
3 10
10°
NOX Concentration (ppb)
Figure 5-5. Calculated steady state concentrations in the free troposphere as a function
of NOX. Conditions described by the authors.
Source: White and Dietz (1984).
August 1991
5-6 DRAFT-DO NOT QUOTE OR CITE
-------
of the source. At typical ambient levels of NO, reaction (5-2) represents the main
mechanism by which NO is converted to NO2. Reaction (5-3) shows that, during the
daytime, NO2 absorbs sunlight at wavelengths less than 430 nm and decomposes to NO and a
triplet-P oxygen atom. This highly reactive oxygen atom forms O3 upon collision with O2 as
in reaction (5-4), where the M represents a third molecule such as nitrogen (N2), O2, etc.,
that absorbs excess vib rational energy from the newly formed O3 molecules.
The net effect of reactions (5-2), (5-3), and (5-4) is an equilibrium in which NO, NO2
and O3 concentrations are interdependent.
NO2 + O2 + hi/ « NO + O3
In the absence of competing reactions, the NO, NO2 and O3 are expected to reach a steady
state condition with their concentrations defined by the following relationship:
1 3Jss "
kJNO]
where JNO is the photolysis frequency for reaction 5-3 and k2 is the rate constant for
reaction (5-2). ~:
In all but the very cleanest atmospheres, sufficient quantities of hydroperoxy and
various organic peroxy radicals are normally present that can compete with O3 for converting
NO to NO2. Thus reactions (5-5) and (5-6) presented below increase the [NO2]/[NO] ratio,
which leads to an increase in O3 levels. In other words, NO is converted to NO2 without
destroying an O3 molecule as happens in reaction (5-2); and consequently, O3 accumulates in
the atmosphere. The peroxy radicals in reactions (5-5) and (5-6) can be initially formed by
photolysis of aldehydes and subsequently from other reactions associated with the
photoxidation of HC.
HO2 + NO •* NO2 + OH (5-5)
RO2 + NO •* NO2 + RO (5-6)
These peroxy radical reactions oxidize NO to NO2 without destroying O3. In effect NO and
NO2 are cycled catalytically with the original NOX concentration remaining essentially
unchanged, but the O3 concentration builds up.
August 1991 . ... • . ... 5-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 The amount of O3 formed is dependent on the concentration of NOX as well as the
2 amounts and reactivity of hydrocarbons available. In urban plumes, O3 concentrations in
3 excess of 200 parts per billion (ppb) are not uncommon in many areas of the United States.
4 However, in NOX rich plumes, such as those eminating from fossil fuel burning power
5 plants, O3 build-up is seldom observed. This is also true at the opposite extreme, where NOX
6 levels are very low. Each of these scenarios will be discussed briefly in the succeeding
7 sections.
8
9 5.2.1 NOx-Rich Chemistry
10 During the early morning hours, the troposphere in many large cities is typically rich in
11 a multitude of HC and NOX (primarily NO). Between 6:00-9:00 A.M., nonmethane
12 hydrocarbon levels generally average between 250 and 1,000 ppb carbon at the surface.
13 During the same period, NOX concentrations fall in the range of 20 to 150 ppb, which leads
14 to typical hydrocarbon to NOX ratios of 7 to 15. Ozone levels in the early morning urban air
15 mass are near zero due to scavenging by NO reaction (5-2). Since reaction (5-2) is very
16 rapid, O3 concentrations cannot rise until most of the NO has been converted to NO2. This
17 will occur as sunlight energy becomes sufficient to generate peroxy radicals. Reactions such
18 as (5-5) and (5-6) then increase the NO2 to NO ratio.
19 Details of the complex HC reactions, which have been reviewed by many authors
20 (Atkinson, 1990), are beyond the scope of this chapter. The key point is that HC serve as a
21 source of free radicals that propagate the series of chemical reactions leading to oxidant
22 production. The process is initiated by the reaction of a HC with an oxidizing species present
23 in the atmosphere. This process is depicted in Figure 5-6.
24 Of the three oxidizing species shown in Figure 5-6, the OH radical is considered to be
25 the most important. Three different production mechanisms for OH have been recognized.
26 Nitrous acid, which accumulates in urban areas during the nighttime hours (Harris et al.,
27 1982), photolyzes at sunrise yielding OH radicals:
28
29 HONO + hv -*• •OH + NO (5-7)
August 1991 =5-8 DRAFT-DO NOT QUOTE OR CITE
-------
CH&CH(OH)CHCH8
|02(2)
CH3CH(OH)CH(CH5)o6
| N0(3)
CHSCH(OH)CH(CH ,)O+NO
I (4)
CH3CH(OH)+CHaCHO
02(5b)
O
x\
CH8=CHCH 8
HOa+CHjCHO CH3CH(OH)OO
N0(6a) i NO (6b)
CH8CH(OH)0+N02
I 7(b)
CH3+ HCOOH
CH3CHO+CH
Figure 5-6. Hydrocarbon oxidation in the atmosphere.
The photolysis of aldehydes also leads to OH radical production. Formaldehyde
(CH2O) is usually present in significant quantities during the morning hours. Photolysis of
CH2O generates OH radicals through the sequence shown below:
5
5
7
3
?
D
1
August 1991
CH2O + hi/ •* H + HCO
H + O2 + M •* HO2 + M
HCO + O2 •* HO2 + CO
HO2 + NO •* NO2 + »OH
5-9
(5-8)
(5-9)
(5-10)
(5-5)
DRAFT-DO NOT QUOTE OR CITE
-------
1 The third source of OH radicals is via photolysis of O3 and subsequent reaction of the O(*D)
2 atoms that are produced with atmospheric water vapor.
3
4 O3 + hi/ •* O(1D) + O2 (5-11)
5
6 O(XD) + H2O H. 2OH (5-12)
7
8
9 This source is probably not very important during the early hours of urban plume chemistry
10 because 03 levels are typically very low. However, as the day progresses and O3 levels
11 build up, it certainly becomes more important.
12 The amount of O3 produced in an urban plume is dependent on the absolute amounts of
13 HC, NOX, and sunlight. In a particular urban area where the HC/NOX ratio remains
14 relatively constant under conditions of high solar intensity, the greater the concentrations of
15 HC and NOX the greater will be the production of O3. However, care must be exercised
16 when comparing oxidant production in different urban areas. Simulations of the atmospheric
17 chemistry of O3 formation have shown that the HC-to-NOx ratio can be quite important in O3
18 formation. The idealized curve in Figure 5-7 illustrates the dependence of O3 production on
19 the precursor ratios. At very low ratios where NO is present in relatively high
20 concentrations, O3 can't build up because of scavenging by NO (reaction 5-2). At high HC
21 to NOX ratios, there is insufficient NOX to propagate the radical reactions that lead to O3
22 production. Urban HC/NOX ratios are generally in the range of 5 to 15, which corresponds
23 to the central portion of the curve shown in Figure 5-7,
24
25 5.2.2 Ozone Production in NOx-Poor Environments
26 Ozone producing reactions identical to those described in near source NOx-rich
27 environments occur in far-field environments, provided there are sufficient quantities of HC
28 and NOX present. Because the magnitude of natural hydrocarbon emissions is considerably
29 larger than natural NOX emissions, there are generally sufficient quantities of HC present,
30 with the result being that the production of O3 is limited by the NOX concentrations. Rural
31 environments can be characterized according to their NOX concentrations.
August 1991 5-10 DRAFT-DO NOT QUOTE OR CITE
-------
HC/NCX
Figure 5-7, Idealized dependence of O3 production on HC/NOX.
Source:
1
2
3
4
5
6
7
8
9
0
1
2
3
1) Regions in which NOX levels average less than 50 parts per trillion (ppt)—
This condition is generally limited to clean maritime areas, most of the free
troposphere and possibly some remote continental locations in the southern
hemisphere.
2) Areas with NOX concentrations ranging from about 50 to 200 ppt—Clean
continental regions in the northern hemisphere comprise this category.
3) Semi-polluted environments with average NOX concentrations ranging
between 1 and 10 ppb—This condition is typical of much of the eastern
United States and probably rural areas in the industrialized countries of
Europe and Asia.
August 1991
5-11
DRAFT-DO NOT QUOTE OR CITE
-------
1 In the cleanest atmospheres, methane (CH^ and carbon monoxide (CO) are the
2 carbon-containing species present in highest concentrations. Consequently, they, along with
3 the NOX, are important participants in the photochemical reactions that control trace gas
4 concentrations. Crutzen (1988) and Penkett (1991) have pointed out that O3 can be produced
5 and destroyed by the CH4 and CO oxidation cycles, depending on the concentrations of NO
6 present. For example, the CO oxidation cycle can proceed by either of the reaction
7 sequences shown below.
8
9 CO + OH -* CO2 + H Al
10
11 H + O2 + M -* HO2 + M A2
12
13 HO2 + NO -* OH + NO2 A3
14
15 NO2 + hv -* NO + O A4
16
17 O + O2 + M -» Q3+M A5
18
19 *Net: CO + 2O2 -* CO2 + O3 (Nl)
20
21
22 CO + OH •* CO2 + H Bl
23
24 H + O2 + M -* HO2 + M B2
25
26 HO2 + O3 -»• OH + 2O2 B3
27
28 *Net: CO + O3 •* CO2 + O2 (N2)
29
30
31 Whether reaction series A or B dominates is dependent on the NO concentration: The rate
32 constant for reaction A3 is about 5,000 times larger than that for reaction B3. Thus, at
33 [O3]/[NO] ratios less than 5,000, the O3 production sequence (A) is more important, while
34 the O3 destruction (B) predominates when NO levels are extremely low. At background O3
35 levels of 20 to 40 ppb, the 5,000 ratio corresponds to NO levels of 10 ppt or less. Similarly,
36 for the oxidation of CH4, the following net reactions for NO-rich and NO-poor environments
37 can be derived (Crutzen, 1988):
38
August 1991 5-12 DRAFT-DO NOT QUOTE OR CITE
-------
In NO-Rich Environments
CH4 + 4O2 •* CH2O + H2O + 2O3 (N3)
CH2O + 4O2 •* CO + 2OH + 2O3 (N4)
CH2O + 2O2 -> CO + H2O + O3 (N5)
(N4) results from a series of reactions that are initiated by photodissociation of CH2O and
subsequent oxidation of NO to NO2 via hydroperoxy radicals. Reaction (N5) involves nearly
the same sequences of steps but is initiated by reaction of CH2O with OH radicals. It is
apparent from net reactions (N3), (N4), (N5), and (Nl) that oxidation of CH2 to CO and
then to carbon dioxide (CO^, in NO-rich environments will yield a gain in O3 molecules.
In NO-Poor Environments
CH4 + OH + OH2 •* CH2O + 2H2O (N6)
CH2O + 2O3 -»• CO + 2O2 + 2OH (N7)
CH2O + O3 •* CO + H2O + O2 (N8)
Based on the net reactions (N6 to N8) and (N2), a loss of O3 is expected to occur at very low
NO concentrations. However, it is expected that in all but the very cleanest of environments
(NO < 10 ppt) O3 production will occur during the daytime hours when conditions are
conducive for photochemical processes.
In the context of this document, it is of greatest interest to study the NO-rich
environments in order to establish whether possibly a quantitative relationship exists between
NOX levels and the amount of O3 that is photochemically produced.
Research studies at a mountain site in Colorado (Niwot Ridge) provide an understanding
of NOX-O3 relationships in rural areas (Parrish et al., 1986). When winds are from the west,
the Niwot site is fumigated by clean air that is devoid of recent anthropogenic emissions.
NOX levels in these westerly air masses are typical of those associated with clean continental
environments (category 2)—namely between 10 and 200 ppt. However, the Niwot site is less
than 100 km from the Denver-Boulder metropolitan areas. Occasionally, upslope winds
advect polluted air that originates in this urban region to the mountain site. Under upslope
flow conditions, NOX levels generally exceed 800 ppt and range up to a few ppb. Figures
August 1991 5-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 5-8a and 8b show the relationship between low NOX levels and O3 at Niwot Ridge. The two
2 lines in each figure represent morning and afternoon relationships. The best fit lines during
3 the morning hours include data collected during the 7:00-11:00 A.M. time periods when the
4 nocturnal inversion has burned off but photochemical O3 production has not fully developed.
5 It is evident that there is little dependence of O3 on NOX during the winter and only a slight
6 dependence during the morning hours in the summer. The slope of the summer morning best
7 fit line is 1.9 +2.8. In contrast to the morning behavior, a large dependence of O3 on NOX
8 is seen in the afternoon summertime data. The slope of the least-squares fit is 16.8 + 2.6.
9 Thus, the daily photochemical production of O3 amounts to about 17 parts per billion volume
10 (ppbv) O3 per day per ppbv NOX.
11 Kelly et al. (1984), made similar measurements at sites in South Dakota, Missouri and
12 Louisiana. Their data provided a value of 6 ppbv O3 per ppbv NOX for the daily
13 photochemical production of O3. This is considerably less than the value of 17 found at
14 Niwot Ridge and could be due to the fact that the Kelly et al. data covered the time period
15 between 10:00 and 2:00 P.M., when the daily solar flux was not.at its maximum and
16 excluded much of the afternoon when solar intensity, and hence O3 production tend to be
17 higher. Also, the average NOX levels were higher (2 to 9 ppbv) at the three sites monitored
18 by Kelly et al. Based on Niwot Ridge data, O3 production per unit NOX becomes smaller at
19 NOX concentrations above 1 ppb. Figure 5-9 shows the summertime O3 mixing ratio
20 measured during the afternoon hours (3:00-8:00 P.M.) versus the concurrently measured NOX
21 mixing ratio. Between about 0.5 and 3 ppb NOX, O3 values exhibit a general increase.
22 Above and below this range there is no apparent dependence of O3 on NOX levels in the
23 ambient data from Niwot Ridge.
24 Recent photochemical modeling results of Liu et al. (1987) agree fairly well with the
25 O3 - NOX relationships derived from ambient data at Niwot Ridge. Figure 5-10 compares
26 calculated O3 production values with ambient measurement data at the Colorado site. The
27 model included nonmethane hydrocarbon (NMHC) chemistry, surface deposition of trace
28 gases and the dilution effect of trace gases due to changes in the daily inversion height. The
29 solid lines in Figure 5-10 represents the model calculated dependence of O3 on NOX. Several
30 scenarios were examined in the model with that represented by the NMHC-PO line being
31 deemed the most appropriate for the Niwot site. At low NOX concentrations, the model
August 1991 5-14 DRAFT-DO NOT QUOTE OR CITE
-------
0 0.2 0.4 0.6 0.8 1.0
30
60
SO
CL
*»
O
40
30
C
(b)
WINTER
,, 7T ,. £ T " ,- -J T ~
T T- ^T V-T- 75T ~r~ O -p- 1 I nVnrrr
'^ — 3: |----ft--r^4^— -11
" -U-T
I ill!
) 0.2 0.4 0.6 0.8 1.0
NO* ppbv
Figure 5-8. (a) Summertime (June 1 to August 31) and (b) wintertime (December 1 to
February 28) Oj mixing ratio vs. NOX mixing ratio during the morning
and afternoon. Filled circles are used for the morning values; open circles
are used for the afternoon values. Each point is an average of [O3] for all
[NOX] values in a ppbv interval. The vertical error bars give the 95%
confidence limits for the average deduced from the standard deviation of
the measurements and the number of measurements in each NOX interval.
The lines give the linear, least squares fit to the data which the average
comprise. For clarity, the morning points and the linear fits have been
offset horizontally as indicated by the second abscissas.
Source: Parrish et al. (1986). .
overestimates the O3 build up. This is suspected to result from an overestimation of odd
hydrogen radical concentrations. While the modeled results always exceed the measured
values, the agreement becomes better at higher NOX concentrations and the general shape of
August 1991 5-15 DRAFT-DO NOT QUOTE OR CITE
-------
IUU
80-
|_ 60-
a.
I 40:
20-
n
3:00-8:00 P.M.
i
I -
I '
*
i
ii
'hi
' fr '
^
i
i
i
>
^
t
M
j
{ n
i M'
Ml *•
H
1
i
r*"1"
,01
.1
1 10
[NO j (ppbv)
Figure 5-9. Summertime O3 mixing ratio vs. NOX mixing ratio measured during the
afternoon hours.
Source: Parrish et al. (1986),
1 the calculated and measured curves are very comparable. An important feature of the net
2 daily O3 change shown in Figure 5-10 is the nonlinear relationship with NOX. Both
3 calculations and measurements indicate that O3 production increases more rapidly at low
4 concentrations of NCX.. This is demonstrated in more detail in Figure 5-11. The two curves
A.
5 in Figure 5-11 show the calculated average daily O3 production per unit concentration of
6 NOX (AP) versus the NOX concentration for summer and winter conditions. The shape of the
7 two curves is similar for the two seasons, however, the summertime daily O3 production
8 values are approximately a factor of 10 larger. This is due to the higher photochemical
9 activity in the summer. The decline in daily O3 production reported by Liu et al. (1987) for
August 1991
5-16
DRAFT-DO NOT QUOTE OR CITE
-------
o.
ex
120
100
80
60
40
20
0
-20
CO-CH4
345
NOX, ppbv
8
Figure 5-10. Model calculated daytime change in O3 (from sunrise to 4:30 P.M.) for
summer clear sky conditions is compared to the observed difference
between the afternoon (2:00-7:00 P.M.) and the morning (7:00-11:00
A.M.) for clear sky conditions. The dashed line is calculated from a
model without NMHC. The shaded area represents calculated values from
a model with anthropogenic NMHC. The lower envelope of the shaded
area is calculated by assuming no overnight retention of secondary HC
(NMHC-PO), while the upper envelope assumes buildup of secondary HC
to their steady state values (NMHC-FO).
Source: Liu et al. (1987).
NOX concentrations larger than 1 ppb is consistent with other modeling results.
Photochemical smog models suggest that the degree of nonlinearity is a function of the ratio
of NMHC to NOX and the relative abundance of various HC.
As pointed out by Liu et al. (1987) the nonlinearity of the NOx-dependent O3 produced
may have important implications for regional and global O3 budgets. Clearly, as atmospheric
August 1991
5-17
DRAFT-DO NOT QUOTE OR CITE
-------
50
40
£ 30
ex.
CL
Q-
DC
20
10-2
WINTER x
SUMME.R
10
10-1
100
NO*, ppbv
101
102
Figure 5-11. O3 production per unit NOX per day, (AP) from the NMHC-PO model are
plotted as function of NOX mixing ratios. A constant NMHC-to-NOx ratio
is assumed; see text for detail. The solid line gives summer values. The
dashed line gives the winter values mutiplied by 10.
Source: Liu et al. (1987).
10
turbulence and advection dilute NOX emissions, the efficiency of O3 production will be
enhanced. Furthermore, the dependence of the O3 production rate on NOX suggests that the
average daily O3 produced at a rural station may be predicted if the NOX concentration is
known. For the United States, Liu and coworkers have estimated an average summer column
O3 production rate due to reactions involving anthropogenic NOX and NMHC that is 20 times
larger than the downward flux from the stratosphere. If O3 produced from natural NOX
emissions is considered as well, the proportion of O3 produced in the eastern and central
regions of the United States that is associated with human activities amounts to 50 to 80%.
These findings are supported in a recent report by Trainer et al. (1987) which compared
model predicted O3 build up with observed values at a rural site in central Pennsylvania.
These authors concluded that the photochemistry of NOX, which is mainly of anthropogenic
origin, and isoprene front biogenic sources can lead to high O3 levels. Consequently, in
August 1991
5-18
DRAFT-DO NOT QUOTE OR CITE
-------
order to reduce O3 levels in rural areas of the eastern United States, consideration must be
given to a reduction in anthropogenic NOX emissions.
5.3 ODD NITROGEN SPECIES
Up to this point there has been little discussion of the nitrogen containing products that
are produced from NOX in tropospheric photochemical reactions. These oxidation products,
which, are commonly referred to as odd nitrogen species (NO ), include HNO3, peroxynitric
acid (HO2NO2), nitrous acid (HNO^, peroxyacylnitrates (RC(O)O2NO2), dinitrogen
pentoxide (N2O5), nitrate radical (NO3) and organic nitrates (RONC^. Many of these
species are detrimental to human health and welfare, and therefore it is important to
understand their production mechanisms from NOX and their fate in the atmosphere.
5.3.1 Nitric Acid
Nitric acid is a strong mineral acid that contributes to acidic deposition problems in the
United States. According to recent estimates, HNO3 accounts for roughly one-third of the
total acidity deposited in the eastern United States (Calvert, 1983). In terms of atmospheric
photochemistry, HNO3 is a major sink for active nitrogen. During the daytime hours, HNO3
is formed by the reaction of NO2 with the OH radical.
NO2 + OH + M -* HONO2 + M (5-13)
Reaction (5-13) serves as a chain terminating step in the photochemistry that produces urban
smog. This is a relatively fast reaction that can produce significant amounts of HNO3 over a
period of a few hours. During nighttime, the heterogeneous reaction between gaseous N2O5
and liquid water is thought to be a source of HNO3. The sequence of reactions that produces
N2O5 and subsequently HNO3 is as follows:
August 1991 5-19 DRAFT-DO NOT QUOTE OR CITE
-------
NO2 + 03 •+ NO3 + 02 (5-14)
NO3 + NO2 -*• N2O5 (5-15)
5 N2O5 + H2O(1) -" 2HONO2 (5-16)
This pathway to HNO3 production is not viable in the daytime because NO3" photolyzes
rapidly and therefore is not present in sufficient quantities to react with NO2. The ,JNQ3~ can
10 abstract a hydrogen atom from certain organics (such as aldehydes, dicarbonyls, and cresols)
to provide another nighttime source of HNO3.
RC(O)-H + NO3 -»• HNO3 + other products (5-17)
15
The importance of this reaction pathway to HNO3 is not well understood at the present time
(see Section 5.3.5).
Logan (1983) has estimated a lifetime of 1 to 10 days for HNO3 in the lower
troposphere. This variability results from the fact that the primary removal mechanism is
20 deposition. The loss of HNO3 by rainout is subject to precipitation frequency while the dry
depositional loss varies with the surface cover and atmospheric mixing characteristics of the
boundary layer. Atmospheric mixing determines the rate of delivery to the surface, and is of
high importance since, in general, natural surface cover (e.g., vegetation) is nearly a perfect
sink for HNO3 (Huebert and Robert, 1985). Chemical destruction mechanisms for HNO3 do
25 exist; however, their importance is not well understood or suspected to be minor in the lower
troposphere. For example, HNO3 can be destroyed through photolysis or reaction with OH
radicals.
HNO3 + OH -*• NO3 + H2O (5-18)
30
HNO3 + hi/ -*• NO2 + OH (5-19)
Reactions (5-18) and (5-19) are slow and, probably cannot compete with the depositional
35 losses of HNO3 in the boundary layer. Neutralization of HNO3 through reaction with
gaseous ammonia (NH3) is another potential sink for HNO3.
August 1991 5-20 DRAFT-DO NOT QUOTE OR CITE
-------
HNO3 + NH3 « NH4NO3 (5-20)
The importance of reaction (5-20) as a removal mechanism for HNO3 is not known. The
5 interaction with NH3 has been reported to influence surface fluxes of HNO3 (Luke and
Huebert, 1987).
5.3.2 Nitrous Acid
Little is known about the distribution and concentration of HNO2 in various ambient
.0 atmospheres. There have been a few measurements in urban environments (Harris et al.,
1982; Winer et al., 1987). During the daytime HNO2 levels are expected to be low because
it photolyzes rapidly.
HNO2 + hv - *OH + NO (5-21)
.5
This reaction likely serves as a source of OH radicals during the morning in urban regions
where HNO2 may accumulate during the nighttime hours. The most likely production
mechanisms for HNO2 include:
:0
•OH + NO + M -*' HNO2 + M (5-22)
NO + NO2 + H2O •* 2HNO2 (5-23)
:5
Reaction (5-22) will only lead to a build up of HNO2 during the late afternoon and evening
hours when sunlight intensities are low but some OH radicals are still present. The latter
reaction (5-23), which is heterogeneous in nature, can produce HNO2 throughout the
nighttime hours.
0
5.3.3 Peroxynitric Acid
While this oxidized form of HNO3 has never been measured, it is expected to be
present in the upper troposphere. Models suggest concentrations in the 10 to 100 parts per
trillion (ppt) range at altitudes above 6 kilometers (Singh, 1987; Logan, 1983). Peroxynitric
5 acid is thermally unstable, consequently, boundary layer concentrations are expected to be
August 1991 5-21 DRAFT-DO NOT QUOTE OR CITE
-------
extremely low (< 1 ppt). Peroxynitric acid is formed through the combination of a
hydroperoxy radical with NO2.
HO2 + NO2 + M •* HO2NO2 + M (5-24)
5
In the upper troposphere, HO2NO2 is destroyed by photolysis or by reaction with OH
radicals.
10 HO2NO2 + hv •* HO2 + NO2 (5-25>
HO2 NO2 + »OH •* products (5-26)
15 5.3.4 Peroxyacylnitrates
Peroxyacetylnitrate is the most abundant of this family of nitrates. The second most
abundant homolog, peroxypropionylnitrate (PPN), is generally less than 10% of the PAN
concentration, with higher molecular weight species such as peroxybenzoylnitrate (PBzN)
expected to be present at even lower levels. Peroxyaeetylnitrate is the only member of this
20 family of compounds that has been extensively studied. It is a strong oxidant and therefore
can have adverse effects on human health and can cause plant damage if ambient
concentrations become high enough. Of greatest interest to this chapter is the role PAN plays
in atmospheric chemistry. Based on its primary means of formation, it might be expected
25 CH3C(O)OO + NO2 •+ CH3C(O)O2NO2 (5-27)
that PAN would serve as a sink for the NOX. This is not true, however, because PAN is
thermally unstable and is much more likely to produce NO2 through the reverse of reaction
30 (5-27) than to be removed by depositional process. Due to thermal decomposition, the
lifetime of PAN is expected to be approximately one hour at 25 °C, two days at O °C, five
months at -23 °C and 42 years at -43 °C. Based on these estimated lifetimes, it was
suggested that PAN could be the principal form of reactive nitrogen in the upper troposphere
(Singh and Hanst, 1981). In reality, since PAN reacts with OH radical and photolyzes, its
35 mean lifetime cannot exceed three months.
August 1991 5-22 DRAFT-DO NOT QUOTE OR CITE
-------
CH3C(O)O2NO2 + hv •* products (5-28)
CH3C(O)O2NO2 + *OH •+ products (5-29)
5 •
Reactions (5-28) and (5-29) are slow compared to the thermal degradation of PAN at
temperatures above 0 °C and therefore are unimportant in determining the summertime
boundary layer fate of this reactive nitrogen compound.
There is no question that PAN can serve as a storage reservoir for NO2. In the
3 summertime boundary layer, PAN concentrations will be decreased somewhat by dry
depositional losses over land (Vd ~ 0.25 centimeter per second [cm s'1]) but it is very likely
that a significant fraction of the PAN produced in urban plumes can be transported into the
regional environment. For example, PAN lifetimes of about 5 and 20 h have been calculated
at 20 °C and 10 °C, respectively. In transport layers above a nighttime surface inversion,
5 PAN could be transported several hundred kilometers. Even during transport conditions
where mixing occurs down to the surface, PAN is expected to persist because it is continually
being produced.
5.3.5 Nitrate Radical
3 The NO3 radical is a short-lived NOX that is formed by the reaction of NO2 with O3.
NO2 + O3 -»• »NO3 + O2 (5-30)
5 Other sources of NO3 radicals exist (Wayne et al., 1991), however reaction (5-30)
serves as the primary tropospheric production mechanism for the NO3 radical. Photolysis of
the NO3 radical is rapid, resulting in a lifetime of about 5 s at midday. Nitrate radicals react
rapidly with NO, which limits their lifetime both during the daylight and nighttime hours. At
NO concentrations of 320 pptv, the lifetime of NO3 radicals due to reaction with NO is
3 similar to that for photolysis (—5 s). Thus, in urban regions where NO concentrations will
normally exceed 300 pptv, the reaction with NO will control the NO3 radical lifetime.
At night, concentrations on the order of 0.3 ppt have been measured in clean
tropospheric air, with recorded levels ranging up to 430 ppt in urban areas (Biermann et al.,
1988). After sunset, the concentration of NO3 radicals is expected to be governed by the
August 1991 5-23 DRAFT-DO NOT QUOTE OR CITE
-------
availability of O3 and NO2, with the latter serving as both a source (reaction [5-30]) and a
sink for NO3 radicals:
•NO3 + NO2 •* NO + NO2 + O2 (5-31)
5
NO2 + »NO3 •* N2O5 (5-15)
N2O5 + H2O(1) •* 2HNO3 (5-16)
10
In clean background environments, it has been reported that measured NO3 radical
levels are significantly less than predicted from a consideration of reactions (5-30), (5-31),
(5-15), and (5-16) alone. This implies that some additional loss mechanism must be
occurring. Speculation has centered around four different loss processes (Platt et al., 1981;
15 Noxonetal., 1980):
(1) heterogeneous losses of NO3 radicals and/or N2O5 on particle surfaces,
(2) reactions of NO3 radicals with H2O vapor,
(3) reaction of NO3 radicals with NO, and
(4) NO3 radical reaction with organic compounds.
20 Based on modeling results, Heikes and Thompson (1983), suggest that low NO3 radical
concentrations could result from the reaction of NO3 radicals with NO provided sufficient
quantities of NO are present at night. In the absence of NO, heterogenous loss of NO3
radicals and N2O5 could account for lower than expected NO3 radical levels provided their
sticking coefficients are greater than 10"3.
25 Wayne et al. (1991) have studied the nighttime chemistry of NO3 radicals and conclude
that simple analyses are useful but are generally insufficient for interpreting NO3 radical
behavior. They suggest that a numerical simulation is required in order to accurately assess
NO3 radical observations in individual data sets. Perner et al. (1991) have carried out such a
modeling exercise, and their results suggest that nighttime concentrations of NO3 radicals can
30 be reduced by the presence of naturally emitted monoterpenes. Thus, in regions when
reactive organic compounds are present in nighttime air masses, lower than anticipated NO3
radical concentrations may be due to scavenging by organic species.
The reactions of NO3 radicals with organic species have garnered considerable interest
in recent years. Kinetic studies have shown NO3 radicals to be very reactive toward a variety
August 1991 5-24 DRAFT-DO NOT QUOTE OR CITE
-------
of organics such as olefins and aldehydes. For example, at NO3 radical levels of
approximately 100 ppt, the lifetime of monoterpene HC due to NO3 radical oxidation is less
than 10 min. This, then, might be an important nighttime sink for both biogenic HC and the
NO3 radical (Winer et al,, 1984). At this time, little is known about the mechanisms or
products that result from the reaction of organics with the NO3 radical. It is expected that
hydrogen abstraction from aldehydes will yield HNO3 and an organic radical:
RCHO + *NO3 •* HNO3 + RCO (5-17)
. . . . . . .. .. . .,-.....
With olefins and cresols, which are products of aromatic HC oxidation, organic nitrates are
expected to dominate,
R-CH=CH2 + •NO3 -* RCH-CH2ONO2 •* products (5-32)
Ar-OH + 'NOg •* ArQ» + HNO3 •* Ar(OH)NO2 (5-33)
5.3.6 Dinitrogen Pentoxide
Dinitrogen pentoxide is the anhydride of HNO3. As indicated in the previous section, it
is formed from NO3 and NO2. Since NO3 is present only at night, N2O5 is primarily a
nighttime species as well. Dinitrogen pentoxide is thermally unstable, decomposing to NO3
and NO2. At high altitudes in the troposphere, where temperatures are low, N2O5 can act as
a temporary reservoir for NO3. Dinitrogen pentoxide also photolyzes at wavelengths less
than 330 mm to give once again NO3 and NO2. This provides the major source of NO3
production in the stratosphere. ,
Dinitrogen pentoxide reacts heterogeneously with water on the surface of hydrometers to
form HNO3. This serves as the main nighttime production mechanism for HNO3, and since
HNO3 is readily deposited by dry and wet deposition, it provides an important method for
removal of NOX from the atmosphere. The importance of the gas phase reaction between
N2O5 and water vapor is not well understood (Logan, 1983).
N2°5 +, H2°(§)- :* 2HNO3 (5-34)
August 1991 -... •;,,.:. , . 5-25 DRAFT-DO NOT QUOTE OR CITE
-------
An upper limit of 1.3 x 10~21 cm3 moleeule'V1 for the rate coefficient has been reported
(Tuazon et al.» 1983). Wayne et al. (1991) have pointed out that even with this low rate
coefficient, the gas phase reaction of N2O5 and water vapor could contribute significantly to
the atmospheric formation of HNC>3. However, until the rate constant is established with
5 greater certainty, the importance of this reaction as a source of HNO3 will remain uncertain.
There are a few reports of N2O5 reacting with aromatic HC such as naphthalene and pyrene
(Pitts et al,, 1985; Atkinson et al., 1986). Nitroarenes appear to be the product of the
reaction.
10 5.3.7 Summary of NOy Species
The total reactive odd nitrogen species are referred to as NO . It is expected that NO,
NO2, PAN, and HNO3 comprise the bulk of NOy present in the ambient atmosphere. This
has been tested by measuring these individual compounds at the same time as a total NOy
measurement is performed (Fahey et al., 1986). The sum of the individual species should
15 equal the NO concentration if NO, NO2, HNO3, and PAN are the only nitrogen compounds
present. Figure 5-12 shows a plot of the (NOJ/NOy ratio versus NOy. (NOy)j is the sum
of measured odd nitrogen compounds. A ratio of 1.0 implies that the sum of the individual
species is equal to the total NOy. These were all summertime measurements at Point Arena
on the California coast, Niwot Ridge in Colorado, and Scotia Range in central Pennsylvania.
20 Both at the Point Arena and Niwot sites, there is a significant NOy shortfall. Approximately
45% of the NOy are unaccounted for at the Niwot site. During the winter months, the
(NOy)j shortfall is not nearly as large (—5%), which implies that the unknown component is
most likely photochemically produced. Organic nitrates, such as methyl nitrate (CH3ONO2)
and higher homologs of PAN, have been suggested as the missing components. Calvert and
25 Madronich (1987) have recently reported that organic nitrates should be important to
photochemistry. While the importance of organic nitrates is recognized, evidence for their
existence in the atmosphere is sparse and often only circumstantial. Therefore, it is unknown
whether or not they constitute the missing NOy fraction at Niwot Ridge and elsewhere.
August 1991 5-26 DRAFT-DO NOT QUOTE OR CITE
-------
1.2
1.0
0.8
Q 0.6
z
0.4
0.2
0
1 I
— o Point Arena
DNiwot Ridge
• Scotia Range
i i i i i i i i
10'
103
NOy,pptv
10*
Figure 5-12. NOy shortfall.
1 5.3.8 Amines, Nitrosamines, and Nitramines
2 The concentration of amines (RNH2, R2NH, RgN) in the atmosphere is thought to be
3 low, although there are few data available to confirm this hypothesis. Highest concentrations
4 would be expected in the vicinity of sources such as cattle feed lots, sewage treatment
5 facilities, waste incinerators, and industrial plants that utilize or produce amines.
6 Nitrosamines (RjNNO) and nitramines (RgNNO^ are produced in the troposphere
7 through photochemical reactions involving alkyl amines and the NOX (NO and NO^. Since
8 both R2NNO and R2NNO2 have proven to be carcinogenic in animals, considerable interest
9 has centered around the tropospheric sources and distribution of these organo-nitrogen
0 compounds. In the preceding NOX criteria document (U.S. Environmental Protection
August 1991
5-27
DRAFT-DO NOT QUOTE OR CITE
-------
1 Agency, 1982), three chemical mechanisms were described for the formation of R2NNO in
2 the atmosphere:
3 * reaction of gaseous amines with NOX and HNO2 (non-photochemical),
4 • photochemical reactions of amines with NOX in the gas phase, and
5 • heterogeneous formation processes involving atmospheric aerosols.
6 All three of these pathways involve reactions of amines with the NOX and/or HNO2.
7 The dark reaction (non-photochemical) and heterogeneous mechanism are poorly understood.
8 In the former case, there are conflicting reports concerning reaction rates and yields of ;
9 It^NNO. It is quite likely that at least part of the dark reactions (5-23 and 5-35) that are
10 believed to convert amines to R2NNO occur on the walls of chambers used to study these
11 processes (Finlayson-Pitts and Pitts, 1986).
12 NO + NO2 + H2O -+ 2HNO2 (5-23)
13
14 R2NH + HNO2 -»• R2NNO + H2O (5-35)
15
16 Therefore, caution must be exercised when extrapolating laboratory smog chamber studies to
17 the real atmosphere. Laboratory studies conducted in the late 197Q's (Hanst et al., 1977;
18 Grosjean et al., 1978; Pitts et al., 1978) indicated low yields (3%) of R2NNO produced in
19 the dark. If this is true, boundary layer concentrations of R2NNO should be low during the
20 nighttime hours.
21 Since there do not appear to be any recent studies that clarify the non-photochemical
22 conversion of amines to R2NNO, the reader is referred to the 1982 NOX criteria document
23 for a more thorough discussion of this subject.
24 As is the case with most heterogeneous chemical transformations in the atmosphere,
25 nitrosation of aerosols is highly speculative. The absorption of basic amines by acidic aerosol
26 droplets followed by reaction with nitrite (NO^, HNO2 or other species could theoretically
27 lead to the formation of R2NNO (U.S. Environmental Protection Agency, 1976). Whether or
28 not the R2NNO so produced could withstand photodecomposition and/or further oxidation is
29 unknown.
30 There is good evidence that photolysis of gaseous amines in the presence of NOX will
31 produce R2NNO and R2NNO2. Figure 5-13 shows the concentration time profiles for
32 diethylnitrosamine in photooxidation experiments involving diethyl- and triethylamine in the
33 presence of NO and NO2. Diethylnitrosamine appears shortly after the reactants are mixed in
August 1991 . • , 5-28 DRAFT-DO NOT QUOTE OR CITE
-------
I 60
S
DC
|o
D
FROM(C2HS)2NH
OFROM(C2Hs)aN
1 2
TIME HOURS
DARK
SUNLIGHT
Figure 5-13. Formation and decay of diethylnitrosamine in the dark and in the sunlight
from diethylamine (open squares) and from triethylamine (open circles).
Source: U.S. Environmental Protection Agency (1982).
1 a darkened chamber. This is presumably due to the dark reactions (5-23 and 5-35) discussed
2 previously. As can be seen in Figure 5-13, the yield of triethymitrosamine is lower than the
3 diethyl analog (—1% vs 3%). Irradiation caused the remaining diethyl- and triethylamines to
% disappear and a variety of products appeared including O3,- PAN, acetaldehyde, nitramines
5 [(C2H5)2NNO2, (CH3)2NNO2] and various amides. The diethylnitrosamine that formed in
3 the dark rapidly decayed, while there was clearly additional generation of this compound
7 from triethylamine followed by its decomposition after continued photolysis. Pitts et al.
3 (1978) derived the following sequence of reactions to explain the photochemical
) transformations.
August 1991
5-29
DRAFT-DO NOT QUOTE OR CITE
-------
1 (C2H5)3N + OH -* (C2H5)2NCHCH3 + H2O
2 *02
3
4 i (NO -» NO^
5 02
6 HO2 + 2H5)2NC(O)-CH3 *- (C2H5)2NC(O)HCH3 -* (C2H5)2NCHO + CH3
7 +
8
9 N02
10 CH3CHO + (C2H5)2N •* (C2H5)2NNO2
11
12 4. NO
13
14 (C2H5)2NNO
15
16
17
18
19
20 In the case of diethylamine, hydrogen abstraction can occur from the nitrogen, as well. This
21 probably accounts for the significantly higher yield of R2NNO2 observed from secondary
22 amines compared to tertiary amines.
23
24 R2NH 4- OH •* R2N + H2O (5-36)
25
26 R2N + NO2 -> R2NNO2 (5-37)
27
28
29 The R2NNO and R2NNO2 have limited lifetimes in the atmosphere due to photolytic
30 decomposition and/or reactions with OH radical and O3. Nitrosamines absorb light in the
31 ultraviolet region (325 to 375 mm) efficiently and are rapidly photolyzed. Tuazon et al.
32 (1984) estimated that dimethylnitrosamine has a half-life of about 5 min at Los Angeles
33 latitudes during the mid-summer daytime. Photolysis will control the fate of nitrosamines in
34 the troposphere because the reactions with OH radical and O3 are relatively slow (Tuazon
35 et al., 1984). The lifetime of dimethylnitrosamine due to reaction with OH radical ([OH] =
36 Ixl06cm"3) has been estimated to be four days. At O3 concentrations of 100 ppb, the
37 lifetime of dimethylnitrosamine will exceed one year.
38 The situation is somewhat different with R2NNO2. They have low light absorption
39 cross sections and do not photolyze readily. Nitramines react very slowly with O3 (lifetime
August 1991 5-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 of dimethylnitramine ~ 4 years) but will be removed from the atmosphere through reaction
2 with OH radical. For example, the estimated lifetime for dimethylnitramine is about 3 days
3 at an OH radical concentration of Ixl06cm~3 (Tuazon et al., 1984). Little is known about
4 the mechanism or products formed in the reaction of R2NNO2 with OH radicals.
5
6
7 5.4 TRANSPORT
8 The transport and dispersion of the various NOy species are dependent on both
9 meteorological and chemical parameters. Advection, diffusion, and chemical transformations
.0 combine to dictate the atmospheric residence time of a particular trace gas. Nitrogenous
. 1 species that undergo slow chemical changes in the troposphere and are not readily removed
2 by depositional processes can have atmospheric lifetimes of several months. Gases with
.3 lifetimes on the order of months can be dispersed over continental scales and possibly even
4 over an entire hemisphere. At the other extreme are gases that undergo rapid chemical
5 transformation and/or depositional losses which limit their atmospheric residence times to a
6 few hours or less. Dispersion of these short-lived species may be limited to only a few
7 kilometers from their point of emission.
8 Surface emissions are dispersed vertically and horizontally through the atmosphere by
9 turbulent mixing processes that are dependent to a large extent on the vertical temperature
K) structure and wind speed. On the vertical scale, transport can occur in three separate layers.
:i
\2 (1) The daytime and/or nighttime mixed layer—This layer can extend from
'3 the surface up to a few hundred meters at night or several thousand meters
'A during the daytime.
'5
'.6 (2) A layer that exists during the nighttime above a low level surface inversion
\1 and below the daytime mixing height—This layer generally is situated
!8 between 200 and 2,000 m (AGL).
'9
•0 (3) The free troposphere—This transport zone is above the boundary layer
11 mixing region.
i2
13 During the warm, summertime period when the impact of reactive nitrogen species is
14 the greatest, vertical mixing follows a fairly predictable diurnal cycle. A surface inversion
15 normally develops during the evening hours and persists throughout the nighttime and
August 1991 5-31 DRAFT-DO NOT QUOTE OR GITE
-------
1 morning period until broken by surface heating. While the inversion is in place, surface NOX
2 emissions can lead to relatively high, local concentrations because of restricted vertical
3 dispersion. Following the break up of the nighttime surface inversion, vertical mixing will
4 increase and surface based emissions will disperse to higher altitudes. The extent of the
5 vertical mixing during the daytime is often controlled by synoptic weather features. Elevated
6 temperature inversions associated with high pressure systems are common in many parts of
7 the United States. An elevated inversion in the Los Angeles Basin frequently traps pollutants
8 in the lower 600 m of the atmosphere. In the midwestern and northeastern regions of the
9 United States, summertime afternoon mixing levels normally range between 1,600 and
10 1,800 m (Holzworth, 1967). Horizontal dispersion of trace gases in the mixing layer is
11 caused by horizontal turbulence and vertical wind shear. For mesoscale and synoptic scale
12 transport, mean wind shear is the dominant cause for dispersion.
13 The dispersion processes described above coupled with chemical transformations of a
14 particular reactive nitrogen compound dictate transport distances in the troposphere.
15 A reasonable understanding exists concerning the short-term (daylight hours) fate of NOX
16 emitted in urban'areas during the morning hours. As described in detail in Section 5.2.1,
17 NOX emitted in the early morning hours in an urban area will disperse vertically and move
18 downwind as the day progresses. On sunny summer days, most of the NOX will have been
19 converted to HNO3 and PAN by sunset. Much of the HNO3 will be removed by depositional
20 processes as die air mass moves along. After dusk, an upper portion of the daytime mixed
21 layer will be decoupled from the surface due to formation of a low-level radiation inversion.
22 Transport will continue in this upper level during the nighttime hours and while
23 photochemical processes will cease, dark-phase chemical reactions can proceed. There are no
24 reports of plume measurement studies that have tracked plumes for more than one daylight
25 period. Thus, nothing is known concerning the fate of the remaining nitrogenous species that
26 become entrapped in the layer above the nighttime surface inversion and below a higher
27 subsidence inversion. Peroxyacetyl nitrate and HNO3, if carried along in this layer, could be
28 transported long distances.
29
30
August 1991 5-32 DRAFT-DO NOT QUOTE OR CITE
-------
i 5.4.1 Transport of Reactive Nitrogen Species in Urban Plumes
I The most extensive studies of the fate of nitrogen species in urban plumes have been
B reported by Spicer and eoworkers. They have examined the behavior of reactive nitrogen
1 compounds in plumes eminating from Los Angeles, CA (Spicer et al., 1979); Phoenix, AZ
5 (Spicer et al., 1978); Boston, MA (Spicer, 1982); and Philadelphia, PA (Spicer and
5 Sverdrup, 1981). The nitrogen budget derived from Boston plume measurements provides a
7 good example of the fate of reactive nitrogen compounds in urban plumes. Nitrogen oxide
3 concentrations at short distances from Boston varied from about 30 to 130 ppb. After travel
) times of 4 to 7 h, NOX concentrations in the plume were in the 5 to 10 ppb range. Removal
3 rates of NOX were calculated from measured NOX, PAN, and HNO3 concentrations. Dilution
I effects were accounted for by using tracers of opportunity such as CO, acetylene, and
I CFC13. The plume was monitored by aircraft to distances as far as 150 km east of Boston.
$ This corresponded to reaction times of as long as 7.5 h. The NOX removal rate ranged from
t 0.14 to 0.24 h"1 on four different days. This corresponds to NOX lifetimes (I/removal rate)
5 of 4.2 to 7.1 h (Altshuller, 1986). These lifetimes apply to sunny, summertime, moderately
5 polluted plume conditions with transport mainly over water. Nitrogen oxide depositional
7 losses 'during over-water transport should be very small. In the Boston plume, the loss of
? nitrogen as NOX is equal to the gain in nitrogen in the products HNO3, PAN, and NO3. On
J an August 18, 1978, flight, an overall NOX loss rate of 0.24 h"1 was obtained. During the
) same measurement period, a value of 0.23 h"1 was calculated for the conversion rate of NOX
I to NO3 products.
I Somewhat lower NOX loss rates have been reported from data collected in Los Angeles
5 (Chang et al., 1979). Chang and eoworkers derived a value of 0.04 h"1 as a lower limit for
}• the yearly average daytime NOX removal rate. Calvert (1976) estimated the NOX removal
> rate to be approximately 0.09 h"1 during the midmorning to early afternoon hours. It has
5 been suggested that the values derived from Los Angeles data probably represent only a
7 portion of the true NOX loss rate because of NOX measurement interferences by HNO3 and
$ PAN. However, the 0.09 h"1 value derived by Calvert (1976) agrees well with recent
) estimates in the Detroit metropolitan area (Kelly, 1987). Using a combination of captive air
) outdoor irradiation experiments, photochemical modeling, and ambient measurements, Kelly
i (1987) obtained an NOX removal rate of approximately 0.1 h"1. It was determined that HNO3
August. 1991 5-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 accounted for 67 to 84% of the N-containing products. This led to the conclusion that HNO3
2 formation will control the chemical lifetime of NOX on photochemically active days.
3 Based on a calculated formation rate of HNO3, Kelly (1987) was able to estimate the
4 expected ambient HNO3 concentrations in the Detroit plume. The predicted concentrations
5 always exceeded the measured ambient HNO3 concentrations by a significant amount (3 to
6 4 times). In order to reconcile this difference, he hypothesized that once formed, HNO3 was
7 rapidly removed from the urban plume. Several removal mechanisms were considered with
8 incorporation into coarse atmospheric aerosol deemed to be most likely. Based on his
9 analysis, approximately 70% of the gaseous NOX in Detroit's morning atmosphere is
10 photochemically converted to nitrogen products by sundown. In the absence of sinks, the
11 product distribution would be ~75% HNO3, -20% PAN, and the remaining 5% as other
12 organic nitrates. However, removal of HNO3 as coarse NO3 leads to a maximum HNO3
13 concentration that is only 20 to 30% of that expected in the absence of the aerosol sink.
14 Since the coarse NO3 aerosol is quickly removed by sedimentation, the majority of the
15 nitrogen containing species emitted and produced in the Detroit urban area are not transported
16 long distances downwind.
17 It should be kept in mind that the studies just described in Boston, Los Angeles, and
18 Detroit addressed only the daytime fate of reactive nitrogen species. The nighttime chemistry
19 of NOy compounds in urban environments is poorly understood. Nighttime emissions of NO
20 will react with O3 to produce NO2 as long as there is sufficient O3 present. Often the O3
21 reservoir that exists aloft over urban areas is decoupled from the surface layer at night by a
22 low level radiation inversion. Under these conditions, the O3 supply is not replenished and
23 once it has been used up, NO will no longer be oxidized to NO2 via reaction with O3. As
24 described earlier, O3 reacts with NO2 to form the NO3 radical. The NO3 radical is very
25 reactive; consequently, its concentration remains low (~ 1-500 ppt). It reacts rapidly with
26 NO, so that as the night progresses and NO levels increase, NO3 radical concentrations will
27 fall.
28
29 NO3 + NO •* 2NO2 (5-38)
30
31
August 1991 5.34 DRAFT-DO NOT QUOTE OR CITE
-------
1 Other sinks for NO3 include the reactions with organic species (reactions 5-17; 5-32 and
2 5-33) and with NO2 to produce N2O5 (reaction 5-15). The N2O5 will react with water to
3 form HNO3 (reaction 5-16). It can also react with polycyclic aromatic HC which may
4 produce significant quantities of nitro-substituted polycyclic aromatics (Pitts, 1987).
5 Chemical models have been employed to try to sort out the nighttime NOX chemistry
6 (StockweU and Calvert, 1983; Jones and Seinfeld, 1983; Russell et al., 1985). Russell et al.
7 (1985) were able to get reasonable agreement between predicted and measured time profiles
8 of NO3, NO2, and O3 during nighttime hours in the Los Angeles Basin. The reactions of
9 importance to nighttime chemistry and the rate constants as employed in the Russell et al.
10 model are listed in Table 5-1. In order to accurately model this nighttime NOX chemistry,
11 atmospheric concentrations of NO, NO2, NO3, O3, water vapor, and organic species must be
12 known. In addition, meteorological factors such as mixing conditions, temperature, etc., are
13 important. An adequate data set is not currently available for modeling purposes.
14 Consequently, concentrations must be estimated for unmeasured species. From the rate
15 constants shown in Table 5-1, it is obvious that NO can play a very critical role in nighttime
16 chemistry. Nitric oxide very rapidly scavenges NO3. For example, Russell et al. (1985)
17 calculate a NO3 concentration of 12 ppt in the LA surface layer when 1 ppb NO is present
18 and over 200 ppt NO3 when NO is negligible. Dinitrogen pentoxide is another species that
19 can significantly influence nighttime chemistry. The magnitudes of the various N2O5 loss
20 processes (see Table 5-1) are not well understood. Due to the transient character of N2O5, it
21 has been difficult to determine the homogeneous gas phase reaction rate constant with water
12 vapor, the deposition velocity, and heterogeneous interactions with ambient particulate matter.
23
24 5.4.2 Transport and Chemistry in NOx-Rieh Plumes
15 Interest in the NOX chemistry of power plant plumes increased significantly following
26 the report by Davis et al. (1974) that O3 could be generated through a series of reactions
27 involving sulfur and nitrogen constituents within this type of plume. In subsequent studies, it
28 was pointed out that the well known photochemical reactions involving NOX and HC is the
19 more likely mechanism of O3 build up in power plant plumes (Miller et al., 1978). Since the
30 HC-to-NOx ratio in these plumes is very low, the HC must be mixed into the power plant
August 1991 5-35 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-1. MAJOR REACTIONS IN THE NO3-N2O5 SYSTEM AT NIGHT
Rate Constant 298 K
Reaction (ppm min units)
NO2
NO -
NO2
NA
NA
NO3
NO3
NO,
NO2
NA
NO3
+ 03
f NO3
+ NO3
+ H2O
+ HCHO
+ RCHO
+ OLE
•f NO3
+ aerosol
+ aerosol
7
8
44
45
46
53
54
— >
56
— >
57
— >
— >
— >
k7 = 0.05
NO3
kg = 29560
2NO2
k44 = 2510
NA
NO2 + NO3
k« = 1.9 x 10-*
2HN03
k53 = 0.86
HNO3 + HO2 + CO
k* = 3.6
RCO3 + HN03
k56 = 12.4
RPN
kj7 = 0.59
NO + NO, + O2
2HN03 k*
N203
aerosol k*
NO3
Reference and
Comments
1
1
2
1,3
4 ; "•
5
5,6
5, 7, 8, 9
10
11
11
(1) Baulch et al. (1982); (2) Tuazon et al. (1984); (3) Malko and Troe (1982); (4) Tuazon et al. (1983);
(5) Atkinson et al. (1984); (6) The rate constant used for the NO3 reaction with high aldehydes is that measured
for aeetaldehyde; (7) The value used for the rate constant of the NO3 reaction with olefins is that measured for
the NO3 reaction wit; (8) The ultimate products of reaction (56) are reported to be nitroxyperoxyalkyl nitrates
and dinitrates (Bandow et al., 1980); (9) Bandow et al. (1980); (10) Atkinson and Lloyd (1984); (11) Russell
et al. (1985)
Source: Russell et al. (1985).
1 plume as it moves downwind. Excess O3 concentrations of 20 to 50 ppb above ambient have
2 been reported in plumes after several hours of downwind transport. However, an O3 build
3 up is not found in all power plant plumes (Hegg et al., 1977; Ogren et al., 1977; White,
4 1977). The single most important ingredient appears to be the availability of reactive HC in
5 the dilution air.
August 1991
5-36
DRAFT-DO NOT QUOTE OR CITE
-------
Hegg et al. (1977) have reported measurements in four different NOx-rich power plant
1 plumes. Out to distances of 90 km and travel times up to 4 h, no O3 enhancements were
» observed in any of these plumes. Measured NO2/NO ratios in these plumes were generally
• in the 1 to 2 range. This is considerably below the value of 10 which has been shown in
• laboratory experiments to be the minimum ratio at which appreciable O3 generation can
i occur. Aircraft measurements have shown little enhancement of inorganic and particulate
' NO3 concentrations in power plant plumes (Easter et al., 1978; Hegg and Hobbs, 1979).
! Because of the difficulties associated with tracking and making accurate measurements in a
> narrow power plant plume, little, if any, other information is available concerning the fate of
I reactive nitrogen species in these plumes. As a consequence, smog chamber and modeling
studies have been employed to study NOX transformation rates. Smog chamber experiments
1 under a variety of conditions expected in real atmospheres yield NOX lifetimes varying from
i about 1 to 7 h (Spicer et al., 1981). The fastest conversion times were observed when
HC/NOX ratios were high (~ 15) and the HC mix included those species typically found in
I urban atmospheres. Nitric acid and PAN were the major nitrogen-containing products
i observed in the chamber reactions. Generally, they accounted for between 70 and 90% of
' the nitrogenous species present at the end of the final irradiation period. The PAN/HNO3
> ratio varied depending on the initial HC/NOX ratio. Less PAN was produced in cases where
> organic levels were initially low.
) The smog chamber studies imply that the lifetime of NOX in a power plant plume can
vary from a few hours to more than a day, depending on environmental conditions. Under
'. conditions of low HC levels (e.g., in rural areas or aloft above a surface inversion) the NOX
> lifetime will be sufficiently long to allow NOX input to regional air masses.
i
r
i 5.4.3 Regional Transport
i Transport of reactive nitrogen species in regional air masses can involve several
' mechanisms. Mesoscale phenomena such as land-sea breeze circulations or mountain-valley
> wind flows will transport pollutants over distances of 10's to 100's of km. On a larger scale,
> synoptic weather systems such as the migratory highs that cross the eastern United States in
) the summertime influence air quality over many hundreds of kilometers. The accumulation
and fate of nitrogen compounds will differ somewhat between the mesoscale and synoptic
August 1991 5-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 systems. Mountain-valley and land-water transport mechanisms have dual temporal scales
2 due to their dependence on solar heating. However, in the larger scale synoptic systems,
3 reactive nitrogen species can build up over multiday periods. The residence time of air
4 parcels within a slow moving high pressure system can be as long as six days (Vukovich
5 etal., 1977).
6 In many cases, the transport mechanisms mentioned above are interrelated. For
7 example, slow moving high pressure systems that migrate across the eastern United States are
8 characterized by weak pressure gradients. Thus, mountain-valley or land-water breezes can
9 dictate pollutant transport in the immediate vicinity of sources but the eventual fate of
10 reactive nitrogen species will be distribution into the synoptic system.
11 Combined studies of air quality and meteorology along the western shore of Lake
12 Michigan have clearly documented this relationship (Lyons and Cole, 1976; Westberg et al.,
13 1981). Data shown in Table 5-2 and Figure 5-14 were collected near Kenosha, WI, during
14 the period August 14-22, 1976. During this time, a strong, slow-moving high-pressure
15 system traveled across the Great Lakes Region. Degradation of air quality in southeastern
16 Wisconsin was clearly associated with both synoptic transport and mesoscale (lake breeze)
17 advection during this nine day period. August 14th was the first day during which the effects
18 of the advancing high were observed in southeastern Wisconsin. Northerly flow associated
19 with the leading edge of the anticyclone persisted through August 16th. As can be seen in
20 Table 5-2 pollutant levels were low during the period with northerly winds. From
21 August 18-22, meteorology along the western shore of Lake Michigan was controlled by
22 synoptic features characteristic of the trailing edge of an anticyclone and the local lake breeze
23 phenomenon. Thus, gradient winds were from the southwest, but during the afternoon hours
24 a shift to southeasterly flow occurred as the lake breeze front moved inland. Figure 5-14
25 shows pollutant profiles recorded about five miles inland on the afternoons of August 18 and
26 19. Pollutant levels increased dramatically following passage of the lake breeze front on both
27 of these days. These high pollutant levels were most likely the result of emissions from the
28 Chicago-Hammond-Gary urban complex. During the night and early morning hours, the
29 plume from this industrial region drifted in a northerly direction over the lake. Morning
30 sunlight served to initiate photochemical processes in the contaminated air mass over Lake
31 Michigan. High levels of secondary pollutants such as O3 and NO2 had developed by early
August 1991 5-38 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-2. AVERAGE AFTERNOON BACKGROUND POLLUTANT
CONCENTRATIONS MEASURED AT KENOSHA, WI
North wind
South wind
03
(ppb)
37
94
NO
(ppb)
2
3
NO2
(Ppb)
3
7
NMTHC
(ppbC)
123
213
CO
(ppm)
0.4
0.6
CFC13
(PPO
163
229
Source: Westberg et al. (1981).
600
Lake breeze
front past trailer
1200
14 16
August 18
Lake breeze
front past trailer
7 . .-.- .•:..-.". : * *-::
181200
Time of day
14 16
August 19
18
Figure 5-14. Pollutant levels at the Kenosha sampling site before and after passage of
the lake breeze front.
Source: Westberg et al. (1981).
August 1991
5-39
DRAFT-DO NOT QUOTE OR CITE
-------
1 afternoon when the air mass was transported on shore by the lake breeze. The pollutants
2 associated with the local, mesoscale lake breeze system got incorporated into the larger scale
3 synoptic circulation and contribute to the increased levels associated with southwesterly flow
4 (see Table 5-2). The distribution of NO within these anticyclones will depend on rates of
5 chemical conversion and deposition.
6 The studies just described provide little detailed information about NOX behavior under
7 various regional transport scenarios. At the time these studies were conducted, sophisticated
8 NOX and NOy monitoring instrumentation was not generally available. Consequently, we
9 only know that NOX levels were near the detection limit of the instrumentation (~1 to 2 ppb)
10 in synoptic air masses advected into the United States from Canada and that NOX
11 concentrations increased significantly after a period of residence over industrialized regions of
12 the United States. The composition of NOy in the aged air mass is not well known.
13 However, more recent studies being conducted in conjunction with the NAPAP and WATOX
14 programs may be able to provide better insight into NOX chemistry in regional air masses
15 (Fahey et al., 1986; Luke and Dickerson, 1987). For example, Luke and Dickerson (1987)
16 have reported NOX and NOy measurements off the east coast of the United States. Sampling
17 flights were conducted during the period January 3 to 11, 1986. The investigators subdivided
18 the atmosphere along their flight tracks into two-dimensional boxes and calculated the flux of
19 nitrogen species through each box. A calculated gross nitrogen flux of 0.5 Tg/year for the
20 eastern coast of the United States was derived using this methodology. Luke and Dickerson
21 (1987) emphasize that this flux should be viewed with caution since it results from the
22 combination of an annually averaged wind field and a NOy data base of limited time
23 resolution.
24 Even though the Luke and Dickerson (1987) flux number is subject to considerable
25 uncertainty, it is interesting to compare it to earlier flux estimates that were derived by less
26 direct methods. Logan (1983) calculated a nitrogen flux of 1.7 Tg/year by balancing the
27 regional nitrogen budget of eastern North America. Galloway et al. (1984) estimated that
28 1.1 to 3.2 Tg/year are transported eastward off the Atlantic coast. More recently, Galloway
29 and Whelpdale (1987) have reduced their flux estimate downward to 0.8 to 1.2 Tg/year.
30 This flux corresponds to approximately 25% of the NOX emitted to the atmosphere of eastern
31 North America based on Logan's (1983) NOX emission estimate of 4.5 Tg/year.
August 1991 • 5,40 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.5 OXIDES OF NITROGEN AND THE GREENHOUSE EFFECT
2 Except for N2O, the oxygenated nitrogen species typically assumed to comprise the
3 NOX and NOy families, as discussed in previous sections of this chapter, do not absorb
4 infrared radiation, and therefore, do not contribute to direct radiative "greenhouse" forcing.
5 Nitrogen dioxide is an efficient absorber of visible radiation, and it has been proposed as a
6 possible source of additional climatic influences, assuming atmospheric concentrations were to
7 become sufficiently large (Wuebbles, 1989). The previously described NOX species can,
8 however, contribute indirectly to the greenhouse process through the photochemical
9 production of O3, a known greenhouse gas. Additionally, N2O, which is chemically inert in
0 the troposphere, readily absorbs infrared radiation and is among the more significant non-CO2
.1 greenhouse gases.
2
3 5.5.1 NOx-Related Ozone Greenhouse Effects
4 On a per mole basis, Rodhe (1990) estimated tropospheric O3 to be approximately
5 2,000 times more effective at absorbing infrared radiation than CO2. The contribution of O3
6 to the greenhouse effect at present tropospheric levels (10 to 50 ppbv) has been calculated to
7 be ori the order of 8% (Rodhe, 1990). Measured background tropospheric O3 concentrations,
8 particularly in the northern hemisphere, have shown an apparent increase over the past few
9 decades, although the uncertainties are generally large (Logan, 1985; Oltmans and Kohmyr,
0 1986; Angell, 1988). Rodhe (1990) estimated the current annual rate of increase in global
1 tropospheric O3 to be 0,5%, Using a one-dimensional model, Lads et al, (1990) showed the
2 direction of the O3-induced radiative forcing to be sensitive to the vertical O3 distribution.
3 Tropospheric O3 increases, as well as stratospheric decreases, could both lead to surface
4 warming. Ozone concentration changes in the upper troposphere and lower stratosphere,
5 where temperatures are at a minimum compared to surface temperatures, are the most
6 effective in producing surface layer temperature changes (Wuebbles, 1989). It should be
7 noted, however, that Lads et al. (1990) calculated a net O3-induced 0.05 °C (±0.05 °C)
8 surface cooling for mid-latitude regions during the 1970s, Based on limited O3 observation
9 column data, the modeled surface cooling caused by decreases in stratospheric O3 outweighed
0 warming effects brought on by tropospheric O3 increases.
August 1991 5-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 The photochemical relationship of NOX to the formation of tropospheric O3 has been
2 previously described in Section 5.2. It is generally regarded that due to the relatively short
3 atmospheric lifetime of O3» local (urban) sources of O3 do not contribute significantly to the
4 upper tropospheric global O3 levels (Machta, 1983). Sources of greenhouse-important
5 tropospheric O3 are presumably an equal combination of downward injection of stratospheric
6 O3 and O3 precursors (NOX) during intrusion episodes and upper tropospheric photochemical
7 production (Liu et al., 1980; Fishman, 1985; Wuebbles et al., 1989). Of particular interest
8 then are the mechanisms by which NOX species can be transported and dispersed sufficiently
9 throughout the mid to upper troposphere to result in the formation of upper-level O3. The
10 processes of mesoscale and synoptic transport (see Section 5.4.3) in combination with
11 transformation of NOX to reservoir species, such as peroxy- and organic nitrates, could serve
12 to relay the O3 precursor species to the remote troposphere. Singh et al. (1985) submitted
13 that PAN, in particular, would be an effective long-range transport mechanism for boundary
14 layer NOX, thereby influencing background levels of O3, as well as other important oxidation
15 compounds (i.e., OH radicals).
16 Another proposed anthropogenic source of mid and upper tropospheric, as well as
17 stratospheric, O3 precursors (NOX and HC) is via the exhaust of high-flying jet aircraft (Liu
18 et al., 1980; Kinnison et al., 1988). Liu et al. (1980) estimated a potential 7 to 15% increase
19 in upper tropospheric O3 over the northern hemisphere due to high-flying subsonic aircraft
20 for the 1970s. They compared this to an observed average 8% increase over the same
21 hemisphere from 1966 to 1977. Nitrogen oxides are also known to be produced in
22 conjunction with lightning discharges (Logan, 1983). While investigating lightning as a
23 potential source for stratospheric NO , Ko et al. (1986) showed that significant levels of
24 tropospheric NOy can be produced and transported upwardly into the stratosphere, especially
25 in the tropical latitudes.
26 As briefly mentioned earlier, O3 perturbations within the stratosphere, particularly the
27 lower regions, can produce surface temperature changes of equal or greater magnitude than
28 O3 perturbations within the troposphere (Lacis et al., 1990). In the lower to mid-
29 stratosphere, at middle latitudes, the destruction of stratospheric O3 proceeds via a series of
30 complex, catalyzed reactions involving the HOX and C1X families of free radicals. These
31 reactions as summarized by Johnston (1982) are as follows:
32
August 1991 5-42 DRAFT-DO NOT QUOTE OR CITE
-------
NO + O3 •* NO2 + O2
direct
catalyzed by HOX (5"39'
catalyzed by C1X
O3 + hv •* O + O2 (5-40)
>
NO2 + O •* NO + O2 (5-41)
) where, because reaction 5-41 is the rate determining step, the gross rate of stratospheric O3
'•) depletion through reactions of NOX can be approximated by
7
I d[03] = 2 ks..42 [O] [N02] (5-42)
)
) In polar (high latitude) regions, especially in the Antarctic where favorable conditions
often exist, additional NCL/C1X heterogeneous reactions on the particulate surfaces of polar
\ stratospheric clouds can further enhance stratospheric O3 depletion (see Section 5.6), thereby
> increasing the likelihood of greenhouse forcing.
5.5.2 Nitrous Oxide Greenhouse Contributions
Chemically unreactive in the troposphere, N2O readily absorbs infrared radiation, and
is estimated to be responsible for approximately 4 to 5% of the theorized greenhouse effect
(Hansen et al., 1989; Rodhe, 1990). At present atmospheric mixing ratios (307 to
310 ppbv), N2O, on a per mole basis, is given to be 200 times more effective than CO2 as an
absorber of heat radiation (Wuebbles, 1989; Rodhe, 1990). Nitrous oxide in the troposphere
is thought to originate predominately through soil denitrifieation (McElroy, 1980; Machta,
1983). Additionally, anthropogenic sources, especially high-temperature combustion, may
also release significant amounts of N2O. Wuebbles (1989) approximated as much as 40% of
atmospheric N2O to be a product of anthropogenic processes (fossil fuel combustion «21%,
biomass burning «5%, fertilized soils «5%, cultivated natural soils ~IQ%). Other researchers
(Hao et al., 1987; Muzio and Kramlich, 1988) have proposed that the contribution of fossil
fuel combustion may by significantly less. Thiemens and Trogler (1991) calculated the
commercial manufacture of nylon releases, on a global basis, approximately 6.6 x 108 kg of
August 1991 5-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 N2O annually. This would account for about 0.03% of the current tropospheric levels, or
2 about 10% of the observed annual increase.
3 Although there are still unanswered questions on the partitioning of N2O sources, the
4 atmospheric levels of N2O have been observed to be increasing at an annual rate of 0.2 to
5 0.3% (Machta, 1983; Wuebbles et al., 1989; Rodhe, 1990). With an atmospheric lifetime of
6 150 years (Rodhe, 1990; Levander, 1990), it can be seen that given an increase in
7 atmospheric loading, N2O will become an increasingly more important greenhouse gas.
8 Levander (1990), using a 0.2% annual increase, predicted the increased atmospheric mixing
9 ratio of N2O after 50 years would result in an even greater efficiency (350x) in infrared
10 absorption compared to expected CO2 levels.
11 Nitrous oxide does not decompose until it is transported to the stratosphere. The main
12 sink for stratospheric N2O is its reaction with O(*D) (Johnston, 1982; Logan, 1983;
13 Wuebbles, 1989). The products of this reaction are the primary source of stratospheric NOX,
14 which, as previously mentioned, catalytically react to deplete stratospheric O3 and partially
15 supply NOX species to the upper troposphere during intrusion episodes.
16
17
18 5.6 STRATOSPHERIC OZONE DEPLETION BY OXIDES OF
19 NITROGEN
20 Oxides of nitrogen, from anticipated high-altitude supersonic aircraft exhaust, were first
21 proposed as a destruction mechanism for stratospheric O3 in the early 1970's (Crutzen, 1970;
22 Johnston, 1971). Since then, the relationship between stratospheric NOy, this group of
23 nitrogen oxides (NO + NO2 + NO3 + HNO3 + C1NO3 + N2O5 + HNO4), has received
24 considerable attention.
25 The primary source of stratospheric NOy is thought to be via the reaction of O(*D) and
26 N2O (Johnston, 1982; Logan, 1983; Wuebbles, 1989).
27
28 O(JD) + N2O - 2NO (5-43)
29
30
31 The significant sources of N2O have been discussed in Section 5.5.2. Jackman et al. (1980)
32 list other possible, less significant, sources of stratospheric odd nitrogen as NOy produced by
August 1991 5-44 DRAFT-DO NOT QUOTE OR CITE
-------
I lightning in the troposphere followed by upward transport, nuclear bomb blasts,
I thermospheric NOy produced with downward transport, and NOy produced during ionizing
I events (i.e., solar proton episodes). Additionally, recent renewed interest in stratospheric,
!• intercontinental passenger jet aircraft has likewise restimulated research into the sensitivity of
i stratospheric O3 to potential aircraft exhaust (Kinnison et al., 1988; Kinnison and Wuebbles,
5 1989).
' In the mid-latitude and lower to mid-stratospheric regions, as mentioned in
! Section. 5.5.1, NO cycles through NO2 for a net destruction of two molecules of O3:
1 -•.••. '
I NO + O3 -* NO2 + O2 (5-39)
: O3 + hv '-* O + O2 (5-40)
i ,
NO2 + O -* NO + O2 (5-41)
*Net: 203 -* 3O2
1 It is known, however, that catalytic reactions involving chlorinated compounds within
i the stratosphere can be even more effective as an O3 depletion mechanism. These reactions,
although not specifically involving NOX in the O3 destruction reaction, are an important
component of the overall reaction system. The homogeneous CIO dimer reaction sequences,
as proposed by Molina and Molina (1987), are as follows.
2C1O + M -* C12O2 + M ' (5-44)
C12O2 + hv -* Cl + C1OO (5-45)
C1OO + M -* Cl + O2 + M , (5-46)
2C1 + 2O3 •+ 2C10 + 2O2 (5-47)
*Net: 2O3 -* 3O2
Fahey et al. (1989) showed a similar CIO dimer reaction sequence, differing only by the self-
decomposition of dichlorine dioxide (C12O2) and the photolysis of diatomic chloride. The
actual O3 depletion mechanism (reaction 5-47) is the same in both reaction schemes.
August 1991 .5-45 DRAFT-DO NOT QUOTE OR CITE
-------
1 2C1O + M •* C12O2 + M (5-44)
2
3 C12O2 + M •* C12 + O2 + M (5-48)
4
5 C12 + hi/ •* 2C1 (5-49)
6
7 2C1 + 2O3 •* 2C1O + 2O2 (5-47)
8
9 *Net: 2O, -f 3O9
*j i-
10
11
12 McElroy et al. (1986a) also proposed the following chlorine/bromine oxidation cycles as
13 important stratospheric O3 depletion mechanisms.
14
15 Br + O3 •* BrO + O2 (5-50)
16
17 Cl + O3 •* CIO + O2 (5-51)
18
19 CIO + BrO •+ Cl + Br + O2 (5-52)
20
21 *Net: 2O3 •* 3O2
22
23
24 The role of reactions (5-44) to (5-52) in stratospheric O3 depletion are generally limited in
25 the mid-latitude regions by the reaction of chlorine monoxide (CIO) and BrO with NO2 which
26 forms the unreactive chlorine nitrate (C1NO3) and bromine nitrate (BrNO3) (McElroy et al.,
27 1986a).
28
29 CIO + NO2 + M -»• C1NO3 + M (5-53)
30
31 BrO + NO2 + M •+ BrNO3 + M (5-54)
32
33
34 In the atmospheric regions mentioned above, where oxygenated nitrogen compounds are
35 generally much more prevalent than chlorine compounds, reactions (5-53) and (5-54) act as
36 important sinks for CIO and BrO, but as insignificant sinks for NO2 (McElroy and Salawitch,
37 1989). Therefore, in order for reactions (5-44) to (5-52) to become significant, NO2 must be
38 removed from the reaction sequences or transformed into a less reactive species.
August 1991 5-46 DRAFT-DO NOT QUOTE OR CITE
-------
Several heterogeneous reactions, believed to occur on the ice surfaces of polar
stratospheric clouds (PSCs), have been proposed which act to sequester or remove (by
deposition) reactive NOy. This then initiates the more effective O3 depletion Cl and Cl/Br
cycles. The proposed heterogeneous reactions involve the reactions of C1NO3 and N2O5 with
hydrogen chloride (HC1) and water, presumably in the solid phase, on PSCs particulate
surfaces (Molina et al,, 1987; Tolbert et al., 1987; Tolbert et al., 1988a).
C1NO3 + HC1 - C12 + HNO3 (5-55)
C1NO3 4- H2O -*• HOC1 + HNO3 (5-56)
N2O5 + HC1 -*• C1M)2 + HNO3 (5-57)
N2O5 + H2O •* 2HNO3 . (5-16)
Leu (1988) also suggested recombination of CIO on the surface of PSCs.
CIO + CIO •* C12 + O2 (5-58)
However, there still remains much uncertainty in the mechanism and in the importance of
reaction 5-58 to the heterogeneous reactions associated with stratospheric O3 depletion.
Polar stratospheric clouds are commonly formed at altitudes in the 10 to 20 km range
over the Antarctic, and to a lesser extent over the Arctic, during the respective winter months
(McCormick et al., 1982). Model calculations which have included both heterogeneous
reactions (5-55) to (5-57), and (5-16) and homogeneous mechanisms have simulated the
observed Antarctic O3 depletion patterns reasonably well (McElroy et al., 1986b; Solomon
et al., 1986; Wofsy et al., 1988; Fahey et al., 1989). Additionally, the model calculations of
Douglass and StolarsM (1989) have demonstrated that such heterogeneous reactions may have
a noticeable impact on the O3/NOy/ClX chemistry of the Arctic stratosphere, even though
arctic PSCs are less common and less persistent.
McElroy et al. (1986a) and others have suggested that the HNO3 formed in the above
heterogeneous reactions could condense, along with water, resulting in the formation of PSCs
at temperatures warmer than the ice point of water. The embellished formation of PSCs
would, in turn, enhance the efficiency of the heterogeneous reactions of the polar
August 1991 5.47 DRAFT-DO NOT QUOTE OR CITE
-------
1 stratospheric O3 depletion mechanisms. Laboratory experiments by Hanson and ,
2 Mauersberger (1988) showed the HNO3 trihydrate (HNO3-3H2O) formed at stratospheric
3 conditions condensed approximately 7 K above the ice point.
4 Recent investigations have suggested there is still considerable uncertainty in the.precise
5 mechanisms of the theorized polar stratospheric heterogeneous chemistry. Wolff et al. (1989)
6 presented data which contradicted the findings of earlier investigators (Molina et al., 1987;
7 Wofsy et al., 1988) pertaining to the incorporation and movement of HC1 within ice crystals.
8 Wolff et al. (1989) found HC1 is not readily incorporated into ice crystals, but rather strongly
9 partitioned along grain boundaries. They proposed that HC1 may be present in some form
10 other than solid (i.e., liquid surface film, grain boundary liquid, chemisorbed to the ice
11 surface, reactant on super-cooled droplet), thereby supplying the HC1 as required by current
12 theories.
13 Other mechanisms for the release of active chlorine from reservoir species, which would
14 initiate the homogeneous O3 depletion cycle, have been presented by Tolbert et al. (1988b)
15 and Finlayson-Pitts et al. (1989). Laboratory studies by Tolbert et al. (1988b) suggests that
16 the heterogeneous reactions suspected to occur on PSCs may also occur on atmospheric
17 sulfuric acid aerosols. This would have the effect of extending the latitude and lowering the
18 altitude at which the proposed heterogeneous reactions could occur. Finlayson-Pitts et al.
19 (1989) found that C1NO3 and N2O5 react with sodium chloride (NaCl) particles at 298 K
20 similarly to the previously discussed polar stratospheric reactions.
21
22 C1NO3 + NaCl(s) •* C12 + NaNO3(s) (5-59)
23
24 N2O5 + NaCl(s) -* C1NO2 + NaNO3(s) (5-60)
25
26
27 These additional mechanisms, which result in photochemically active chlorine (O3 depletion)
28 species, indicate that reactions similar to the stratospheric heterogeneous reactions may also
29 impact the tropospheric chemistry. Additional research may also result in the inclusion of
30 these mechanisms in polar, and possibly global, stratospheric chemistry.
31 •.•.-.--.
32
33
August 1991 5-48 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.7 DEPOSITION OF NITROGEN OXIDES
2 Nitrogen oxides and related nitrogen containing species can be dry and/or wet
3 deposited. Dry deposition consists of transfer of a gaseous species from the atmosphere to an
4 underlying surface where the gas is chemically or biologically assimilated. Wet deposition
5 requires incorporation of NOX into cloud and/or precipitation droplets or ice crystals followed
5 by delivery to the earth's surface. Wet deposition rates are highly variable since they depend
7 on atmospheric advection and mixing processes, storm dynamics, atmospheric chemical
3 transformations and physicochemical processes in the cloud environment.
3 .
3 5.7.1 Dry Deposition of Nitrogen Oxides
I The general procedure for calculating dry deposition fluxes is to multiply the deposition
I velocity (Vd) for a particular trace gas by its air concentration (C) at some reference height
i above the surface.
t F = VdC
5
5 Deposition velocity is an experimentally determined parameter that depends on meteorological
i conditions and surface and trace gas characteristics. In order to apply the Vd to a diversity of
>
^ conditions, the Vd can be broken down into the reciprocal sum of three individual resistances.
) . Vd = 1/Ra + 1/Rb + 1/Rc
)
Ra is the aerodynamic resistance related to the atmospheric turbulence above the surface. Rb
1 accounts for resistance associated with the thin boundary layer of viscous flow that exists very
i close to the surface, and Re defines the sink capacity of the surface itself (surface uptake
• resistance).
i The Ra is a function of a number of physical and meteorological parameters, including
i friction velocity, height above the surface, surface roughness length, and turbulence class. In
' most cases, Ra cannot be calculated directly, and approximations must be invoked. Various
! formulations have been described by Sinclair et al. (1976) and Hicks et al. (1987).
1 Molecular diffusion becomes the primary transport pathway in the thin layers close to
1 the depositional surface. Therefore, Rb is a function of the molecular diffusivity of the
depositing gas. As with Ra, it is generally difficult to precisely define Rb, because surface
roughness is highly variable and, consequently, difficult to parameterize in deposition models.
August 1991 5-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 For many gases, the Re is the most difficult to evaluate. The surface uptake resistance
2 is dependent on the sink capacity of the depositional surface, which is a function of numerous
3 physical, chemical, and biological processes. For example, transport through stomatal
4 openings on leaf surfaces is a function of solar radiation, leaf temperature, leaf water
5 potential, etc.
6
7 5.7.2 Methods For Determining Vd
8 Deposition velocities are defined as the ratio of surface fluxes to air concentrations at
9 some height above the surface (e.g., 2 m). To determine the Vd, the vertical flux (rate of
10 transport per unit area) and fluctuations of a trace gas must be measured. The most common
11 methods employed include eddy correlation, vertical gradients, and enclosure-based systems.
12
13 5.7.2.1 Eddy Correlation
14 TMs is the most direct method for measuring vertical fluxes. It requires fast response
15 and sensitive detection of trace gases, along with the simultaneous measurement of vertical
16 velocity. Vertical velocity is generally determined with a, sonic anemometer. The limiting
17 feature associated with the eddy correlation technique is the availability of fast response
18 chemical sensors. Of the nitrogen containing trace gases, only NO and NO2 have detection
19 systems of sufficient speed to utilize the eddy correlation method.
20 The fast response limitation is eliminated in a variation of eddy correlation known as
21 eddy accumulation. Vertical air movement is still monitored with a sonic anemometer; but
22 instead of having a co-located fast response chemical sensor, a pump is used to fill two air
23 sampling devices. Air is pumped into one of the containers when movement is upward (as
24 sensed by the sonic anemometer) and into the second container when air is subsiding. The
25 mass of the trace gas of interest in each container is then determined using conventional
26 analytical techniques. As long as sample integrity is maintained in the collection device,
27 eddy accumulation could be used to measure the flux of such nitrogenous gases as PAN and
28 PPN, organic nitrates and, possibly, HNO3.
29
30
August 1991 5-50 DRAFT-DO NOT QUOTE OR CITE
-------
5.7.2.2 Vertical Gradient Methods
These methods involve the measurement of trace gas concentrations at several levels
above the surface. The vertical concentration profile is proportional to the flux of the trace
gas of interest.
F = Kz(dc/dz)
The proportionality constant (K^ is normally estimated from concurrently measured
meteorological parameters or by assuming that it is the same for another quantity as for the
trace gas of interest. For example, sensible heat flux can be measured by eddy correlation.
This allows calculation of the transfer (diffusion) coefficient (ly which can then be
combined with concurrent vertical profile measurements of a trace gas to determine its flux.
The main limitation of this latter procedure is the requirement that sources and sinks of the
trace gas and heat flux are uniformly (equally) distributed.
5.7.2.3 Chamber Methods
A volume of air is enclosed above a depositional surface (soil, water, etc.) by one of
two types of chambers. In closed circulation chambers, trace gas fluxes are determined by
periodically collecting samples from the chamber and calculating the change in concentration
with time. In the closed or static mode, the chamber remains over the surface of interest
only long enough to make the measurement. By contrast, dynamic or open chambers
generally are kept in place for several hours or even days. In this latter case, a continuous
flow of air passes through the enclosure. When chamber methods are utilized for"
determining NOX depositional fluxes, special care must be exercised to eliminate the
following problems (Mosier, 1989).
* Uptake or production of NO and NO2 by chamber wall material-
corrections must be made when fluxes of NO and NO2 are very small.
* Changes in aerodynamic mixing close to the surface—air flow through
the chamber should mimic the natural environment as closely as possible.
• If reaction times for species of interest are similar to the enclosure
residence time, corrections must be made. This is generally true for
open chambers where NO, NO2, and O3 are present.
August 1991 5-51 DRAFT-DO NOT QUOTE OR CITE
-------
1 • For NO, the net exchange above the surface has been shown to be
2 dependent on the concentration. Thus, a "compensation point" exists
3 above which there will be uptake of NO and below which NO emission
4 will occur. Therefore, in order to get representative information, NO
5 concentrations in chamber air must be as close as possible to those in
6 ambient air.
7
8
9 5.7.3 NOX Deposition
10 Nitrogen oxide dry deposition fluxes are still very uncertain. Results from many
11 micrometeorological studies exhibit a large scatter in NOX fluxes. This is due both to
12 analytical problems (i.e., interference from other nitrogen containing species) and the fact
13 that NOX exchange includes simultaneous emission and deposition. It is generally believed
14 that NO emission exceeds NO deposition and that NO2 deposition is greater than NO
15 deposition. Deposition velocities for NO have been reported to range from less than 0.1 to
16 approximately 0.2 cm s"1. For NO2, reported Vd values generally fall in the range 0.3 to
17 0.8 cm s"1. Since NO normally constitutes a small fraction (—10%) of the atmospheric NOX
18 concentration, the dry deposition of NO to terrestrial surfaces can be neglected as a sink for
19 removal of atmospheric NOX.
20
21 5.7.4 HNO3 Deposition
22 The surface uptake resistance for HNO3 deposition to terrestrial surfaces has been
23 shown to be, very small (Huebert and Robert, 1985). Therefore, when aerodynamic and
24 diffusional processes bring gaseous HNO3 in contact with a surface, the HNO3 molecules will
25 deposit at nearly 100% efficiency. Consequently, Vd values are larger than those for NO and
26 NO2. Terrestrial Vd values reportedly range between 0.5 and 3.0 cm s"1. Nitric acid Vd
27 values over water surfaces are somewhat lower, falling in the 0.3 to 0.7 cm s"1 range.
28
29 5.7.5 PAN Deposition
30 Very little information exists concerning PAN deposition rates. A Vd of 0.25 cm s"1
31 has been reported over a grass and soil surface (Garland and Penkett, 1976). In a study of
32 the photochemistry of biogenic emissions in the Amazon Basin, Jacob and Wofsy (1988)
33 assigned PAN a Vd value of 2 cm s"1. A high value was selected because of the large
August 1991 5-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 surface area associated with the tropical vegetation canopy. It is expected that the deposition
2 velocity over water surfaces would constitute the other extreme with values as low as
3 0.01 cm s"1 proposed (Andreae et al., 1989). Deposition velocities for PAN will probably
4 remain uncertain due to the difficulties associated with making accurate PAN measurements.
5
6 5.7.6 Wet Deposition of Nitrogen Oxides
7 Wet deposition is not a significant atmospheric removal mechanism for NO and NO2.
3 These two gases are minimally soluble in water and, therefore, must be transformed to more
? highly oxidized forms for wet removal to become effective. The reaction of NO2 with OH
3 radical to produce HNO3 appears to be the main source of the nitrate ion that is measured in
1 precipitation. It is estimated that about one third of the NOX emissions in the United States
I are removed by wet deposition processes (Hicks et al., 1990).
3
t
5 5.8 SUMMARY AND CONCLUSIONS
5 Nitrogen oxides are important chemical species in the planetary boundary layer, as well
7 as in the free troposphere and the stratosphere. Nitrogen oxides play important roles (1) in
3 the control of concentrations of radicals in the clean troposphere; (2) in the production of
) tropospheric O3; and (3) directly or indirectly, in the production and deposition of acidic
) species.
I
I 5.8.1 Ozone Production
5 Combustion processes emit a variety of nitrogen compounds, but chiefly NO, which is
I- rapidly oxidized to NO2 in ambient air, primarily by O3. Photolytic decomposition of NO2
5 then leads to regeneration of NO, producing also an excited oxygen atom, O(3P), that reacts
> with molecular oxygen to form O3. In the absence of competing reactions, NO, NO2, and
i O3 reach an equilibrium described by the steady-state equation.
> Competing reactions exist, however, so that free radicals (hydroperoxy, organic peroxy)
> generated from the the oxidative degradation of volatile organic compounds (VOCs) oxidize
) NO to NO2 without destroying O3. Thus, the amount of O3 formed in ambient air is
August 1991 5-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 dependent upon the concentration of NOX present as well as the concentrations and
2 reactivities of VOC species.
3
4 5.8.2 Production of Odd Nitrogen Species
5 Photochemical processes that include the coupled reactions of NOX, oxygen species, and
6 free radicals produce not only O3 but nitrogen-containing products as well. These oxidation
7 products, termed NOy, include HNO3, HO2NO2, HNO2, RC(O)O2NO2, N2O5, and
8 inorganic and organic nitrates.
9 Nitric acid is a major sink for active nitrogen and is a contributor to acidic deposition.
10 It has been estimated to account for roughly one-third of the total acidity deposited in the
11 eastern United States (Calvert, 1983). Potential physical and chemical sinks for HNO3
12 include wet and dry deposition, photolysis, reaction with OH radicals, and neutralization by
13 gaseous ammonia.
14 Peroxyacyl nitrates are formed from the combination of organic peroxy radicals with
15 NO2. Peroxyacetyl nitrate is the most abundant member in the lower troposphere of this
16 homologous series of compounds. It can serve in the troposphere as a temporary reservoir
17 for reactive nitrogen species and can be regionally transported; but it cannot function as a
18 true sink in the lower troposphere because of its thermal instability. In the upper
19 troposphere, where temperatures are colder, the lifetime of PAN is longer but is only about
20 3 mo, since PAN is photolyzed and also reacts with OH radicals.
21 The NO3 radical is a short-lived NOX that is formed in the troposphere primarily by the
22 reaction of NO2 with O3. In daylight, NO3 undergoes rapid photolysis or reaction with NO.
23 After sunset, accumulation of NO3 can occur and is expected to be controlled by the
24 availability of NO2 and O3 plus chemical destruction mechanisms involving the formation of
25 N2O5 and HNO3.
26 Dinitrogen pentoxide, the anhydride of HNO3, is primarily a nighttime constituent of
27 ambient air since it is formed from the reaction of NO3 (itself a nighttime species) and NO2.
28 Dinitrogen pentoxide is thermally unstable, but at the lower temperatures of the upper
29 troposphere it can serve as a temporary reservoir of NO3. In ambient air, N2O5 reacts
30 heterogeneously with water to form HNO3, which in turn is deposited out; thus, N2O5
31 provides an important removal mechanism for HNO3.
August 1991 5-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 Amines, nitrosamines, and nitramines are thought to exist in ambient air but at low
2 concentrations. Both nitrosamines and nitramines have short lifetimes in ambient air because
3 they are photolytically decomposed (nitrosamines) and/or react with OH radicals and O3
4 (nitramines and nitrosamines).
5
6 5.8.3 Transport
7 General Features
8 The transport and dispersion of the various nitrogenous species are dependent on both
9 meteorological and chemical parameters. Adveetion, diffusion, and chemical transformations
3 combine to dictate the atmospheric residence time of a particular trace gas. In turn,
1 atmospheric residence times help determine the geographic extent of transport of given
2 species. Surface emissions are dispersed vertically and horizontally through the atmosphere
3 , by turbulent mixing processes that are dependent to a large extent on the vertical temperature
4- structure and wind speed.
5 As the result of meteorological processes, NOX emitted in the early morning hours in an
S urban area will disperse vertically and move downwind as the day progresses. On sunny
7 summer days, most of the NOX will have been converted to HNO3 and PAN by sunset.
3 Much of the HNO3 is removed by deposition as the air mass is transported, but HNO3 and
} PAN carried in layers aloft (above the nighttime inversion layer but below a higher
) subsidence inversion) can potentially be transported long distances.
I
I Transport of Reactive Nitrogen Species in Urban Plumes
? Studies of the fate of reactive nitrogen species in daytime urban plumes indicate removal
t rates ranging from 0.04 hf* in Los Angeles (Chang et al., 1979), to 0.1 hf1 in Detroit (Kelly,
5 1987), to 0.14 to 0.24 hf1 (for 4 different, nonconsecutive days) in Boston (Spicer, 1982).
5 In the Detroit study, HNO3 accounted for 67 to 84% of the nitrogenous transformation
7 products, but still fell short of predicted HNO3 levels. Removal by incorporation into coarse
I atmospheric aerosol was postulated as a major sink for HNO3 and the cause of the
) discrepancy between measured and predicted levels (Kelly, 1987).
August 1991 5-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 The nighttime chemistry of NOy is poorly understood. Nighttime concentrations of
2 NO3 show a typical pattern of increase until O3 is no longer available, followed by a
3 decrease as NO emissions cannot be oxidized by O3 to NO2 but react instead with the NO3.
4
5 Transport and Chemistry in NOx-Rich Plumes
6 Ozone build up in power plant plumes appears to be the result of mixing of VOCs into
7 the plume as it moves downwind, since the VOC/NOX ratio in these plumes is quite low. An.
8 O3 build up is not found in all power plant plumes, however (e.g., Hegg et al., 1977; Ogren
9 et al., 1977; White, 1977); and the most important factor in the in-plume formation of O3
10 appears to be the availability of reactive VOC in the dilution air.
11 Little information is available on the fate of reactive nitrogen species in NOx-rich
12 plumes. Aircraft measurements have shown little increase in inorganic and paniculate NO3
13 concentrations in power plant plumes (Hegg and Hobbs, 1979). Chamber and modeling
14 studies indicate that in NOx~rieh but VOC-poor plumes the NOX lifetime will be long enough
15 to allow NOX to be incorporated into regional air masses (Spicer et al., 1981).
16
17 Regional Transport
18 Transport of reactive NOX in regional air masses can occur via several mechanisms:
19 (1) mesoscale phenomena, such as mountain-valley wind flow or land-sea breeze circulations
20 (transport for 10's to 100's of km ); (2) synoptic weather systems such as the migratory highs
21 that cross the Eastern United States in the summertime (transport for many 100's of km); and
22 (3) mesoscale phenomena coupled with slow-moving high-pressure systems having weak
23 pressure gradients. In the latter interrelated phenomena, mountain-valley or land-water
24 breezes can govern pollutant transport in the immediate vicinity of sources but the ultimate
25 fate of reactive NOX species will be distribution into the synoptic system.
26 Information remains sparse on NOX species and their concentrations in synoptic
27 transport systems. Calculated fluxes for the northeastern, Atlantic coast area (Luke and
28 Dickerson, 1987; Galloway et al., 1984; Galloway and Whelpdale, 1987) correspond to about
29 25% of the NOX emitted to the atmosphere of eastern North America, using Logan's (1983)
30 emission estimate of 4.5 X 109 g/year.
31
August 1991 5-56 DRAFT-DO NOT QUOTE OR CITE
-------
I 5.8.4 Oxides of Nitrogen and the Greenhouse Effect
I Except for N2O, the reactive nitrogen species comprising the NOX and NOy families in
5 the atmosphere do not absorb infrared radiation and therefore do not contribute directly to
\ radiative "greenhouse" forcing. They can, however, contribute indirectly to greenhouse
5 processes through the photochemical production of 03 in the troposphere. Ozone absorbs
5 infrared radiation about 2,000 times more effectively per mole than CO2 does, but at its
7 present tropospheric levels O3 contributes only about 8% of the total theorized greenhouse
§ effect (Rodhe, 1990).
J Nitrous oxide, which is chemically inert in the troposphere, readily absorbs infrared
3 radiation and is among the more significant non-CO2 greenhouse gases. Absorption of visible
I radiation by NO2 could make this compound a possible source of other climatic influences if
I atmospheric concentrations become sufficiently large (Wuebbles, 1989).
J
\ Nitrous Oxide Greenhouse Contributions
> Nitrous oxide is thought, on a per mole basis, to be 200 times more effective than CO2
5 as an absorber of heat radiation (Wuebbles, 1989; Rodhe, 1990). It is estimated at present
J levels to be responsible for about 4 to 5 % of the theorized greenhouse effect (Hansen et al.,
5 1989; Rodhe, 1990). Assuming a 0.2% per year increase, Levander (1990) predicted that the
) increased atmospheric mixing ratio of N2O after 50 years would result in an even greater
) efficiency (350 times) in infrared radiation absorption compared to expected CO2 levels.
I
I Stratospheric Ozone Depletion by Oxides of Nitrogen
5 In mid-latitudes and lower to mid-stratospheric regions, cyclic reactions initiated by the
\ oxidation of NO by O3 lead to the net destruction of two molecules of O3. Among the
> stratospheric O3-depletion mechanisms that have been proposed, however, are much more
> important reactions involving the dimerization of CIO in the presence of a third body, M, and
7 subsequent sequences in which the monomer is regenerated and two O3 molecules are
5 destroyed (Fancy et al., 1989; Molina and Molina, 1987). McElroy et al. (1986) also
) proposed Cl and Br oxidation cycles as an important stratospheric O3~depletion mechanism.
) In this mechanism, the dimerizaton and regeneration of CIO, coupled with the oxidation of Br
and Cl by O3, are limited by the reaction of CIO and BrO with NO2 to form the unreactive
August 1991 5-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 C1NO3 or BrNO3 (McEIroy et al., 1986a). Reactions of CIO and BrO with NO2 are
2 important sinks for the halogen oxides but are insignificant sinks for NO2 (McEIroy and
3 Salawitch, 1989).
4 Sequestration of reactive NOy by heterogeneous reactions on the ice surfaces of polar
5 PSCs has been proposed as a means of removing NO2. Its removal allows other O3-depleting
6 cycles to proceed (Molina et al., 1987; Tolbert et al., 1987; Tolbert et al., 1988a).
7 Dinitrogen pentoxide has been implicated in these heterogeneous reactions,
8
9 5.8.5 Deposition of Nitrogen Oxides
10 Both wet and dry deposition of NOX and other nitrogen species occur, but wet
11 deposition is not a significant removal mechanisms for NO or NO2, since both gases are
12 minimally soluble in water. Transformation to more highly oxidized forms is necessary for
13 effective wet depositon of NOX; and the reaction of NO2 with the OH radical to form HNO3
14 appears to be the main source of NO3 in precipitation. About one-third of the emissions of
15 NOX in the United States is estimated to be removed by wet deposition (NAPAP, 1990).
16 Dry deposition fluxes for NOX are highly uncertain, mainly because of analytical
17 problems and the simultaneous occurrence of emission and deposition of NOX. Data available
18 indicate, however, that NO emission exceeds NO deposition and that NO2 deposition exceeds
19 NO deposition. Reported Vd values for respective nitrogen species are: < 0.1 to
20 -0.2 cm s'1 for NO; 0.3 to 0.8 cm s'1 for NO2; 0.5 to 3.0 cm s'1 for HNO3 over land and
21 0.3 to 0.7 cm s"1 for HNO3 over water (Huebert and Robert, 1985). The few data that exist
22 show deposition rates for PAN of 0.01 cm s"1 over water (Andreae et al., 1988) and
23 0.25 cm s"1 (Garland and Penkett, 1976) to 2 cm s"1 (Jacob and Wofsy, 1988) over land.
August 1991 5-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Altshuller, A. P. (1986) The role of nitrogen oxides in nonurban ozone formation in the planetary boundary layer
3 over N America, W Europe and adjacent areas of ocean. Atmos. Environ. 20: 245-268. '
4
5 Andreae, M,; Delany, A. C.; Liu, S.; Logan, S.; Steele, L. P.; Westberg, H.; Zika, R. (1989) Key aspects of
6 species related to global biogeochemical cycles. In: Lenschow, D. H.; Hicks, B. B., eds. Global
7 tropospheric chemistry: chemical fluxes in the global atmosphere. Boulder, CO: NCAR.
8
9 Angell, J. K. (1988) An update through 1985 of the variations in global total ozone and north temperate
0 layer-mean ozone. J. Appl. Meteorol. 27: 91-97.
1 • ......
2 Atkinson, R. (1990) Gas-phase tropospheric chemistry of organic compounds: a review. Atmos. Environ. Part A
3 24: 1-41.
4
5 Atkinson, R.; Carter, W, P. L. (1984) Kinetics and mechanisms of the gas-phase reactions of ozone with organic
6 compounds under atmospheric conditions. Chem. Rev. 84: 437-470.
7
8 Atkinson, R.; Lloyd, A. C. (1984) Evaluation of kinetic and mechanistic data for modeling of photochemical
9 smog. J. Phys. Chem. Ref. Data 13: 315-344.
0
1 Atkinson, R.; Pitts, J. N,, Jr.; Aschmann, S. M. (1984) Tropospheric reactions of dimethyl sulfide with NO3 and
2 OH radicals. J. Phys. Chem. 88: 1584-1587.
3
4 Atkinson, R.; Winer, A. M.; Pitts, J. N,, Jr. (1986) Estimation of night-time N2O5 concentrations from ambient
5 NO2 and NO3 radical concentrations and the role of N2OS in night-time chemistry. Atmos. Environ.
6 20: 331-339.
7 .
8 Bandow, H.; Okuda, M.; Akimoto, H. (1980) Mechanism of the gas-phase reactions of C)H6 and NO3 radicals.
9 J. Phys. Chem. 84: 3604-3608.
0
1 Baulch, D. L.; Cox, R. A.; Crutzen, P. J.; Hampson, R. F.; Kerr, J. A.; Troe, J.; Watson, R. T. (1982)
2 Evaluated kinetic and photochemical data for atmospheric chemistry: supplement 1. J. Phys. Chem. Ref.
3 Data 11: 327-496.
4
5 Biermann, H. W.; Wallington, T. J.; Pitts, J. N., Jr.; Winer, A. M. (1988) Measurement of trace nitrogenous
6 species in photochemical smog: implications for night-time atmospheric chemistry of NO3 and HONO.
7 Atmos. Environ.: [submitted].
8
9 Calvert, J. G. (1976) Hydrocarbon involvement in photochemical smog formation in Los Angeles atmosphere.
0 Environ. Sci. Technol. 10: 256-262.
1
2 Calvert, J. G.; Madronich, S. (1987) Theoretical study of the initial products of the atmospheric oxidation of
3 hydrocarbons. J. Geophys. Res. [Atmos.] 92: 2211-2220.
4
5 Chang, T. Y.; Norbeck, J. M.; Weinstock, B. (1979) An estimate of the NOX removal rate in an urban
6 atmosphere. Environ. Sci. Technol. 13: 1534-1537.
7
8 Crutzen, P. J. (1970) The influence of nitrogen oxides on the atmospheric ozone content. Q. J. R. Meteorol.
9 Soc. 96: 320-325.
0
1
August 1991 5-59 DRAFT-DO NOT QUOTE OR CITE
-------
1 Crutzen, P. J. (1988) Tropospheric ozone: an overview. In: Isaksen, I. S. A., ed. Tropospheric ozone - regional
2 and global scale interactions: proceedings of the NATO advanced workshop on regional and global ozone
3 interaction and its environmental consequences; June 1987; Lillehammer, Norway. Dordrecht, The
4 Netherlands: D. Reidel Publishing Company; pp. 3-32. (NATO advanced science institutes studies:
5 series C, mathematical and physical sciences v. 227).
6
7 Davis, D, D.; Smith, G.; Klauber, G. (1974) Trace gas analysis of power plant plumes via aircraft measurement:
8 O3, NOX, and SO2 chemistry. Science (Washington, DC) 186: 733-736.
9
10 Douglass, A. R.; Stolarski, R. S. (1989) Impact of heterogeneous reactions on stratospheric chemistry of the
11 Arctic. Geophy. Res. Lett. 16: 131-134.
12
13 Easter et al. (1978)
14
15 Fahey, D. W.; Hubler, G.; Parrish, D. D.; Williams, E. J.; Norton, R. B.; Ridley, B. A.; Singh, H. B.; Liu,
16 S. C.; Fehsenfeld, F. C. (1986) Reactive nitrogen species in the troposphere: measurements of NO, NO2,
17 HNO3, particulate nitrate, peroxyacetyl nitrate (PAN), O3, and total reactive odd nitrogen (NOy) at Niwot
18 Ridge, Colorado. J. Geophys. Res. [Atmos.] 91: 9781-9793.
19
20 Fahey, D. W.; Murphy, D. M.; Kelly, K. K.; Ko, M. K. W.; Proffitt, M. H.; Eubank, C. S.; Ferry, G. V.;
21 Loewenstein, M.; Chan, K. R, (1989) Measurements of nitric oxide and total reactive nitrogen in the
22 Antarctic stratosphere: observations and chemical implications. J. Geophys. Res. [Atmos.] 94:
23 16665-16681.
24
25 Finlayson-Pitts, B, J.; Pitts, J. N., Jr. (1986) Atmospheric chemistry: fundamentals and experimental techniques.
26 New York, NY: John Wiley & Sons.
27
28 Finlayson-Pitts, B. J.; Ezell, M. J.; Pitts, J. N., Jr. (1989) Formation of chemically active chlorine compounds
29 by reactions of atmospheric NaCl particles with gaseous N2OS and C1ONO2. Nature (London) 337:
30 241-244.
31
32 Fishman, J. (1984) Ozone in the troposphere. In: Whitten, R. C.; Prasad, S., eds. Stratospheric ozone. New
33 York, NY: Van Nostrand Reinhold.
34
35 Galloway, J. N.; Whelpdale, D. M. (1987) WATOX-86 overview and western North Atlantic Ocean S and N
36 atmospheric budgets. Global Biogeochem. Cycles 1: 261-281.
37
38 Gallosvay, J. N.; Whelpdale, D. M.; Wolff, G. T. (1984) The flux of S and N eastward from North America.
39 Atmos. Environ. 18: 2595-2607.
40
41 Garland, J, A.; Penkett, S. A. (1976) Absorption of peroxy acetyl nitrate and ozone by natural surfaces. Atmos.
42 Environ. 10: 1127-1131.
43
44 Grosjean, D.; Van Cauwenberghe, K.; Schmid, J.; Pitts, J. N., Jr. (1978) Formation of nitrosamines and
45 nitramines by photooxidation of amines under simulated atmospheric conditions. In: Proceedings of the
46 fourth joint conference on sensing of environmental pollutants; November 1977; New Orleans, LA.
47 Washington, DC: American Chemical Society; pp. 196-199.
48
49 Hansen, J.; Lacis, A.; Prather, M. (1989) Greenhouse effect of chlorofluorocarbons and other trace gases.
50 J. Geophys. Res. [Atmos.] 94: 16417-16421.
51
52 Hanson and Mauersberger. (1988)
53
August 1991 5-60 DRAFT-DO NOT QUOTE OR CITE
-------
1 Hanst, P. L.; Spence, J. W,; Miller, M. (1977) Atmospheric chemistry of JV-nitroso dimethylamine. Environ,
2 Sci. Technol. 11: 403-405.
3 i '
4 Hao, W. M.; Wofsy, S. C,; McElroy, M. B.; Beer, J. M.; Toqan, M. A. (1987) Sources of atmospheric nitrous
5 oxide from combustion. J. Geophys. Res. [Atmos.] 92: 3098-3104.
6
7 Harris, G. W.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr.; Platt, U.; Perner, D. (1982) Observations of
8 nitrous acid in the Los Angeles atmosphere and implications for predictions of ozone-precursor
9 relationships. Environ. Sci. Technol. 16: 414-419.
LO
11 Hegg, D. A.; Hobbs, P. V. (1979) Some observations of particulate nitrate concentrations in coal-fired power
12 plant plumes. Atmos. Environ. 13: 1715-1716.
13
14 Hegg, D.; Hobbs, P. V.; Radke, L. F.; Harrison, H. (1977) Ozone and nitrogen oxides in power plant plumes.
15 In: Dimitriades, B., ed. International conference on photochemical oxidant pollution and its control:
16 proceedings, v. I; September 1976; Raleigh, NC, Research Triangle Park, NC: U. S. Environmental
17 Protection Agency, Environmental Sciences Research Laboratory; pp. 173-183; EPA report no.
18 EPA-600/3-77-001a. Available from: NTIS, Springfield, VA; PB-264232.
19
10 Heikes, B. G.; Thompson, A. M. (1983) Effects of heterogeneous processes on NO3, HONO, and HNO3
II chemistry in the troposphere. J. Geophys. Res, C: Oceans Atmos. 88: 10883-10895.
12 • '
13 Hicks, B. B.; Baldocchi, D. D.; Meyers, T. P.; Hosker, R. P., Jr.; Matt, D. R. (1987) A preliminary multiple
14 resistance routine for deriving dry deposition velocities from measured quantities. Water Air Soil Pollut.
15 36: 311-330.
>6
17 Hicks, B. B.; Draxler, R. R.; Albritton, D. L.; Fehsenfeld, F. C.; Dodge, M,; Schwartz, S. E,; Tanner, R. L.;
>8 Hales, J. M.; Meyers, T. P.; Vong, R. J.; Davidson, C. L; Wesely, M. L.; Bates, T. S.; Lindberg,
19 S. E.; Liu, S. C.; Gillette, D. A.; Westberg, H. H.; Lamb, B.; Sigmon, J. T.; Mueller, S. F. (1990)
50 NAPAP report 2: atmospheric processes research and process model development. In: Acidic deposition:
51 state of science and technology, v. I, emissions, atmospheric processes and deposition. Washington, DC:
52 National Acid Precipitation Assessment Program; state of science and technology report no. 2.
53 •
54 Holzworth, G. C. (1967) Mixing depths, wind speeds and air pollution potential for selected locations hi the
55 United States. J. Appl. Meteorol. 6: 1039-1044.
56
17 Huebert, B. J.; Robert, C. H. (1985) The dry deposition of nitric acid to grass, J, Geophys. Res. [Atmos.]
58 90: 2085-2090.
59
tO Hunt, W. F.; Curran, T. C.; Faoro, R. B.; Frank, N. H.; Freas, W. (1986) National air quality and emissions
H trends report, 1984. Research Triangle Park, NC: U. S. Environmental Protection Agency, Office of Air
\1 Quality Planning and Standards; report no. EPA/450/4-86/001. Available from: NTIS, Springfield, VA;
t3 PB86-192044.
W
15 Jackman, C. H.; Frederick, J. E.; Stolarski, R. S. (1980) Production of odd nitrogen in the stratosphere and
1-6 mesosphere: an intercomparison of source strengths. J. Geophys. Res. C: Oceans Atmos. 85: 7495-7505.
\1
18 Jacob, D. J,; Wofsy, S. C. (1988) Photochemistry of biogenic emissions over the Amazon forest. J. Geophys.
[9 Res. [Atmos.] 93: 1477-1486.
JO
)1 Johnston, H. (1971) Reduction of stratospheric ozone by nitrogen oxide catalysts from supersonic transport
)2 exhaust. Science (Washington, DC) 173: 517-522.
53
August 1991 5-61 DRAFT-DO NOT QUOTE OR CITE
-------
1 Johnston, H. S. (1982) Odd nitrogen processes. In: Bower, F. A.; Ward, R. B., eds. Stratospheric ozone and
2 man: v. I. Boca Raton, FL: CRC Press, Inc.; pp. 87-140.
3 ,
4 Jones, C. L.; Seinfeld, J. H. (1983) The oxidation of NO2 to nitrate-day and night, .Atmos. Environ.
5 17: 2370-2373.
6
7 Kelly, N. A. (1987) The photochemical formation and fate of nitric acid in the metropolitan Detroit area:
8 ambient, captive-air irradiation and modeling results. Atmos. Environ. 21: 2163-2177.
9
10 Kelly, N. A,; Wolff, G. T.; Ferman, M. A. (1984) Sources and sinks of ozoae in rural areas. Atmos. Environ.
11 18: 1251-1266.
12
13 Kinnison, D. E.; Wuebbles, D. J. (1989) Preventing depletion of stratospheric ozone: implications of future
14 aircraft emissions - revision 1, UCRL-99926-Rev 1. Available from: NTIS, Springfield, VA;
15 DE89-013779/WEP.
16
17 Kinnison, D, E.; Wuebbles, D. J.; Johnston, H. S. (1988) A study of the sensitivity of stratospheric ozone to
18 hypersonic aircraft emissions. UCRL-88314. Available from: NTIS, Springfield, VA; DE89-002957.
19
20 Ko, M. K. W.; McElroy, M. B.; Weisenstein, D. K.; Sze, N. D. (1986) Lightning: a possible source of
21 stratospheric odd nitrogen. J. Geophys. Res. [Atmos.] 91: 5395-5404.
22
23 Lacis, A. A,; Wuebbles, D. J.; Logan, J. A. (1990) Radiative forcing of climate by changes in the vertical
24 distribution of ozone. J. Geophys. Res. [Atmos.] 95: 9971-9981.
25
26 Leu, M.-T. (1988) Heterogeneous reactions of N2Os with H2O and HC1 on ice surfaces: implications for
27 Antarctic ozone depletion. Geophys. Res. Lett. 15: 851-854.
28
29 Levander. (1990)
30
31 Liu, S. C.; Kley, D.; McFarland, M.; Mailman, J. D.; Levy, H., n. (1980) On the origin of tropospheric
32 ozone. J. Geophys. Res. C: Oceans Atmos. 85: 7546-7552.
33
34 Liu, S. C.; Trainer, M.; Fehsenfeld, F, C.; Parrish, D. D.; Williams, E. J.; Fahey, D. W.; Huebler, G.;
35 Murphy, P. C. (1987) Ozone production in the rural troposphere and the implications for regional and
36 global ozone distributions. J. Geophys. Res. [Atmos.] 92: 4191-4207.
37
38 Logan, J, A. (1983) Nitrogen oxides in the troposphere: global and regional budgets. J. Geophys. Res.
39 C: Oceans Atmos. 88: 10785-10807.
40
41 Logan, J, A. (1985) Tropospheric ozone: seasonal behavior, trends, and anthropogenic influence. J. Geophys.
42 Res. [Atmos.] 90: 10463-10482.
43
44 Luke, W. T.; Dickerson, R. R. (1987) The flux of reactive nitrogen compounds from eastern North America to
45 the western Atlantic Ocean. Global Biogeochem. Cycles 1: 329-343.
46
47 Luke, W. T.; Heubert, B. (1987) EOS.
48
49 Lyons, W. A.; Cole, H. S. (1976) Photochemical oxidant transport: mesoscale lake breeze and synoptic-scale
50 aspects. J. Appl. Meteorol. 15: 733-743.
51
52 Machta, L. (1983) Effects of non-CO2 greenhouse gases. In: Changing climate: report of the carbon dioxide
53 assessment committee. Washington, DC: National Academy Press; pp. 285-291.
54
August 1991 5-62 DRAFT-DO NOT QUOTE OR CITE
-------
1 Malko, M. W.; Troe, J, (1982) Analysis of the unimolecular reaction N2O5 + M = NO2 + NO3 + M. Int. J.
2 Chem. Kinet. 14: 399-416.
3
4 McCormiek, M. P.; Steele, H. M.; Hamill, P.; Chu, W. P.; Swissler, T. J. (1982) Polar stratospheric cloud
5 sightings by SAM II. J. Atmos. Sci. 39: 1387-1397.
6
7 McElroy, M. B. (1980) Sources and sinks for nitrous oxide. Washington, DC: U. S. Department of
8 Transportation; report no. FAA-EE-80-20.
9 '
0 McElroy, M. B.; Salawitch, R. J. (1989) Stratospheric ozone: impact of human activity. Planet. Space Sci.
1 37: 1653-1672.
2 '
3 McElroy, M. B.; Salawitch, R. J.; Wofsy, S. C.; Logan, J. A. (1986a) Reductions of Antarctic ozone due to
4 synergistic interactions of chlorine and bromine. Nature (London) 321; 759-762.
5 '
6 McElroy, M. B.; Salawitch, R. J.; Wofsy, S. C. (1986b) Antaractic O3 chemical mechanisms for the spring
7 decrease. Geophys. Res. Lett. 13: 1296-1299.
8
9 Miller, D. Ft; Alkezweeny, A. J.; Hales, J. M.; Lee, R. N. (1978) Ozone formation related to power plant
0 emissions. Science (Washington, DC) 202: 1186-1188.
1
2 Molina and Molina. (1987)
3
.4 Molina, M. J.; Tso, T.-L.; Molina, L. T.; Wang, F. C.-Y. (1987) Antarctic stratospheric chemistry of chlorine
•5 nitrate, hydrogen chloride, and ice: release of active chlorine. Science (Washington, DC) 238:
,6 1253-1257.
,7 '
.8 Mosier, A. R. (1989) Chamber and isotope techniques. In: Andreae, M. O.; Schimel, D. S., eds. Exchange of
;9 trace gases between terrestrial ecosystems and the atmosphere. New York, NY: John Wiley and Sons.
•o . ,
• I Muzio, L. J.; Kramlich, J. C. (1988) An artifact in the measurement of N2O from combustion sources. Geophys.
>2 Res. Lett. 15: 1369-1372.
13
>4 National Research Council. (1983) Acid deposition: atmospheric processes in eastern North America, a review of
15 current scientific understanding. Washington, DC: National Academy Press.
16
i7 Noxon, J. F.; Norton, R. B.; Marovich, E. (1980) NO3 in the troposphere. Geophys. Res. Lett. 7: 125-128.
18
>9 Ogren, J. A.; Blumenthal, D. L.; Vanderpol, A. H. (1977) Oxidant measurements in western power plant
i-0 plumes. Volume I: technical analysis. Volume II: data. Palo Alto, CA: Electric Power Research Institute;
1-1 report no. EPM EA-421. Volume II available from: NTIS, Springfield, VA; DE82-901024.
12
1-3 Oltmans, S. J.; Komhyr, W. D. (1986) Surface ozone distributions and variations from 1973-1984 measurements
14 at the NOAA geophysical monitoring for climatic change baseline observatories. J. Geophys. Res.
£ [Atmos.] 91: 5229-5236.
«
1-7 Parrish, D. D.; Fahey, D. W.; Williams, E. J.; Liu, S. C.; Trainer, M.; Murphy, P. C.; Albritton, D. L.;
t8 Fehsenfeld, F. C. (1986) Background ozone and anthropogenic ozone enhancement at Niwot Ridge,
t9 Colorado. J. Atmos. Chem. 4: 63-80.
>0
51 Penkett, S. A. (1991) Changing ozone: evidence for a perturbed atmosphere. Environ. Sci. Technol.
)2 25: 631-635.
53
August 1991 5-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 Pemer, D.; Greenberg, J.; Noxon, J. F.; Sefameltekopf, A.; Winkler, P.; Zimmennan, P. (1991) The interaction
2 of natural hydrocarbons and NO3 above the canopy. J. Atmos, Chem.: in press.
3
4 Peters and CannichaeL (1983)
5
6 Pitts, J. N., Jr. (1987) Nitration of gaseous polycyclic aromatic hydrocarbons in simulated and ambient urban
7 atmospheres: a source of mutagenic nitroarenes. Atmos. Environ. 21: 2531-2547.
8
9 Pitts, J. N., Jr.; Grosjean, D.; Van Cauwenberghe, K.; Schmid, J. P.; Fitz, D. R. (1978) Photooxidation of
10 aliphatic amines under simulated atmospheric conditions: formation of nitrosamines, nitramines, amides,
11 and photochemical oxidant. Environ. Sci. Technol. 12: 946-953.
12
13 Pitts, J. N,, Jr.; Winer, A. M.; Aschmann, S. M.; Carter, W. P. L.; Atkinson, R. (1985) Experimental protocol
14 for determining hydroxyl radical reaction rate constants: estimation of atmospheric reactivity. Research
15 Triangle Park, NC: U. S. Environmental Protection Agency; EPA report no. EPA-600/3-85-000.
16
17 Platt, U.; Pemer, D.; Schroeder, J.; Kessler, C.; Toennissen, A. (1981) The diurnal variation of NO3.
18 J. Geophys. Res. C: Oceans Atmos. 86: 11965-11970.
19
20 Rodhe, H. (1990) A comparison of the contribution of various gases to the greenhouse effect. Science
21 (Washington, DC) 248: 1217-1219.
22
23 Russell, A. G.; McRae, G. J.; Cass, G. R. (1985) The dynamics of nitric acid production and the fate of
24 nitrogen oxides. Atmos. Environ. 19: 893-903.
25
26 Sinclair et al. (1976)
27
28 Singh, H. B. (1987) Reactive nitrogen in the troposphere: chemistry and transport of NOX and PAN. Environ.
29 Sci. Technol. 21: 320-327.
30
31 Singh, H. B.; Hanst, P. L. (1981) Peroxyacetyl nitrate (PAN) in the unpolluted atmosphere: an important
32 reservoir for nitrogen oxides. Geophys. Res. Lett. 8: 941-944.
33
34 Singh, H. B.; SaJas, L. J.; Ridley, B. A.; Shelter, J. D.; Donahue, N. M.; Fehsenfeld, F. C.; Fahey, D. W.;
35 Parrish, D. D.; Williams, E. J.; Liu, S. C.; Hubler, G.; Murphy, P. C. (1985) Relationship between
36 peroxyacetyl nitrate and nitrogen oxides in the clean troposphere. Nature (London) 318: 347-349.
37
38 Solomon, S.; Garcia, R. R.; Rowland, F. S.; Wuebbles, D. J. (1986) On the depletion of Antarctic ozone.
39 Nature (London) 321: 755-758.
40
41 Spicer, C. W. (1982) The distribution of oxidized nitrogen in urban air. Sci. Total Environ. 24: 183-192.
42
43 Spicer, C. W. (1982) Nitrogen oxide reactions in the urban plume of Boston. Science (Washington, DC)
44 215: 1095-1097.
45
46 Spicer, C. W.; Sverdrup, G. M. (1981) Trace nitrogen chemistry during the Philadelphia oxidant data
47 enhancement study (1979). Research Triangle Park, NC: U. S. Environmental Protection Agency, Office
48 of Air Quality Planning and Standards; contract no. 68-02-3338.
49
50 Spicer, C. W.; Joseph, D. W.; Stickel, P. R.; Sverdrup, G. M.; Ward, G. F. (n.d.) Reactions and transport of
51 nitrogen oxides hi the Los Angeles atmosphere. U. S. Environmental Protection Agency; contract no.
52 68-02-2439.
53
August 1991 5-64 DRAFT-DO NOT QUOTE OR CITE
-------
1 Spicer, C. W.; Joseph, D. W.; Ward, G. F. (1978) Investigations of nitrogen oxides within the plume of an
2 isolated city. New York, NY: Coordinating Research Council, Inc.; report no.
3 CRC-APRAC-CAPA-9-77. Available from: NTIS, Springfield, VA; PB-290107.
t
5 Spicer, C. W.; Joseph, D. W.; Sticksel, P. R.; Ward, G. F. (1979) Ozone sources and transport in the
5 northeastern United States. Environ. Sci. Technol. 13: 975-985.
7
3 Spicer, C. W.; Sverdrup, G. M.; Kuhlman, M. R. (1981) Smog chamber studies of NOX chemistry in power
) plant plumes. Atmos. Environ. 15: 2353-2365.
)
I Stockwell, W. R.; Calvert, J. G. (1983) The mechanism of NO3 and HONO formation in the nighttime chemistry
1 of the urban atmosphere. J. Geophys. Res. C: Oceans Atmos. 88: 6673-6682.
*
\ Thiemens, M. H.; Trogler, W. C. (1991) Nylon production: an unknown source of atmospheric nitrous oxide.
5 Science (Washington, DC) 251: 932-934.
r
7 Tolbert, M. A.; Rossi, M. J.; Malhotra, R.; Golden, D. M. (1987) Reaction of chlorine nitrate with hydrogen
3 chloride and water at Antarctic stratospheric temperatures. Science (Washington, DC) 238: 1258-1260.
)
) Tolbert, M. A.; Rossi, M. J.; Golden, D. M. (1988a) Antarctic ozone depletion chemistry: reactions of N2O5
I with H2O and HC1 on ice surfaces. Science (Washington, DC) 240: 1018-1021.
I
5 Tolbert, M. A.; Rossi, M. J.; Golden, D. M. (1988b) Heterogeneous interactions of chlorine nitrate, hydrogen
t chloride, and nitric acid with sulfuric acid surfaces at stratospheric temperatures. Geophys. Res. Lett.
5 15: 847-850.
5
7 Trainer, M.; Williams, E. J.; Parrish, D. D.; Buhr, M. P.; Allwine, E. J.; Westberg, H. H.; Fehsenfeld, F. C.;
3 Liu, S. C. (1987) Models and observations of the impact of natural hydrocarbons on rural ozone. Nature
) (London) 329: 705-707.
)
1 Tuazon, E. C.; Atkinson, R.; Plum, C. N.; Winer, A. M.; Pitts, J. N., Jr. (1983) The reaction of gas phase
I N2O5 with water vapor. Geophys. Res. Lett. 10: 953-956.
3
t Tuazon, E. C.; Carter, W. P. L.; Atkinson, R.; Winer, A. M.; Pitts, J. N., Jr. (1984) Atmospheric reactions of
5 TV-nitrosodimethylamine and dimethylnitramine. Environ. Sci. Technol. 18: 49-54.
5
7 U. S. Environmental Protection Agency. (1976) Assessment of scientific information on nitrosamines: report of
3 an ad hoc study group of the Science Advisory Board Executive Committee.
)
3 U. S. Environmental Protection Agency. (1982) Air quality criteria for oxides of nitrogen. Research Triangle
1 Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
I Office; EPA report no. EPA-600/8-82-026. Available from: NTIS, Springfield, VA; PB83-131011.
3
\ Vukovich, F. M.; Bach, W. D., Jr.; Crissman, B. W.; King, W. J. (1977) On the relationship between high
5 ozone in the rural surface layer and high pressure systems. Atmos. Environ. 11: 967-983.
5
7 Wayne, R. P.; Barnes, I.; Biggs, P.; Burrows, J. P.; Canosa-Mas, C. E.; Hjorth, J.; Le Bras, G.; Moortgat,
3 G. K.; Perner, D.; Poulet, G.; Restelli, G.; Sidebottom, H. (1991) The nitrate radical: physics,
? chemistry, and the atmosphere. Atmos. Environ. Part A 25A: 1-203.
3
1 Westberg, H.; Sexton, K.; Roberts, E. (1981) Transport of pollutants along the western shore of Lake Michigan.
I 3. Air Pollut. Control Assoc. 31: 385-388.
3
August 1991 . 5-65 DRAFT-DO NOT QUOTE OR CITE
-------
1 White, W. H. (1977) NO^-Oj photochemistry in power plant plumes: comparison of theory with observation.
2 Environ. Sci. Technol. 11: 995-1000.
3
4 White, W, H.; Dietz, D. (1984) Does the photochemistry of the troposphere admit more than one steady state?
5 Nature (London) 309: 242-244.
6
7 Winer, A. M.; Atkinson, R.; Pitts, J. N., Jr. (1984) Gaseous nitrate radical: possible nighttime atmospheric sink
8 for biogenic organic compounds. Science (Washington, DC) 224: 156-159,
9
10 Winer, A. M.; Atkinson, R.; Arey, J,; Biermann, H. W.; Harger, W. P.; Tuazon, E. C.; Zielinska, B. (1987)
11 The role of nitrogenous pollutants in the formation of atmospheric mutagens and acid deposition.
12 Sacramento, CA: California Air Resources Board; report no. ARB-R-87/308. Available from: NTIS,
13 Springfield, VA; PB87-222949.
14
15 Wofsy, S. C.; Molina, M. J.; Salawitch, R. J.; Fox, L. E.; McEIroy, M. B. (1988) Interactions between HC1,
16 NO,, and H2O ice in the Antarctic, stratosphere: implications for ozone. J. Geophys. Res. [Atmos.]
17 93: 2442-2450.
18
19 Wolff, E. W.; Mulvaney, R.; Gates, K. (1989) Diffusion and location of hydrochloric acid in ice: implications
20 for polar stratospheric clouds and ozone depletion. Geophys. Res. Lett. 16: 487-490.
21
22 Wuebbles, D. J. (1989) On the mitigation of non CO2 greenhouse gases. UCRL-101523. Available from: NTIS,
23 Springfield, VA; DE89-0166S9.
24
25 Wuebbles, D, J.; Grant, K. E.; Connell, P. S.; Penoaer, J. E, (1989) The role of atmospheric chemistry in
26 climate change. JAPCA 39: 22-28.
27
August 1991 5-66 DRAFT-DO NOT QUOTE OR CITE
-------
i 6. SAMPLING AND ANALYSIS FOR
2 OF NITROGEN AND RELATED SPECIES
3
4
5 6.1 INTRODUCTION
6 This chapter addresses various methods to measure selected airborne nitrogen-containing
7 species. The focus is on methodologies currently available or in general use for in situ
8 monitoring of airborne concentrations in both ambient and indoor environments. Methods for
9 measuring the species of interest at their respective sources are not considered, and remote
10 sensing technologies are mentioned in only a few cases.
11 Although the primary focus in this document is nitrogen dioxide (NO2), other nitrogen-
12 containing species are also considered. This chapter is organized into several sections with
13 each section devoted to a different species. The species under consideration are nitric oxide
14 (NO), nitrogen dioxide (NO^, oxides of nitrogen (NOX), total reactive odd nitrogen (NOy),
15 peroxyacetyl nitrate (PAN) and other organic nitrates, nitric acid (HNO3), nitrous acid
16 (HNO2), dinitrogen pentoxide (N2O5), the nitrate radical (NO3), particulate nitrate (NO3"),
17 and nitrous oxide (N2O).
18 Where possible, discussions of sampling and analysis methods for each species address
19 the pertinent characteristics for each method. These'topics include: method type (i.e., in
20 situ, remote, active, passive), description, status (i.e., concept, laboratory prototype,
21 commercially available), interferences, time resolution, sensitivity, and precision and
22 accuracy. A good overview of many of the currently available methods for measuring
23 nitrogen-containing species is the proceedings of a recent National Aeronautics and Space
24 Administration workshop (NASA, 1983).
25 Methods development usually progresses through several stages: concepts, laboratory
26 prototypes, laboratory evaluations, field tests, field evaluations and comparisons against other
27 "proven" methods, and finally consensus acceptance by the user community. At each stage,
28 modifications may be implemented to improve or resolve weaknesses that have been revealed.
29 This is usually a winnowing process. As a result of limitations discovered during this
30 process, many candidate methods may be abandoned in favor of other methods. At some
31 stage near the end of the process, commercialization may occur. In the current document,
August 1991 6-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 those methods that have successfully progressed to the final stages of development are
2 emphasized.
3
4
5 6.2 NITRIC OXIDE (NO)
6 Although NO2 rather than NO is the primary focus of this document, the most
7 commonly used method of measuring NO2 does not detect the NO2 molecule directly.
8 Instead, the method relies on a chemiluminescent reaction of the NO molecule after NO2 has
9 been converted to NO. Thus, to provide a background for subsequent discussions of
10 measurement methods for NO2 and other nitrogen-containing species, NO rather than NO2 is
11 the first species that is addressed.
12 Airborne concentrations of NO can be determined by various methods. As noted
13 previously, the most commonly used method is chemiluminescence (CLM). Other methods
14 include laser-induced fluorescence (LIP), absorption spectroscopy, ionization spectroscopy,
15 and passive collection with subsequent wet chemical analysis.
16
17 6.2.1 Chemiluminescence
18 Chemiluminescence can be used to detect several airborne nitrogen-containing species
19 (i.e., NO, NO2, HNO2, HNO3, N2O5, PAN, NOX, NOy, NH3, and NO3"). Among these
20 compounds, only NO is detected directly, while the other compounds must be converted in
21 some manner to NO prior to detection.
22 The principle is based on the detection of the light emitted following the reaction of NO
23 with ozone (O3). Excess O3 is added to an air sample containing NO which is passing
24 through a darkened reaction vessel with infrared-reflective walls and a window for viewing
25 by a photomultiplier (PM) tube. The light-emitting species is electronically-excited nitrogen
jtj
26 dioxide, NO2*, a product of the reaction of NO and O3. The NO2 relaxes by photon
27 emission that ranges in wavelength well beyond 600 nm and is centered near 1,200 nm.
28 Light is detected by a red-sensitive PM tube fitted with optical filters to prevent radiation
29 below 600 nm produced by ozonalysis of other materials from interfering. The intensity of
30 the measured light is proportional to the concentration of NO in the air sample, and the
31 concentration can be determined by calibration with atmospheres of known composition.
August 1991 6-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 In applications to detect other airborne nitrogen-containing species, the air sample is
2 preconditioned prior to entering the reaction vessel to convert some or all of these species to
3 NO, and that signal is compared to the signal for an unconditioned sample. The signal from
4 the unconditioned sample represents NO, while that from the preconditioned sample
5 represents the sum of the originally present NO along with the NO resulting from conversion
6 of the other nitrogen species. Signal differencing permits determination of the other nitrogen-
7 containing species. The specificity of the preconditioning process may be controversial and is
8 discussed in subsequent sections.
9 Chemiluminescence is designated by the U. S. Environmental Protection Agency (EPA)
0 as the Reference Method for determining NO2 in ambient air (see Section 6.3). As a result,
1 commercial instruments for measuring NO and NO2 are available. Detection limits of
2 approximately 5 parts per billion (ppb) with response times on the order of minutes are
3 claimed by suppliers. While these performance parameters are adequate for monitoring NO
4 and NO2 in relatively polluted urban and suburban environments, they may be inadequate in
5 less polluted remote areas. Efforts have been reported by several researchers to improve the
6 sensitivity and response of CLM NO measurement technology to permit deployment in
7 remote locations both on ground-based and airborne platforms. Delany et al. (1982),
8 Dickerson et al. (1984), Tanner et al. (1983), and Kelly (1986) reported techniques for
9 modifying commercially available NOX detectors to achieve improved sensitivity and response
0 times. Modifications that can be employed include: (a) operation at a low pressure, high
1 flow rate and increased O3 supply; (b) addition of a prereactor where sample air and O3
2 flows are mixed out of view of the PM tube to obtain a more stable background signal;
3 (c) use of a larger more efficient reaction vessel, with highly reflective walls, that promotes
4 the reaction close to the PM tube; (d) use of pure oxygen as the O3 source; (e) cooling the
5 PM tube to reduce noise in the dark current; and (f) change of the electronics to employ
6 photon counting techniques rather than analog signal processing.
7 Kelly (1986) has provided instructions for application of the first three modifications to
8 the Thermo Electron Model 14-B and the Monitor Labs Model 8840. Post modification
9 detection limits of 0.1 to 0.2 ppb and 90% response times of 5 to 10 s were claimed for these
0 instruments. Dickerson et al. (1984) applied modifications (a), (b), (c), and (e) noted above
August 1991 6-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 to a Thermo Electron Model 14 B/E and report minimum detection limits (MDLs) for NO of
2 10 ppt with an 1/e response time of 20 s.
3 Other workers have developed and built highly sensitive research grade instruments for
4 the CLM determination of NO (Ridley and Hewlett, 1974; Kley and McFarland, 1980; Helas
5 et al., 1981; Drummond et al., 1985; Carroll et al., 1985; Torres, 1985; Kondo et al.,
6 1987). Such devices have been used to measure NO at the earth's surface, from airborne
7 platforms in the troposphere, and from balloon-borne platforms in the stratosphere. These
8 instruments generally employ those features listed above. Minimum detection limits of 5 ppt
9 or less, response times of 2 to 60 s, and accuracy of 10 to 20% have been claimed.
10 Chemiluminescence NO instruments appear to be specific for NO. Water vapor may
11 act to quench excited NO2 efficiently (Matthews et al., 1977; Folsom and Courtney, 1979).
12 Operation at reduced pressure reduces this problem. With a commercial analyzer, a 7%
13 reduction in the NO signal was reported for 81% relative humidity (RH) versus dry air
14 (MacPhee et al., 1976). Recent tests of eight commercial analyzers have not shown a water
15 vapor interference with the NO2 signal (Michie et al., 1983). With various research grade
16 instruments, interference due to varying humidities has been reported to be negligible below
17 20 ppm H2O, increase by less than 10% with RH up to 2.5%, and show no change between
18 2 and 100% RH (Fahey et al., 1985a; Drummond et al., 1985). Using commercial CLM
19 instruments, no or very small (i.e., less than 2%) interferences have been reported for six
20 chlorine-containing species (Joshi and Bufalini, 1978), 14 sulfur-containing species (Sickles
21 and Wright, 1979), seven nitrogen-containing species, and three sulfur-containing species
22 (Grosjean and Harrison, 1985b), Zafiriou and True (1986), however, do report interferences
23 from hydrogen sulfide (H2S) and from gases purged from anoxic waters that may have
24 contained sulfur compounds. Using research grade CLM instruments, Fahey et al. (1985a)
25 found no NO interference for NO2, HNO3, N2O5, and PAN and negligible responses for
26 NH3, HCN, N2O, CH4, nine chlorine-containing and three sulfur-containing compounds.
27 These findings are consistent with those of Drummond et al. (1985) who report no or
28 negligible NO interferences from NO2, HNO3, PAN, HO2NO2, H2O2, C3H6, H2O, and
29 aerosols using a research grade instrument with a humidified O3 source.
August 1991 6-4 DRAFT-DO NOT QUOTE OR CITE
-------
I From an operational perspective, aerosols can accumulate on the glass filter separating
I the reaction chamber from the PM tube causing a reduction in sensitivity (Klapheck and
J Winkler, 1985). Cleaning the filter is reported to restore the original sensitivity.
I Whereas most of the CLM methods discussed above are continuous, CLM has also been
) used to analyze nitrogen species collected in integrated samples. Gallagher et al. (1985) have
5 taken cryosamples (4 K) of whole stratospheric air. Samples were analyzed following
7 desorption in the laboratory for NO and NO2 using modified commercial CLM instruments.
? Braman et al. (1986) have employed a series of hollow denuder tubes coated with chemicals
) chosen to preconcentrate various NOX. The collected nitrogen species are thermally
3 desorbed, and detected as NO with a commercial CLM instrument. The coating materials
I used to preconcentrate the various target species in sequence are: tungstic acid removes
Z HNO3; potassium-iron oxide removes HNO2; copper (I) iodide removes NO2; and cobalt
3 (HI) oxide removes NO. Future field testing is needed to demonstrate the adequacy of this
t method.
5
S 6.2.2 Laser-Induced Fluorescence
7 Laser-induced fluorescence techniques may incorporate single-photon (SP), two-photon
3 (TP), or photofragmentation (PF) schemes. Although SP-LEF has been used for NO
3 (Bradshaw et al., 1982), TP-LIF represents an advancement in the state of the art (Bradshaw
3 et al., 1985) and is discussed below. The TP-LIF detection principle requires that a molecule
1 have more than one bonding excited state and can be sequentially pumped into the highest
2 state. If the lifetime of the excited state is short compared to collisional deactivation, the
3 excited molecule will decay to a more stable state by a fluorescence process. The
4 fluorescence wavelength is shifted relative to that of the pumping wavelengths and thus
5 overcomes noise problems associated with background nonresonant fluorescence. For
6 application to NO, pulsed ultraviolet (UV) and infrared (IR) laser light sources are used.
7 Ground state X n NO is excited to the A22 electronic level using UV of 226-nm wavelength.
8 Then, using IR wavelengths of 1.06 to 1.15 ^m, the molecule is further pulsed to the D22
9 level. The fluorescence resulting from the D22 to X2II transition is monitored at 187 to
0 220 nm. By using long-wavelength blocking filters with solar blind PM tubes, this type of
1 detector discriminates against noise and becomes signal, rather than signal-to-noise, limited.
August 1991 6-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 Photon counting and grated-eharge integrators are used for signal processing. The intensity
2 of the light is related to the concentration of NO in the air sample by calibration with
3 atmospheres of known concentration.
4 Since the TP-LIF instrument is signal limited, the sensitivity is defined by the
5 integration time (e.g., 1 ppt for 5 min and 10 ppt for 30 s) (Davis et al., 1987). Propagation
6 of error analysis has been used to place 90% confidence limits on the accuracy of TP-LIF
7 NO field measurements performed on an aircraft at +16%.
8 The TP-LIF technique is expected to be highly specific, since it has two levels of
9 spectroscopic selectivity. If some trace atmospheric compound were to produce an NO
10 molecule by interaction with the 226-nm beam, then the opportunity for interference exists.
11 The potential interference from HNO3, CH3NO2, C2H5NO2, CH3ONO2, NO2, PAN,
12 HNO2, SO2, and CH3ONO has been evaluated. Only the last compound was found to show
13 potential interference, and arguments have been given to neglect its influence when sampling
14 tropospheric air (Davis et al., 1987).
15
16 6.2.3 Absorption Speetroseopy
17 Absorption Spectroscopy encompasses techniques that measure the change in radiance
18 from a source that occurs as a result of absorption by analyte molecules over a known path
19 length. Several techniques including Fourier Transform Infrared Spectroscopy (FTIR), long-
20 path absorption, and infrared Tunable-Diode Laser Spectroscopy (TOLAS) have been
21 employed for measuring the concentration of various NOX in the atmosphere (National
22 Aeronautics and Space Administration, 1983). Among these techniques, the TOLAS is a
23 well developed technique that has been applied to NO as well as NO2 and HNO3. Similar.
24 sensitivities have been reported for both remote sensing applications using open air path
25 lengths and in situ application using multipass cells (Cassidy and Reid, 1982). The latter
26 configuration has found broader application for ambient measurements of NOX, and the use of
27 the White cell avoids atmospheric turbulence-related errors that can affect open air
28 application. As a result, in situ TOLAS is the primary focus of this section.
29 Tunable-diode laser Spectroscopy employs a tunable diode laser to scan over a narrow
30 wavelength region around a particular absorption line or feature of the gas of interest. High
31 sensitivity is achieved by the high spectral radiance of the diodes and the rapid tunability of
August 1991 6-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 the laser. With rapid scanning back and forth across an absorption line, the absorption
2 appears as an AC signal at twice the tuning frequency and can be sensitively detected by
3 sychronous demodulation. System sensitivities sufficient to measure signal changes of 10~5
4 permit the detection of concentrations of less than 10 ppt with a 1-km path length. For
5 analyte molecules that have resolved absorption spectra that are not coincident with other
6 atmospheric constituents, TDLAS is highly specific. Additional information on the operating
7 principles and hardware for TDLAS is provided by Schiff et al. (1983) and National
8 Aeronautics and Space Administration (1983).
9 For optically thin systems, the Beer Lambert Law suggests that the fraction of the
0 power transmitted through an absorbing medium is proportional to the concentration of the
1 absorbing molecule. However, since the total laser power is not measured, it is usually
2 necessary to calibrate the TDLAS by introducing a known concentration of the target gas and
3 determining the proportionality between signal and concentration.
4 Using a 40-m path length near 1,850 cm"1 the MDL for NO is 0.5 ppb (Schiff et al.
5 1983). At a sampling rate giving a 4 s residence time in the White cell, stable NO signals
6 are achieved in approximately 1 min. Linearity has been demonstrated between the signal
7 and NO concentration at levels between 7 and 175 ppb. Since the NO calibration gas is
8 introduced directly into the sampling line, surface losses are compensated for automatically.
9 As noted previously, the measurement of NO using TDLAS is highly specific.
0 A newly developed method, Two-Tone Frequency Modulated Spectroscopy (TTFMS),
1 has shown great promise in the laboratory for the measurement of NO, NO2, PAN, HNO3,
2 N2O, and other atmospheric trace gases (Hansen, 1989). Two-Tone Frequency Modulated
3 Spectroscopy uses a diode laser light source that is modulated simultaneously at two arbitrary
4 but closely spaced frequencies. The beat tone between these two frequencies is monitored as
5 the laser carrier and associated sidebands are tuned through an absorption line. The method
6 is fast, specific, and extremely sensitive. Using a low pressure (20-torr) multiple-reflection
7 optical cell with a 100-m path length and 1-min signal averaging time, the projected MDL for
8 NO is 4 ppt, and the projected MDLs for NO2, PAN, HNO3, and N2O are lower.
9 Additional development of this laboratory prototype is needed to demonstrate its performance
0 in the field.
1
August 1991 6-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.2.4 Passive Samplers
2 Whereas the previous methods are focused primarily on low NO concentrations
3 representative of ambient air; passive samplers are focused on atmospheres having higher
4 concentrations such as those found indoors or in the workplace. They are used to obtain data
5 at a large number of sites averaged over a long period of time. The Palmes tube is a passive
6 sampler that relies on diffusion of an analyte molecule through a quiescent diffusion path of
7 known length and cross-sectional area to a reactive surface where the molecule is captured by
8 chemical reaction (Palmes et al., 1976). After exposure durations ranging from hours to
9 days, the reactive surface is analyzed and the integrated loading of the reaction product is
10 used to infer the average gas concentration. A quiescent diffusion path is required to ensure
11 that sampling is diffusion controlled and as a result is relatively constant. This permits the
12 average ambient concentration to be related directly to the ratio of the reaction product
13 loading to the exposure time. This proportionality factor is analogous to the reciprocal of a
14 sampling rate where sampling rate is the product of the diffusivity of the analyte gas and the
15 area of the opening through which the molecules diffuse divided by the distance they must
16 travel to be collected.
17 Palmes tubes are fabricated in a range of measurements from tubing (Palmes and
18 Tomczyk, 1979). The dimensions are chosen to provide a ratio of sampling area to diffusion
19 distance of 0.1 and thus ensure diffusion controlled sampling. Reactive grids are secured and
20 sealed at one end of the tubing segment using a plastic cap. The opposite end of the tube is
21 sealed with a similar cap. The capped sampler is stored until the sample is to be collected.
22 A sample is collected by placing the tube in the appropriate location (e.g., for personal
23 sampling the tube may be attached to a worker's lapel), removing the end cap opposite the
24 gridded end with the open end facing down, sampling for the appropriate period, recording
25 the time, recapping the tube, and returning the sampler to the laboratory for analysis.
26 The Palmes tube passive sampler does not measure NO directly. Two tubes are
27 required: one has reactive grids coated with triethanolamine (TEA) to collect NO2. The
28 second tube is similar but has an additional reactive surface coated with chromic acid to
29 convert NO to NO2 which is in turn collected by the TEA coated grids. The NO
30 concentration is determined by subtraction after correction for differences in sampling rates
31 caused by differences in diffusivities of the two molecules. To ensure reliable results, contact
August 1991 6-8 DRAFT-DO NOT QUOTE OR CITE
-------
I between the chromic acid coated surface and the TEA coated grids for longer than 24 h must
I be avoided.
5 Analysis is accomplished by extracting the grids in solution and analyzing the extract for
!• nitrite ion (NO2~). This analysis may be performed by adding a solution of water,
) sulfanilamide, and N-1-naphthylethylene-diamine-dihydrochloride (NEDA) reagent directly
5 into the tube and determining the concentration using a spectrophotometer at 540 nm (Palmes
7 et al., 1976). Increased sensitivities are claimed by analyzing the solution using ion
$ chromatography (1C) with a concentrator column (Miller, 1984). The colorimetric analysis is
) calibrated by dilution of gravimetrically prepared nitrite solutions.
) The sampling rate (i.e., 0.02 cm3 s"1) for NO is reported to be independent of pressure,
I to increase approximately 1% for each 5.5 °C increase in temperature and increase by 3% for
I each 5 cm s"1 increase in wind velocity (Palmes and Tomczyk, 1979). Linear response was
$ found for loadings between 2 and 120 ppm-h. This method was proposed for sampling
1- occupational exposures where the dosage is not to exceed 25 ppm for 8 h (i.e., 200 ppm-h.).
> This method cannot be used for sampling periods longer than 24 h. The reliability of this
j method in the field at both ppb and ppm levels remains to be demonstrated.
7 A badge-type sampler similar to the Palmes tube has been devised by Yanagisawa and
$ Nishimura (1982). Their device uses a series of 12 layers of CrO3-impregnated glass fiber to
) oxidize NO to NO2. The filters also act as a diffusion barrier between the ambient air and a
) TEA coated cellulose fiber filter. Nitric oxide is oxidized to NO2 on the oxidizing filters and
I collected along with NO2 that has diffused from ambient air through the filters to the TEA
I coated collection surface. The TEA coated filter is extracted and analyzed for NO2". Either
$ a colorimetric or 1C analytical finish may be employed., The analytical finish is calibrated by
I dilution of gravimetrically prepared nitrite solutions. An effective proportionality factor (i.e.,
) calibration factor) for the badge is provided by the supplier. This technique is claimed to be
5 more sensitive than the Palmes tube and to have a. lower limit dosage of 0.07 ppm-h.
7
$ 6.2.5 Calibration
) Chemiluminescence, TP-LIF, and TDLAS NO measurement systems all employ
) calibration cylinders containing known,concentrations of NO in nitrogen (N2) at nominal
August 1991 6-9 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations of 1 to 50 ppm (Carroll et al., 1985; Bradshaw et al., 1985; Schiff et al.,
2 1983). Calibrations are performed using dynamic dilution with air.
3 In the calibration procedure for the measurement of NO2 by CLM, the EPA specifies
4 the use of an NO concentration standard (Code of Federal Regulations, 1987a). This
5 standard is a cylinder of compressed gas containing between 50 and 100 ppm NO in nitrogen.
6 The concentration must be traceable according to a certification protocol to a National
7 Institute of Standards and Technology (NIST) (formerly National Bureau of Standards)
8 Standard Reference Material (SRM) or an NIST/EPA-approved commercially available
9 Certified Reference Material (CRM). National Institute of Standards and Technology
10 provides 10 NO SRMs at nominal concentrations between 5 and 3,000 ppm (National Bureau
11 of Standards, 1988). Shores and Smith (1984) have demonstrated that aluminum calibration
12 cylinders containing 10 to 150 ppm NO in nitrogen were stable over time, and for
13 103 cylinders, the average change was less than 1% over an 18-month period. Commercially
14 supplied cylinders from 11 producers containing certified concentrations at nominal values of
15 70 and 400 ppm were evaluated for accuracy (Wright et al., 1987). In all cases, the certified
16 and auditor-measured concentrations were within 5%, and in over two-thirds of the cases the
17 agreement was within 2%.
18 Passive NO samplers do not employ full calibration of sampling and analysis operations
19 (Palmes etal., 1976). Only the analysis portion of the procedure is calibrated. Calibration
20 standards for colorimetric or 1C determination of nitrite are prepared similarly. Dilution of
21 gravimetrically prepared liquid solutions of nitrite is used to produce calibration standards that
22 cover the working range of analysis.
23
24 6.2.6 Intel-comparisons
25 Several intercomparisons of the performance of research grade NO instrumentation have
26 been conducted recently (Walega et al., 1984; Hoell et al., 1985; Hoell et al., 1987;
27 Fehsenfeld et al., 1987). Walega et al. (1984) have reported comparisons of NO
28 measurements made with a highly sensitive CLM instrument and a TDLAS system.
29 Measurements of NO-spiked synthetic air were made in the laboratory and field. In addition,
30 measurements were made of ambient and downtown Los Angeles air. Good agreement was
31 found for all test conditions.
August 1991 6-10 DRAFT-DO NOT QUOTE OR CITE
-------
Tests to compare the performance of several instruments at measuring trace in the
troposphere have been performed as part of NASA's Global Tropospheric Experiment/
Chemical Instrumentation Test and Evaluation 1 and 2 (GTE/CITE 1 and 2). Hoell et al.
(1985 and 1987) and Gregory et al. (1990a) have reported comparisons of NO measurements
made with two highly sensitive CLM instruments and a TP-LIF instrument. The first
intercomparison was a ground-based study performed at Wallops Island, VA. The second
intercomparison was an airborne study comprised of two missions performed on a Convair
CV-990 flown out of California and Hawaii, The third intercomparison involved 13 flights
sampling tropical, nontropical, maritime, and continental air masses at altitudes between
150 and 5,000 m. The two CLM instruments were of similar design (Kley and McFarland,
1980) with the main differences being the injection of water vapor to the airstream entering
the reaction chamber of one instrument to minimize the background variability caused by
changing ambient humidity and to suppress an O3-related background signal. In the first
study, measurements of ambient NO concentrations ranged from 10 to 60 ppt and of NO-
enriched ambient air ranged from 20 to 170 ppt. Agreement among the techniques at the
95% confidence level was ±30%, and no artifact or species interferences were identified. In
the second and third studies, NO concentrations ranged from below 5 ppt to above 100 ppt
with the majority below 20 ppt. At NO concentrations below 20 ppt, measurements agreed
to within stated instrument precision and accuracy (i.e., to within 15 to 20 ppt). Good
correlation was observed between CLM and TP-LIF measurements. The authors concluded
that equally valid measurements of ambient NO can be expected from either instrument.
A field intercomparison of instruments designed to measure NO, NOX, and NOy was
conducted near Boulder, CO (Fehsenfeld et al., 1987). The study was performed to compare
the performance of instruments that employed different approaches to reduce NOX or NOy to
NO prior to detection by CLM. In several tests, both zero air and ambient air spiked with
NO were measured. Excellent agreement was found among the measurements of the three
tested instruments. These results confirm the equivalence of CLM NO detection systems.
August 1991 6-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.2.7 Sampling Considerations for NO and Other Nitrogen-Containing
2 Species
3 Nitric oxide reacts rapidly with O3 to form NO2. In the presence of sunlight, NO2 will
4 photolyze to form NO and atomic oxygen (O) which will combine with atmospheric oxygen
5 (O^ to form O3. Thus, under daylight conditions NO, O3, and NO2 can exist
6 simultaneously in ambient air in a condition known as photostationary state where the rate of
7 photolysis of NO2 is nearly equal to the rate of reaction between NO and O3 to form NO2.
8 The relative amounts of the three species at any time are influenced by the intensity of the
9 sunlight present at that moment. When a sample is drawn into a dark sampling line,
10 photolysis ceases, while NO continues to react with O3 to form NO2. As a result, long
11 residence times in sampling lines must be avoided to insure a representative sample.
12 Sampling requirements for a given error in tolerance were discussed by Butcher and Ruff
13 (1971). Figure 6-1 shows the absolute error in NO2 introduced for a 10 s residence time in a
14 dark sampling line in the presence of NO and O3 at various concentrations.
15 In addition to sampling time considerations, sampling surfaces should be considered.
16 Oxides of nitrogen are in general reactive species. As a result, the most nearly inert
17 materials (i.e., glass and Teflon™) are recommended for use in sampling trains. If water
18 molecules accumulate on sampling train surfaces and influence sample integrity, then species
19 solubility may be one indicator of the susceptibility of a species to surface effects.
20 Solubilities at 25 °C, expressed as Henry's Law Coefficients (M atm"1), for selected nitrogen
21 containing species are: NO, 2 x 10'3; NO2, 1 x 10'2; N2O, 3 x 10'2; PAN, 4; HNO2, 50;
22 NH3, 60; and HNO3, 2 x 105 (Schwartz, 1983). This suggests that of the NOX species, NO
23 may be the least susceptible to surface effects while surface effects may be very important in
24 the sampling of HNO3.
25
26
27 6.3 NITROGEN DIOXIDE (NO2)
28 Among the NOX, NO2 is the only criteria pollutant and the only species to have
29 sampling and analysis methodologies specified by the EPA for determining ambient airborne
30 concentrations. As a result, methods for sampling and analysis of NO2 are emphasized in
31 this document. Airborne concentrations of NO2 can be determined by several methods
August 1991 6-12 DRAFT-DO NOT QUOTE OR CITE
-------
Q.
Q.
DC
H
2
LU
O
o
O
CO
O
0.002 L
0.001
0.001 0.01 0.10
NO CONCENTRATION, ppm
1.0
Figure 6-1. Absolute error in the NO2 (A NO^ for 10 s in the dark sampling line.
Source: Butcher and Ruff (1971).
including CLM, LIF, absorption spectroscopy, and bubbler and passive collection with
subsequent wet chemical analysis.
6.3.1 Chemiluminescence, NO + O3 (CLM)
Instruments discussed in this section sample continuously, employ the CLM reaction of
NO and O3, but do not detect NO2 directly. Instead, they rely on the direct detection of NO,
the conversion of some or all of the NO2 in the air sample to NO, reaction with O3, and the
appropriate signal processing to infer the NO2 concentration. The CLM NO detection
principle and hardware are described in Section 6.2.1. To measure NO2, a CLM NO
detector, a converter, plumbing modifications, and changes in signal processing are required.
August 1991
6-13
DRAFT-DO NOT QUOTE OR CITE
-------
1 Several methods have been employed to reduce NO2 to NO (Kelly, 1986). They
2 include catalytic reduction using heated molybdenum or stainless steel, reaction with CO over
3 a gold catalyst surface, reaction with iron sulfate (FeSO4) at room temperature, reaction with
4 carbon at 200 °C, and photolysis of NO2 to NO at 320 to 400 nm.
5 Since CLM is designated by the EPA as the Reference Method for NO2 in ambient air
6 (Code of Federal Regulations, 1987a), CLM instruments for the determination of NO2 are
7 readily available commercially. As noted previously, these instruments are used to measure
8 both NO and NO2. Nominal detection limits of approximately 5 ppb and response times on
9 the order of minutes are claimed by suppliers. Field evaluation of 9 instruments has shown
10 the MDLs to range from 5 to 13 ppb (Michie et al., 1983; Holland and McElroy, 1986).
11 Recent field and laboratory evaluation of two commercial instruments operated on a 0.00 to
12 0.05 ppm range revealed detection limits of between 0.5 and 1.0 ppb and operating precision
13 estimates of ±0.3 ppb (Rickman etal., 1989). While these performance parameters are :
14 adequate for monitoring NO and NO2 in urban and suburban environments, they may be
15 inadequate in less polluted remote areas. As noted in Section 6.2.1, efforts have been
16 reported by several researchers to improve sensitivity and response of CLM NO measurement
17 technology to permit deployment in remote locations on ground-based and airborne platforms.
18 Since the research grade instruments employed by these workers included NO2-to-NO
19 converters and were designed to measure both NO and NO2, instrument performance for the
20 determination of NO2 is also improved substantially over that of commercially available
21 instruments. Typically reported performance parameters for NO2 response using research
22 grade CLM instruments are MDLs of 10 to 25 ppt, response times of 1 to 100 s and accuracy
23 of 30 to 40% (Helas et al., 1987; Fehsenfeld et al., 1987).
24 Different converters may not be specific for NO2 and may convert several nitrogen-
25 containing compounds to NO, giving rise to artificially high values for NO2. Using
26 commercial instruments, Winer et al. (1974) found over 90% conversion of PAN, ethyl
27 nitrate, and ethyl nitrite to NO with a molybdenum converter and similar responses to PAN
28 and w-propyl nitrate with a carbon converter. With a stainless steel converter at 650 °C,
29 Matthews et al. (1977) reported 100% conversion for NO2, 86% for NH3, 82% for
30 CH3NH2, 68% for HCN, 1% for N2O, and 0% for N2. Using a commercial instrument
31 Joseph and Spicer (1978) found quantitative conversion of HNO3 to NO with a molybdenum
August 1991 6-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 converter at 350 °C. Similar responses to PAN, methyl nitrate, w-propyl nitrate, «-butyl
2 nitrate, and HNO3, substantial response to nitrocresol and no response to PBzN were reported
3 with,a commercial instrument using a molybdenum converter at 450 °C (Grosjean and
4 Harrison, 1985b). These results were confirmed for PAN and HNO3 by Rickman and
5 Wright (1986) using commercial instruments with a molybdenum converter at 375 °C and a
6 carbon converter at 285 °C.
7 Interferences from species that do not contain nitrogen have also been reported. Joshi
3 and Bufalini (1978), using a commercial instrument with a carbon converter, found
3 significant apparent NO2 responses to phosgene, trichloroacetyl chloride, chloroform,
3 chlorine, hydrogen chloride, and photochemical reaction products of a perchloroethylene-NOx
1 mixture. Grosjean and Harrison (1985b) reported substantial responses to photochemical
2 reaction products of C12-NOX and Cl2-methanethiol mixtures and small negative responses to
3 methanethiol, methyl sulfide, and ethyl sulfide. Sickles and Wright (1979) using a
t commercial instrument with a molybdenum converter at 450 °C found small negative
5 responses to 3-methylthiophene, methanethiol, ethanethiol, ethyl sulfide, ethyl disulfide,
S methyl disulfide, hydrogen sulfide, 2,5-dimethylthiophene, methyl sulfide, methyl ethyl
7 sulfide, and negligible responses to thiophene, 2-methylthiophene, carbonyl sulfide, and
3 carbon disulfide.
5 With a research grade instrument, Bollinger et al. (1983) reported that NO2, HNO3,
3 n-propyl nitrate, and N2O5 are reduced to NO by a gold-catalyzed reaction with CO. Fahey
1 et al. (1985a) using a similar instrument with 3,000 ppm CO over a gold converter at 300 °C
I reported conversion efficiencies exceeding 90% for NO2, HNO3, N2O5, and PAN. Although
J at a converter temperature of 300 °C negligible response to HCN and NH3 was found in the
1 presence of water vapor, complete conversion was noted at 700 °C.
5 A room temperature NO2-to-NO converter using FeSO4 has been suggested by Winfield
5 (1977) and adopted in research grade instruments by Helas et al. (1981), Kondo et al. (1983),
7 and Dickerson et al. (1984). A reduction in conversion efficiency has been reported under
3 dry conditions, and conversion of PAN, HNO2, and other nitrogen-containing species to NO
) has been noted (Fehsenfeld et al., 1987). Nonspecificity of the FeSO4 converter has been
) observed by Fehsenfeld et al. (1987) in measuring NO and NO2 in a remote environment.
I At NOX levels below 1 ppb, results from the FeSO4 converter were biased high in the
August 1991 K 6-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 measurement of NO2 (i.e., a factor of two at 0.1 ppb). Airborne measurements of NO2 at
2 concentrations below 0.2 ppb showed high biases of factors of 2 to 3 for a CLM instrument
3 with an FeSO4 converter (Gregory etal., 1990b).
4 In another research grade instrument, Kley and McFarland (1980) used an Xe arc lamp
5 to photolyze a portion of the NO2 in sampled air to NO and determine the NO2 concentration
6 from the increase in NO. A fractional conversion was established using a calibration source.
7 Interferences with the photolytic converter approach (CLM-PC) are expected from HNO2,
8 NO3, HO2NO2, and N2O5, but not from HNO3, n-propyl nitrate, and PAN. A detailed
9 description of the operation (including minimization of interferent decomposition and both
10 homogeneous and heterogeneous oxidation of NO) and performance of a CLM-PC instrument
11 is given by Ridley et al. (1988). An artifact identified with this method is caused by nitrate-
12 containing aerosols deposited on the surface of the photolysis tube which release NO and
13 NO2 upon irradiation (Bellinger et al., 1984). This interference is eliminated by filtering
14 sampled air and periodic cleaning of tube surfaces.
15 The methods discussed above employ CLM detection of NO and are continuous. Other
16 researchers have employed various methods of integrated sampling followed by a CLM
17 instrument for measuring NO and NO2 in the desorbed sample. Gallagher et al. (1985) have
18 used cryosampling of stratospheric whole air samples, and Braman et al. (1986) have used
19 copper (I) iodide coated denuder tubes to sample NO2 in ambient air.
20
21 6.3.2 Chemiluminescence (CLM), Luminol
22 A method for the direct CLM determination of NO2 was reported by Maeda et al.
23 (1980). This method is based on the CLM reaction of gaseous NO2 with a surface wetted
24 with an alkaline solution of luminol (5-amino-2,3-dihydro-l,4-phthalazinedione). The
25 emission is strong at wavelengths between 380 and 520 nm. The intensity of the measured
26 light is proportional to the NO2 concentration in the sampled air, and the concentration can
27 be determined by calibration with atmospheres of known concentration.
28 Since the introduction of the luminol method by Maeda et al. (1980), improvements
29 have been made to develop an instrument suitable for use in the field (Wendel et al., 1983),
30 and additional modifications have been made recently to produce a continuous commercial
31 instrument (Schiff et al., 1986). Detection limits of 5 to 30 ppt and a response time of
August 1991 6-16 DRAFT-DO NOT QUOTE OR CITE
-------
t seconds have been claimed based on laboratory tests (Wendel et al., 1983; Schiff et al.,
I 1986). Recent laboratory evaluation of two instruments has revealed a detection limit (i.e.,
} twice the standard deviation of the clean air response) of 5 ppt and 95 % rise and fall times of
!• 110 and 15 s (Rickman et al., 1988). Field tests of the same instruments have shown an
5 operating precision of ±0.6 ppb.
5 The original method showed no interferences from NO, N2O, NH3, CO,
7 1,2-dichloroethylene, and propylene, but positive interferences from O3 and SO2 and a
I negative interference from CO2 (Maeda et al., 1980). More recently, the luminol solution
> has been reformulated containing water, luminol, NaOH, Na2SO3, and alcohol in proportions
) chosen to enhance the sensitivity and minimize interferences from O3, SO2, and CO2
(Wendel et al., 1983; Schiff et al., 1986). At concentrations below 100 ppb, no interferences
\ were reported for NO, HNO3, NH3, HCN, H2O2, CO, CO2, and SO2. The instrument has
I shown sensitivity to PAN (Wendel et al., 1983; Sickles, 1987) and HNO2 (Rickman et al.,
!• 1989), and nonlinear response to NO2 at concentrations below 10 ppb (Schiff et al., 1986;
i Kelly et al., 1990). The method has shown appreciable sensitivity to an operating
5 temperature which can be resolved by controlling the temperature of the reaction cell or by
' signal processing (Schiff et al., 1986; Bubacz et al., 1987). Recent tests of the CLM
! (luminol) instrument have demonstrated the need to correct results for pressure (as might be
• seen in airborne applications), nonlinearity of response below 3 ppb NO2, interferences from
> O3 and PAN, and the age-dependent sensitivity of the luminol solution (Kelly et al., 1990).
A manufacturer-supplied O3 scrubber designed to eliminate the O3 interference was also
: found to remove appreciable amounts of NO2.
i
6.3,3 Photofragmentation/Two Photon-Laser-Induced Fluorescence
(PF/TP-LIF)
• Two NO2 sensors based on the measurement of the fluorescence of excited NO2 have
been reported by Fincher et al. (1978). One device employs a small high pressure Xe arc
: flash lamp to excite NO2. This device has a sensitivity of 10 ppb for an 80-s (i.e.,
1,024 flash lamp pulses) integration time. The other device uses a 442-nm LIF and a PM
1 tube with photon counting of light above 600 nm. The sensitivity of this device is 1 ppb for
a comparable integration time. A major drawback of these devices for broad ambient
August 1991 6-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 applications is limited sensitivity associated with background signals. This has been
2 overcome with PF/TP-LIF instrumentation.
3 A NO2 sensor incorporating PF/TP-LEF has recently been developed and deployed in
4 the field (Rodgers et al., 1980; Davis, 1988). With this method, NO is measured in one cell
5 using TP-LDF (see Section 6.2.2). An XeF excimer laser with output at 353 nm is used in a
6 second cell to photolyze NO2. The total NO signal in the second cell resulting from ambient
7 and photofragment NO is measured as NO using TP-LIF. The NO2 concentration is
8 determined from the difference in signals of the two reaction cells and the fractional
9 photolysis of NO2. The NO2 fluorescence cell is calibrated using calibration sources of NO
10 and NO2.
11 Since the PF/TP-LEF instrument is signal-limited, the sensitivity is defined by
12 integration time. The detection limit for NO2 for a 2-min integration time is 12 ppt (Davis,
13 1988). The accuracy of PF/TP-LEF NO2 determinations is likely to be similar to the ±16%
14 reported for TP-LIF NO measurements (Davis et al., 1987). At 15 ppt NO and 50 ppt NO2,
15 the precision of NO2 determinations is given as ±17% (Gregory et al., 1990b).
16 The PF/TP-LIF technique is expected to be highly specific for NO2. In addition to
17 those potential NO interferents with TP-LIF that are discussed in Section 6.2.2, other species
18 that could photolyze or otherwise decompose to produce NO or NO2 have been considered
19 (Davis, 1988). Arguments were given to dismiss HNO3, HO2NO2, N2O5, CH3ONO, and
20 CH3ONO2 as possible interfering species.
21
22 6.3.4 Absorption Spectroscopy
23 Absorption Spectroscopy is discussed in a previous section for NO (Section 6.2.3).
24 Absorption methods may measure the absorption of light in the UV, visible, or infrared
25 regions of the electromagnetic spectrum. They may employ closed cells for in situ
26 measurements (e.g., TOLAS) or open paths for remote sensing. Absorption methods require
27 a source of radiation. Active methods utilize artificial light from a source such as an
28 incandescent lamp or a laser; whereas, passive methods use natural light from the sun or
29 moon. Laser sources offer advantages of scanning, narrow spectral width, high intensity, and
30 as a result usually provide better sensitivity than nonlaser sources. In this section, several
31 absorption techniques are addressed including the in situ methods, TOLAS, photometry, and
August 1991 6-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 TTFMS, as well as remote sensors employing long-path absorption differential optical
2 absorption spectroscopy (DOAS), and differential absorption lidar (DIAL).
3 Tunable-diode laser spectroscopy is a well developed technique that has been used to
4 measure NO2 as well as other species in the atmosphere. Descriptive information about the
5 operating principle is given in Section 6.2.3 and the references therein. -With a 150-m path
6 length near 1,600 cm"1 the MDL is 0.1 ppb and the accuracy is ±15% (Mackay and Schiff,
7 1987). For a 40-m path length, the MDL for NO2 is 0.5 ppb (Schiff et al., 1983). At a
8 sampling rate giving a 4-s residence time in the White cell, stable NO2 signals are achieved
9 in approximately 1 min. Linearity has been demonstrated between the signal and NO2
0 concentration at levels between 35 and 175 ppb. Surface losses are compensated for
1 automatically since the NO2 calibration gas, typically from a permeation tube, is introduced
2 directly into the sampling line. Tunable-diode laser spectroscopy is a spectroscopic
3 technique; as a result, the measurement of NO2 using this method is highly specific.
4 A prototype instrument using an in situ absorption technique to measure NO2 was
5 recently reported (Jung and Kowalski, 1986). This technique employs a modified commercial
6 O3 photometric analyzer to measure the absorption of visible light by NO2 at wavelengths
7 longer than 400 nm. The signal obtained in a 1.12-m absorption cell from unscrubbed
8 ambient air is compared with that from ambient air scrubbed of NO2. A microcomputer uses
9 Beer's Law and an absorption coefficient derived from a NO2 calibration source to determine
0 the NO2 concentration. Interferences from NH3, NO, O3, SO2, and PAN have been shown
1 to be negligible. Comparisons with commercially available CLM analyzers monitoring smog
2 chamber experiments and ambient air have shown good agreement when NO2 was expected
3 to be present. The CLM signal was found to exceed that of the photometer when
4 photochemical reaction products such as PAN were believed to be present. Although noise of
5 less than 3 ppb and linear response have been demonstrated between 100 and 700 ppb,
6 additional development and evaluation are needed to permit routine use of this technique to
7 ambient monitoring applications,
8 Cryogenic sampling at 77 K combining the matrix-isolation technique (i.e., solid CO2)
9 with FTLR spectroscopy has shown promise for the sensitive determination of NO2, PAN,
0 HNO3, HNO2, and N2O5 (Griffith and Schuster, 1987). A theoretical MDL of 5 ppt was
1 claimed for NO2 in 15-L integrated samples of ambient air.
August 1991 " 6-19 DRAFT-DO NOT QUOTE OR CITE
-------
1 A laboratory prototype method, TTFMS, has been developed recently and is described
2 in Section 7,2.3 (Hansen, 1989). This method has a projected MDL for NO2 of 0.3 ppt.
3 Long-path absorption is a remote sensing technique which is typically operated over
4 open terrain with optical path lengths of up to 10s of km. The absorption cross section for
5 each species of interest is determined in the laboratory. This information is used to convert
6 optical densities measured in the field to concentration data. For 1-h observation periods an
7 MDL of 20 ppt has been reported using an artificial light source and a 9.2-km path length
8 (Johnston and McKenzie, 1984). Noxon (1978) used the sun as a light source and the
9 structure in the NO2 absorption spectrum near 440 nm to obtain measures of the abundance
10 of NO2 in both the troposphere and the stratosphere. Platt and Perner (1983) have reported
11 the application of differential optical absorption spectroscopy (DOAS) to the determination of
12 several nitrogen-containing species. A Xe-high pressure or an incandescent lamp was used
13 with a 1- to 10-km path length. Selected applicable compounds, detection limits, and target
14 wavelengths are: NO, 400 ppt and 226 nm; NO2, 100 ppt and 363 nm; HNO2, 20 ppt and
15 354 nm; and NO3, 0.5 ppt and 662 nm. The DOAS technique has been recently adapted to
16 employ a 25-m multipass open reflection system with a path length of up to 2 km (Biermann
17 et al., 1988). Using a 0.8-km path length and 12-min averaging times, MDLs and accuracies
18 for NO2, HNO2, and NO3 of 4 ppb (±10%), 0.6 ppb (±30%), and 20 ppt (±15%) have
19 been reported, respectively.
20 Common remote sensing techniques employ light detection and ranging (LIDAR).
21 Differential absorption lidar (DIAL) is a long-path absorption technique. This method
22 employs light of two wavelengths propagated over a given distance at a given intensity. The
23 concentration of the gas species of interest is related to the difference in intensities of the two
24 wavelengths at the receiver. Differential absorption lidar techniques have been applied to the
25 ambient measurement of both NO (Alden et al., 1982) and NO2 (Fredriksson and Hertz,
26 1984; Edner et al., 1987). Baumgartner et al. (1979) report a 5 ppb sensitivity for NO2, and
27 Staehr et al. (1985) report 10 ppb sensitivity for NO2 using a laser source and a 6-km path
28 length. Differential absorption lidar methods are in the development stages for monitoring
29 NO2 in ambient atmospheres.
30
31
August 1991 , 6-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.3.5 Wet Chemical Methods
2 Most wet chemical methods for measuring NO2 involve the collection of NO2 in
3 solution followed by a colorimetric finish using an azo dye. Many variations of this method
4 exist including both manual and automated versions. Szonntagh (1979) traced the history of
5 azo dye methods for NO2 sampling and analysis. Nitrogen dioxide, first collected in aqueous
6 solution, is thought to form HNO2. An aromatic amine is used in the presence of an acid to
7 react with HNO2 and form a diazonium salt. The salt then rearranges and couples with
8 another organic amine that has been added to form a red azo dye. The intensity of the color
9 is proportional to the NO2 collected and is measured using a spectrophotometer. A good
0 overview of wet chemical methods for sampling and analysis of NO2 is given by Purdue and
1 Mauser (1980).
2
3 6.3.5.1 Griess-Saltzman Method
4 In this method, air is sampled for no longer than 30 min through a fritted bubbler that
5 contains the Griess-Saltzman reagent (Saltzman, 1954). This reagent is a solution of
6 sulfanilic acid, NEDA, and acetic acid. Color development is complete within 15 min and is
7 measured at 550 nm within an hour. Interferences from SO2 and PAN have been noted but
8 are usually too low in ambient air to cause significant error. Concentrations of NO2 ranging
9 from 20 ppb to 800 ppb for 30-min sampling periods may be determined using this method.
0 An MDL of 2 ppb and a precision of ± 11 % have been reported as well as a positive bias of
1 18% for spiked ambient air (Saltzman, 1980; Purdue and Hauser, 1980).
2 Calibration is usually performed statically using dilute solutions of sodium nitrite.
3 Saltzman (1954) reported that 0.72 moles of nitrite were formed for each mole of NO2
4 absorbed. Values of this "stoichiometric factor" ranging from 0.5 to 1.0 have been reported
5 (Crecelius and Forwerg, 1970). The method can, however, be calibrated dynamically using
6 calibrated NO2 gas standards.
7
8 6.3.5.2 Continuous Saltzman Method
9 The measurement principle is based on the Griess-Saltzman reaction. Ambient air is
D sampled continuously through a gas-liquid contactor, and NO2 is collected on contact with an
1 absorbing solution containing diazotizing-coupling reagents. After the color has developed,
August 1991 6-21 DRAFT-DO NOT QUOTE OR CITE
-------
1 the absorbance of the solution is measured continuously with a spectrophotometer. Ozone
2 was found to act as a negative interferent and eliminated this method as a candidate for
3 designation as an Equivalent Method by EPA. The MDL is 10 ppb, the bias ranges from
4 +3 to +15% at NO2 concentrations between 30 and 150 ppb, and the precision is
5 approximately ±12% (Purdue and Hauser, 1980). Although calibration can be performed
6 using nitrite solutions or calibrated NO2 gas standards, the latter approach is recommended.
7
8 6.3.5.3 Alkaline Guiacol Method
9 Various alternatives to the Griess-Saltzman Method have been proposed. Recently,
10 Baveja et al. (1984) proposed a method using alkaline o-methoxyphenol (guaiacol) as both the
11 absorbing medium and coupling reagent. Samples are collected using fritted bubblers
12 containing an alkaline guaiacol solution. After sampling, p-nitroaniline is added, the pH is
13 adjusted with HC1 and later with NaOH, the resulting dye is extracted in amyl alcohol, and
14 the absorbance is read at 545 nm. Collection efficiency of 98% and a stoichiometric factor
15 of 0.74 were reported.
16
17 6.3.5.4 Jacobs-Hochheiser Method
18 This method is a modified version of the Griess-Saltzman method to permit 24-h
19 sampling and delay in analysis times (Jacobs and Hochheiser, 1958). Samples are collected
20 by bubbling ambient air through a 0.1 N aqueous NaOH solution using a fritted bubbler.
21 The collected nitrite is then reacted with sulfanilamide and NED A in acid media to form an
22 azo dye which is measured with a spectrophotometer at 540 nm. As with the Griess-
23 Saltzman method, dilute nitrite solutions are used for calibration.
24 This method was the original Reference Method designated by the EPA for NO2
25 (Purdue and Hauser, 1980). Testing of the method showed that the originally specified NO2
26 collection efficiency of 35% was not constant and that it varied nonlinearly with NO2
27 concentration. In addition, interferences from NO and combinations of NO and NO2 were
28 found. As a result, in 1973 this method was withdrawn and considered unacceptable for air
29 sampling and analysis.
30
31
August 1991 6-22 DRAFT-DO NOT QUOTE OR CITE
-------
I 6.3.5.5 Sodium Arsenite Method (Manual and Continuous)
I This method has been designated by the EPA as an Equivalent Method in both the
5 manual and continuous forms (Federal Register, 1986). The manual method is a 24-h
t integrated method similar to the Jacobs-Hochheiser method, except that sodium arsenite is
> added to the aqueous NaOH absorbing solution, and an orifice bubbler is used. The nitrite is
5 reacted with sulfanilamide and NEDA in acid media to form an azo dye which is determined
7 with a spectrophotometer^ The continuous method employs the same measurement principle
I but uses hardware to permit continuous determination of NO2 in a manner similar to that for
> the Continuous Saltzman Method.
) The overall NO2 recovery is 82%. Although NO and CO2 may act as interferents, their
impact is minimal at typical ambient levels. Sulfur dioxide has not been tested as an
', interferent. The MDL is 5 ppb, the bias is -3% independent of concentration, and the
i precision at NO2 concentrations between 30 and 150 ppb is +6 ppb (Purdue and Hauser,
1980). Recently HNO2 was found to respond equivalently to NO2 (Braman et al., 1986).
: This interference is likely to be appreciable in urban environments during nighttime hours
i where concentrations above 5 ppb have been observed (Rodgers and Davis, 1989; Appel
etal., 1990).
6.3.5.6 Triethanolamine-Guaiacol-Sulfite (TGS) Method
This method has been designated by the EPA as an Equivalent Method (Federal
Register, 1986). It is a manual 24-h integrated method. Samples are collected using orifice
bubblers and a solution of triethanolamine (TEA), guaiacol and sodium metabisulfite. The
resulting nitrite is reacted with sulfanilamide and 8-amino-l-naphthalene-sulfonie acid
ammonium salt (ANSA) and the resulting azo dye is determined at 550 nm with a
spectrophotometer. No interferences were found in tests with NH3, CO, HCHO, NO,
phenol, O3, and SO2. The overall NO2 recovery is 93%, the MDL is 8 ppb, the bias is -5%
independent of concentration, and the precision at NO2 concentrations between 30 and
150 ppb is ±6 ppb (Purdue and Hauser, 1980).
August 1991 6-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 6,3.5.7 Triethanolamine (TEA) Method
2 This method is a manual 24-h integrated method (Ellis and Margeson, 1974). Samples
3 are collected using an aqueous solution of TEA and fritted bubblers. As with the Equivalent
4 Methods, the resulting nitrite may be determined with a spectrophotometer after reaction with
5 sulfanilamide and either NEDA or ANSA. Although recoveries of 80 to 90% were found at
6 NO2 concentrations between 20 and 350 ppb using fritted bubblers, only 50% recovery was
7 found using orifice bubblers. Since the requirement of a fritted rather than an orifice bubbler
8 was considered to be a major disadvantage by the EPA, the development of this method as an
9 Equivalent Method was terminated (Purdue and Hauser, 1980).
10 Triethanolamine has been used as the collection medium for many active and passive
11 techniques to sample NO2. Although colorimetry may be used as the analytical finish,
12 recently 1C appears to be the method of choice (Miller, 1984). Vinjamoori and Ling (1981)
13 have used an aqueous solution of TEA, ethylene glycol, and acetone on 13X molecular sieves
14 to sample air in the workplace for NO2. Passive devices employing TEA have been used for
15 industrial hygiene and indoor air quality sampling (Palmes et al., 1976; Wallace and Ott,
16 1982) as well as for ambient applications (Sickles and Michie, 1987; Mulik and Williams,
17 1987). Recently a method using Whatman GF/B filters coated with an aqueous solution of
18 TEA, ethylene glycol, and acetone has been developed for extended sampling of both NO2
19 and SO2 from ambient air (Sickles et al., 1990). This method using 1C to determine the
20 collected NO2 as nitrite and nitrate showed no interferences from NO, NH3, O3, H2S,
21 CH3SH, and CS2 but major interferences from PAN and HNO2. Recovery of NO2 averaged
22 87% in laboratory tests at concentrations between 5 and 400 ppb. Using two coated 47 mm
23 diameter filters in series, filter temperatures above 5 °C, and flow rates below 2 L min"1 are
24 required to insure good NO2 collection efficiency. This method has been incorporated into
25 the design of a prototype sampler known as a Transition Flow Reactor (TFR) system for
26 measuring acidic deposition components (Knapp et al., 1986).
27
28 6.3.6 Other Active Methods
29 Several other methods for the determination of NO2 have been reported. These
30 methods include ionization spectroscopy, mass spectrometry, photothermal detection,
August 1991 6-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 denuders, and solid sorbents. Since they are in the early stages of development or are not
2 being used widely, they are mentioned only briefly.
3 lonization spectroscopy is a new and sensitive in situ laser technique that is currently
4- under investigation for tropospheric measurements of NO and NO2 (National Aeronautics and
5 Space Administration, 1983). This method is in the early stages of development.
S Atmospheric pressure ionization mass spectrometry (API/MS) has been investigated for
7 the continuous measurement of NO2 and SO2 in ambient air (Benoit, 1983). An MDL of
8 approximately 0.5 ppb was reported. .,.,.*
3 Methods employing photothermal detection of NO2 have been reported (Poizat and
3 Atkinson, 1982; Higashi et al., 1983; Adams et al., 1986). Detection is accomplished by
1 selectively exciting transitions of NO2 with a chopped continuous wave or pulsed laser
2 source. At pressures near atmospheric, collisional de-excitation converts the absorbed energy
3 into translational energy leading to a temperature rise along the beam and expansion of the
4- gas. The resulting change in refractive index of the thermal lens can be detected by
5 spectroscopic means (Higashi et al., 1983), or the resulting pressure change can be detected
6 with a microphone (Poizat and Atkinson, 1982). Minimum detection limits of 2 to 5 ppb
7 have been reported. Development of photoacoustic methods for continuous application may
8 be limited by acoustic noise associated with vibration and flow.
9 A tube or channel with its walls coated with a chemical chosen to remove a gaseous
0 species of interest from a sample drawn through the tube under laminar conditions is known
1 as a denuder. The concentration of the species of interest is determined by measuring the
2 amount of the species collected on the walls or by comparing the signal strengths in the
3 presence and absence of the denuder. Possanzini et al. (1984) recommended using a coating
4 containing KI to collect NO2. Subsequent tests showed the collection efficiency of this
5 material to be dependent on the humidity of the sample (Sickles, 1987). An alkaline guaiacol
6 coating on annular denuders has shown high collection efficiency for NO2 (Buttini et al.,
7 1987). After extraction using DI water, 1C analysis showed quantitative recovery as NO2".
8 The reported MDL was 0.13 ppb for aim3 sample (i.e., 16 L min"1 for 1 h), and the
9 median precision for 14 paired samples was 3.9% expressed as relative standard deviation
0 (RSD). A direct interference was noted with HNO2, none was found with NO or PAN, and
1 no humidity effects were observed between 20 and 80% RH. No interference tests were
August 1991 6-25 DRAFT-DO NOT QUOTE OR CITE
-------
1 performed with O3; however, comparison of 4-h results with those of a commercial CLM
2 (NO + O3) instrument sampling air near Rome showed good, correlation (i.e., r = 0.92).
3 Although alkaline guaiacol solutions degrade with time, no degradation effects were found for
4 72-h presampling or 24-h postsampling denuder storage. Longer duration storage tests (e.g.,
5 2 weeks pre- and postsampling) and additional field evaluations are needed before this method
6 is ready for routine application.
7 Recently, Adams et al. (1986) found activated MnO2 to be effective under unspecified
8 RH for removing NO2 in denuder applications. Additional are needed under conditions
9 representative of the ambient environment before this approach is ready for routine
10 application.
11 Lipari (1984) has used a commercially available cartridge containing the sorbent Florisil
12 (magnesium silicate) coated with diphenylamine to sample NO2 in ambient and indoor
13 environments. The NO2 reacts with the diphenylamine to form nitro- and nitroso-derivatiyes
14 which are subsequently detected by HPLC-UV. Although no interference from NO, O3,
15 SO2, HNO3, and water vapor were found, PAN produced a 50% positive interference, and
16 HNO2 was expected to interfere quantitatively. Sorbent temperature must be held below
17 32 °C to prevent volatilization of the diphenylamine and nonquantitative sampling. An MDL
yj
18 of 0.1 ppb for a 2.0 nr air sample was claimed. The method shows promise for the
19 sensitive determination of NO2 under conditions where the noted interferents and temperature
20 sensitivity do not pose problems.
21
22 6.3.7 Passive Samplers
23 Passive samplers are frequently used in industrial hygiene, indoor air, and personal
24 exposure studies and are less frequently used in ambient air sampling. Passive NO samplers
25 are described in Section 6.2.4. Namiesnik et al. (1984) have provided a good overview of
26 passive samplers. The basis for all passive samplers is the same. The gaseous analyte
27 molecule is transported from the bulk air to a reactive surface where the molecule impinges
28 and is captured by chemical reaction. After exposure periods ranging from hours to days, the
29 'reactive surface is analyzed and the integrated loading of the reaction product is used to infer
30 the average gas concentration present during the sampling period. When the transport of
31 analyte molecules to the reactive surface is diffusion-controlled, the average ambient
August 1991 6-26 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentration may be related directly to the ratio of the product loading to the sampling
2 duration. This proportionality is defined as a sampler calibration factor or alternatively as the
3 reciprocal of the sampler sampling rate.
4 One type of passive NO2 sampler for ambient applications is the nitration plate. It is
5 essentially an open petri dish containing TEA-impregnated filter paper. Thus, there is no
6 diffusion barrier between the ambient air and the NO2 collection surface. Nitrogen dioxide
7 reacts with the TEA and is retained primarily as NO2" which can be extracted and determined
8 with a spectrophotometer or by 1C. A single calibration factor is provided by the
9 manufacturer. A recent study indicated that the calibration factors determined experimentally
0 for the device are extremely sensitive to wind speed, NO2 concentration, and temperature
1 (Sickles and Michie, 1987). Triethanolamine is expected to collect not only NO2 but HNO2,
2 HNO3, and PAN (Sickles, 1987). These results suggest that nitration plates may be useful
3 only as qualitative indicators of ambient levels of NOX.
4 Another open surface device has been proposed by Kosmus (1985) for ambient
5 applications. This device uses chromatographic paper in the shape of a candle that is coated
6 with diphenylamine and is continually impregnated with a KSCN-glyeerin catalyst. Nitrogen
7 oxides, presumably NO2 and NO to some extent, are collected by a catalyzed reaction with
8 diphenylamine to form the nitrosamine. After extraction, the nitrosamine may be determined
9 with a spectrophotometer or by differential pulse polarigraphy. Interference by iron oxide
0 particles was noted, and interference from both PAN and HNO2 is expected (Lipari, 1984).
1 Sensitivity to wind speed as noted previously for the nitration plate is also expected.
2 Collocated sampling was performed using four candles and a CLM NO/NO2 analyzer at each
3 of 14 stations (Kosmus, 1985). The nitrosamine loadings were highly correlated but in a
4 nonlinear manner with the sum of 100% NO2 plus 90% NO from the CLM instruments.
5 These results also suggest that open surface passive samplers may be useful as qualitative
6 indicators of ambient levels of NOX.
7 Mulik and Williams (1986) have adapted the nitration plate concept by adding diffusion
8 barriers in their design of a passive sampling device (PSD) for NO2 in ambient and personal
9 exposure applications. The physical configuration employs a TEA-coated cellulose filter that
0 uses two 200 mesh stainless steel diffusion screens and two stainless steel perforated plates on
1 each side of the coated filter to act as diffusion barriers and also permits NO2 collection on
August 1991 6-27 DRAFT-DO NOT QUOTE OR CITE
-------
1 both faces of the filter. After sampling, the filter is removed from the PSD, extracted in
2 water, and analyzed for NO2~ by 1C. A sensitivity of 0.03 ppm-h and a sampling rate of
3 2.6 cm3 s"1 were claimed. Comparison of PSD results with CLM determinations of NO2 in
4 laboratory tests at concentrations between 10 and 250 ppb were linearly related and highly
5 correlated (i.e., r = 0.996). The device exhibited increased sampling rates of approximately
6 50% as the wind speed increased from 20 to 45 cm s"1, but displayed a relatively constant
7 sampling rate at wind speeds between 45 and 300 cm s"1 (Mulik and Williams, 1987).
8 Interferences from PAN and HNO2 are expected (Sickles, 1987). Results of TOLAS and
9 triplicate daily PSD NO2 measurements in a recent 13-day field study showed good
10 agreement between the study average values but a correlation coefficient for daily results of
11 only 0.47 (Mulik and Williams, 1987; Sickles et al., 1990). Further development and testing
12 of the PSD appears warranted.
13 The Palmes tube is a passive device that has been used to sample air in the workplace
14 and indoor environments to assess personal exposure to NO2 (Palmes et al., 1976; Wallace
15 and Ott, 1982). This device and its operation are described in detail in Section 6.2.4. It
16 consists of a tube, open at one end with a TEA-coated interior surface on the closed end.
17 Nitrogen dioxide diffuses through the 7.1-cm diffusion length where it is captured by TEA.
18 Analysis is accomplished by extracting the TEA-coated surface and analyzing the extract for
19 NO2". This may be done directly by adding an aqueous solution of sulfanilamide and NEDA
20 to the tube and determining the NO2" concentration using a spectrophotometer at 540 nm
21 (Palmes et al., 1976). A stoichiometric factor of unity, a linear response for dosages
22 between 1 and 30 ppm-h, and a sampling rate of 0.02 cm3 s"1 are reported. An improvement
23 in sensitivity from 0.3 ppm-h to 0.03 ppm-h is claimed if the aqueous TEA extract is
24 analyzed by 1C using a concentrator column (Mulik and Williams, 1986). Absorption and
25 desorption of NO2 by the internal walls of the acrylic tube have been reported to limit
26 applications to exposures of 0.1 ppm-h (Miller, 1988). This sensitivity can be improved to
27 0.03 ppm-h by using stainless steel rather than acrylic tubes. The device exhibited sampling
28 rates increased by up to 15% as wind speed was increased from 50 to 250 cm s"1 (Palmes
29 et al., 1976) and decreased by 15% as temperature was reduced from 27 to 15 °C (Girman
30 et al., 1984). Interferences from PAN and HNO2 are expected (Sickles, 1987). The
August 1991 6-28 DRAFT-DO NOT QUOTE OR CITE
-------
1 calibration factor for the Palmes tube is theoretically derived, and the analytical finish is
2 calibrated by dilution of gravimetrically prepared nitrite solutions.
3 The performance of the Palmes tube was compared with that of two commercially
4 available passive personal samplers, the DuPont Pro-Tek and the MDA Chronotox System
5 (Woebkenberg, 1982). The Palmes tube displayed greater sensitivity than either of the
6 commercial samplers and displayed adequate precision and accuracy at loadings between
7 1 and 80 ppm-h. Since the commercial devices may be used at only moderate to high
8 loadings (i.e., above 5 ppm-h), they are not sufficiently sensitive for most ambient or
9 personal exposure applications. As a result, they are not discussed further in this document.
0 A badge-type NO2 personal sampler has been devised by Yanagisawa and Nishimura
1 (1982). Their device uses a series of 5 layers of hydrophobic Teflon™-type filter material as
2 a diffusion barrier between ambient air and a TEA-coated cellulose fiber filter. Nitrogen
3 dioxide diffuses through the hydrophobic filters to the TEA-coated surface, where it is
4 collected. Following extraction of the TEA-coated filter in a solution of sulfanilic acid,
5 phosphoric acid, and NEDA, a colorimetric finish at 540 nm is employed. A sensitivity of
6 0.07 ppm-h, a sampling rate of 1.4 cm3 s"1, and an accuracy of +20% are claimed. The
7 device exhibited increased sampling rates of up to 30% as the wind speed was increased from
8 15 to 400 cm s"1. Interferences from PAN and HNO2 are expected (Sickles, 1987). The
9 calibration factor for the sampler is provided by the supplier, and the analytical finish is
'0 calibrated by dilution of gravimetrically prepared nitrite solutions.
11 A variation on the above approach has been proposed by Cadoff and Hodgeson (1983).
'2 The sampler is comprised of a Nuclepore 47 mm filter holder with a capped base containing a
13 TEA-coated glass fiber filter and a 0.8 ^im pore size polycarbonate filter. The polycarbonate
14 filter and the air space between this filter and the TEA-coated filter act as a diffusion barrier.
'.5 A colorimetric analytical finish is employed. The performance was tested at NO2 loadings
16 between 0.06 and 1 ppm-h, and a sampling rate of 1.9 cm3 s"1 was claimed.
17 West and Reiszner (1978) proposed a passive NO2 sampler using a silicone membrane
18 as a diffusion barrier between ambient air and an alkaline thymol blue NO2 collection
19 solution. Collected NO2 is converted to NO2" and determined colorimetrically. Results of a
>0 field comparison with the EPA-designated TGS method were not favorable and showed the
August 1991 6-29 DRAFT-DO NOT QUOTE OR CITE
-------
1 proposed device to yield results approximately a factor of three higher than the TGS method.
2 As a result, this method is not recommended.
3
4 6.3.8 Calibration
5 Two methods have been designated by the EPA (Code of Federal Regulations, 1987a)
6 as alternative calibration procedures for the measurement of NO2 in the atmosphere. These
7 methods use permeation tubes or gas phase titration (GPT) to generate known amounts of
8 NO2. Calibrations are performed using dynamic dilution with air.
9 A permeation tube is a porous, inert tube usually made of Teflon™ that has been
10 partially filled with liquid NO2 and sealed. Permeation of NO2 through the porous tube will
11 occur at a constant rate if the temperature of the tube and the NO2 concentration gradient
12 across the tube are held constant. The tube is maintained at a constant temperature
13 (±0.1 °C), and a measured flow of a dry carrier, usually nitrogen, is passed over it. The
14 NO2 permeating through the porous wall and entering the carrier stream is diluted with zero
15 air to produce calibration NO2 atmospheres of known concentrations. The permeation tube is
16 calibrated gravimetrically by measuring the weight loss of the tube over time. The National
17 Bureau of Standards (NBS) provides SRM permeation tubes that emit NO2 at a nominal rate
18 of 1 /xg min"1 (National Bureau of Standards, 1988). Additional information on the
19 performance of NO2 permeation tubes is given by Hughes et al. (1977). A recent report by
20 Braman and de la Cantera (1986) indicates that permeation sources of NO2 can produce
21 atmospheres that are contaminated with other NOX including HNO3, HNO2, and NO.
22 Further work appears warranted to define the conditions where permeation devices may be
23 used to provide an unambiguous source of NO2.
24 Gas phase titraton employs the rapid, quantitative gas phase reaction between NO,
25 usually from a standard gas cylinder, and O3, from a stable O3 generator, to produce one
26 NO2 molecule for each NO molecule consumed by reaction. When O3 is added to excess
27 NO in a titration system, the decrease in NO (and O3) is equivalent to the NO2 produced.
28 Different amounts of NO2 may be produced by adding different amounts of O3. When the
29 NO concentration and the flow rates entering the dynamic titration system are known
30 accurately, the NO2 concentration leaving the system can be determined accurately. The
31 accuracy and stability of NO standard gas cylinders are described in Section 6.2.5.
August 1991 6-30 DRAFT-DO NOT QUOTE OR CITE
-------
I A third source of NO2 sometimes used for calibration is the cylinder of compressed gas
I containing NO2 usually in N2 (Fehsenfeld et al., 1987; Davis, 1988), Calibrations are
$ subsequently performed using dynamic dilution with zero air. These cylinders are
t commercially available, and the NO2 concentration should be referenced to an accepted
) standard. Bennett (1979) has shown that of 26 aluminum cylinders initially containing
5 supplier-certified concentrations of NO2 in N2 between 100 and 300 ppb, 10 showed modest
7 declines in NO2 concentration during the first 3 months after preparation. The NO2 levels in
I all 26 cylinders declined substantially over the 10-month study period. Schiff et al. (1983)
) have noted problems handling trace concentrations of NO2 from a cylinder. A cylinder
) containing 9 ppm NO2 in N2 gave 15% higher readings for NO2 when analysis was
I performed by CLM than by TOLAS (Walega et al., 1984). This discrepancy may have been
I due to an impurity (e.g., HNO3) in the cylinder that could act as an interference with the
5 CLM but not the TDLAS determination of NO2. Davis (1988) examined a cylinder
I containing 44 ppm NO2 in air at regular intervals over 3 years and observed a 16% change in
5 concentration. In view of these findings, caution should be exercised if a cylinder containing
3 NO2 in N2 or air is to be employed as a calibration source of NO2.
1 •
5 6.3.9 Intercomparisons
) Several intercomparisons of research grade NO2 instrumentation have been conducted
) (Helas et al., 1981; Walega et al., 1984; Sickles et al., 1990; Fehsenfeld et al., 1987, 1990;
I Gregory et al., 1990b) and are described in this section. The performance of EPA-
l Designated Methods based on intercomparisons and other studies is discussed in
5 Section 6.3.10.
I- Helas et al. (1981) report the results of a field intercomparison of several NO2 methods
5 conducted in April 1979 at Deuselbach, Germany. Good agreement between a highly
j sensitive CLM instrument and long-path absorption was found over the 1 to 8 ppb range of
f observed NO2 concentrations.
I Walega et al. (1984) report comparisons of NO2 measurements from a highly sensitive
) CLM instrument using a thermal NO2 to NO converter with NO2 measurements from a
) TDLAS system. Measurements of NO2-spiked synthetic air conducted both in the laboratory
I and in the field showed good agreement. Measurements were made of ambient and captive
August 1991 6-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 air in downtown Los Angeles and showed maximum respective concentrations of 100 and
2 600 ppb. Chemiluminescence results were appreciably higher than those of the TDLAS: this
3 difference averaged 18% in the ambient air studies and 15% in the captive air studies. In the
4 latter studies the agreement was generally within 10% in the morning but by the end of the
5 day, could be as large as 80%. This behavior was attributed to the reaction of NO2 and the
6 accumulation of photochemically produced CLM interferents such as PAN that occurred
7 during the day.
8 Daily NO2 concentrations determined by TDLAS, CLM (luminol), and PSDs were
9 reported recently from a 13-day study conducted at Research Triangle Park, NC in the Fall of
10 1986 (Bubacz et al., 1987; Mulik and Williams, 1987; Sickles et al., 1990). Collocated
11 sampling was performed using a TDLAS system, two CLM (luminol) instruments, and
12 triplicate daily PSDs. Daily average results were computed for the TDLAS, each CLM
13 (luminol) instrument, and the PSDs. The 13-day average values from the CLM (luminol)
14 instruments and the TDLAS system agreed to within 2 ppb; the average daily ratios of NO2
15 by CLM (luminol) to TDLAS were 1.01 ± 0.11 (SD) and 1.19 ± 0.17; the respective
16 correlation coefficients were high, 0.94 and 0.91; and while the results of one CLM
17 (luminol) instrument showed no bias, results of the other were biased higher than those of the
18 TDLAS system. The 13-day average values from the PSDs and TDLAS system agreed to
19 within 1 ppb; the average daily ratio of NO2 by PSD to TDLAS was 1.08 ± 0.32. There
20 was no apparent bias but the correlation coefficient was only 0.47.
21 A field intercomparison of instruments designed to measure NO, NOX, and NOy was
22 conducted near Boulder, CO (Fehsenfeld et al., 1987). In addition, an intercomparison of
23 NO2 measurements was performed using two different NO2-to-NO converters prior to NO
24 detection by CLM. The two CLM detection systems were tested and found to be equivalent.
25 One instrument used a photolytic NO2-to-NO converter while the other used a FeSO4 • 7H2O
26 surface converter. In spiking tests, the instrument with the FeSO4 converter responded to
27 NO2, PAN, and n-propyl nitrate but not to HNO3 or NH3. The CLM-PC instrument
28 responded to NO2 but not significantly to HNO3, NH3, n-propyl nitrate, or PAN. For
29 measurements of NO2 + NO in ambient air, results from the two instruments agreed at
30 concentrations above 1 ppb. However, results from the instrument with the surface converter
31 were biased higher than those from the photolytic converter at lower NO + NO2
August 1991 6-32 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations. This discrepancy was a factor of 2 at 0.1 ppb. These results suggest that
2 surface converters sufficiently active to Convert NO2 to NO can convert other NOX species
3 such as PAN to NO. While the use of CLM with surface NO2 to NO converters may not
4 pose a problem in many urban and suburban areas where NO and NO2 are expected to be the
5 dominant NOX, results cited here and elsewhere in this section suggest that surface converters
6 are unsuitable for the interference-free measurement of NO2 in ambient air containing PAN
7 and similar compounds.
8 A ground-based intercomparison of NO2 measurements using CLM-PC, CLM
9 (luminol), and TDLAS research grade instruments was performed near Boulder, Colorado
10 (Fehsenfeld et al., 1990). Ambient concentrations ranged from 0.02 to 4 ppb. The potential
11 interferences of H2O2, HNO3, «-propyl nitrate, PAN, and O3 were examined in spiking
12 tests. Only the CLM (luminol) instrument displayed appreciable interferences, and they were
13 with O3 (0.6%) and PAN (24%). At ambient NO2 concentrations above 2 ppb all three
14 instruments gave similar results. Below 2 ppb, interferences from O3 and PAN provided
15 high biases to the CLM (luminol) results, but they could be corrected with measured O3 and
16 PAN results at NO2 levels above 0.3 ppb. An O3 scrubber added to a second CLM
17 (luminol) instrument removed the O3 interference but failed to remove PAN and appeared to
18 remove substantial amounts (i.e., 50%) of NO2. Removal of NO2 in the manufacturer-
19 supplied O3 scrubber has also been reported by Kelly et al. (1990). Tunable-diode laser
20 spectroscopy results compared favorably with CLM-PC at relatively high NO2 levels (i.e.,
21 >0.4 ppb) but displayed a high bias (i.e., factor of 5) at lower NO2 concentrations. No
22 interferences or artifacts were found for the CLM-PC results,
23 An airborne intercomparison (i.e., CITE 2) of NO2 measurements was conducted by
24 NASA using TDLAS, PF/TP-LIF, CLM-PC, and CLM (with FeSO4 converter) research
25 grade instruments (Gregory et al,, 1990b). Sampling flights were performed primarily in the
26 free troposphere, and NO2 concentrations were below 200 ppt and generally below 100 ppt.
27 High biases (i.e., factors of 2 to 3) apparently resulting from PAN interferences were present
28 in results from the CLM instrument with the FeSO4 converter, and results from this
29 instrument were not considered in subsequent analyses. At concentrations below 200 ppt,
30 results from the remaining three instruments were highly correlated (i.e., correlation
31 coefficients ranged from 0.84 to 0.95) and displayed a general level of agreement to within
August 1991 6-33 DRAFT-DO NOT QUOTE OR CITE
-------
1 30 to 40%. The PF/TP-LIF results were higher than those of the CLM-PC, and the TDLAS
2 results were the lowest. At concentrations below 50 ppt the results were poorly correlated,
3 although the PF/TP-LEF and CLM-PC results agreed to within 20 ppt. Below 50 ppt TDLAS
4 results were much higher than those of the other two instruments. This bias, similar to that
5 observed in the ground-based intercomparison of Fehsenfeld et al. (1990), was enhanced at
6 low NO2 concentrations by an error in the data reduction protocol employed in both studies.
7
8 6,3.10 Designated Methods
9 Sampling and analysis methodologies for NO2 have been specified by the EPA (Code of
10 Federal Regulations, 1987b). These designated methods are termed "Reference" or
11 "Equivalent." In 1973 the original Federal Reference Method for NO2, the Jacobs-
12 Hochheiser Technique, was withdrawn because of technical deficiencies (Purdue and Hauser,
13 1980). In 1976, the measurement principle and the associated calibration procedure on which
14 Reference Methods for NO2 must be based were specified. The measurement principle is gas
15 phase chemiluminescence and the calibration procedure may employ either GPT of an NO
16 standard with O3 or an NO2 permeation device (Code of Federal Regulations, 1987a). Since
17 only the measurement principle and calibration procedures applicable to NO2 Reference
18 Methods were specified, different analyzers can be built and designated as Reference Methods
19 provided they meet the performance specifications shown in Table 6-1 (Code of Federal
20 Regulations, 1987b).
21 To be designated as an Equivalent Method, the candidate method must be based on
22 measurement principles different from the Reference Method and meet certain performance
23 specifications (Code of Federal Regulations, 1987b). An Equivalent Method may be either
24 manual or automated. To be designated as Equivalent a candidate manual method must
25 demonstrate comparability, as shown in Table 6-2, with the Reference Method when applied
26 simultaneously to a real atmosphere. A candidate automated method must meet the
27 performance specifications shown in Table 6-1 and demonstrate comparability as shown in
28 Table 6-2 with the Reference Method when applied simultaneously to a real atmosphere.
29 Methods designated by the EPA as Reference and Equivalent are identified in Table 6-3
30 (Federal Register, 1986). Detailed descriptions of these as well as other methods for NO2
August 1991 6-34 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 6-1. PERFORMANCE SPECIFICATIONS FOR
NITROGEN DIOXIDE AUTOMATED METHODS
Performance Parameter Units NO
'2
Range ppm 0-0,5
Noise
0% upper range limit ppm 0.005
80% upper range limit ppm 0.005
Lower detectable limit ppm 0.01
Interference equivalent
Each interferant (SO2, NO, NH3, H2O) ppm ±0.02
Total interferant ppm <0.04
Zero drift, 12 and 24 h ppm ±0.02
Span drift, 24 h , .
20% of upper range limit ' % ±20.0
80% upper range limit % ±5.0
Lag time min 20
Rise time min 15
Fall time . . mj^ ^5
Precision
20% of upper range h'mil ppm 0.02
80% of upper range limit ppm 0.03
»
Source: Code of Federal Regulations (1987b).
1 are given in previous subsections. Studies were conducted to provide a basis for the
2 designation of methods by the EPA. Tests were performed to compare the performance of
3 CLM, continuous eolorimetric, manual sodium arsenite, and manual TGS methods (Purdue
4 and Hauser, 1980). The methods were compared by measuring NO2 in spiked and unspiked
5 ambient air simultaneously. Quadruplicate samples were taken for the two manual methods
August 1991 6-35 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 6-2. COMPARABILITY TEST SPECIFICATIONS
FOR NITROGEN DIOXIDE
Concentration Range,
ppm NO2
Maximum Discrepancy
Specification, ppm .,
Low 0.02 to 0.08
Medium 0.10 to 0.20
High 0.25 to 0.35
0.02
0.02
0.03
Source: Code of Federal Regulations (1987b).
TABLE 6-3. AND EQUIVALENT METHODS FOR NITROGEN
DIOXIDE DESIGNATED BY THE EPA
Method
NO2 Manual Methods (Equivalent Methods)
Sodium aresnite
Sodium aresnite/Technicon n
TGS-ANSA
NO2 Analyzers (Reference Methods)
Beckman 952A
Bendix 8101-B
Bendix 8101-C
CSI 1600
Meloy NA53OR
Monitor Labs 844OE
Monitor Labs 8840
Philips PW9762/02
Thermo Electron 14B/E
Thermo Electron 14D/E
Designation Number
EQN-1277-026
EQN-1277-027
EQM-1277-028
RFNA-0179-034
RFNA-0479-038
RFNA-0777-022
RFNA-0977-025
RFNA- 1078-031
RFNA-0677-021
RFNA-0280-042
RFNA-0879-040
RFNA-0179-035
RFNA-0279-037
Method Code
'
026
027
028
034
038
022
025
031
021
042
040
• 035
037
Source: Federal Register (1986).
1 and duplicate analyzers were used for the two continuous methods. The NO2 spikes were
2 varied randomly from day to day over the sampling schedule and ranged from 0 to 430 ppb.
3 Agreement both within and between methods was good: the average difference was neyer
4 greater than 4 ppb. Correlation coefficients for between-method comparisons exceeded
5 0.985 in all cases. No between-method differences could be attributed to concentrations of
August 1991
6-36
DRAFT-DO NOT QUOTE OR CITE
-------
1 NO, CO2, O3, total sulfur, or total suspended partieulate matter in the ambient air.
2 Significant negative interference in the continuous colorimetric method was found at NO2
3 concentrations of 40 and 53 ppb in the presence of 180 and 340 ppb O3. At O3
4 concentrations of 50 ppb no interference was detected. No interference was detected with the
5 manual sodium arsenite method at NO concentrations as high as 250 ppb. The performance
6 of the CLM analyzers was judged to be superior to that of the continuous colorimetric
7 analyzers with respect to zero drift, span drift, response times, and overall operation. Of the
8 two manual methods, the performance of the sodium arsenite method was judged superior to
9 the TGS method.
0 Eight of the Reference Methods have undergone extensive post-designation testing in the
1 laboratory and field (Michie et al., 1983). Performance test results have been reported and
2 were found to meet the specifications shown in Table 6-1. Based on the field test results,
3 minimum detection limits were defined as three times the precision. These MDL results
4 ranged from 5 to 13 ppb with an average of 9 ppb. An independent analysis of this data by
5 Holland and McElroy (1986) also showed similar results.
6 Interrogation of the National Aerometric Data Bank (NADB) records for 1985 revealed
7 that NO2 data were archived from 40 states (Hustvedt, 1987). Of the 335 data sets, CLM
8 was being employed in 291 cases, and manual methods in the remainder. Of the manual
9 methods, 42 employed the sodium arsenite method with either orifice or fritted bubblers.
0 Interrogation of the Precision Accuracy Reporting System (PARS) data base for the State and
1 Local Air Monitoring Stations (SLAMS) network for fourth quarter 1986 and first quarter
2 1987 records revealed that data were archived from tests of 114 CLM analyzers (Rhodes,
3 1987). Of these, 43% were Bendix 8101C, 40% were CSI 1600, 15% were Monitor
4 Labs 8840 and 8440E, 2% were Meloy NA 530R, and 1% were Beckman 952A.
5 To illustrate the precision and accuracy of the designated methods in field applications,
6 PARS data were examined for 1983 through 1986 (Rhodes and Evans, 1988). The results,
7 shown in Table 6-4 as 95 % probability limits suggest that the precision of continuous NO2
8 analyzers falls in the range of +10 to 15% while the manual methods are much worse at
9 ±20 to 50%. It should be noted that the manual precision results show a recent worsening.
0 This trend may reflect the phasing out of manual methods in the network which was
1 completed by the end of 1986.
August 1991 6-37 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 6-4. NATIONAL PRECISION AND ACCURACY
PROBABILITY LIMIT VALUES EXPRESSED AS PERCENT
FOR CONTINUOUS AND MANUAL METHODS FOR NO2
NO2 Method
Continuous
Precision
Accuracy
Manual
Precision
Accuracy
1983a
-13 + 12
(9,299)b
-19 + 15
(680)c
-19 + 21
(l,324)d
-5 + 6
(348)c
1984.
-14 + 13
(8,653)b
-21 + 20
(613)c
-21 + 27
(691)d
-6 + 7
(175)c
1985
-12 + 12
(7,695)b
-20 + 21
(573)c
-27 + 29
(469)d
-3 + 5
(161)c
1986
-11 + 11
(6,686)b
-21 + 20
(529)c
-48 + 45
(174)d
-4 + 5
(92)c
•Calculated differently for 1983 than for 1984 through 1986.
bNumber of precision checks.
"Number of audits; manual at 0.074 to 0.083 ppm; continuous at 0.03 to 0.08 ppm.
dNumber of collocated samples.
Source: Rhodes and Evans (1988).
1 The tabulated probability limits for accuracy of continuous NO2 analyzers are ±20%,
2 while for the manual methods they are at ±3 to 7% (Rhodes and Evans, 1988). The
3 accuracy results reflect audits of the analysis portion of the manual methods and audits of
4 both sampling and analysis for the continuous methods. Thus, the apparent difference in
5 accuracy may be reflecting differences in the auditing procedures employed.
6
7
8 6.4 OXIDES OF NITROGEN (NOX)
9 For the purposes of this document, NOX are considered to be the sum of NO and NO2.
10 No widely accepted methods are available for determining NOX except by determining NO
11 and NO2 individually and summing, or by converting NO2 to NO and determining NOX as
12 the total NO. Sections 6.2 and 6.3 describe methods for the determination of NO and NO2.
August 1991 £-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 Commercial CLM NOX analyzers catalytically convert NO2 to NO and measure NOX as
2 the sum of the originally present NO and the converted NO. As noted in Section 6,3.1,
3 ,.NO2-to-NO converters used may not be specific for NO2. Heated molybdenum converters,
4 typically used in commercial analyzers, have been shown to convert PAN, HNO3, and other
5 nitroxy compounds to NO, giving rise to artificially high values for NO2 and NOX. In
6 research grade CLM NOX analyzers, FeSO4 converters have been shown to overestimate NOX
7 by a factor of 2 to 3 at concentrations of 0.2 ppb (Fehsenfeld et aL, 1987; Gregory et al.,
8 1990b).
9 The catalytic conversion approach will permit an accurate measure of NOX as long as
0 the nitroxy compounds present in the sampled atmosphere are limited to NO and NO2.
1 Atmospheric concentrations of potential interferences are generally low relative to NO2 (Code
2 of Federal Regulations, 1987a). There are cases, however, where compounds other than NO
3 and NO2 contribute substantially to the atmospheric nitroxy burden. Examples include urban
4 atmospheres such as Los Angeles where both PAN and HNO3 levels may reach appreciable
5 levels (Tuazon et al., 1981) and remote environments where PAN may comprise a significant
6 fraction of the airborne nitroxy reservoir (Fehsenfeld et al., 1987; Gregory et al., 1990c).
7 A prototype method employing CLM has been suggested to measure NOX (Fontijn
8 et al., 1980). This method uses the reaction between atomic hydrogen and NO2 to give NO
9 along with the subsequent CLM reaction between atomic hydrogen and NO. The emission
,0 occurs between 628 and 800 nm, and the intensity is measured by a PM tube at 640 to
1 740 nm. At a constant atomic hydrogen concentration, the light intensity is proportional to
.2 the NOX concentration. The instrument was developed for application to automotive exhaust
3 gas. Significant interferences were noted for O2 and ethene but not for H2O, toluene,
4 isopentane, CO, CO2, NH3, and HCN. Response was linear within 2% from 6 to
.5 3,000 ppm. Significant development is expected if the limit of detection for this technique is
:6 to be extended from 6 ppm to the ppt to ppb range appropriate for ambient air monitoring
.7 applications,
,8
.9
•0 •
August 1991 6-39 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.5 TOTAL REACTIVE ODD NITROGEN (NOy)
2 In the present document total reactive odd nitrogen is represented by NOy. Individual
3 components comprising NOy are NO, NO2, NO3, N2O5, HNO2, HNO3, HO2NO2, PAN,
4 other organic nitrates, and particulate NO3".
5 Although no single instrument has been devised to measure NOy, researchers have
6 combined highly sensitive research grade CLM NO detectors with catalytic converters that are
7 sufficiently active to reduce most of the important gas phase NO species to NO for
8 subsequent detection (Helas et al., 1981; Dickerson, 1984; Fahey et al., 1986). Calibrations
9 are performed using dynamic dilution with air. Two standards are usually employed:
10 a cylinder of compressed gas containing NO in nitrogen at an NBS-traceable concentration;
11 and an NO2 permeation tube. The NO cylinder is used to calibrate the instrument for NO,
12 and NO2 from the permeation tube is used as a surrogate to calibrate the instrument for NOy.
13 Two types of heated converters have been employed: molybdenum and gold. As noted
14 in Section 6.3.1, heated molybdenum has been shown to convert NO2, HNO3, PAN, methyl
15 nitrate, ethyl nitrate and nitrite, n-propyl nitrate, and n-butyl nitrate to NO with high
16 efficiency. Dickerson (1984) also reports that NO3 and N2O5 are converted to NO on heated.
17 molybdenum, while CH3CN, HCN, and NH3 are not. Dickerson (1984) has coupled this
18 converter with a sensitive CLM NO detector and reported a detection limit for NOy of 25 ppt
19 for a 20 s integration time and an accuracy of ±40% at levels well above the detection limit
20 (Fehsenfeld et al., 1987).
21 A gold catalyst operated at 300 °C in the presence of 3,000 ppm CO has been reported
22 to reduce NOy to NO (Bellinger et al., 1983; Fahey et al., 1985a). Converter efficiencies
23 near 100% were found for NO2, HNO3, N2O5, and PAN. Interferences in the presence of
24 water vapor were found to be negligible for O3, NH3, N2O, HCN, CH4, and various
25 chlorine- and sulfur-containing compounds. Fahey et al. (1986) coupled this converter with a
26 sensitive CLM NO detector and reported a detection limit of 10 ppt for a 10 s integration
27 time and an accuracy of ±15%.
28 A field intercomparison of the two instruments described above was conducted near
29 Boulder, CO (Fehsenfeld et al., 1987). In this study, ambient NOy concentrations ranged
30 from 400 ppt to over 100 ppb. Both instruments gave similar estimates of NOy
August 1991 6-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations under conditions that varied from representing urban to continental background
2 air.
3 Using the instrument described above with a gold converter, Fahey et al. (1986)
4 compared NOy measurements with the sum of the component species measured individually,
5 The NOy levels systematically exceeded the sum. The difference was attributed to the
6 presence of one or more unmeasured organic nitrate species that are similar to PAN and may
7 be of photochemical origin.
8
9
0 6.6 PEROXYACETYL NITRATE (PAN)
1 Several methods have been used to measure the concentration of PAN in ambient air.
2 Stephens (1969) and Roberts (1990) have provided a good overview of many of these
3 methods. Peroxyacetyl nitrate was first measured by using long-path infrared spectrometry;
4 however, insufficient sensitivity by this technique prompted the development of other
5 methods (Darley et al., 1963). A ground-based FTIR system with a 1-km cell has reported
6 detection limits of 4 ppb for PAN near 790 and 1,160 cm'1 (Tuazon et al., 1978). The
7 limited sensitivity and the complexity of FTIR systems have generally limited ambient
8 applications of the FTIR to the relatively high concentrations associated with the Los Angeles
9 basin. More recently, cryogenic sampling and matrix-isolation FTIR has been used to
0 measure PAN in 15-L integrated samples of ambient air with a theoretical MDL of 50 ppt
1 (Griffith and Schuster, 1987). A laboratory prototype method, TTFMS, has a projected
2 MDL for PAN of 2 ppt (Hansen, 1989). Gas chromatography with flame ionization
3 detection (GC-FID) may be employed to measure PAN, but this method is only practical for
4 concentrated mixtures above 10 ppm using a 1-mL sample loop (Meyrahn et al., 1987). The
5 most common method is gas chromatography using electron capture detection (GC-ECD)
6 (Darley et al., 1963; Smith et al., 1972; Stephens and Price, 1973; Singh and Salas, 1983).
7
8 6.6.1 Gas Chromatography-Electron Capture Detection (GC-ECD)
9 Both manual and automated integrated sampling methods using GC-ECD have been
0 employed (Stephens and Price, 1973; Lonneman et al., 1976). Relatively low column and
August 1991 6-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 detector temperatures (below 50 and 100 °C, respectively) have been used to minimize
2 thermal decomposition of PAN. Short packed columns coated with polyethylene glycol-type
3 stationary phases (e.g., Carbowax 400) have normally been used. Recently, improved
4 precision and sensitivity have been achieved using silica capillary columns (Helmig et al.,
5 1989; Roberts et al., 1989). Although sampling intervals are limited by the elution times of
6 the chromatographic system, intervals of 10 to 15 min have been employed (Helmig et al.,
7 1989; Nieboer and van Ham, 1976). Using packed columns, detection limits of 10 ppt have
8 been reported using direct sampling with a 20-mL sample loop (Vierkorn-Rudolph et al.,
9 1985), and 1 to 5 ppt using cryogenic enrichment of samples (Vierkorn-Rudolph et al., 1985;
10 Singh and Salas, 1983). Capillary columns offer the potential for considerable (i.e., factor of
11 20) enhancement in sensitivity (Roberts et al., 1989). Accuracy estimates of ±20 to 30%
12 have been claimed.
13 A comparison of two similar GC-ECD methods for airborne PAN measurements was
14 performed (Gregory et al., 1990c). Both methods employed cryogenic enrichment of
15 samples, used packed GC columns, and claimed detection limits below 5 ppt. Results of this
16 study showed that at PAN concentrations below 100 ppt, agreement was approximately
17 17 ppt, and at higher concentrations (i.e., 100 to 300 ppt) the measurements agreed to within
18 25% (expressed as a percent difference). These findings are generally consistent with
19 accuracy claims noted earlier.
20
21 6,6.2 Alkaline Hydrolysis
22 Alkaline hydrolysis in 5% NaOH has been shown by Nicksic et al, (1967) to convert
23 PAN quantitatively to nitrite and acetate. This permits sampling with a bubbler containing
24 5% NaOH and subsequent analysis for nitrite or acetate. Nitrogen dioxide is usually present
25 in ambient air with PAN. It can interfere with PAN determination as nitrite since NO2 may
26 be collected as nitrite in alkaline solution. Acetate particles or acetic acid can interfere with
27 PAN determination as acetate. A method involving alkaline hydrolysis followed by 1C
28 determination of acetate has been used to measure PAN in photochemical systems (Grosjean
29 and Harrison, 1985a). The results compare favorably with those of a CLM method
30 employing difference in NOX signals measured upstream and downstream of an alkaline
31 bubbler. In addition to NaOH, other alkaline salts (e.g., KOH and Na2CO3) have been used
August 1991 6-42 DRAFT-DO NOT QUOTE OR CITE
-------
I to coat filters, cartridges, and annular denuders (Grosjean and Parmar, 1990; Williams and
I Grosjean, 1990). Peroxyacetyl nitrate collection efficiencies ranged from 10 to 100%
5 depending on the type and amount of the alkaline salt, the flow rate, and the collection device
!• employed.
5
j 6.6.3 Gas Chromatography (GC)-Alternate Detectors
7 As noted in Section 6.3.1, PAN is readily reduced to NO. Meyrahn et al. (1987) have
5 coupled a GC to separate PAN, NO, and NO2; a molybdenum converter; and a CLM NO
) analyzer to measure PAN as NO. Using a 10 mL sample loop, a detection limit of 10 ppb
) was reported.
The luminol-based detector has shown sensitivity to PAN (see Section 6.3.2).
5 Burkhardt et al. (1988) have used gas chromatography and a commercially available luminol-
\ based instrument (i.e., Scintrex LMA-3 Lummox) to detect both NO2 and PAN. Using a
\ sampling interval of 40 s, linear response was claimed from 0.2 to 170 ppb NO2 and from
> 1 to 65 ppb PAN. Although the PAN calibration .was nonlinear below 1 ppb, and MDL of
j 0.12 ppb was reported. Drummond et al. (1989) have slightly modified the above approach
1 by converting the PAN from the GC column to NO2 and measuring the resulting NO2 with a
> CLM (luminol) instrument.
) 6.6.4 Peroxyacetyl Nitrate Stability
Peroxyacetyl nitrate is an unstable gas and is subject to surface-related decomposition as
! well as thermal instability. Peroxyacetyl nitrate exists in a temperature-sensitive equilibrium
i with the peroxyacetyl radical and NO2 (Cox and Roffey, 1977). Increased temperature
favors the peroxyacetyl radical and NO2 at the expense of PAN. Added NO2 should force
i the equilibrium toward PAN and enhance its stability. In the presence of NO, peroxyacetyl
i radicals react rapidly to form NO2 and acetoxy radicals which decompose in O2 to radicals
' which also convert NO to NO2. As a result, the presence of NO acts to reduce PAN stability
\ and enhance its decay rate (Lonneman et al., 1982). Stephens (1969) reported that
1 appreciable PAN loss in a metal sampling valve was traced to decomposition on a silver-
l soldered joint. Meyrahn et al. (1987) reported that PAN decayed according to first order
kinetics at a rate of 2 to 4% h"1 in glass vessels that had been previously conditioned with
August 1991 , 6-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 PAN. They employed 200 ppm PAN in glass vessels and the noted first order decay as the
2 basis for one proposed method of in-field PAN calibration. In contrast, Holdren and Spicer
3 (1984) found that without NO2 added, 20 ppb PAN decayed in Tedlar bags according to first
4 order kinetics at a rate of 40% h"1. The addition of 100 ppb NO2 acted to stabilize the PAN
5 (20 ppb) in the Tedlar bags.
6 A humidity-related difference in GC-ECD response has been reported (Holdren and
7 Rasmussen, 1976). Low responses observed at humidities below 30% and PAN
8 concentrations of 10 and 100 ppb but not 1,000 ppb were attributed to sample-column
9 interactions. This effect was not observed by Lonneman (1977). Watanabe and Stephens
10 (1978) conducted experiments at 140 ppb and did not conclude that the reduced response was
11 from faults in the detector or the instrument. They concluded that there was no column-
12 related effect, and they observed surface-related sorption by PAN at 140 ppb in dry acid-
13 washed glass flasks. They recommended that moist air be used to prepare PAN calibration
14 mixtures to avoid potential surface-mediated effects.
15 Another surface-related effect has been reported for PAN analyses of remote marine air
16 (Singh and Viezee, 1988). Peroxyaeetyl nitrate concentrations were found to increase by 20
17 to 170 ppt, an average factor of 3.2, when the sample was stored in a glass vessel for 1 to
18 2 min prior to analysis. This effect remains to be explained.
19
20 6.6.5 Calibration
21 Since PAN is unstable, the preparation of reliable calibration standards is difficult.
22 Several methods have been employed. The original method used the photolysis of ethyl
23 nitrite in pure oxygen (Stephens, 1969). When pure PAN is desired, the reaction mixture
24 must be purified, usually by chromatography, to remove the major by-products, acetaldehyde,
25 and methyl and ethyl nitrate (Stephens et al., 1965). For GC calibration, purification is
26 unnecessary; the PAN concentration in the reactant matrix is established from the IR
27 absorption spectrum and subsequently diluted to the ppb working range needed for calibration
28 purposes (Stephens and Price, 1973).
29 Static mixtures of molecular chlorine, acetaldehyde, and NO2 in the ratio of 2:4:4 can
30 be photolyzed in the presence of a slight excess NO2 to give a near stoichiometric yield of
31 PAN (Gay et al., 1976). This method was adapted by Singh and Salas (1983) and later by
August 1991 6-44 DRAFT-DO NOT QUOTE OR CITE
-------
Grosjean et al. (1984) using photolytic reactors to provide continuous PAN calibration units
at concentrations between 2 and 400 ppb. In the former approach, the PAN concentration is
established by measuring the change in acetaldehyde concentration across the reactor. In the
latter, the PAN concentration is established by measuring the acetate in an alkaline bubbler
where PAN is hydrolyzed.
A static technique involving the photolysis of acetone in the presence of NO2 and air at
250 nm has been reported to produce a constant concentration of PAN (Meyrahn et al.,
1987). A Penray Hg lamp is inserted into a mixture of 10 ppm N02 and 1% acetone and
irradiated for 3 min to yield 8.9 ± 0.3 ppm PAN.
Peroxyacetyl nitrate can be synthesized in the condensed phase by the nitration of
peracetic acid in hexane (Helmig et al., 1989), heptane (Nielsen et al., 1982), octane
(Holdren and Spicer, 1984), or n-tridecane (Gaffney et al., 1984). Purification of PAN in
the liquid phase is needed using the first two methods. The resulting PAN-organic solution
can be stored at -20 to -80 °C with losses of less than 3.6% month"1 and can be injected
directly into a vessel containing air to produce a calibration mixture. The PAN concentration
is normally established by FTIR analysis of the solution or the resulting PAN-air mixture.
As noted in Section 6.3.1, PAN is readily reduced to NO, and CLM NOX analyzers
have near quantitative response to PAN. Thus under some circumstances, CLM NOX
response can be used for PAN calibration. One method uses the difference in NOX signal
measured upstream and downstream of an alkaline bubbler (Grosjean and Harrison, 1985a).
Joos et al. (1986) have coupled a CLM NOX analyzer with a GC system to permit calibration
of the BCD response by reference to the CLM NOX analyzer that has been calibrated by
traditional methods.
As noted previously, NO in the presence of PAN is converted to NO2. Approximately
4 molecules of NO can react per molecule of PAN. Lonneman et al, (1982) have devised a
PAN calibration procedure based on the reaction of PAN with NO in the presence of
benzaldehyde which is added to control unwanted radical chemistry and improve precision.
Using this approach and an initial NO to PAN ratio of between 10 and 20 to one, the change
in NO concentration is monitored with a CLM NO analyzer, the change in PAN GC-ECD
response is monitored, and the resulting ratio (i.e., ANO/APAN) is divided by the
stoichiometric factor of 4.7 to arrive at a calibration factor for the BCD.
August 1991 6-45 DRAFT-DO NOT QUOTE OR CITE
-------
1 Peroxyaeetyl nitrate and n-propyl nitrate (NPN) have similar BCD responses. Serial
2 dilution of the more stable compound, NPN, has been used for field operations (Vierkorn-
3 Rudolph et al., 1985). This approach is not recommended for primary calibration, however,
4 since it does not permit verification of quantitative delivery of PAN to the detector (Stephens
5 and Price, 1973).
6
7 6.6.6 Other Organic Nitrates
8 Other organic nitrates (e.g., alkyl nitrates, peroxypropionyl nitrate (PPN) and
9 peroxybenzoyl nitrate (PBzN)) are also present in the atmosphere but usually at lower
10 concentrations than PAN (Fahey et al., 1986). In general, similar methods for sampling,
11 analysis, and calibration may be used for other organic nitrates as are used for PAN
12 (Stephens, 1969). Both FTIR and GC-EC may be used to measure these compounds.
13 With MDLs of 0.1 to 0.4 ppb, inspection of 3,000 GC-ECD chromatograms recorded at
14 five to nine sites during the 1987 Southern California Air Quality Study yielded only seven
15 possible (but nonprobable) observations of methyl nitrate (Grosjean et al., 1990). Roberts
16 et al. (1989) have reported separation of PAN, PPN, and Cj to C4 alkyl nitrates and the
17 potential increase in sensitivity by a factor of 20 using fused silica coated capillary columns
18 rather than the more conventional coated packed columns. Atlas (1988) has used two 5-mg
19 charcoal traps in series to collect C3 to C7 alkyl nitrates from 12-to-300 L samples at 200 to
20 400 mL min"1 in remote atmospheres. The traps are extracted in small volumes of benzene
21 and analyzed using capillary GC-ECD. Concentrations as low as 1 ppt were reported.
22 Peroxybenzoyl nitrate may be collected as methyl benzoate using bubblers containing
23 methanol-NaOH solutions (Appel, 1973). The resulting methyl benzoate is solvent extracted
24 and analyzed by packed column GC-FID with an MDL of 70 ppt. Recently, a collection
25 method using aqueous alkaline hydrolysis of PBzN to the benzoate ion followed by IC-UV
26 analysis was reported to have a detection limit of 30 ppt in a 60 L sample (Fung and
27 Grosjean, 1985). Using this method a median PBzN level of 0.32 ppb was reported for Los
28 Angeles air samples.
29
30
31
August 1991 6-46 DRAFT-DO NOT QUOTE. OR CITE
-------
1 6.7 NITRIC ACID (HNO3)
2 Several methods are available for the determination of airborne concentrations of
3 HNO3. Among them are filtration (OMta et al., 1976; Spicer et al., 1978b), denuder tubes
4 (Forrest et al., 1982; DeSantis et al., 1985; Perm, 1986), CLM (Joseph and Spicer, 1978),
5 absorption spectroscopy (Tuazon et al., 1978; Schiff et al., 1983; Biermann et al., 1988),
6 and microcoulometry (Spicer et al., 1978b). As a result of its 2 ppb detection limit and long
7 response time, microcoulometry has been largely replaced by other methods. Consequently,
3 only the first four methods listed above are described in this section.
}
3 6.7.1 Filtration
1 Filtration techniques generally employ dual filters that rely on the collection of
I particulate NO3" on the first filter and gaseous HNO3 as NO3" on the second filter. This
3 method is sometimes called the filter pack (FP) method. Typically, filtration is used in
1- conjunction with instrumental detection or subsequent chemical analysis of the material
> collected on the filter media. Filter extracts are usually analyzed for NO3~ using 1C.
'•> Efficient HNO3 collection has been found with nylon filters (Spicer et al., 1978b) and with
1 filters impregnated with NaCl or NaF (OMta et al., 1976; Forrest et al., 1980; Fuglsang,
\ 1986). The HNO3 capacity of 47-mm diameter NaCl- coated filters (500 jig cm'2) far
) exceeds that of nylon (30 /tg cm"2) (Anlauf et al., 1986). This advantage may be offset,
) since the presence of the chloride ion in the NaCl-coated filter extract may hamper 1C
[ determination of NO3". With a 47-mm diameter nylon filter sampling at 1 m3 h"1 at a
I nominal HNO3 level of 5 ptg m"3 (2 ppb), the capacity is sufficient for just over 4 days of
I sampling. The sensitivity of filtration and other integrative methods depends on the detection
!• limit of the analytical finish, the variability and magnitude of the blank level, collection and
> extraction efficiencies, and the volume of air sampled. As an example, under an assumed 1C
> detection limit for NO3" of 0.05 #g mL"1 a filter extraction volume of 10 mL, negligible
j-»
' blank, quantitative collection and extraction, and a sampled air volume of 24 m (i.e., flow
> rate 1 m3 h"1 for 1 day), the minimum sensitivity is 0.02 /tg m""3 (8 ppt). A precision of
> ±10% and an accuracy of +20 to -40% are claimed for FP containing Teflon1™ and nylon
> filters (Fahey et al., 1986).
August 1991 6-47 DRAFT-DO NOT QUOTE OR CITE
-------
1 Although the HNO3 determination by FP methods is desirable due to low cost,
2 simplicity, and high sensitivity, there is great difficulty in distinguishing between gaseous and
3 particulate forms of nitrate. Errors in the measurement of gaseous HNO3 may be in the form
4 of positive artifacts due to volatilization of collected aerosol nitrates on the prefilter to form
5 gaseous HNO3 (i.e., NH4NO3 «» NH3 + HNO3) (Appel et al., 1980); reaction of collected
6 particulate nitrates on the prefilter with strong acids, resulting in the release of HNO3 (i.e.,
7 H2SO4 + 2NH4NO3 •* (NH^SC^ + 2HNO3) (Appel and Tokiwa, 1981); or formation of
8 HNO3 on the collection medium by interaction with other NOX species (e.g., HNO2 or NO^
9 (Eatough et al., 1988; Spicer and Schumacher, 1979). Negative HNO3 artifacts may result
10 from retention of HNO3 by the prefilter collection medium (Appel et al., 1984); retention of
11 HNO3 by collected particles on the prefilter (Appel et al., 1980); a low capacity for HNO3
12 on the collection medium; or losses of HNO3 by volatilization or by displacement by other
13 acids.
14 Inert prefilter materials, such as Teflon™, should not collect appreciable amounts of
15 HNO3 (Appel et al., 1979); this, however, does not preclude the possibility of HNO3
16 reaction with aerosol particles collected on Teflon™ prefilters. In addition, some types of
17 Teflon™ may sorb HNO3 to a larger extent than others (Appel et al., 1988). This
18 underscores the importance of using "inert" materials for all surfaces coming into contact
19 with HNO3 to insure representative sampling.
20
21 6.7.2 Denuders
22 To avoid some of the artifact problems associated with the use of filters, denuder tube
23 samplers were introduced. In general, a denuder is a tube or channel that has its walls coated
24 with or fabricated from a substance that removes the gaseous species of interest, in this case
25 HNO3 (also see Section 6.3.6). The HNO3 molecules diffuse to and impact the surface while
26 the sample is drawn through the channel. The flow conditions are usually laminar
27 (Re < 2,000) and by taking advantage of differences in diffusivities permit particles to pass
28 through the denuder relatively undisturbed. Using the sampled air volume, the concentration
29 of HNO3 is calculated from the measured amount of NO3" collected on the denuder walls
30 (Perm, 1986) or from the difference of NO3" collected downstream in the presence and
31 absence of the denuder (Shaw et al., 1982; Forrest et al., 1982).
August 1991 6-48 DRAFT-DO NOT QUOTE OR CITE
-------
Denuder tubes have employed MgO (Shaw et al., 1982), Na2CO3 (Perm, 1986), nylon
(Mulawa and Cadle, 1985), A12(SO4)3 (Lindqvist, 1985), MgSO4, BaSO4 (Klockow, 1989),
and tungstic acid (H2WO4) (McClenny et al., 1982) to retain HNO3. A coating of MgO is
frequently used with the denuder difference (DD) approach where one coated denuder and
one uncoated denuder are followed by nylon or NaCl-coated filters. The difference in NO3"
on the two downstream filters is attributed to HNO3 (Shaw et al., 1982). The Na2CO3
coating is used in methods employing the DD or direct analysis approach. In the latter, the
denuder is extracted, and the extract is analyzed for NO3" which is attributed to HNO3
(Perm, 1986).
To maintain laminar flow, the flow rate through many conventional open channel
denuders of reasonable dimensions is limited to approximately 1 to 2 1 min"1. As a result,
long duration samples may be required to provide sufficient analyte for quantitation. A new
type of denuder, the annular denuder (AD), has been developed where the same equivalent
diameter permits the flow rate to be increased by a factor of 12 (Possanzini et al., 1983). In
the AD, ambient air is passed through the annular space of two concentric tubes. The outside
of the inner tube and the inside of the outer tube are coated with a specific gas-absorbing
substance. For collecting HNO3, Na2CO3 has been used (DeSantis et al., 1985). In cases
where appreciable HNO2 is present it may be cocollected on Na2CO3 and the resulting NO2"
oxidized to NO3" over extended sampling periods by atmospheric oxidants (e.g., O3), two or
more denuders are used to permit resolution of HNO3 and HNO2 (Febo et al., 1986; Perrino
et al., 1990). The first denuder is coated with NaCl or NaF to collect HNO3 as NO3~ and
the downstream denuder(s) is coated with an aqueous solution of Na2CO3 and glycerol to
collect HNO2 as the sum of NO2" and NO3". For aim3 h"1 1-day sample under the same
assumptions given earlier for filtration, the MDL for the AD is 0.02 /ig irf3 (8 ppt). Median
precision estimates of 8 and 5% RSD have been reported for thirteen 22-h and twelve 1-wk
duration samples (Sickles et al., 1989; Sickles, 1987).
Partial denuders have been fabricated of nylon filter material (Mulawa and Cadle,
1985). These denuders operated under laminar flow conditions have relied on a mathematical
description of molecular diffusion to a perfect wall sink along with HNO3 deposition
measured along the length of the denuder to infer the sampled, ambient HNO3 concentration.
A refinement in the data treatment has been offered recently that considers interferent nitrate
August 1991 6-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 on nylon partial denuders (Febo et al., 1988). While these denuders are operated under
2 laminar flow conditions, a recently introduced technique employs a nylon partial denuder that
3 is operated under transition flow conditions (Re «2,600) (Knapp et al., 1986). This
4 approach, called a transition flow reactor (TFR), uses a piece of nylon filter material rolled
5 into a cylindrical shape and placed in a Teflon™ tube. The sample is drawn through the tube
6 and denuder under transition flow conditions where a constant fraction of the HNO3 is
7 claimed to be collected. The tube is followed by a Teflon™ and a nylon filter. The HNO3 is
8 calculated by analyzing the denuder extract for NO3" and applying the constant collection
9 fraction. The paniculate NO3" is determined algebraically using the NO3" measured in the
10 extracts of the Teflon™ and nylon filters. For 8.5% collection efficiency and 1 m3 h"1 1-day
11 sample under the same assumptions given earlier for filtration, the MDL for the TFR is
12 0.2 fj,g m-3 (0.1 ppb). A median precision of 7% RSD has been determined for seven 1-wk
13 duration samples (Knapp et al,, 1986).
14 Automated systems using coated denuders with thermal desorption employ Al2(SO4)3,
15 MgSC>4, BaSO4, or tungstic acid (TA) to preconcentrate HNO3 for subsequent delivery to an
16 instrumental detection system. In the first case, HNO3 from a 30-L sample is collected on an
17 Al2(SO4)3-coated denuder, thermally desorbed, thermally converted to NO, and analyzed by
18 gas chromatography with a photoionization detector (Lindqvist, 1985). A nominal MDL of
19 5 ppt and precision estimate of ±10% were claimed.
20 Klockow et al. (1989) have used MgSO4- and BaSO4-coated denuders to collect HNO3.
21 The sample is thermally desorbed and measured with a CLM NOX analyzer. For a 30-min
22 sample at 5 L min"1, a nominal MDL of 0.1 /*g/m3 (40 ppt) and precision estimate of ±5%
23 were claimed.
24 A TA-coated denuder has been used to collect HNO3 for analysis on an automated basis
25 with a 40-min cycle time (McClenny et al., 1982). The collected HNO3 is thermally
26 desorbed as NO2, thermally converted to NO, and measured with a commercial CLM NOX
27 analyzer. A nominal MDL of 70 ppt and a precision estimate of +10% were claimed.
28 Recent claims for a similar device with a 20-min cycle time have included MDL of 20 ppt,
29 accuracy of 15 to 20%, and precision of 8% (Gregory et al., 1990d). Tungstic acid-coated
30 denuders have drawbacks: they are difficult to prepare, have low capacities and are subject
August 1991 6-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 to unknown atmospheric interferences (Fellin et al., 1984; Eatough et al., 1985; Roberts
2 et al., 1987).
3
4- 6.7.3 Chemiluminescence (CLM)
5 As noted in Section 6.3.1, HNO3 is readily reduced to NO in NO2-to-NO converters
3 used in commercial and many research grade CLM NOX analyzers (Joseph and Spicer, 1978;
7 Bollinger et al., 1983; Fahey et al., 1985a; Grosjean and Harrison, 1985b; Rickman and
3 Wright, 1986). Nitric acid measurements that employ CLM generally use a NOX analyzer to
) measure the NOX in a sampled air stream in the presence and absence of a HNO3 scrubber.
) The difference in these NOX signals is attributed to HNO3. Nylon filters have been used as a
I HNO3 scrubber with both commercial and research grade CLM NOX analyzers. The
I instrumental performance for HNO3 is similar to that for NO2 with the same instruments
} (Joseph and Spicer, 1978; Kelly et al., 1979).
1- In other methods employing CLM and described in Section 6.7.2, TA, MgSO4, or
5 BaSO4 are used as regenerable HNO3 scrubbers (McClenny et al., 1982; Klockow et al.,
y 1989). With these methods, HNO3 is collected on a coated denuder, the collected HNO3 is
7 thermally desorbed as NO2 (regenerating the scrubber), the desorbed NO2 is thermally
\ converted to NO, and the resulting NO is measured with a commercial CLM analyzer.
) 6.7.4 Absorption Spectroscopy
Absorption spectroscopy is discussed in Sections 6.2.3 and 6.3.4. Although FTIR,
I TDLAS, and potentially TTFMS techniques may be used to measure ambient levels of
S HNO3, poor sensitivity limits ambient applications of FTIR. A ground-based 23-m multipass
I- FTIR system with a 1-km path length has reported detection limits of 4 ppb near 900 cm"1
I (Tuazon et al., 1978; Biermann et al., 1988). A theoretical MDL for HNO3 of 10 ppt has
5 been claimed for 15-L integrated samples of ambient air using cryogenic sampling and
7 matrix-isolation FTIR (Griffith and Schuster, 1987).
! Cassidy and Reid (1982) report an expected MDL of 0.4 ppb using TDLAS near
> 1,330 cm"1. For a 40-m path length near 1,720 cm"1, the MDL is 0.4 ppb (Schiff et al.,
) 1983). With a 150-m path length Mackay and Schiff (1987) report an MDL of 0.1 ppb and
an accuracy of ±20%. Although the volumetric residence time in the White cell of the
August 1991 6-51 DRAFT-DO NOT QUOTE OR CITE
-------
1 TOLAS is 4 s, sample-surface interactions limit the response time to changes in HNO3
2 concentration to about 5 min. As described in Section 6.2.3, a laboratory prototype method,
3 TTFMS, has been developed (Hansen, 1989). The projected MDL for HNO3 is 0.3 ppt.
4
5 6.7.5 Calibration
6 Nitric acid is a highly polar material and consequently interacts readily with many
7 surfaces (Goldan et al., 1983; Appel et al., 1988). This reactivity prevents the preparation of
8 stable calibration mixtures in cylinders of compressed gases. Two methods, permeation
9 devices and diffusion tubes, are generally employed to generate calibration atmospheres of
10 HNO3 (Schiff et al., 1983; Goldan et al., 1983). Permeation tubes are described for NO2 in
11 Section 6.3.8. Permeation tubes for HNO3 are available with various emission rates from
12 commercial suppliers. An alternate permeation device may be fabricated in the laboratory by
13 passing carrier gas through a length of Teflon™ tubing that is immersed in a reservoir of
14 HNO3 and H2SO4 (Mackay and Schiff, 1987).
15 Diffusion tubes are generally fabricated in the laboratory (Schiff et al., 1983). A liquid
16 mixture of HNO3 and H2SO4 is held in a reservoir that is connected to a clean air dilution
17 manifold by a capillary tube. The HNO3 diffusion rate depends on the length and area of the
18 capillary as well as the temperature of the reservoir.
19 Nominal HNO3 emission rates for permeation tubes are provided by the supplier and are
20 calculated for diffusion tubes (Nelson, 1971). Although it is common to calibrate permeation
21 tubes gravimetrically, it has been reported that non-HNO3 species (i.e., NO2) are also
22 released and may account for 10 to 15% of the observed weight loss (Goldan et al., 1983).
23 Since the emission rate estimate for diffusion tubes is also an approximation, the independent
24 measurement of HNO3 emission rates from permeation and diffusion tubes is recommended.
25 This measurement may be accomplished by pH titration or by using nylon filters, NaCl-
26 coated filters or caustic bubblers to collect and quantify the HNO3 as NO3". Since caustic
27 bubblers may also collect NO2 to some extent, their use could overestimate the HNO3
28 emission rate, and a filtration technique is preferred. Alternative, but more elaborate
29 methods of confirming the HNO3 emission rate are with FTIR, TDLAS, and the CLM NOX
30 analyzer that uses photolysis to convert NO2 to NO.
31
August 1991 6-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.7.6 Intercomparisons
2 Several field studies have been conducted that have permitted the comparison of
3 different techniques for the measurement of HNO3 (Spicer et al., 1982; Walega et al,, 1984;
4 Anlauf et al., 1985; Roberts et al., 1987; Hering et al., 1988; Solomon et al., 1988; Benner
5 et al., 1987; Tanner et al., 1989; Sickles et al., 1990; Gregory et al., 1990d; Dasch et al.,
6 1989). Results from these studies suggest that the FP overestimates the HNO3 concentrations
7 and that coated denuder thermal desorption techniques in various tested configurations may
8 not provide reliable measurements of HNO3.
9 An intercomparison of HNO3 measurement methods was conducted in Claremont, CA,
0 in August and September of 1979 (Spicer et al., 1982). Ten methods were compared: 5 FP,
1 2 DD, 2 CLM, and 1 FTIR. The results of 5 methods (i.e., 2 FP, 1 DD, 1 CLM, and
2 1 FTIR) were in excellent agreement with median results.
3 Walega et al. (1984) report comparisons of CLM and TOLAS HNO3 measurements of
4 ambient and captive air performed during October and November of 1981 in Los Angeles,
5 CA. The CLM gave erratic HNO3 results for ambient air. Although CLM and TOLAS
6 measurements of HNO3 in captive air samples were highly correlated, linear regression
7 analysis indicated significant biases,
8 Measurements of HNO3 were made during June 1982 at a rural site in Ontario using
9 FP, TOLAS, and TA techniques (Anlauf et al., 1985). For daytime measurements, the FP
0 and TA measurements were 16% lower than the TDLAS results. Nighttime TA results
1 exceeded those from the FP by a factor of 2. Roberts et al. (1987) compared FP and TA
2 measurements of HNO3 made at a rural site in the Colorado mountains. The TA results were
3 a factor of 3 higher than those of the FP. It was concluded that there are unknown
4 atmospheric species that interfere with TA measurements of HNO3.
5 Another HNO3 intercomparison study was conducted in Claremont, CA in September
6 1985 (Hering et al., 1988). The methods compared include FP, DD, AD, TFR, TA, FTIR,
7 and TDLAS. For the whole study, comparison of method means against mean of methods
8 showed the FP to be 36% high, the DD to be 1% low, the AD to be 21% low, and the
9 TDLAS to be 13% low. Comparison of TFR means against DD means showed the TFR to
0 be 9% high. Tunable-diode laser spectroscopy gave lower daytime and higher nighttime
1 readings than the DD. In those few cases where the HNO3 concentrations were sufficiently
August 1991 6-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 high to be detected by the FTIR, agreement within reported uncertainties was observed
2 between the FTIR and the FP, DO, AD, and TOLAS. Results from the TA technique were
3 high at night and low during the day, and in view of large systematic differences, they were
4 not included in many of the reported analyses.
5 During 1986, HNO3 data were collected using DD and FP techniques for 24-h periods
6 every 6 days at 8 sites in the Los Angeles basin and at one background site (Solomon et al.,
7 1988). The annual average DD basin-wide estimate of HNO3 was 4.6 ^g mf3. The
8 corresponding FP estimate exceeded that of the DD by 3.4 ^g m"3 or by approximately 80%.
9 A study was conducted in January and February 1986 near Page, AZ (Benner et al.,
10 1987). Twelve hour gaseous HNO3 concentrations were measured with FP and AD. The
11 mean HNO3 concentration measured with the FP, 1.1 ^g m"3, exceeded that measured with
12 the AD by 10%.
13 A study was conducted in July 1986 on Long Island, New York, to compare HNO3
14 measurements resolved to a 6-h basis using high-volume FP, DD, real-time two-channel (i.e.,
15 nylon filter versus no nylon filter) CLM, and Al2(SO4)3-coated denuder thermal desorption-
16 to-CLM (Tanner et al., 1989). The FP results were highly correlated with those of the DD.
17 The daytime real-time CLM results were correlated with those of the DD, but nighttime real-
18 time CLM results exceeded DD results. This may have been caused by the retention of
19 nighttime HNO2 on the nylon filter. Results with the A^SO^-coated denuder were
20 scattered, mostly lower, and poorly correlated with the other methods.
21 Daily measurements of HNO3 were made in the Research Triangle Park, NC during
22 13 days in September and October, 1986 (Sickles et al., 1990). Comparisons of the TOLAS
23 results with those of the AD, FP, and TFR revealed significant differences at the 0.05 level
24 for the comparison between the TOLAS and TFR results. Significant differences were not
25 apparent in the other two cases. Comparisons of the study wide means of daily ratios of AD,
26 FP, and TFR to TOLAS results showed the AD to be 5% low, the FP to be 8% high, and
27 the TFR to be 36 to 76% high.
28
29
30
August 1991 6-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 6.8 NITROUS ACID (HNO2)
2 The measurement of HNO2 in ambient atmospheres is receiving increased recent
3 attention. Currently available techniques employ denuders (Perm and Sjodin, 1985), AD
4 (DeSantis et al., 1985) CLM (Braman et al., 1986), PF/LIF (Rodgers and Davis, 1989) and
5 absorption spectroscopy (Biermann et al., 1988; Tuazon et al., 1978).
6 - • .
7 6.8.1 Denuders
8 See Sections 6.3.6 and 6.7.2 for additional discussions of denuders. As noted, in
9 Section 6.2.1, Braman et al. (1986) have employed a series of open channel denuders coated
0 with materials that act to preconcentrate HNO3, HNO2, NO2, and NO from sampled ambient
1 air. Nitrous acid is collected using a potassium-iron oxide-coated denuder located
2 downstream of a TA-eoated denuder that removes HNO3. The HNO2 is thermally desorbed
3 from the potassium-iron oxide-coated denuder and detected as NO with a CLM NO analyzer.
4 Although sub-ppb sensitivity is claimed, field testing is needed to demonstrate the adequacy
5 of this method.
6 Nylon filter material has also been used as an open channel denuder to collect HNO2
7 (Benner et al., 1988). Recent studies, however, have indicated that HNO2 may not be
8 retained quantitatively by nylon filters (Sickles and Hodson, 1989; Perrino et al., 1988).
9 Perm and Sjodin (1985) have used two conventional open channel Na2CO3-eoated
3 denuders in series for the determination of HNO2 in ambient air. Nitrous acid is collected
1 quantitatively on the first denuder, while interferent artifacts from PAN and other NOX
2 species (i.e., NO2) are collected in approximately equal amounts on both denuders. Each
3 denuder is extracted and the extract is analyzed for NO2" using spectrophometry or 1C. To
4 correct for interferent artifacts, the difference in NO2" found on the two denuders is attributed
5 to HNO2.
5 Annular denuders have also been used to measure HNO2 using a similar approach
7 (DeSantis et al., 1985; Sickles, 1987; Sickles et al., 1988; Eatough et al., 1988; Vossler
B et al., 1988; KoutraMs et al., 1988; Dasch et al., 1989; Appel et aJL, 1990; Perrino et al.,
? 1990). The MDL for a 1-day AD sample operating at 1 m3 h"1, assuming an extract volume
3 of 10 mL, negligible blank, and an 1C detection limit of 0.05 /ig NO2" mL~l, is 0.02 /ig m"3
I (10 ppt). Estimates of precision for 1-day AD samples range from 5 to 15% (Sickles et al.,
August 1991 6-55 DRAFT-DO NOT QUOTE OR CITE
-------
1 1989; Vossler et al., 1988). In those cases where denuder sampling is performed over
2 extended periods in the presence of oxidants (i.e., O3), the collected NO2" may be oxidized
3 to NO3" (Febo et al., 1986; Sickles et al., 1989; Sickles and Hodson, 1989; Perinno et al.,
4 1988). To avoid this potential for sampling artifacts an initial denuder coated with NaCl or
5 NaF is added to collect HNO3 as NO3" and pass HNO2. The difference in the sums of NO2"
6 and NO3" on the two downstream Na2CO3-coated denuders is attributed to HNO2 (Febo
7 etal., 1986; Perrinoetal., 1990).
8
9 6.8.2 Chemiluminescenee (CLM)
10 It has been shown that HNO2 may be measured nonspecifically as NOX with a CLM
11 NOX analyzer (Cox, 1974; Sickles and Hodson, 1989). As noted in the previous section,
12 Brarnan et al. (1986) have used a system of selective denuders to collect HNO2 as well as
13 HNO3, NO2, and NO for subsequent thermal desorption and detection as NO with a CLM
14 NO analyzer.
15
16 6.8.3 Photofragmentation/Laser-Induced Fluorescence (PF/LIF)
17 Photofragmentation/laser-induced fluorescence is discussed in Section 6.3.3 for the
18 measurement of NO2. In its present application, HNO2 is photofragmented to NO and OH
19 radical using radiation at 355 nm from a Nd: YAG laser (Rodgers and Davis, 1989).
20 Appreciable amounts of NO are also produced by the photolysis of NO2 which is generally
21 present along with HNO2 in ambient air. As a result, the current method is based on the
22 detection of OH radical using single photon LBF. With this technique the resulting OH is
23 excited to the A22+ state using laser radiation at 282 nm, and the fluorescence at 310 nm
24 that accompanies the A to X transition of the excited OH radical is monitored. Detection
25 limits in the low 10s of ppt for 15-min integration times are claimed.
26
27 6.8.4 Absorption Spectroscopy
28 Although HNO2 is potentially detectable (i.e., MDL of 4 ppb) using a 23-m multipass
29 FTIR system with a 1-km path length, FTIR has not been used to measure the concentration
30 of HNO2 in ambient air (Tuazon et al., 1978). A theoretical MDL for HNO2 of 10 ppt has
31 been claimed for 15-L integrated samples of ambient air using cryogenic sampling and
August 1991 6-56 DRAFT-DO NOT QUOTE OR CITE
-------
1 matrix-isolation FTIR (Griffith and Schuster, 1987). Long-path UV/visible DOAS has been
2 used to determine HNO2 as well as other trace atmospheric constituents (see Section 6,3.4),
3 Using a 25-m multipass open system with a 0.8-km path length at wavelengths near 354 nm,
4 an MDL of 0.6 ppb is claimed (Biermann et al., 1988). Using a single-pass open system
5 with a 10-km path length an MDL of 20 ppt has been reported (Platt and Perner, 1983).
6
7 6.8.5 Calibration
8 The preparation of reliable calibration mixtures containing known concentrations of
9 HNO2 is difficult. Atmospheres containing HNO2 as well as NO2 and NO may be produced
0 by acidifying solutions of NaNO2 with H2SO4 (Cox, 1974). Another more convenient
1 method uses a sublimation source where HNO2 is produced by subliming oxalic acid onto
2 solid NaNO2 at 30 to 60% RH (Braman and de la Cantera, 1986). Small concentrations of
3 HNO3, NO2, and NO may also be generated using the latter technique. Both methods
4 require independent and periodic determination of the HNO2 concentration, since the source
5 strengths are not necessarily constant.
6 In November and December 1987 a study was conducted in Long Beach, California,
7 where simultaneous HNO2 measurements were made using AD and DOAS on 6 days (Appel
8 et al., 1990). The AD samples were integrated over 4 and 6 h, while the 15-min DOAS
9 results were averaged to permit comparison with the AD results. The HNO2 concentrations
0 ranged from less than 1 to approximately 15 ppb, and the AD results were highly correlated
1 with those of the DOAS. Except at the low HNO2 levels that occurred during the mid-day
2 periods where the AD exceeded the DOAS results, the AD results were 7% lower than the
3 DOAS results. This difference is within the ±30% uncertainty of the DOAS results in the
4 study.
5
6
7 6.9 DINITROGEN PENTOXIDE (N2O5) AND NITRATE RADICALS
8 (N03)
9 The NO3 radical photolyzes rapidly, and as a result ambient concentrations are low
0 during daylight hours. Dinitrogen pentoxide exists in a thermally sensitive equilibrium with
1 NO2 and the NO3 radical and can also react heterogeneously with water vapor to produce
August 1991 6-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 HNO3. In addition, the NO3 radical reacts rapidly with NO to produce NO2. In spite of
2 their low ambient concentrations, N2O5 and the NO3 radical may have important roles in
3 both tropospheric and stratospheric NOX chemistry.
4 Although N2O5 has not been measured in the troposphere, it has been observed in the -
5 stratosphere using spectroscopic methods (Roscoe, 1982). In the troposphere nighttime N2O5
6 concentrations of up to 15 ppb have been inferred under the assumption of equilibrium using
7 measured NO2 and NO3 radical concentrations (Atkinson et al., 1986). At concentrations
8 above 5 ppb, measurement of N2O5 with FTIR spectrometry appears feasible using a 1-km
9 path length near 1,250 cm"1. A theoretical MDL for N2O5 of 20 ppt has been claimed for
10 15-L integrated samples of ambient air using cryogenic sampling and matrix-isolation FTIR
11 (Griffith and Schuster, 1987).
12 Dinitrogen pentoxide is readily reduced to NO at temperatures above 200 °C and, as
13 noted in Section 6.3.1, may be measured nonspecifically as NOX with CLM NOX analyzers
14 (Bellinger et al., 1983; Fahey et al., 1985a). A N2O5 calibration system has been devised
15 using a crystalline sample at -80 °C, thermal dissociation of gaseous N2O5, scavenging of the
16 dissociation product (i.e., the NO3 radical) with added NO to produce NO2, and a CLM NO
17 detector (Fahey et al., 1985b). This calibration technique focuses on the loss of NO, and an
18 accuracy of ±15% is claimed.
19 Ambient concentrations of the NO3 radical have been made using DOAS, and
20 concentrations between 1 and 430 ppt have been observed (Atkinson et al., 1986).
21 Additional information on absorption spectroscopy is given in Section 6.3.4. Using a 25-m
22 multipass open system with 0.8 km path length, an MDL of 20 ppt is claimed (Biermann
23 et al., 1988). Using an optical path length of 17 km and a wavelength of 662 nm, the
24 reported detection limit for the NO3 radical is 1 ppt (Platt et al., 1984). Noxon (1983) using
25 a passive absorption spectroscopic method with the moon as the light source reports a NO3
26 concentration of 0.25 ppt measured at a 3 km altitude from Mauna Loa, Hawaii.
27
28
29 6.10 PARTICULATE NITRATE (PN)
30 Atmospheric aerosols are chemically heterogeneous and occur in sizes ranging
31 nominally from <0.01 to 100 jwn. Many methods are available for sampling ambient
August 1991 6-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 aerosols, including impactors, filtration, and filtration coupled with devices to remove
2 particles larger than a specified size (e.g., elutriators, impactors, and cyclones). The method
3 of choice usually depends on the particle size range and the chemical composition of the
4 aerosol of interest. As an example, ambient concentrations of particles are subject to
5 National Ambient Air Quality Standards (Code of Federal Regulations, 1987a). This
6 standard focuses on the concentration of particulate mass for all particles less than 10 /am in
7 equivalent aerodynamic diameter, rather than the individual chemical species (e.g., nitrates)
8 comprising the collected particles.
9 The particle size distribution of ambient PN is bimodal (Kadowaki, 1977; Wolff, 1984;
10 Yoshizumi, 1986; Wall et al., 1988). Particulate nitrate is concentrated in the coarse size
LI (i.e., greater than 2.5 jum) in marine environments, where ambient HNO3 reacts with, the
12 coarse suspended sea salt (i.e., NaCl) to form NaNO3. Under other circumstances, the size
13 distribution of PN will be determined by environmental conditions and the relative presence
14 of precursors, including HNO3, NH3, and acidic aerosols. For example, in the eastern
15 United States, during the summer, when the concentration of acidic sulfates is high, the
i.6 temperature is high, and the NH3 emissions are low, the NH4NO3 = NH3 + HNO3
17 equilibrium is shifted to the right. This and metathetical reactions with acidic aerosols and
• 8 gases make gaseous HNO3 available for reaction with and retention by coarse soil-derived
.9 particles, giving rise to high concentrations of PN in the coarse size range. In contrast, some
$ western urban areas (e.g., Los Angeles and Denver) have low SO2 emissions and adequate
II NH3 emissions to neutralize acidic aerosols. These conditions favor the concentration of PN
12 in the fine size range, presumably as NH4NO3.
13
>4 6.10.1 Filtration
15 Particulate nitrates are generally collected by filtration techniques for subsequent
16 analysis. Using ambient dust, John and Reischl (1978) found the filtration efficiencies of
17 Nuclepore (polycarbonate, 0.8 /*m pore) and Whatman 41 (cellulose) filters to be less than
t8 90%. Efficiencies exceeded 99% for Gelman GA"1 (cellulose acetate), Gelman Spectrograde
19 (glass fiber), Gelman A (glass fiber), MSA (glass fiber) and EPA grade (glass fiber) filters.
10 For polytetrafluoroethylene (Teflon™) membrane filters, the efficiencies exceeded 99% for
August 1991 6-59 DRAFT-DO NOT QUOTE OR CITE
-------
1 Fluoropore (1 jim pore), Ghia (1-3 /urn pore), and Ghia (2-4 /*m pore) filters but did not
2 exceed 99% for some tests with the Fluoropore (3 foa. pore) and Ghia (3-5 /nm pore) filters.
3 The integrity of PN collected on filters may depend on storage conditions and other
4 factors. Highsmith et al. (1986) have attributed weight loss observed on quartz high volume
5 and Teflon™ dichotomous filters to particle loss and volatilization during handling and
6 shipment. Smith et al. (1978) report a 73% loss of NO3" from Gelman AB (glass fiber) high
7 volume filters stored for 15 months in the open at room temperature. In contrast, filters
8 stored for 2 years in containers at -28 °C showed no loss of NO3". Witz et al. (1990) have
9 observed 19% loss of nitrates on PM-10 samples collected on Whatman QM-A (quartz and
10 glass fiber) filters after 1 week of room temperature storage. Nitrate losses ranged from
11 28 to 50% after 1 month. Dunwoody (1986), using acid-treated Whatman QM-A (quartz and
12 glass fiber) high volume filters found NO3" losses of 86% after 6 to 8 months of dry room
13 temperature storage, while refrigerated filter extracts were stable over this period. Witz et al.
14 (1990) found nitrate losses to increase with decreasing filter alkalinity, increasing acidity of
15 the aerosol deposit, and increasing storage temperature. Dunwoody (1986) found that filters
16 spiked with solutions containing KNO3 and other salts showed no NO3" losses over 60 days
17 of storage. In contrast, filters spiked with HNO3 lost 70 to 90% of the NO3" over a period
18 of 3 days.
19 Using filters to collect PN can also result in both positive and negative biases that occur
20 during the sampling process. Some of the difficulties encountered with filtration techniques
21 for distinguishing between paniculate and gaseous nitrate are also discussed in Section 6.7.1.
22 Gas-filter interactions may lead to one type of positive bias. Glass fiber filters have been
23 employed to collect particles including PN from ambient air, and at one point glass fiber
24 filters were specified by the EPA for sampling Total Suspended Particulate matter (Code of
25 Federal Regulations, 1987a). Glass fiber filters can retain gaseous HNO3 and to a lesser
26 extent promote the oxidation of gaseous NO2 leading to the formation of artifact nitrates and
27 the resulting positive biases (Appel et al., 1979; Spicer and Schumacher, 1979). Substantial
28 positive biases from HNO3 have been reported by Appel et al. (1979) for Gelman A, GA"1,
29 and Spectrograde; Whatman 41; MSA 1106 BH (glass fiber); and "EPA grade" filters, by
30 Spicer and Schumacher (1979) for Millipore Nylon; Gelman E, A, AE, AA, and
31 Spectrograde (glass fiber); MSA 1106 BH; Millipore (cellulose acetate); Gelman Microquartz
August 1991 6-60 DRAFT-DO NOT QUOTE OR CITE
-------
1 (quartz fiber); and Pallflex E 70-2075W (quartz fiber) filters, and by Appel et al. (1984) for
2 Gelman "EPA grade"; Schleicher and Schuell (S & S) (glass fiber); S & S 1 HV (glass
3 fiber); Whatman EPM 2600 (glass fiber); Whatman EPM 1000 (glass fiber); Whatman
4 QM-A; Pallflex 2500 QAST (quartz fiber); Gelman Microquartz; and Gelman ADL (quartz
5 fiber) filters. Witz and Wendt (1981) report that the magnitudes of artifact nitrates on high
6 volume sampler filters were ordered as follows: Whatman EPM 1000 > Gelman AE (acid
7 washed glass fiber) > Gelman Microquartz > Pallflex TX40H120 (Teflon™-coated glass
8 fiber) > Pallflex 2500 QAO (quartz fiber). Artifact nitrates on Gelman A and Pallflex
9 TX40H120 filters based on laboratory tests exceeded the amount found on Pallflex QAST
0 filters by factors of 8.6 and 3.4 (Mueller and Hidy, 1983). Substantial amounts of artifact
1 nitrates have been reported based on field studies using Gelman A and Pallflex TX40H120
2 filters (Pierson et al., 1980). Higher ambient PN measurements were reported using S & S
3 and Whatman QM-A filters than with Gelman Microquartz, Pallflex 2500 QAST or
4 Membrana/Ghia Zefluor (Teflon™) filters (Rehme et al., 1984). Small biases were also
5 reported for EPA/ADL (quartz fiber) and Pallflex QAST filters (Spicer and Schumacher,
6 1979). Negligible artifact nitrates were reported for Fluoropore (Teflon™) (Appel et al.,
7 1979; Mueller and Hidy, 1983) and Ghia Zefluor filters (Appel et al., 1984), and no artifact
8 nitrates were reported for Nuclepore (0.8 pm pore) and Millipore Mitex (Teflon™) filters
9 (Spicer and Schumacher, 1979). Good agreement was reported for PN collected on acid-
0 treated Pallflex 2500 QAO and Fluoropore filters (Forrest et al., 1982) in contrast to tests
1 where PN on Pallflex QAST exceeded that on Ghia Zefluor filters by 33% (Appel et al.,
2 1984).
3 A second source of positive bias in using filtration for the collection of PN is the
4 retention of gaseous HNO3 by paniculate matter collected on the filter. Appel et al. (1980)
5 have reported increased retention of HNO3 with mass loading of particulate matter on Ghia
6 Zefluor filters.
7 Negative biases may arise from at least two sources. Particulate nitrates may react with
8 cocollected acidic aerosols or gases to release HNO3 from the particulate catch leading to one
9 type of negative bias. In the laboratory, the separate introduction of H2SO4 aerosols and
0 gaseous HC1 each resulted in appreciable losses and downstream recovery as nitrate of
1 preloaded NH4N63 from Ghia Zefluor filters (Appel and Tokiwa, 1981). Harker et al.
August 1991 6-61 DRAFT-DO NOT QUOTE OR CITE
-------
1 (1977) reported that nitrates collected during chamber experiments on Gelman Spectrograde
2 filters were displaced by .sulfate-containing, and presumably acidic, aerosols according to a
3 metathetical reaction. The introduction of H2SO4 aerosols to ambient particles preloaded on
4 acid-treated Pallflex 2500 QAO filters resulted in appreciable PN losses (Forrest et al.,
5 1980). Pierson et al. (1980) reported similar observations for PN on both glass and quartz
6 filters. Negative correlations have been reported between the fraction of PN measured on
7 acid gas-denuded Teflon™ filters and both ammonia-denuded measurements of strong acid on
8 Teflon™ filters (Appel and Tokiwa, 1981) and measurements of strong acid on acid-treated
9 Pallflex 2500 QAO filters (Forrest et al., 1982).
10 A second source of a negative bias with filtration for the collection of PN is the
11 volatilization of NH4NO3. Nominal 40 to 50% losses of nitrate due to volatilization have
12 been reported where laboratory air free of HNO3 and NH3 was drawn through Ghia Zefluor
13 filters preloaded with NH4NO3. Ambient air drawn through acid-treated Pallflex 2500 QAO
14 filters preloaded with NH4NO3 has shown nitrate volatilization losses ranging between 0 and
15 72% (Forrest et al., 1980).
16 Artifact nitrate formation depends to a large extent on the composition of the filter
17 material (e.g., glass versus Teflon™). Artifact nitrate formation also increases with relative
18 humidity and decreases with temperature (Appel et al., 1979; Forrest et al., 1980). To
19 insure efficient particle collection and minimize artifact nitrate formation, Teflon™ membrane
20 or selected quartz fiber filters are preferred over glass, Teflon™-coated glass, cellulose,
21 cellulose acetate or polycarbonate filters. Quartz fiber filters permit sampling at high flow
22 rates with modest pressure drops; however, they are fragile, require care in handling (Rehme
23 et al., 1984), are subject to some positive bias from artifact nitrate formation (Appel et al.,
24 1984), and may require pretreatment to insure low blank levels (Leahy et al., 1980).
25 Teflon™ membrane filters are the most nearly inert but in contrast are subject to clogging
26 with increased mass loadings (Rehme et al., 1984).
27
28 6.10.2 Denuders/Filtration
29 Many of the previously mentioned biases may be eliminated by deploying a combination
30 of denuders and filters (Appel et al., 1981). For additional information on denuders see
31 Section 6.7.2. Biases involving interaction of HNO3 with the filter or collected sample have
August 1991 6-62 DRAFT-DO NOT QUOTE OR CITE
-------
been circumvented to some degree through the use of denuders. During sampling, gaseous
HNO3 diffuses to the surface of the denuder and is collected by reaction with the denuder
surface, while particles pass through the device uncollected. An inert filter material (e.g.,
Teflon™ membrane or quartz fiber) is then used to collect the particulate matter. Since the
inert particulate filter is still subject to the negative biases of liberated HNO3 and volatilized
NH4NO3, the resulting NO3" must be collected on a backup filter. Nylon, Na2CO3-coated
NaCl-coated, or NaF-coated filters may be used as backup filters. Although pressure drop
and capacity considerations favor coated filters for high flow rate applications, due to the
presence of the Cl" or F" ions, this collection method may not be compatible with an 1C
finish for determining NO3".
o
As noted in Section 6.7.1, a minimum sensitivity of 0.02 /ng m may be calculated
under the assumptions of 0.05 /ng ml"1 analytical detection limit, 10 mL extraction volume,
negligible blank, quantitative collection and extraction, and a sampled air volume of 24 m3
(i.e., 1 m3 h"1 for 1 day). Thus, for a combination of a denuder and two filters, the PN
o
sensitivity should be approximately 0.04 /ng m . Median precision estimates of 4 to 16%
RSD have been reported for 22-h duration samples of fine PN (Vossler et ah, 1988; Sickles,
1987), and a median precision of 4% RSD was reported for twelve 1-wk duration samples of
fine PN (Sickles, 1987).
6.10.3 Analysis
After the collection of PN by the techniques discussed in the previous sections, samples
are analyzed directly or indirectly for nitrate. Several methods have been used including ion
chromatography (1C) (Mulik et ah, 1976), colorimetry (Mullin and Riley, 1955),
derivatization/GC (Tesch et ah, 1976), HPLC (Kamiura and Tanaka, 1979), voltametry
(Bodini and Sawyer, 1977), ion specific electrode (ISE) (Driscoll et ah, 1972), FTIR (Bogard
et ah, 1982), and CLM (Yoshizumi et ah, 1985). Ion chromatography and colorimetry are
the methods most commonly used as analytical finishes for the determination of PN.
Many of the methods for determining PN require the extraction of nitrates prior to
analysis. Extraction of nitrate spiked onto nylon filters showed quantitative recovery using
1C eluent solution or basic (i.e., 0.003N NaOH) solution but not using water (Hering et ah,
1988). Similar tests with spiked Teflon™ membrane filters showed essentially quantitative
August 1991 6-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 recovery in each extraction medium. Other tests have shown good NO3" recoveries from
2 spiked and ambient nitrates on cellulose, glass fiber, and Teflon™ filters using 1C eluent and
3 ultrasonication, boiling deionized water, and sequential extraction in warm 1C eluent and
4 deionized water (Jenke, 1983), :
5 In recent years 1C has become a method of choice for the determination of many anions
6 and cations in solution. Ion chromatography uses conductimetric detection and a. combination
7 of resin columns to separate the ions of interest and strip or suppress the eluent from the
8 background (Small et al., 1975; Mulik et al., 1976). The bromide and phosphate
9 interferences noted by Mulik et al. (1976) generally do not present problems with
10 environmental samples of PN. In some cases where filter or denuder extracts are analyzed by
11 1C, to permit good resolution of various peaks, care must be taken to prevent excessive
12 concentrations of chloride or H2O2 must be added to oxidize sulfite to sulfate. One recent
13 study reported a precision estimate of 1 % for replicate NO3" measurements in extracts of
14 ambient samples where the concentration was above 0.15 ptg mL"1 (Sickles et al., 1988a).
15 Detection limits for NO3" of 0.025 to 0.1 p&g mL"1 have been reported using 1C with a
16 0.5 mL sample loop (Anlauf et al., 1988; Mulik et al., 1976). As noted in Section 6.7.1, a
17 0.05 peg mL"1 analytical detection limit for NO3" corresponds to an ambient concentration of
18 0.02 p.g m"3 on a single filter sampling at 1 m3 h"1 for 1 day. Although nonsuppressed 1C
19 has poorer detection limits than the previously described suppressed approach, successful
20 application to the analysis of nitrates in ambient aerosols: has also been reported for
21 nonsuppressed 1C (Willison and Clark, 1984).
22 Various colorimetric methods for NO3~ have been used. In one widely automated
23 method, NO3" in the extract is reduced to NO2" which is diazotized and determined at
24 550 nm using the Griess-Saltzman Method (see Section 6.3.5) (Saltzman, 1954). The
25 reduction may be accomplished using a copper-cadmium (Cu-Cd) reductor column
26 (Technicon, 1972) or using hydrazine sulfate with copper as a catalyst under slightly basic
27 conditions (Mullin and Riley, 1955; Kamphake et al., 1967). The detection limits of
28 0.001 to 0.006 peg mL"1 claimed with these methods are somewhat more sensitive than those
29 previously cited for the 1C method. A comparison of the performance of 1C with this
30 colorimetric method for PN collected on Fluoropore filters showed excellent agreement (Fung
31 etal., 1979).
August 1991 6-64 DRAFT-DO NOT QUOTE OR CITE
-------
1 Another colorimetric technique, the Brucine Method, involves the reaction of NO3" with
2 brucine under acidic conditions (Kothny et al., 1972). The color is measured at 410 nm, and
3 a detection limit of 0.4 /j.g ml/1 is reported. Other colorimetric methods involve the
4 nitration of 2,4-xylenol in the presence of sulfuric acid followed by steam distillation and
5 absorbance measurement at 435 nm (Saltzman et al., 1972), or the nitration of toluene in the
6 presence of sulfuric acid followed by extraction into toluene and absorbance measurement of
7 the nitrotoluene-toluene complex at 284 nm (Bhatty and Townshend, 1971). The sensitivity
8 of both methods is marginal (i.e., 1 fj.g mL"1), and they are subject to interferences (Saltzman
9 et al., 1972; Appel et al., 1977; Norwitz and Keliher, 1978; Kamiura and Tanaka, 1979;
10 Bhatty and Townshend, 1971).
11 Using reactions similar to those described above, Tesch et al. (1976) have reacted NO3"
12 with benzene or other aromatic compounds in the presence of sulfuric acid and measured the
13 resulting nitroaromatic compound using GC-EC. A sensitivity of 0.1 /^g mL"1 and
14 applicability to determining NO3" in saliva, blood, drinking water, and airborne particles
15 were claimed. The above method has been modified by Tanner et al. (1979) using electron
16 capture-sensitive fluoroaromatic derivatizing agents and a more effective catalyst (i.e.,
17 trifluoromethanesulfonic acid). With a sensitivity of 0.01 fj.g mL"1, this method has been
18 applied to ^L-sized samples and the analysis of PN.
19 High-performance liquid chromatography (HPLC) coupled with UV detection at 210 nm
20 has been used to measure NO3" in the aqueous extracts of PN from glass fiber filters
21 (Kamiura and Tanaka, 1979). No interferences were reported, and a detection limit of
22 0.1 ng mL"1 was claimed.
23 A voltammetric technique for the measurement of NO3" in solution has been reported by
24 Bodini and Sawyer (1977). The technique is based on the reduction of NO3" by a Cu-Cd
25 catalyst that is formed on the surface of a pyrolytic graphite electrode. The detection limit is
26 0.06 fj.g mL"1, but NO2" is a direct interferent. Favorable comparison was reported between
27 the results of this method and those of the Technicon (1972) colorimetric method for
28 determining NO3" in extracts of PN.
29 Ion specific electrodes (ISE) have been used to measure NO3" in extracts of PN with a
30 detection limit of 1 /^g mL"1 (Simeonov and Puxbaum, 1977). Ion specific electrodes suffer
31 from poor sensitivity, potential drifts caused by variable agitation speed, frequent need for
August 1991 6-65 DRAFT-DO NOT QUOTE OR CITE
-------
1 restandardization, and interferences by other ions (Driscoll et al., 1972). Spicer et al.
2 (1978a) evaluated a NO gas-sensing electrode for the indirect measurement of NO3" in
3 solution. Since the electrode responds to NO2" in solution, the approach was to measure
4 NO2" in solution before and after reducing the NO3" to NO2" and attribute the difference to
5 NO3". Although the gas-sensing electrode was both sensitive (i.e., 0.1 ng mL"1) and specific
6 for NO2", difficulties in the reduction step prevented further development.
7 A dry technique using FTIR for measuring NO3~ incorporated in a KBr matrix from
8 samples of ambient PN has been reported (Bogard et al., 1982). Absorbance bands for NO3"
9 occur at 2,430, 1,384, and 840 cm"1. Using the 1,384 cm'1 band a detection limit of 0.1 ^g
10 NO3" per sample was reported, although the nearby NH4+ band under same circumstances
11 may not permit distinct resolution of NO3". Techniques have been developed recently that
12 permit FTIR detection of nitrates and other species in samples of ambient aerosols collected
13 by filtration on thin Teflon™ membrane filters using direct transmission or by impaction using
14 attenuated total internal reflection (Johnson and Kumar, 1987). This method is currently in
15 the research prototype stage of its development.
16 The decomposition of NO3" followed by the CLM detection of the resulting NOX (see
17 Sections 6.2.1 and 6.3 1) has been used to determine NO3" in PN samples. Thermal
18 decomposition can be applied to NO3" either on filters or in liquid extracts (Spicer et al.,
19 1985). With this technique, NO3" is decomposed by rapid heating to 425 °C in a N2
20 atmosphere, and the resulting NOX (i.e., NO and NO^ is measured (i.e., integrated) using a
21 conventional CLM analyzer. Particulate nitrite is a direct interferent. A detection limit of
22 0.7 fig mL"1 is claimed. Comparison of CLM and 1C analyses of spiked and ambient
23 samples showed good agreement, although the 1C was more precise especially at low
24 concentration levels. Yoshizumi et al. (1985) have modified a method developed by Cox
25 (1980) to reduce NO3" and NO2" in solution and measure the evolving NO using a
26 commercially-supplied CLM instrument. The method of Yoshizumi et al. (1985) uses a flow
27 system, does not distinguish NO2" from NO3" (i.e., NO2" is a direct interferent), but has a
28 NO3~ detection limit of 0.001 jig mL"1. The method of Cox (1980), although using a batch
29 approach, does distinguish NO2" from NO3" and has respective detection limits of 0.00005
30 and 0.05 /tg mL'1.
August 1991 6-66 DRAFT-DO NOT QUOTE OR CITE
-------
1 It has been suggested that volatile and nonvolatile nitrates may be distinguished by
2 taking advantage of their different temperatures of volatilization (Yoshizumi and Hoshi,
3 1985). Samples of atmospheric particles collected by filtration or impaction are heated in a
4 furnace to the optimum volatilization temperature of ammonium nitrate (i.e., 160 °C). The
5 volatilized nitrate is then collected in water for subsequent determination (e.g., by 1C).
6 A similar principle has been used in thermal dehuders (Klockow et al., 1989). In this case,
7 HNO3 is collected at ambient temperature on a MgSO4-coated annular denuder, and NH4NO3
8 is collected at 150 °C on a similar downstream denuder. After sampling, denuders are
9 heated in turn to 700 °C to liberate NOX for determination by a CLM NOX analyzer. Sturges
0 and Harrison (1988) have reported several potential interferences with the volatilization
1 approach. Nitric acid from volatilized NH4NO3 in the presence of NaCl, for example, will
2 cause displacement of the chloride as HC1 and formation of nonvolatile NaNO3. Differences
3 in the thermal stabilities of ammonium sulfate/nitrate double salts were also demonstrated.
4 These observations cast doubt on the feasibility of thermal speciation of PN.
5
6
7 6.11 NITROUS OXIDE (N2O)
8 Ambient N2O levels have been measured by several methods. These methods include
9 infrared spectroscopy (both absorption and emissions spectra), mass spectrometry,
0 manometry, and gas chromatography coupled with thermal conductivity, flame ionization,
1 ultrasonic phase shift, helium ionization, and electron capture detectors (Pierotti and
2 Rasmussen, 1977). The most commonly used method employs GC-EC with a detection limit
3 of 20 ppb (Thijsse, 1978) and a precision of ±3% at the background level of 330 ppb
4 (Cicerone et al., 1978). Cassidy and Reid (1982) also report an expected MDL of 20 ppb for
5 . N2O using TOLAS near 1,150 cm'1 (see Section 6.2.3 for more on TOLAS). As described
6 in Section 7.2.3, a laboratory prototype method, TTFMS, has been developed with a
7 projected MDL for N2O of 3 ppt (Hansen, 1989).
8 Calibration can be performed using commercially supplied cylinders of compressed gas
.9 (Thijsse, 1978), dilution of pure N2O, N2O permeation tubes (Cicerone et al., 1978) or
•0 gravimetric preparation of calibration mixtures (Komhyr et al. 1988). Standard reference
August 1991 6-67 DRAFT-DO NOT QUOTE OR CITE
-------
1 material mixtures of N2O and CO2 in air are also available at nominal N2O concentrations of
2 300 and 330 ppb (National Bureau of Standards, 1988).
3
4
5 6.12 SUMMARY
6 Since the publication in 1971 of the original version of Air Quality Criteria for Nitrogen
7 Oxides, changes have occurred in the technology associated with the sampling and analysis
8 for ambient NOX and related species. During the 1970s, roughly the period between
9 publication of the original Criteria Document and its first update and revision, several events
10 occurred that focused on the determination of NO2 in ambient air. In 1973 the original
11 Reference Method was withdrawn because of unresolvable technical difficulties. Major
12 methods development efforts over the next 3 to 4 years yielded both automated and manual
13 methods that were suitable for the determination of NO2 in ambient air. As a result, EPA
14 designated a new Reference Method and Equivalent Methods for NO2. The Reference
15 Method specifies a measurement principle and calibration procedures, namely gas phase
16 chemiluminescence (GP-CLM) wifli calibration using either gas phase titration of NO with O3
17 or a NO2 permeation device. The Sodium Arsenite Method in both the manual and
18 continuous forms and the TGS Method were also designated as Equivalent Methods.
19 Subsequently, commercial GP-CLM instruments were designated as Reference Methods. The
20 sensitivity of these devices was in the low ppb range and while the GP-CLM instruments
21 were recognized as being susceptible to interferences by other nitroxy species, it was believed
22 that the atmospheric concentrations of these compounds were generally low relative to NO2.
23 In the 1980s additional developments have occurred. Information from air quality
24 monitoring networks is now readily available and has shown the GP-CLM instruments to
25 have nominal precision and accuracy of ±10 to 15% and 20%, respectively, and to have
26 replaced manual methods to a large extent in network applications. Heightened interest in the
27 research community on the speciation of atmospheric trace gases and specifically nitrogen-
28 containing species has prompted a new wave of methods development. While the basic
29 design and performance of the commercial instruments have remained essentially unchanged,
30 researchers have improved GP-CLM measurement technology and refined other instrumental
31 methods to permit the determination of NO, NO2, and NO in the low ppt range. Although
August 1991 6-68 DRAFT-DO NOT QUOTE OR CITE
-------
I GP-CLM NO detectors coupled with catalytic NO2-to-NO converters are still not specific for
I NO2, they have proven useful for measuring NOy, and GP-CLM NO detectors coupled with
5 photolytic NO2-to-NO converters have shown improved specificity for NO2.
I A continuous liquid phase CLM device for sensitively detecting NO2 has been
5 developed and may be suitable to measure NO2 if interference problems can be overcome.
5 Passive samplers for NO2 have been used, primarily for workplace and indoor applications
1 but hold promise for ambient measurements as well. Gas chromatography with electron
i capture detection is useful in the determination of PAN, other organic nitrates, and N2O,
J Laser-induced fluorescence has been introduced to detect NO, NO2, and HNO2 with
) high sensitivity and specificity. Tunable-diode laser spectroscopy has been used to detect
i NO, NO2, and HNO3. Long path spectroscopy has also been used to detect NO, NO2,
I HNO2, and NO3. Two-tone frequency modulated spectroscopy holds promise for the
I sensitive measurement of NO, NO2, PAN, HNO3, and N2O. These spectroscopic methods
I- are research tools and are not yet easily or econpmically suited for routine monitoring.
> Interest in acidification of the environment has resulted in the development of methods
> for HNO2 and HNO3. Integrative methods using denuders have been introduced to permit.
7 sensitive determination of these and other species. In recent years, the potential for artifacts
I in using filters for sampling particulate matter and specifically partieulate nitrate has been
> recognized. This has given rise to careful characterization of filter media for potential
) artifacts and the use of combinations of denuders and filters to permit more specific
1 determination of nitrogen-containing gases and particulate nitrates in ambient air.
August 1991 6-69 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2
3 Adams, K. M.; Japar, S. M.; Pierson, W. R. (1986) Development of a MnO2-coated, cylindrical denuder for
4 removing NO2 from atmospheric samples. Atmos. Environ. 20: 1211-1215,
5
6 Alden, M.; Edner, H.; Svanberg, S. (1982) Laser monitoring of atmospheric NO using ultraviolet
7 differential-absorption techniques. Opt. Lett. 7: 543-545.
8
9 Anlauf, K. G.; Fellin, P.; Wiebe, H. A.; Schiff, H. I.; Mackay, G. L; Braman, R. S.; Gilbert, R. (1985)
10 A comparison of three methods for measurement of atmospheric nitric acid and aerosol nitrate and
11 ammonium. Atmos. Environ. 19: 325-333.
12
13 Anlauf, K, G.; Wiebe, H. A.; Fellin, P. (1986) Characterization of several integrative sampling methods for
14 nitric acid, sulphur dioxide and atmospheric particles. 3. Air Pollut. Control Assoc. 36: 715-723.
15
16 Anlauf, K. G.; MacTavish, D. C.; Wiebe, H. A.; Schiff, H. L; Mackay, G. I. (1988) Measurement of
17 atmospheric nitric acid by the filter method and comparisons with the tuneable diode laser and other
18 methods. Atmos. Environ. 22: 1579-1586.
19
20 Appel, B. (1973) A new and more sensitive procedure for analysis of peroxybenzoyl nitrate. J. Air Pollut.
21 Control Assoc. 23: 1042-1044.
22
23 Appel, B. R,; Tokiwa, Y. (1981) Atmospheric particulate nitrate sampling errors due to reactions with particulate
24 and gaseous strong acids. Atmos. Environ. 15: 1087-1089.
25
26 Appel, B. R.; Hoffer, E. M.; Kothny, E. L.; Wall, S. M. (1977) Interference in 2,4-xylenol procedure for
27 nitrate determination hi atmospheric aerosols. Environ. Sci. Technol. 11: 189-190.
28
29 Appel, B. R.; Wall, S. M.; Tokiwa, Y.; Haik, M. (1979) Interference effects in sampling particulate nitrate in
30 ambient air. Atmos. Environ. 13: 319-325.
31
32 Appel, B. R.; Wall, S. M.; Tokiwa, Y.; Haik, M. (1980) Simultaneous nitric acid, particulate nitrate and acidity
33 measurements in ambient air. Atmos. Environ. 14: 549-554.
34
35 Appel, B. R.; Tokiwa, Y.; Haik, M. (1981) Sampling of nitrates in ambient air. Atmos. Environ. 15: 283-289.
36
37 Appel, B, R.; Tokiwa, Y.; Haik, M.; Kothny, E. L. (1984) Artifact particulate sulfate and nitrate formation on
38 filter media. Atmos. Environ. 18: 409-416.
39
40 Appel, B. R.; Povard, V.; Kothny, E. L. (1988) Loss of nitric acid within inlet devices intended to exclude
41 coarse particles during atmospheric sampling. Atmos. Environ. 22: 2535-2540.
42
43 Appel, B. R.; Winer, A. M.; Tokiwa, Y.; Biermann, H. W. (1990) Comparison of atmospheric nitrous acid
44 measurements by annular denuder and differential optical absorption systems. Atmos. Environ. Part A
45 24: 611-616.
46
47 Atkinson, R,; Winer, A. M.; Pitts, J. N., Jr. (1986) Estimation of night-time N2OS concentrations from ambient
48 NO2 and NO3 radical concentrations and the role of N2O5 in night-time chemistry. Atmos. Environ. 20:
49 331-339.
50
51 Atlas, E. (1988) Evidence for kQ alkyl nitrates in rural and remote atmospheres. Nature (London) 331: 426-428.
52
53 Baumgartner, R. A.; Fletcher, L. D.; Hawley, J. G. (1979) A comparison of lidar and air quality station NO2
54 measurements. J. Air Pollut. Control Assoc. 29: 1162-1165.
August 1991 6-70 DRAFT-DO NOT QUOTE OR CITE
-------
1 Baveja, A. K.; Chaube, A.; Gupta, V. K. (1984) Extractive spectrophotometric method for the determination of
2 atmospheric nitrogen dioxide. Atmos. Environ. 18: 989-993.
3
V Benner, C. L.; Eatough, D. J.; Eatough, N. L.; Bhardwaja, P. (1987) Evaluation of an annular denuder method
5 for the collection of atmospheric nitrogenous species in the southwest desert. Presented at: 80th annual
5 meeting of the Air Pollution Control Association; June; New York, NY. Pittsburgh, PA: Air Pollution
7 Control Association; paper no. 87-63.6.
3
) Benner, C. L.; Eatough, N. L.; Lewis, E. A.; Eatough, D. J.; Huang, A. A.; Ellis, E. C. (1988) Diffusion
) coefficients for ambient nitric and nitrous acids from denuder experiments in the 1985 nitrogen species
I methods comparison study. Atmos. Environ. 22: 1669-1672.
I
} Bennett, B. L (1979) Stability evaluation of ambient concentrations of sulfur dioxide, nitric oxide, and nitrogen
\ dioxide contained in compressed gas cylinders. Research Triangle Park, NC: U. S. Environmental
5 Protection Agency, Environmental Monitoring and Support Laboratory; EPA report no. EPA
5 600/4-79-006. Available from: NTIS, Springfield, VA; PB-292749.
7 -
S Benoit, F. M. (1983) Detection of nitrogen and sulfur dioxides in the atmosphere by atmospheric pressure
) ionization mass spectrometry. Anal. Chem. 55: 2097-2099.
)
I Bhatty, M. K.; Townshend, A. (1971) Spectrophotometric determination of small amounts of nitrate and nitrite
I by conversion to nitrotoluene and extraction into toluene. Anal. Chim. Acta 56: 55-60.
5
}• Biermann, H. W.; Tuazon, E. C.; Winer, A. M.; Wallington, T. J.; Pitts, J. N., Jr. (1988) Simultaneous
5 absolute measurements of gaseous nitrogen species in urban ambient air by long pathlength infrared and
5 ultraviolet-visible spectroscopy. Atmos. Environ. 22: 1545-1554.
7
3 Bodini, M. E.; Sawyer, D. T. (1977) Voltammetric determination of nitrate ion at parts-per-billion levels. Anal.
) Chem. 49: 485-489.
)
I Bogard, J. S.; Johnson, S. A.; Kumar, R.; Cunningham, P. T. (1982) Quantitative analysis of nitrate ion in
I ambient aerosols by Fourier-transform infrared spectroscopy. Environ. Sci. Technol. 16: 136-140.
5
}• Bollinger, M. J.; Sievers, R. E.; Fahey, D. W.; Fehsenfeld, F. C. (1983) Conversion of nitrogen dioxide, nitric
) acid, and «-propyl nitrate to nitric oxide by gold-catalyzed reduction with carbon monoxide. Anal. Chem.
5 55: 1980-1986.
7
S Bollinger, M. J.; Hahn, C. J.; Parrish, D. D.; Murphy, P. C.; Albritton, D. L.; Fehsenfeld, F. C. (1984) NOX
) measurements in clean continental air and analysis of the contributing meteorology. J. Geophys. Res.
) [Atmos.] 89: 9623-9631.
I
I Bradshaw, J. D.; Rodgers, M. O.; Davis, D. D. (1982) Single photon laser-induced fluorescence detection of
} NO and SO2 for atmospheric conditions of composition and pressure. Appl. Opt. 21: 2493-2500.
*
) Bradshaw, J. D.; Rodgers, M. O.; Sandholm, S. T.; KeSheng, S.; Davis, D. D. (1985) A two-photon
S laser-induced fluorescence field instrument for ground-based and airborne measurements of atmospheric
7 NO. J. Geophys. Res. [Atmos.] 90: 12861-12873.
3
) Braman, R. S.; de la Cantera, M. A. (1986) Sublimation sources for nitrous acid and other nitrogen compounds
) in air. Anal. Chem. 58: 1533-1537.
I
1 Braman, R. S.; de la Cantera, M. A.; Han, Q. X. (1986) Sequential, selective hollow tube preconcentration and
} chemiluminescence analysis system for nitrogen oxide compounds in air. Anal. Chem. 58: 1537-1541.
August 1991 6-71 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bubacz, D. K.; Daughtrey, E. H.; Pleil, J. D.; Kronmiller, K. G. (1987) Luminox measurements of ambient
2 NO2. In: Proceedings of the 1987 EPA/APCA symposium on measurement of toxic and related air
3 pollutants; May; Research Triangle Park, NC. Pittsburgh, PA: Air Pollution Control Association;
4 pp. 398-403; EPA report no. EPA-600/9-87-010. Available from: NTIS, Springfield, VA; PB88-113402.
5 (APCA publication VIP-8).
6
7 Burkhardt, M. R.; Maniga, N. I.; Stedman, D. H.; Paur, R. J. (1988) Gas chromatographic method for
8 measuring nitrogen dioxide and peroxyacetyl nitrate in air without compressed gas cylinders. Anal.
9 Chem. 60: 816-819.
10
11 Butcher, S. S.; Ruff, R. E. (1971) Effect of inlet residence time on analysis of atmospheric nitrogen oxides and
12 ozone. Anal. Chem. 43: 1890-1892.
13
14 Buttini, P.; DiPalo, V.; Possanzini, M. (1987) Coupling of denuder and ion chromatographic techniques for NO2
15 trace level determination in air. Sci. Total Environ. 61: 59-72.
16
17 Cadoff, B. C.; Hodgeson, J. (1983) Passive sampler for ambient levels of nitrogen dioxide. Anal. Chem. 55:
18 2083-2085.
19
20 Carroll, M. A.; McFarland, M.; Ridley, B. A.; Albritton, D. L. (1985) Ground-based nitric oxide measurements
21 at Wallops Island, Virginia. J. Geophys. Res. [Atmos.] 90: 12853-12860.
22
23 Cassidy, D. T.; Reid, J. (1982) Atmospheric pressure monitoring of trace gases using tunable diode lasers. Appl.
24 Opt. 21: 1185-1190.
25
26 Cicerone, R. J.; Shetter, J. D.; Stedman, D. H.; Kelly, T. J.; Liu, S. C. (1978) Atmospheric N2O:
27 measurements to determine its sources, sinks, and variations. J. Geophys. Res. C: Oceans Atmos. 83:
28 3042-3050.
29
30 Code of Federal Regulations. (1987a) National primary and secondary ambient air quality standards. C. F. R. 40:
31 § 50.
32
33 Code of Federal Regulations. (1987b) Ambient air monitoring reference and equivalent methods. C. F. R. 40:
34 § 53.
35
36 Cox, R. A. (1974) The photolysis of gaseous nitrous acid. J. Photochem. 3: 175-188.
37
38 Cox, R. D. (1980) Determination of nitrate and nitrite at the parts per billion level by chemiluminescence. Anal.
39 Chem. 52: 332-335.
40
41 Cox, R. A.; Roffey, M. J. (1977) Thermal decomposition of peroxyacetylnitrate in the presence of nitric oxide.
42 Environ. Sci. Technol. 11: 900-906.
43
44 Crecelius, H.-J.; Forwerg, W. (1970) Investigations of the "Saltzman factor". Staub Reinhalt. Luft 30: 23-25.
45
46 Darley, E. F.; Kettner, K. A.; Stephens, E. R. (1963) Analysis of peroxyacyl nitrates by gas chromatography
47 with electron capture detection. Anal. Chem. 35: 589-591.
48
49 Dasch, J. M.; Cadle, S. H.; Kennedy, K. G.; Mulawa, P. A. (1989) Comparison of annular denuders and filter
50 packs for atmospheric sampling. Atmos. Environ. 23: 2775-2782.
51
52 Davis, D. D. (1988) Atmospheric nitrogen oxides, their detection and chemistry. In: Third year report to
53 Coordinating Research Council. Atlanta, GA: Georgia Institute of Technology; pp. 1-13.
54
August 1991 6-72 DRAFT-DO NOT QUOTE OR CITE
-------
1 Davis, D. D.; Bradshaw, J. D.; Rodgers, M. O.; Sandholm, S. T.; KeSheng, S. (1987) Free tropospheric and
1 boundary layer measurements of NO over the central and eastern North Pacific Ocean. J. Geophys. Res.
3 [Atmos.] 92: 2049-2070.
\ '
5 De Santis, F.; Febo, A.; Perrino, C.; Possanzini, M.; Liberti, A. (1985) Simultaneous measurements of nitric
S acid, nitrous acid, hydrogen chloride and sulfur dioxide in air by means of high-efficiency annular
7 denuders. In: Proceedings of the ECE workshop on advancements in air pollution monitoring and
3 procedures; June; Freiburg, Federal Republic of Germany. Bonn: Federal Ministry of the Interior;
? pp. 68-75.
3
I Delany, A. C.; Dickerson, R. R.; Melchior, F. L., Jr.; Wartburg, A. F. (1982) Modification of a commercial
I NOX detector for high sensitivity. Rev. Sci. Instrum. 53: 1899-1902.
3
\ Dickerson, R. R. (1984) Measurements of reactive nitrogen compounds in the free troposphere. Atmos. Environ.
5 18: 2585-2593.
S
7 Dickerson, R. R.; Delany, A. C.; Wartburg, A. F. (1984) Further modification of a commercial NOX detector
3 for high sensitivity. Rev. Sci. Instrum. 55: 1995-1998.
?
3 Driscoll, J. N.; Berger, A. W.; Becker, J. H.; Funkhouser, J. T.; Valentine, J. R. (1972) Determination of
t oxides of nitrogen in combustion effluents with a nitrate ion selective electrode. J. Air Pollut. Control
I Assoc. 22: 119-122.
3
% Drummond, J. W.; Volz, A.; Ehhalt, D. H. (1985) An optimized chemiluminescence detector for tropospheric
5 NO measurements. J. Atmos. Chem. 2: 287-306.
r
7 Drummond, J. W.; Schiff, H. L; Karecki, D. R.; Mackay, G. I. (1989) Measurements of NO2, NOX, O3, PAN,
3 HNO3, H2O2, and H2CO during the Southern California air quality study. Presented at: 82nd annual
? meeting of the Air and Waste Management Association; June; Anaheim, CA. Pittsburgh, PA: Air and
3 Waste Management Association.
1
I Dunwoody, C. L. (1986) Rapid nitrate loss from PM10 filters. J. Air Pollut. Control Assoc. 36: 817-818.
3
I Eatough, D. J.; White, V. F.; Hansen, L. D.; Eatough, N. L.; Ellis, E. C. (1985) Hydration of nitric acid and
5 its collection in the atmosphere by diffusion denuders. Anal. Chem. 57: 743-748.
f
7 Eatough, N. L.; McGregor, S.; Lewis, E. A.; Eatough, D. J.; Huang, A. A,; Ellis, E. C. (1988) Comparison of
3 six denuder methods and a filter pack for the collection of ambient HNO3(g), HNO2(g), and SO2(g) in the
? 1985 NSMC study. Atmos. Environ. 22: 1601-1618.
3
1 Edner, H.; Fredriksson, K.; Sunesson, A.; Svanberg, S.; Uneus, L.; Wendt, W. (1987) Mobile remote sensing
1 system for atmospheric monitoring. Appl. Opt. 26: 4330-4338.
3
1 Ellis, E. C.; Margeson, J. H. (1974) Evaluation of triethanolamine procedure for determination of nitrogen
5 dioxide in ambient air. Research Triangle Park, NC: U.S. Environmental Protection Agency, Quality
5 Assurance and Environmental Monitoring Laboratory; EPA report no. EPA-650/4-74-031. Available
7 from: NTIS, Springfield, VA; PB-237348/8BE.
3
} Fahey, D. W.; Eubank, C. S.; Hubler, G.; Fehsenfeld, F. C. (1985a) Evaluation of a catalytic reduction
3 technique for the measurement of total reactive odd-nitrogen NOy in the atmosphere. J. Atmos. Chem. 3:
1 435-468.
I
3 Fahey, D. W.; Eubank, C. S.; Huebler, G.; Fehsenfeld, F. C. (1985b) A calibrated source of N2O5. Atmos.
4 Environ. 19: 1883-1890.
August 1991 6-73 DRAFT-DO NOT QUOTE OR CITE
-------
1 Fahey, D. W.; Hubler, G.; Parrish, D. D,; Williams, E. J.; Norton, R. B.; Ridley, B. A.; Singh, H. B.; Liu,
2 S. C.; Fehsenfeld, F. C. (1986) Reactive nitrogen species in the troposphere: measurements of NO, NO2,
3 HNO3, particulate nitrate, peroxyacetyl nitrate (PAN), O3, and total reactive odd nitrogen (NOy) at Niwot
4 Ridge, Colorado. J. Geophys. Res. [Atmos.] 91: 9781-9793.
5
6 Febo, A.; De Santis, F.; Perrino, C. (1986) Measurement of atmospheric nitrous and nitric acid by means of
7 annular denuders. In: Physico-chemical behavior of atmospheric pollutants: proceedings of the fourth
8 European symposium; September; Stresa, Italy. Boston, MA: D. Reidel Publishing Company; ;
9 pp. 121-125.
10
11 Febo, A.; De Santis, F.; Liberti, A.; Perrino, C. (1988) Nitric acid-nitrate aerosol measurements by a diffusion
12 denuder: a performance evaluation. Atmos. Environ. 22: 2062-2064.
13
14 Federal Register. (1986) F. R. (March 19) 51: 9595.
15
16 Fehsenfeld, F. C.; Dickerson, R. R.; Hubler, G.; Luke, W. T.; Nunnermacker, L. J.; Williams, E. J.; Roberts,
17 J. M.; Calvert, J. G.; Curran, C. M.; Delany, A. C.; Eubank, C. S.; Fahey, D. W.; Fried, A.;
18 Gandrud, B. W.; Langford, A. O.; Murphy, P. C.; Norton, R. B.; Pickering, K. E.; Ridley, B, A.
19 (1987) A ground-based intercomparison of NO, NOX, and NOy measurement techniques. J. Geophys. Res.
20 [Atmos.] 92: 14710-14722.
21
22 Fehsenfeld, F. C.; Drummond, J. W.; Roychowdhury, U. K.; Gdvin, P. J.; Williams, E. J.; Buhr, M. P.;
23 Parrish, D. D,; Hubler, G.; Langford, A- O.; Calvert, J. G.; Ridley, B. A.; Grahek, F.; Heikes, B. G.;
24 Kok, G. L.; Shetter, J. D.; Walega, J. G.; Elsworth, C. M.; Norton, R. B.; Fahey, D. W.; Murphy,
25 P. C.; Hovermale, C.; Mohaen, V. A.; Demerjian, K. L.; Mackay, G. L; Schiff, H. I. (1990)
26 Intercomparison of NO2 measurement techniques. J. Geophys. Res. [Atmos.] 95: 3579-3597.
27
28 Fellin, P.; Brown, P. M.; Caton, R. B. (1984) Development of a nitric acid sampler. Downsview, ON, Canada:
29 Concord Scientific Corporation.
30
31 Ferm, M. (1986) A N%CO3-coated denuder and filter for determination of gaseous HNO3 and particulate NO3" in
32 the atmosphere. Atmos. Environ. 20: 1193-1201.
33
34 Ferm, M.; Sjodin, A. (1985) A sodium carbonate coated denuder for determination of nitrous acid in the
35 atmosphere. Atmos. Environ. 19: 979-983.
36
37 Fineher, C, L.; Tucker, A. W.; Birnbaum, M. (1978) Laser and flash lamp fluorescence NO2 monitors: a
38 comparison. In; Schlossberg, H., ed. Laser spectroscopy: applications and techniques, [proceedings];
39 August; San Diego, CA. Bellingham, WA: Society of Photo-Optical Instrumentation Engineers;
40 pp. 137-140. (Proceedings of the Society of Photo-Optical Instrumentation Engineers: v. 158).
41
42 Folsom, B. A.; Courtney, C. W. (1979) Accuracy of chemiluminescent analyzers measuring nitric oxide in stack
43 gases. J. Air Pollut. Control Assoc. 29: 1166-1169.
44
45 Fontijn, A.; Volltrauer, H. N.; Frenchu, W. R. (1980) NOX (= NO + NO^ monitor based on an H-atom direct
46 chemiluminescence method. Environ. Sci. Technol. 14: 324-328.
47
48 Forrest, J.; Tanner, R. L.; Spandau, D.; D'Ottavio, T.; Newman, L. (1980) Determination of total inorganic
49 nitrate utilizing collection of nitric acid on NaCl-impregnated filters. Atmos. Environ. 14: 137-144.
50
51 Forrest, J.; Spandau, D. J.; Tanner, R. L.; Newman, L. (1982) Determination of atmospheric nitrate and nitric
52 acid employing a diffusion denuder with a filter pack, Atmos. Environ. 16: 1473-1485.
53
August 1991 6-74 DRAFT-DO NOT QUOTE OR CITE
-------
Fredriksson, K. A.; Hertz, H. M. (1984) Evaluation of the DIAL technique for studies on NO2 using a mobile
lidar system. Appl. Opt. 23: 1403-1411.
Fuglsang, K. (1986) A filter pack for determination of total ammonia, total nitrate, sulfur dioxide and sulfate in
the atmosphere. Miljoestyrelsen, Roskilde, Denmark: Riso National Laboratory; report no.
MST-LUFT-A-103. Available from: NTIS, Springfield, VA; DE87750099.
Fung, K.; Grosjean, D. (1985) Peroxybenzoylnitrate: measurements in smog chambers and in urban air. Sci.
Total Environ. 46: 29-40.
Fung, K. K.; Heisler, S. L.; Price, A.; Nuesca, B. V,; Mueller, P. K. (1979) Comparison of ion
chromatography and automated wet chemical methods for analysis of sulfate and nitrate in ambient
particulate filter samples. In: Mulik, J. D.; Sawicki, E., eds. Ion chromatographic analysis of
environmental pollutants: volume 2. Ann Arbor, MI: Ann Arbor Science Publishers Inc.; pp. 203-209.
Gafmey, J. S.; Fajer, R.; Senum, G. I, (1984) An improved procedure for high purity gaseous peroxyacyl nitrate
production: use of heavy lipid solvents. Atmos. Environ. 18: 215-218.
Gallagher, C, C,; Forsberg, C. A.; Fieri, R. V.; Faucher, G. A.; Calo, J. M. (1985) Nitric oxide and nitrogen
dioxide content of whole air samples obtained at altitudes from 12 to 30 km. J. Geophys. Res. [Atmos.]
90: 7899-7912.
Gay, B. W., Jr.; Noonan, R. C.; Bufalini, J. J.; Hanst, P. L. (1976) Photochemical synthesis of peroxyacyl
nitrates in gas phase via chlorine-aldehyde reaction. Environ. Sci. Technol. 10: 82-85.
Girman, J. R.; Hodgson, A. T.; Robinson, B. K.; Traynor, G. W. (1984) Laboratory studies of the temperature
dependence of the Palmes NO2 passive sampler. In: Proceedings of the national symposium on recent
advances in pollutant monitoring of ambient air and stationary sources; May; Raleigh, NC. Research
Triangle Park, NC: U. S. Environmental Protection Agency, Environmental Monitoring Systems
Laboratory; pp. 152-166; EPA report no. EPA-600/9-84-019. Available from: NTIS, Springfield, VA;
PB85-144053/REB.
Goldan, P. D.; Kuster, W. C.; Albritton, D. L.; Fehsenfeld, F. C.; Connell, P. S.; Norton, R. B.; Huebert,
B. J. (1983) Calibration and tests of the filter-collection method for measuring clean-air, ambient levels
of nitric acid. Atmos. Environ. 17: 1355-1364.
Gregory, G. L.; Hoell, J. M., Jr.; Torres, A. L.; Carroll, M. A.; Ridley, B. A.; Rodgers, M. O.;
Bradshaw, J.; Sandholm, S.; Davis, D. D. (1990a) An intercomparison of airborne nitric oxide
measurements: a second opportunity. J. Geophys. Res. [Atmos.] 95: 10129-10138.
Gregory, G. L.; Hoell, J. M., Jr.; Carroll, M. A.; Ridley, B. A.; Davis, D. D.; Bradshaw, J.; Rodgers, M. O.;
Sandholm, S. T.; Schiff, H. L; Hastie, D. R.; Karecki, D. R.; Mackay, G. L; Harris, G. W.; Torres,
A. L.; Fried, A. (1990b) An intercomparison of airborne nitrogen dioxide instruments. J. Geophys. Res.
[Atmos.] 95: 10103-10127.
Gregory, G. L.; Hoell, J. M., Jr.; Ridley, B. A.; Singh, H. B.; Gandrud, B.; Salas, L. J.; Shelter, J. (1990c)
An intercomparison of airborne PAN measurements. J. Geophys. Res. [Atmos.] 95: 10077-10087.
Gregory, G. L.; Hoell, J. M., Jr.; Huebert, B. J.; van Bramer, S. E.; LeBel, P. J.; Vay, S. A.; Marinaro,
R. M.; Schiff, H. I.; Hastie, D. R.; Mackay, G. L; Karecki, D. R. (1990d) An intercomparison of
airborne nitric acid measurements. J. Geophys. Res. [Atmos.] 95: 10089-10102.
Griffith, D. W. T.; Schuster, G. (1987) Atmospheric trace gas analysis using matrix isolation-Fourier transform
infrared spectroscopy. J. Atmos. Chem. 5: 59-81.
August 1991 6-75 DRAFT-DO NOT QUOTE OR CITE
-------
1 Grosjean, D.; Harrison, J. (1985a) Peroxyacetyl nitrate: comparison of alkaline hydrolysis and
2 chemiluminescence methods. Environ. Sci. Technol. 19: 749-752.
3
4 Grosjean, D.; Harrison, J. (1985b) Response of chemiluminescence NOX analyzers and ultraviolet ozone analyzers
5 to organic air pollutants. Environ. Sci. Technol. 19: 862-865.
6
7 Grosjean, D.; Parmar, S. S. (1990) Interferences from aldehydes and peroxyacetyl nitrate when sampling urban
8 air organic acids on alkaline traps. Environ. Sci. Technol. 24: 1021-1026.
9
10 Grosjean, D.; Fung, K.; Collins, J.; Harrison, J.; Breitung, E. (1984) Portable generator for on-site calibration
11 of peroxyacetyl nitrate analyzers. Anal. Chem. 56: 569-573.
12
13 Grosjean, D.; Parmar, S. S.; Williams, E. L., II. (1990) Southern California air quality study: a search for
14 methyl nitrate. Atmos. Environ. Part A 24: 1207-1210.
15
16 Hanseo, D. A. (1989) Measuring trace gases with FM spectroscopy. EPRI J. (June): 42-43.
17
18 Harker, A. B.; Richards, L. W.; Clark, W. E. (1977) The effect of atmospheric SO2 photochemistry upon
19 observed nitrate concentrations in aerosols. Atmos. Environ. 11: 87-91.
20
21 Helas, G.; Flanz, M.; Wameck, P. (1981) Improved NOX monitor for measurements in tropospheric clean air
22 regions. Int. J. Environ. Anal. Chem. 10: 155-166.
23
24 Helas, G.; Broil, A.; Rumpel, K.-J.; Warneck, P. (1987) On the origins of night-time NO at a rural
25 measurement site. Atmos. Environ. 21: 2285-2295.
26
27 Helmig, D.; Mueller, J.; Klein, W. (1989) Improvements in analysis of atmospheric peroxyacetyl nitrate (PAN).
28 Atmos. Environ. 23: 2187-2192.
29
30 Hering, S. V.; Lawson, D. R.; Allegrini, I.; Febo, A.; Perrino, C.; Possanzini, M.; Sickles, J. E., II; Anlauf,
31 K. G.; Wiebe, A.; Appel, B. R.; John, W.; Ondo, J.; Wall, S.; Braman, R. S.; Sutton, R.; Cass, G. R.;
32 Solomon, P. A.; Eatough, D. J.; Eatough, N. L.; Ellis, E. C.; Grosjean, D.; Hicks, B. B.; Womack,
33 J. D.; Horrocks, J.; Knapp, K. T.; Ellestad, T. G.; Paur, R. J.; Mitchell, W. J.; Pleasant, M.; Peake,
34 E.; MacLean, A.; Pierson, W. R.; Brachaczek, W.; Schiff, H. I.; Mackay, G. I.; Spicer, C. W.;
35 Stedman, D. H.; Winer, A. M.; Biermann, H. W.; Tuazon, E. C. (1988) The nitric acid shootout: field
36 comparison of measurement methods. Atmos. Environ. 22: 1519-1539.
37
38 Higashi, T.; Imasaka, T.; Ishibashi, N. (1983) Thermal lens spectrophotometry with argon laser excitation source
39 for nitrogen dioxide determination. Anal. Chem. 55: 1907-1910.
40 9
41 Highsmith, V. R.; Bond, A. E.; Howes, J. E., Jr. (1986) Particle and substrate losses from Teflon and quartz
42 filters. Atmos. Environ. 20: 1413-1417.
43
44 Hoell, J. M., Jr.; Gregory, G. L.; McDougal, D. S.; Carroll, M. A.; McFarland, M.; Ridley, B. A.; Davis,
45 D. D.; Bradshaw, J.; Rodgers, M. O.; Torres, A. L. (1985) An intercomparison of nitric oxide
46 measurement techniques. J. Geophys. Res. [Atmos.] 90: 12843-12851.
47
48 Hoell, J. M., Jr.; Gregory, G. L.; McDougal, D. S.; Torres, A. L.; Davis, D. D.; Bradshaw, J.; Rodgers,
49 M. O.; Ridley, B. A.; Carroll, M. A. (1987) Airborne intercomparison of nitric oxide measurement
50 techniques. J. Geophys. Res. [Atmos.] 92: 1995-2008.
51
52 Holdren, M. W.; Rasmussen, R. A. (1976) Moisture anomaly in analysis of peroxyacetyl nitrate (PAN).
53 Environ. Sci. Technol. 10: 185-187.
54
August 1991 6-76 DRAFT-DO NOT QUOTE OR CITE
-------
1 Holdren, M, W,; Spicer, C. W, (1984) Field compatible calibration procedure for peroxyacetyl nitrate. Environ.
2 Sci. Technol. 18: 113-116.
3
4- Holland, D. M.; McElroy, F. F. (1986) Analytical method comparisons by estimates of precision and lower
5 detection limit. Environ. Sci. Technol. 20: 1157-1161.
5
7 Hughes, E. E.; Rook, H. L.; Deardorff, E. R.; Margeson, J. H.; Fuerst, R. G. (1977) Performance of a
3 nitrogen dioxide permeation device. Anal. Chem. 49: 1823-1829.
)
3 Hustvedt, C. (1987) SAROAD quick look report [personal communication]. Research Triangle Park, NC: U. S.
I Environmental Protection Agency.
I
$ Jacobs, M. B.; Hochheiser, S. (1958) Continuous sampling and ultramicrodetennination of nitrogen dioxide in
t air. Anal. Chem. 30: 426-428.
i
5 Jenke, D. R. (1983) Comparison of three methods for the extraction of selected anions from media used for the
7 collection of airborne participates. J. Air Pollut. Control Assoc. 33: 765-767.
}
) John, W.; Reischl, G. (1978) Measurements of the filtration efficiencies of selected filter types. Atmos. Environ.
) 12: 2015-2019.
I Johnson, S. A.; Kumar, R. (1987) Comparison of three aerosol sampling techniques and the differences in the
I nitrate determined by each. In: Proceedings of the 1987 EPA/APCA symposium on measurement of toxic
!• and related air pollutants; May; Research Triangle Park, NC. Pittsburgh, PA: Air Pollution Control
I Association; pp. 183-188; EPA report no. EPA-600/9-87-010. Available from: NTIS, Springfield, VA;
} PB88-113402. (APCA publication VIP-8).
r
Johnston, P. V.; McKenzie, R. L. (1984) Long-path absorption measurements of tropospheric NO2 in rural New
Zealand. Geophys. Res. Lett. 11: 69-72.
Joos, L. F.; Landolt, W. F.; Leuenberger, H. (1986) Calibration of peroxyacetyl nitrate measurements with an
NOX analyzer. Environ. Sci. Technol. 20: 1269-1273.
Joseph, D. W.; Spicer, C. W, (1978) Chemiluminescence method for atmospheric monitoring of nitric acid and
nitrogen oxides. Anal. Chem. 50: 1400-1403.
Joshi, S. B.; Bufalini, J. J. (1978) Halocarbon interferences in chemiluminescent measurements of NOX. Environ.
Sci. Technol. 12: 597-599.
Jung, J.; Kowalski, J. (1986) Direct ambient nitrogen dioxide measurement by visible light absorption. In:
Proceedings of the 1986 EPA/APCA symposium on measurement of toxic air pollutants; April; Raleigh,
NC. Pittsburgh, PA: Air Pollution Control Association; pp. 845-855; EPA report no.
EPA-600/9-86-013. Available from: NTIS, Springfield, VA; PB87-182713. (APCA publication VIP-7).
Kadowaki, S. (1977) Size distribution and chemical composition of atmospheric particulate nitrate in the Nagoya
area. Atmos. Environ. 11: 671-675.
Kamiura, T.; Tanaka, M. (1979) Determination of nitrate in suspended particulate matter by high-performance
liquid chromatography with U.V. detection. Anal. Chim. Acta 110: 117-122.
Kamphake, L. J.; Hannah, S. A.; Cohen, J. M. (1967) Automated analysis for nitrate by hydrazine reduction.
Water Res. 1: 205-216.
August 1991 6-77 DRAFT-DO NOT QUOTE OR CITE
-------
1 Kelly, T. J. (1986) Modifications of commercial oxides of nitrogen detectors for improved response. Upton, NY:
2 U. S. Department of Energy, Brookhaven National Laboratory; report no. BNL-38000. Available from:
3 NTIS, Springfield, VA; DE86010536.
4
5 Kelly, T. J.; Stedman, D. H.; Kok, G. L. (1979) Measurements of H2O2 and HNO3 in rural air. Geophys. Res.
6 Lett. 6: 375-378.
7
8 Kelly, T. J.; Spicer, C. W.; Ward, G. F. (1990) An assessment of the luminol chemiluminescence technique for
9 measurement of NO2 in ambient air. Atmos. Environ. Part A 24: 2397-2403.
10
11 Klapheck, K.; Winkler, P. (1985) Sensitivity loss of a NOx-chemiluminescence analyzer due to deposit formation.
12 Atmos. Environ. 19: 1545-1548.
13
14 Kley, D.; McFarland, M. (1980) Chemiluminescence detector for NO and NO2. Atmos. Technol. 12: 63-69.
15
16 Klockow, D.; Niessner, R.; Malejczyk, M.; Kiendl, H.; von Berg, B.; Keuken, M. P.; Wayers-Ypelaan, A.;
17 Slanina, J. (1989) Determination of nitric acid and ammonium nitrate by means of a computer-controlled
18 thennodenuder system. Atmos. Environ. 23: 1131-1138.
19
20 Knnpp, K. T.; Durham, J. L.; Ellestad, T. G. (1986) Pollutant sampler for measurements of atmospheric acidic
21 dry deposition. Environ. Sci. Technol. 20: 633-637.
22
23 Komhyr, W. D.; Dutton, E. G.; Thompson, T; M. (1988) A general gravimetric dilution technique for preparing
24 trace calibration gases: N2O calibration gas preparation. Environ. Sci. Technol. 22: 845-848.
25
26 Kondo, Y.; Iwata, A.; Takagi, M. (1983) A chemiluminescent NOx-detector for the aircraft measurement.
27 J. Meteorol. Soc. Jpn. 61: 756-762.
28
29 Kondo, Y.; Matthews, W. A.; Iwata, A.; Morita, Y.; Takagi, M. (1987) Aircraft measurements of oxides of
30 nitrogen along the eastern rim of the Asian continent: winter observations. J. Atmos. Chem. 5: 37-58.
31
32 Kosmus, W. (1985) Summation method for monitoring nitrogen oxides. Int. J. Environ. Anal. Chem. 22:
33 269-279.
34
35 Kothny, E. L.; Cook, W. A.; Dimitriades, B.; Ferrand, E. F.; Nifong, G. D.; McDaniel, P. W.; Saltzman,
36 B. E.; Weiss, F. T. (1972) Tentative method of analysis for nitrate in atmospheric particulate matter
37 (brucine method). Health Lab. Sci. 9: 324-326.
38
39 Koutrakis, P.; Wolfson, J. M.; Slater, J. L.; Brauer, M.; Spengler, J. D.; Stevens, R. K.; Stone, C. L. (1988)
40 Evaluation of an annular denuder/filter pack system to collect acidic aerosols and gases. Environ. Sci.
41 Technol. 22: 1463-1468.
42
43 Leahy, D. F.; Phillips, M. F.; Garber, R. W.; Tanner, R. L. (1980) Filter material for sampling of ambient
44 aerosols. Anal. Chem. 52: 1779-1780.
45
46 Lindqvist, F. (1985) Determination of nitric acid in ambient air by gas chromatography/photoionization detection
47 after collection in a denuder. J. Air Pollut. Control Assoc. 35: 19-23.
48
49 Lipari, F. (1984) New solid-sorbent method for ambient nitrogen dioxide monitoring. Anal. Chem. 56:
50 1820-1826.
51
52 Lonneman, W. A. (1977) PAN measurement in dry and humid atmospheres. Environ. Sci. Technol. 11: 194-195.
53
August 1991 6-78 DRAFT-DO NOT QUOTE OR CITE
-------
i Lonneman, W. A.; Bufalini, J. J.; Seila, R, L. (1976) PAN and oxidant measurement in ambient atmospheres.
2 Environ, Sci. Techno!. 10: 374-380.
3
4 Lonneman, W. A.; Bufalini, J. J.; Namie, G. R. (1982) Calibration procedure for PAN based on its thermal
5 decomposition in the presence of nitric oxide. Environ. Sci. Technol. 16: 655-660.
6
7 Mackay, G. L; Schiff, H. I. (1987) Reference measurements of HNO3 and NO2 by tunable diode laser absorption
8 spectroscopy. In: Proceedings of the 1987 EPA/APCA symposium on measurement of toxic and related
9 air pollutants; May; Research Triangle Park, NC, Pittsburgh, PA: Air Pollution Control Association;
10 pp. 367-372; EPA report no. EPA-600/9-87-010. Available from: NTIS, Springfield, VA; PB88-113402.
11 (APCA publication VIP-8).
12
13 MacPhee, R. D.; Higuchi, J. E.; Leh, F. K. V. (1976) Comparison of oxidant measurement methods, ultraviolet
14 photometry, and moisture effects. In: Speciality conference on ozone/oxidants-interactions with the total
15 environment; March; Dallas, TX. Pittsburgh, PA: Air Pollution Control Association; pp. 166-179.
16
17 Maeda, Y.; Aoki, K.; Munemori, M. (1980) Chemiluminescence method for the determination of nitrogen
18 dioxide. Anal. Chem. 52: 307-311.
19
10 Matthews, R. D.; Sawyer, R. F.; Schefer, R. W. (1977) Interferences in chemiluminescent measurement of NO
II and NO2 emissions from combustion systems. Environ. Sci. Technol. 11: 1092-1096.
12
13 McClenny, W. A.; Galley, P. C.; Braman, R. $.; Shelley, T. J. (1982) Tungstic acid technique for monitoring
14 nitric acid and ammonia in ambient air. Anal. Chem. 54: 365-369.
55
16 Meyrahn, H.; Helas, G.; Warneck, P. (1987) Gas chromatographic determination of peroxyacetyl nitrate: two
17 convenient calibration techniques. J. Atmos. Chem. 5: 405-415.
^8
',9 Michie, R. M., Jr.; Sokash, J. A.; Fritschel, B. P.; McElroy, F. F.; Thompson, V. L. (1983) Performance test
>0 results and comparative data for designated reference methods for nitrogen dioxide. Research Triangle
11 Park, NC: U. S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory; EPA
>2 report no. EPA-600/4-83-019. Available from: NTIS, Springfield, VA; PB83-200238.
.3
4 Miller, D. P. (1984) Ion chromatographic analysis of Palmes tubes for nitrite. Atmos. Environ. 18: 891-892.
5
'6 Miller, D. P. (1988) Low-level determination of nitrogen dioxide in ambient air using the Palmes Tube. Atmos.
7 Environ. 22: 945-947.
8
9 Mueller, P. K.; Hidy, G. M. (1983) The sulfate regional experiment (SURE): report of findings. Palo Alto, CA:
0 Electric Power Research Institute; EPRI report no. EA-1901. 3v.
1
2 Mulawa, P. A.; Cadle, S. H. (1985) A comparison of nitric acid and particulate nitrate measurements by the
3 penetration and denuder difference methods. Atmos. Environ. 19: 1317-1324.
4
5 Mulik, J. D.; Williams, D. (1986) Passive sampling devices for NO2. In: Proceedings of the 1986 EPA/APCA
6 symposium on measurement of toxic air pollutants; April; Raleigh, NC. Pittsburgh, PA: Air Pollution
7 Control Association; pp. 61-70; EPA report no. EPA-600/9-86-013. Available from: NTIS, Springfield,
8 VA; PB87-182713. (APCA publication VIP-7).
9
August 1991 6-79 DRAFT-DO NOT QUOTE OR CITE
-------
1 Mulik, J. D.; Williams, D. E. (1987) Passive sampling device measurements of NO2 in ambient air. In:
2 Proceedings of the 1987 EPA/APCA symposium on measurement of toxic and related air pollutants;
3 May; Research Triangle Park, NC. Pittsburgh, PA: Air Pollution Control Association; pp. 387-397; EPA
4 report no. EPA-600/9-87-010. Available from: NTIS, Springfield, VA; PB88-113402. (APCA
5 publication VIP-8).
6
7 Mulik, J.; Puckett, R.; Williams, D.; Sawicki, E. (1976) Ion chromatographic analysis of sulfate and nitrate in
8 ambient aerosols. Anal. Lett. 9: 653-663.
9
10 Mullin, J. B.; Riley, J. P. (1955) The spectrophotometric determination of nitrate in natural waters, with
11 particular reference to sea-water. Anal. Chim. Acta 12: 464-480.
12
13 Namiesnik, J.; Gorecki, T.; Kozlowski, E,; Torres, L.; Mathieu, J. (1984) Passive dosimeters - an approach to
14 atmospheric pollutants analysis. Sci. Total Environ. 38: 225-258.
15
16 National Aeronautics and Space Administration. (1983) Assessment of techniques for measuring tropospheric
17 NxOy: proceedings of a workshop; August 1982; Palo Alto, CA. Hampton, VA: Langley Research
18 Center; NASA conference publication NASA-CP-2292. Available from: NTIS, Springfield, VA;
19 N84-13706.
20
21 National Bureau of Standards. (1988) NBS standard reference materials catalog 1988-89. Washington, DC: U. S.
22 Department of Commerce, Office of Standard Reference Materials; NBS special publication 260.
23 Available from: GPO, Washington, DC; S/N 003-003-02841-0.
24
25 Nelson, G. O. (1971) Controlled test atmospheres: principles and techniques. Ann Arbor, MI: Ann Arbor
26 Science Publishers, Inc.; pp. 126-134.
27
28 Nieksic, S. W.; Harkins, J.; Mueller, P. K. (1967) Some analyses for PAN and studies of its structure. Atmos.
29 Environ. 1: 11-18.
30
31 Nieboer, H.; van Ham, J. (1976) Peroxyacetyl nitrate (PAN) in relation to ozone and some meteorological
32 parameters at Delft in The Netherlands. Atmos. Environ. 10: 115-120.
33
34 Nielsen, T.; Hansen, A. M,; Thomsen, E. L. (1982) A convenient method for preparation of pure standards of
35 peroxyacetyl nitrate for atmospheric analyses. Atmos. Environ. 16: 2447-2450.
36
37 Norwitz, G.; Keliher, P. N. (1978) Inorganic interferences in the 2,4-xylenol spectrophotometric method for
38 nitrate and their elimination. Anal. Chim, Acta 98: 323-333.
39
40 Noxon, J. F. (1978) Tropospheric NO2. J. Geophys. Res. C: Oceans Atmos. 83: 3051-3057.
41
42 Noxon, J. F. (1983) NO3 and NO2 in the mid-Pacific troposphere. J. Geophys. Res. C: Oceans Atmos. 88:
43 11017-11021.
44
45 Okita, T.; Morimoto, S.; Izawa, M.; Konno, S. (1976) Measurement of gaseous and particulate nitrates in the
46 atmosphere. Atmos. Environ. 10: 1085-1089.
47
48 Palmes, E. D.; Tomezyk, C. (1979) Personal sampler for NOX, Am. Ind. Hyg. Assoc. J. 40: 588-591.
49
50 Palmes, E. D.; Gunnison, A. F.; DiMattio, J.; Tomczyk, C. (1976) Personal sampler for nitrogen dioxide. Am.
51 Ind. Hyg. Assoc. J. 37: 570-577.
52
53 Perrino, C.; De Santis, F.; Febo, A. (1988) Uptake of nitrous acid and nitrogen oxides by nylon surfaces:
54 implications for nitric acid measurement. Atmos. Environ. 22: 1925-1930.
August 1991 6-80 DRAFT-DO NOT QUOTE OR CITE
-------
Perrino, C.; De Santis, F.; Febo, A. (1990) Criteria for the choice of a denuder sampling technique devoted to
I the measurement of atmospheric nitrous and nitric acids. Atmos. Environ. Part A 24: 617-626.
!
!• Pierotrt, D.; Rasmussen, R. A. (1977) The atmospheric distribution of nitrous oxide. J. Geophys. Res. 82:
i 5823-5832:
»
' Pierson, W. R.; Braehaczek, W. W.; Komiski, T. J.; Truex, T. J.; Butler, J. W. (1980) Artifact formation of
! sulfate, nitrate, and hydrogen ion on backup filters: Allegheny Mountain experiment. J. Air Pollut.
> Control Assoc. 30: 30-34.
I
Platt, U.; Perner, D. (1983) Measurements of atmospheric trace gases by long path differential UV/visible
absorption spectroscopy. In: Killinger, D. K.; Mooradian, A., eds. Optical and laser remote sensing.
i New York, NY: Springer-Verlag; pp. 97-105. (Springer series in optical sciences: v. 39).
Platt, U. F.; Winer, A. M.; Biermann, H. W.; Atkinson, R.; Pitts, J. N., Jr. (1984) Measurement of nitrate
i radical concentrations in continental air. Environ. Sci. Technol. 18: 365-369.
Poizat, O.; Atkinson, G. H. (1982) Determination of nitrogen dioxide by visible photoacoustic spectroscopy.
Anal. Chem. 54: 1485-1489.
i
Possanzmi, M.; Febo, A.; Liberti, A. (1983) New design of a high-performance denuder for the sampling of
atmospheric pollutants. Atmos. Environ. 17: 2605-2610.
Possanzini, M.; Febo, A.; Cecchini, F. (1984) Development of a KI annular denuder for NO2 collection. Anal.
Lett. 17: 887-896.
Purdue, L. J.; Mauser, T. R. (1980) Review of U.S. Environmental Protection Agency NO2 monitoring
methodology requirements. In: Lee, S. D., ed. Nitrogen oxides and their effects on health. Ann Arbor,
MI: Ann Arbor Science Publishers, Inc.; pp. 51-76.
Rehme, K. A.; Smith, C. F.; Beard, M. E.; Fitz-Simons, T. (1984) Investigation of filter media for use in the
determination of mass concentrations of ambient particulate matter. Research Triangle Park, NC: U. S.
Environmental Protection Agency, Environmental Monitoring Systems Laboratory; EPA report no.
EPA-600/4-84-048. Available from: NTIS, Springfield, VA; PB84-199876.
Rhodes, R. C. (1987) [Personal communication, interrogation of PARS database]. Research Triangle Park, NC:
U. S. Environmental Protection Agency.
Rhodes, R. C.; Evans, E. G. (1988) Precision and accuracy assessments for state and local air monitoring
networks, 1986. Research Triangle Park, NC: U. S. Environmental Protection Agency, Environmental
Monitoring Systems Laboratory; EPA report no. EPA-600/4-88-007. Available from: NTIS, Springfield,
VA; PB88-180138.
Rickman, E. E., Jr.; Wright, R. S. (1986) Interference of nitrogenous compounds on chemiluminescent
measurement of nitrogen dioxide. Research Triangle Park, NC: Research Triangle Institute; RTI report
no. RTI/3180/24-0IF.
Rickman, E. E., Jr.; Green, A. H.; Wright, R. S.; Sickles, J. E., II. (1988) Laboratory and field evaluations of
extrasensitive sulfur dioxide and nitrogen dioxide analyzers for acid deposition monitoring. Research
Triangle Park, NC: Research Triangle Institute; pp. 3-8; RTI report no. RTI/3999/18-02F.
Rickman, E. E., Jr.; Green, A. H.; Wright, R. S.; Sickles, J. E., H. (1989) Laboratory and field evaluations of
extrasensitive sulfur dioxide and nitrogen dioxide analyzers for acid deposition monitoring. Research
Triangle Park, NC: Research Triangle Institute; RTI report no. RTI/3999/18-04F.
August 1991 6-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ridley, B, A.; Hewlett, L. C. (1974) An instrument for nitric oxide measurements in the stratosphere. Rev. Sci.
2 Ihstrum. 45: 742-746.
3
4 Ridley, B. A,; Carroll, M. A,; Gregory, G. L.; Sachse, G. W. (1988) NO and NO2 in the troposphere:
5 technique and measurements in regions of a folded tropopause. J. Geophys. Res. [Atmos.] 93:
6 15813-15830.
7
8 Roberts, J. M. (1990) The atmospheric chemistry of organic nitrates. Atmos. Environ. Part A 24; 243-287.
9
10 Roberts, J. M.; Norton, R. B.; Goldan, P. D.; Fehsenfeld, F. C. (1987) Evaluation of the tungsten oxide
11 denuder tube technique as a method for the measurement of low concentrations of nitric acid in the
12 troposphere. J. Atmos. Chem. 5: 217-238.
13
14 Roberts, J. M.; Fajer, R. W.; Springston, S. R. (1989) Capillary gas chromatographic separation of alkyl nitrates
15 and peroxycarboxylic nitric anhydrides. Anal. Chem. 61: 771-772.
16
17 Rodgers, M. O.; Davis, D. D. (1989) A UV-photofragmentation/laser-induced fluorescence sensor for the
18 atmospheric detection of HONO. Environ. Sci. Technol. 23: 1106-1112.
19
20 Rodgers, M. O.; Asai, K.; Davis, D. D. (1980) Photofragmentation-laser induced fluorescence: a new method
21 for detecting atmospheric trace gases. Appl. Opt. 19: 3597-3605.
22
23 Roseoe, H, K. (1982) Tentative observation of stratospheric N2OS. Geophys. Res. Lett. 9: 901-902.
24
25 Saltzman, B. E. (1954) Colorimetric microdetermination of nitrogen dioxide in the atmosphere. Anal. Chem. 26:
26 1949-1955.
27
28 Saltzman, B, E. (1980) Critique of measurement techniques for ambient nitrogen oxides. In: Lee, S. D., ed.
29 Nitrogen oxides and their effects on health. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.;
30 pp. 31-50
31
32 Saltzman, B. E.; Cook, W. A.; Dimitriades, B.; Ferrand, E. F.; Kothny, E. L.; Levin, L.; McDaniel, P. W.
33 (1972) Tentative method of analysis for nitrate in atmospheric particulate matter (2,4-xylenol method
34 (1)). In: Methods of air sampling and analysis. Washington, DC: American Public Health Association;
35 pp. 322-324.
36
37 Schjff, H. L; Hastie, D. R.; Mackay, G. L; Iguchi, T.; Ridley, B. A. (1983) Tunable diode laser systems for
38 measuring trace gases in tropospheric air: a discussion of their use and the sampling and calibration
39 procedures for NO, NO2, and HNO3. Environ. Sci. Technol. 17: 352A-364A.
40
41 Schiff, H. L; Mackay, G. L; Castledine, C.; Harris, G. W.; Tran, Q. (1986) A sensitive direct measurement
42 NO2 instrument. In: Proceedings of the 1986 EPA/APCA symposium on measurement of toxic air
43 pollutants; April; Raleigh, NC. Pittsburgh, PA: Air Pollution Control Association; pp. 834-844; EPA
44 report no. EPA-600/9-86-013. Available from: NTIS, Springfield, VA; PB87-182713. (APCA
45 publication VTP-7).
46
47 Schwartz, S, E, (1983) Mass-transport considerations pertinent to aqueous-phase reactions of gases in liquid-water
48 clouds. Upton, NY: U. S. Department of Energy, Brookhaven National Laboratory; BNL report no.
49 BNL-34174. Available from: NTIS, Springfield, VA; DE84006478.
50 .
51 Shaw, R. W., Jr.; Stevens, R. K.; Bowermaster, J,; Tesch, J. W.; Tew, E. (1982) Measurements of atmospheric
52 nitrate and nitric acid: the denuder difference experiment. Atmos. Environ. 16: 845-853.
53
August 1991 6-82 DRAFT-DO NOT QUOTE OR CITE
-------
1 Shores, R. C.; Smith, F. (1984) Stability evaluation of sulfur dioxide, nitric oxide and carbon monoxide gases in
2 cylinders. Research Triangle Park, NC: U. S. Environmental Protection Agency, Environmental
3 Monitoring Systems Laboratory; EPA report no. EPA-600/4-84-086. Available from: NTIS, Springfield,
4 VA; PB85-122646.
5 ' ...... ..
6 Sickles, J. E., II. (1987) Sampling and analytical methods development for dry deposition monitoring. Research
7 Triangle Park, NC: Research Triangle Institute; RH report ao. RTI/2823/00-15F.
8
9 Sickles, J. E., II; Hodson, L. L. (1989) Fate of nitrous acid on selected collection surfaces. Atmos. Environ. 23:
0 2321-2324. ,, , . -
1
2 Sickles, J. E., II; Michie, R. M. (1987) Evaluation of the performance of sulfation and nitration plates. Atmos.
3 Environ. 21: 1385-1391.
4
5 Sickles, J. E., II; Wright, R. S. (1979) Atmospheric chemistry of selected sulfur-containing compounds: outdoor
6 smog chamber study - phase 1. Research Triangle Park, NC: U. S. Environmental Protection Agency,
7 Environmental Sciences Research Laboratory; pp. 45-49; EPA report no. EPA-60Q/7-79-227. Available
8 from: NTIS, Springfield, VA; PB81-141525.
9
0 Sickles, J. E., II; Perrino, C.; Allegrini, I.; Febo, A.; Possanzini, M.; Paur, R. J. (1988a) Sampling and
1 analysis of ambient air near Los Angeles using an annular denuder system. Atmos. Environ. 22:
2 1619-1625.
3
4 Sickles, J. E., II; Hodson, L. L.; McClenny, W. A.; Paur, R. J.; Ellestad, T. G.; Mulik, J. D.; Anlauf, K. G.;
5 Wiebe, H. A.; Mackay, G. E.; Schiff, H. I.; Bubacz, D. K. (1988b) Field comparison of methods for
6 the measurement of contributors to acidic dry deposition. Atmos. Environ, (submitted).
7 .
8 Sickles, J. E., ri; Hodson, L. L.; Rickman, E. E., Jr.; Saeger, M. L.; Hardison, D. L.; Turner, A. R.; Sokol,
? C. K.; Estes, E. D.; Paur, R. J. (1989) Comparison of the annular denuder system and the transition
3 flow reactor for measurements of selected dry deposition species. JAPCA 39: 1218-1224.
1
2 Sickles, J. E., II; Grohse, P. M.; Hodson, L. L.; Salmons, C. A.; Cox, K. W.; Turner, A. R.; Estes, E. D.
3 (1990) Development of a method for the sampling and analysis of sulfur dioxide and nitrogen dioxide
\ from ambient air. Anal. Chem. 62: 338-346.
5 •
S Simeonov, V.; Puxbaum, H. (1977) A comparative study on the nitrate determination in airborne dust.
7 Mikrochim. Acta 2: 397-403.
$
? Singh, H. B.; Salas, L. J. (1983) Methodology for the analysis of peroxyacetyl nitrate (PAN) in the unpolluted
D atmosphere. Atmos. Environ. 17: 1507-1516.
I
I Singh, H. B.; Viezee, W. (1988) Enhancement of PAN abundance in the Pacific marine air upon contact with
? ' selected surfaces. Atmos. Environ. 22: 419-422.
t
5 Small, H.; Stevens, T. S.; Bauman, W. C. (1975) Novel ion exchange chromatographic method using
j conductimetric detection. Anal. Chem. 47: 1801-1809.
7
S Smith, R. G.; Bryan, R. J.; Feldstein, M.; Levadie, B.; Miller, F. A.; Stephens, E. R. (1972) Tentative method
) of analysis for peroxyacetyl nitrate (PAN) in the atmosphere (gas chromatographic method). In: Methods
) of air sampling and analysis. Washington, DC: American Public Health Association; pp. 215-219-
[
I Smith, J. P.; Grosjean, D.; Pitts, J. N., Jr. (1978) Observation of significant losses of particulate nitrate and
5 ammonium from high volume glass fiber filter samples stored at room temperature, J. Air Pollut. Control
1- Assoc. 28: 930-933.
August 1991 6-83 DRAFT-DO NOT QUOTE OR CITE
-------
1 Solomon, P. A.; Fall, T.; Salmon, L.; Lin, P.; Vasquez, F.; Cass, G. R. (1988) Acquisition of acid vapor and
2 aerosol concentration data for use in dry deposition studies in the South Coast Air Basin: volume I.
3 Pasadena, CA: California Institute of Technology, Environmental Quality Laboratory; EQL report 25 to
4 CAARB.
5
6 Spieef, C. W.; Schumacher, P. M. (1979) Particulate nitrate: laboratory and field studies of major sampling
7 interferences. Atmos. Environ. 13: 543-552.
8
9 Spicer, C. W.; Schumacher, P. M.; Kouyoumjian, J. A.; Joseph, D. W. (1978a) Sampling and analytical
10 methodology for atmospheric particulate nitrates: final report. Research Triangle Park, NC: U. S,
11 Environmental Protection Agency, Environmental Sciences Research Laboratory; EPA report no.
12 EPA-600/2-78-Q67. Available from: NTIS, Springfield, VA; PB-281933,
13
14 Spicer, C. W.; Ward, G. F.; Gay, B. W., Jr. (1978b) A further evaluation of microcoulometry for atmospheric
15 nitric acid monitoring. Anal. Lett. All: 85-95.
16
17 Spicer, C. W.; Howes, J. E., Jr.; Bishop, T. A.; Arnold, L. H.; Stevens, R. K. (1982) Nitric acid measurement
18 methods: an intercomparison. Atmos, Environ. 16: 1487-1500.
19
20 Spicer, C. W.; Joseph, D. W.; Schumacher, P. M. (1985) Determination of nitrate in atmospheric particulate
21 matter by thermal decomposition and chemiluminescence. Anal. Chem. 57: 2338-2341.
22
23 Stnehr, W.; Lahrnann, W.; Weitkamp, C. (1985) Range-resolved differential absorption lidar: optimization of
24 range and sensitivity. Appl. Opt. 24: 1950-1956.
25
26 Stephens, E. R, (1969) The formatioQ, reactions, and properties of peroxyacyl nitrates (PANs) in photochemical
27 air pollution. In: Pitts, J. N., Jr.; Metcalf, R. L., eds. Advances in environmental science and
28 technology: v. 1. New York, NY: WOey-Interscience; pp. 119-146.
29
30 Stephens, E. R.; Price, M. A. (1973) Analysis of an important air pollutant: peroxyacetyl nitrate. J. Chem.
31 Educ. 50: 351-354.
32
33 Stephens, E. R,; Burleson, F. R.; Cardiff, E. A. (1965) The production of pure peroxyacyl nitrates. J. Air
34 Pollut. Control Assoc. 15: 87-89.
35
36 Sturges, W. T.; Harrison, R. M. (1988) Thermal speciation of atmospheric nitrate and chloride: a critical
37 evaluation. Environ. Sci. Technol. 22: 1305-1311.
38
39 Szonntagh, E. L. (1979) Colorimetric azo dye methods for the atmospheric analysis of nitrogen dioxide;
40 historical development. Period. Polytech. Chem. Eng. 23: 207-215.
41
42 Tanner, R. L.; Fajer, R.; Gaffhey, J. (1979) Determination of parts-per-billion concentrations of aqueous nitrate
43 by derivatization gas chromatography with electron capture detection. Anal. Chem. 51: 865-870,
44
45 Tanner, R. L.; Daum, P. H.; Kelly, T. J. (1983) New instrumentation for airborne acid rain research. Int. J,
46 Environ. Anal. Chem. 13: 323-335.
47
48 Tanner, R. L.; Kelly, T. J.; Dezaro, D. A.; Forrest, J. (1989) A comparison of filter, denuder, and real-time
49 chemilumescence techniques for nitric acid determination in ambient air, Atmos. Environ. 23: 2213-2222.
50
51 Technicon. (1972) Industrial method no. 158-7 IW/tentative nitrate and nitrite in water and seawater. Tarrytown,
52 NY: Technicon Industrial Systems.
53
August 1991 6-84 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tesch, J. W,; Rehg, W. R.; Sievers, R. E. (1976) Mierodetermination of nitrates and nitrites in saliva, blood,
2 water, and suspended participates in air by gas chromatography. J. Chromatogr. 126: 743-755.
3
4 Thijsse, Th, R. (1978) Gas ehromatographic measurement of nitrous oxide and carbon dioxide in air using
5 electron capture detection. Atmos. Environ. 12: 2001-2003.
6
7 Torres, A. L. (1985) Nitric oxide measurements at a nonurban eastern United States site: Wallops instrument
8 results from July 1983 GTE/CITE mission. J. Geophys. Res. [Atmos.] 90: 12875-12880.
9 ...... - ' ...
-------
1 Winer, A. M.; Peters, J. W.; Smith, J, P.; Pitts, J. N., Jr. (1974) Response of commercial chemiluminescent
2 NO-NO2 analyzers to other nitrogen-containing compounds. Environ. Sci. Technol. 8: 1118-1121.
3
4 Winfield, T. W. (1977) A method for converting NO2 to NO by ferrous sulfate prior to ehemilurninescent
5 measurements. In: Proceedings of the American Chemical Society 173rd national meeting; March; New
6 Orleans, LA. Washington, DC: American Chemical Society; pp. 372-374.
7
8 Witz, S.; Wendt, J. G. (1981) Artifact sulfate and nitrate formation at two sites in the South Coast Air Basin. A
9 collaborative study between the South Coast Air Quality Management District and the California Air
10 Resources Board. Environ. Sci, Technol. 15: 79-83.
11
12 Witz, S.; Eden, R. W.; Wadley, M. W,; Dunwoody, C.; Papa, R,; Torre, K. J. (1990) Rapid loss of particulate
13 nitrate, chloride and ammonium on quartz fiber filters during storage. J. Air Waste Manage. Assoc. 40:
14 53-61,
15
16 Woebkenberg, M. L. (1982) A comparison of three passive personal sampling methods for NO2. Am. Ind. Hyg.
17 Assoc. J. 43: 553-561.
18
19 Wolff, G. T. (1984) On the nature of nitrate in coarse continental aerosols. Atmos. Environ. 18: 977-981.
20
21 Wright, R. S.; Tew, E. L.; Decker, C. E.; von Lehmden, D. J.; Barnard, W. F. (1987) Performance audits of
22 EPA protocol gases and inspection and maintenance calibration gases. JAPCA 37: 384-385.
23
24 Yanagisawa, Y.; Nishimura, H. (1982) A badge-type personal sampler for measurement of personal exposure to
25 NO, and NO in ambient air. Environ. Int. 8: 235-242.
26
27 Yoshizumi, K. (1986) Regional size distributions of sulfate and nitrate in the Tokyo metropolitan area in
28 summer, Atmos. Environ. 20: 763-766.
29
30 Yoshizumi, K.; Hoshi, A. (1985) Size distributions of ammonium nitrate and sodium nitrate in atmospheric
31 aerosols. Environ. Sci. Technol. 19: 258-261.
32
33 Yoshizumi, K,; AoM, K.; Matsuoka, T.; Asakura, S. (1985) Determination of nitrate by a flow system with a
34 chemiluminescent NOX analyzer. Anal. Chem. 57: 737-740,
35
36 Zafiriou, O. C.; True, M. B. (1986) Interferences in environmental analysis of NO by NO plus O3 detectors: a
37 rapid screening technique. Environ. Sci. Technol. 20: 594-596.
August 1991 6-86 DRAFT-DO NOT QUOTE OR CITE
-------
i -7. AMBIENT AND INDOOR CONCENTRATIONS
2 OF NITROGEN DIOXIDE
5 7.1 'INTRODUCTION. ...,,. ..... •-...,.' • , . ,.....,.„-..•.
5 The preceding chapters describe the fundamental chemistry of nitrogen oxides, the
7 sources (ambient and indoor), the transformations to other forms that take place during
3 transport in the atmosphere, and the methods of measurement. This chapter summarizes the
? nitrogen dioxide (NO^ concentrations that can be expected, both ambient and indoor, in
) natural and human environments. These concentrations are of critical concern because
t concentration is the main determinant of human exposure, discussed in the next chapter.
I Indeed, all effects on visibility, materials, vegetation, ecosystems, as well as human health,
$ rely on an accurate determination of exposure through knowledge of the NO2 concentrations.
I- For the most part, research on NO2 concentrations is clearly divided between ambient
> air environments and indoor air environments, although some exposure studies use personal
) monitors to measure continuous NO2 concentrations in both situations. Long-term
1 (multiple-year) patterns and trends are available only from stationary ambient monitors; data
I on indoor concentrations are collected predominantly in selected settings during comparatively
> short-term studies. There has been some concern that ambient air concentrations are not
) closely related to human exposure and that the primary determinant of exposure is the indoor,
I residential NO2 concentration because individuals spend more than 75% of their time in an
I indoor residential environment (Ryan et al., 1988). Data from these two categories of
5 environments are reviewed here independently. In Chapter 8, the appropriate apportionment
t and application of these data in quantifying exposure, effects, and subsequent risk assessment
» are discussed.
7.2 AMBIENT AIR CONCENTRATIONS OF NITROGEN DIOXIDE
Most of the information on ambient concentrations of nitrogen dioxide is available from
a network of monitoring stations established to determine compliance with the National
Ambient Air Quality Standard for NO2, which is 0.053 ppm or 100 jig/m3. Information is
August 1991 •' ,- 7-1 • DRAFT-DO NOT QUOTE OR CITE
-------
1 readily available from the database supported by this network through EPA's computerized
2 information system, Aerometric Information Retrieval System (AIRS) (AIRS, 1991).
3 Although much of this information is most closely related to compliance and enforcement, the
4 data also can be used for determining patterns and trends and as inputs to exposure
5 assessment. In some cases, the data can be used to augment existing epidemiological studies
6 where only indoor air data have been collected.
7 The following sections describe this network in greater detail and present some of the
8 results of analyses that focus on specific issues of exposure in this document. These analyses
9 proceed from a national picture of peak annual averages in Metropolitan Statistical Areas
10 through national 10-year and 3-year trends to characteristic seasonal and diurnal patterns at a
11 few selected stations and a brief examination of the incidence of 1-hour levels and associated
12 annual averages. No attempt was made to include information on ambient air concentrations
13 from monitoring sites not included in this network. Likewise, only information on NO2 was
14 extracted from the database, as only sparse information on other nitrogen oxides is available.
15
16 7.2.1 The National Air Monitoring Network
17 The National Air Monitoring Network consists of three types of sites. The National Air
18 Monitoring Station (NAMS) sites are located in areas where the concentrations of NO2 and
19 subsequent potential human exposures are expected to be high. Criteria for these sites have
20 been established by regulation (Federal Register, 1979) to meet uniform standards of siting,
21 quality assurance, equivalent analytical methodology, sampling intervals, and instrument
22 selection to assure consistent data reporting among the reporting agencies. For NO2, NAMS
23 sites are located only in urban areas with populations exceeding 1 million. The other two
24 types of sites are State and Local Air Monitoring Stations (SLAMS) and Special Purpose
25 Monitors (SPM), which meet the same rigid criteria for the NAMS sites but may be located
26 in areas not necessarily directed toward high concentrations and elevated human exposure.
27 For NO2, the sampling interval is one hour and the instrument method is
28 chemiluminescence for all stations. These instruments operate continuously and produce a
29 measurement every hour. In order to produce a valid annual average, at least half of the
30 possible 8,760 hourly readings must be reported.
31
August 1991 7-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.2.2 Peak Annual NO2 Averages in Metropolitan Statistical Areas, 1988-89
2 Table 7-1 lists the highest annual average for 103 Metropolitan Statistical Areas (MSAs)
3 reporting at least one NO2 monitoring station with valid data in 1988 and/or 1989 (U.S.
4 Environmental Protection Agency, 1990, 1991). The MSAs are listed in alphabetical order.
5 Peak annual averages in these MSAs range from 0.007 to 0.061 ppm. Figure 7-1 shows that
6 the collective mode for the peak annual average in this 2-year period is approximately
7 0.02 ppm.
8 . . .
9 7.2.3 Trends in Ambient NO2 Concentrations
0 In order to be included in the 10-year trend analysis in the annual National Air Quality
1 and Emissions Trend Report (U.S. Environmental Protection Agency, 1991) a station must
2 report valid data for at least eight of the last ten years. A companion analysis of the most
3 recent three years requires valid data in all three years. Analyses in the above reports cover
4 the periods 1980-1989 and 1987-1989, respectively; 148 sites met the 10-year requirement,
5 200 met the 3-year requirement. Of the 148 10-year sites, 36 were NAMS sites.
6 For the period 1980-89, there was an indication of a downward trend for the composite
7 annual average NO2 concentration in both the 148 sites and the 36 NAMS subset (U.S.
8 Environmental Protection Agency, 1991). The 1989 composite average was 5% less than the
9 1980 for both sets, but the difference was not statistically significant for either the full 148
0 sites or the NAMS subset (Figure 7-2). The composite annual average is strongly correlated
1 to population size. When sites in MSAs with 250 to 500 thousand population and 500 to
2 1,000,000 were compared to sites with more than 1,000,000, there was a regular pattern that
3 persisted over the full ten years. The sites over 1,000,000 were 0.01 ppm higher than the
4 sites with 250 to 500 thousand, and the mid-population sites were in between, as seen in
5 Figure 7-3. Throughout the ten years, Los Angeles, CA was the only urban area to record
6 violations of the annual NO2 NAAQS of 0.053 ppm.
7
8 7.2.4 Patterns in Ambient NO2 Concentrations
9 For the following analyses, a subset of sites was selected from the AIRS database using
D somewhat more exacting criteria. In this case, four years of continuous data (1986-89) were
1 required, with no more than a month of incomplete data. A month was considered
August 1991 7-3 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-1. MAXIMUM ANNUAL AVERAGE NO2 CONCENTRATIONS
REPORTED IN METROPOLITAN STATISTICAL AJUEAS - 1988,1989
Metropolitan Statistical Area
Albuquerque, NM
Allcntown-Bcthlchcm, PA-NJ
Anaheim-Santa Ana, CA
Atlanta, GA
Auitin, TX
Bakcrsiicld, CA
Baltimore, MD
Baton Rouge, LA
Beaumont-Port Arthur, XX
Beaver Co., PA
Bcrg«n-P««ic, NJ
Boston, MA
Bridgcport-Milford, CT
Buffalo, NY
Burlington, VT
Charleston, WV
Chicago, DL
Chieo, CA
Cincinnati, OH-KY-1N
Cleveland, OH
Dallas, TX
Denver, CO
Detroit, MI
El P*io, TX
Erie, PA
Evansvillc, 1N-KY
Ft. Wayne, IN
Ft. Worth-Arlington, TX
Frcino, CA
Orseniboro- et »L, NC
Harrisburg-Lcbanon-Carlislc, PA
Hartford, CT
Houston, TX
HunUngton-Ashlond, WV-KY-OH
Indianapolis, IN
Jacksonville, PL
Jersey City, NJ
Johnson City- et al. TN-VA
Johnstown, PA
Kansas City, MO-KS
Kenosha, WI
Lancaster, PA
Liltlo Ro«fc-N. Ottle Rock, AR
Loi Angeles-Long Beach, CA
Louisville, KY-OH
Manchester, NH
Memphis, TN-AR-MS
Miami-Hialeah, FL
MIddlesex-Somcrset-Hunterdon, NJ
Milwaukee, WI
MInncapolit-St. Paul, MN-WI
1988
(ppm)
0.018
0.020
0.046
0.030
-
0,032
0.034
0.021
-
0.020
0.036
0.033
0.027
0.022
0.019
0.024
0.032
0.016
0.030
0.031
0.021
0.039
0.023
0.021
0.016
0.022
0.010
0.014
0.032
0.018
0.021
0.020
0.028
0.016
0,024
0,019
0.033
-
0.019
0.014
0.014
0.020
0.010
0.061
0.023
0,024
0.034
0.017
0.025
0.027
0.020
1989
(ppm)
0.019
0.020
0.047
0.029
0.017
0.033
0.035
0.019
0.007
0.020
0.035
0.032
0.026
0.024
0.019
0.021
0.034
0.016
0.030
0.034
0.021
0.040
0.026
0.022
0.015
0.020
0.011
0.013
0.032
0.016
0.022
0.020
0.028
0.013
0.023
0.015
0.031
0.019
0.019
0.015
0.016
0.018
0.009
0.057
-
0.022
0.026
0.018
0.024
0.029
0.009
Metropolitan Statistical Area
Modesto, CA
Nashville, TN
Nassau-Suffolk, NY
New Havcn-Meriden, CT
New Orleans, LA
New York, NY
Newark, NJ
Norfolk- et al, VA
Oakland, CA
Oklahoma City, OK
Orlando, FL
Owensboro, KY
Oximrd-Ventura, CA
Philadelphia, PA-NJ
Pittsburgh, PA
Providence, RI
Provo-Orem, UT
Raleigh-Durham, NC
Reading, PA
Redding, CA
Richmond-Petersburg, VA
Riverside-San Bernardino, CA
Roanoke, VA0.02S
Sacramento, CA
Saginaw-Bay City-Midland, MI
St. Louis, MO-DL
Salinas-Seaside-Monterey, CA
Salt Lake City-Ogden, UT
San Diego, CA
San Francisco, CA
San Jose, CA
Santa Barbara- et al., CA
Santa Cruze, CA
Santa Rosa-Petaluma, CA
Scranton-Wilfces-Barre, PA
Springfield, MO
Springfield, MA
Steubenville-Weirtom, OH-WV
Stockton, CA
Tampa- et al., FL
Tucson, AZ
Tuba, OK
Vallejo-Fairfield-Napa, CA
Visalia-Tulare-Porterville, CA
Washington, DC-MD-VA
West Palm Beach- et al., FL
Wheeling, WV-OH
Wilmington, DE-NJ-MD
Worcester, MA
York, PA
1988'
(ppm)
0.027
0.012
0.033
0.029
0.024
0.041
0.040
0.017
0.026
0.029
-
0.015
0.018
0.039
0.030
-
0.028
•-
0.024
0.013
0.026
0.047
0.016
0.025
-
0.025
,
0.035
0.035
0.026
0.032
0.017
0.008
0,016
0.019
0.010
-
0.021
0.026
0.021
0.017
0.017
0.019
0.023
0.030
0.013
0.018
0.033
0.029
0.023
1989
(ppm)
0,027
0.012
0.029
0.028
0,022
0.049
0.038
0.020
0.025
0.015
0.013
0.014
0.027
0.040
0.028
0.024
0,028
0.012
0.023
0.014
0.025
0.045
0.014
0.025
0,009
0.026
0.014
0.034
0.032
0.026
0.032
0.027
0.009
.. 0.015
0.021
0.010
0.029
0.023
0.026
0.022
0.023
0.020
0.019
0,021
0.031
0.013
0.019
0.034
0,026
0,022
Source! U.S. EPA (1990, 1991).
August 1991
7-4
DRAFT-DO NOT QUOTE OR CITE
-------
13 -
12 -
11 -
10 -
9 -
8 -
7 —
6 -
5 -
4 -
3 -
2 -
1
P
nr
Mr '
0 0
p
r-l
.01
'
0
7
s /
/ s
/ /
/ /
02
•
0.0
p
3
m
n
' -P' P
I l
'i nri nnr n n r
X1 Mr'1 MMr M M r
0.04 0.05 0.06
Peak MSA annual average (ppm), 1988-89
Figure 7-1. Distribution of peak annual NO2 averages in 103 MSAs, 1988-89.
incomplete if less than 75% of the hourly measurements were reported. There were 155 sites
that met these requirements, and it was observed that the vast majority of sites had at least
90% of the data present for each month. The purpose of these selection criteria was to
provide a general selection of all types of stations that could yield some information about
seasonal and diurnal patterns. There was no attempt to select a geographically or
demographically representative set of sites. Of the 155 sites, 70 were in a residential setting,
45 in an industrial setting, 20 were commercial and 20 were miscellaneous, mostly
agricultural.
7.2.4.1 Seasonal Patterns
Seasonal patterns for these 155 sites were examined by compiling the monthly
distributions of hourly values and plotting the 50th, 90th, and 98th percentiles over the period
August 1991
7-5
DRAFT-DO NOT QUOTE OR CITE
-------
0.05 -
2
a.--
a.
z 0.04 -
g'
j|' 0.03 -
fe
UJ
i 0.02 -
8
0.01 -
n.nn -
_ __, Mft A fi"
NAAUo
x 5- — ,T T T •*" T ,,T. T T
1 -1 * I1" i -I- ± i — i ~j
- NAMS SITES (36) • ALL SITES (1 48)
I I I I I I I I I
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 7-2. National trend in the composite annual average nitrogen dioxide
concentrations at both NAMS and all sites with 95% confidence intervals,
1980-1989 (U.S, Environmental Protection Agency, 1991).
1986-1989. The 50th percentile approximates the geometric mean. The sample plots in
Figure 7-4a through 7-4d illustrate the diversity of seasonal patterns that exist in U.S. cities.
The pattern at the Long Beach site shows a broad span of winter months with elevated values
— from about September into March or April. The Denver site exhibits much narrower
winter peaks and broader summer troughs. The pattern at the Cleveland site is altogether
lower and such peaks as there are occur in the spring and summer months and are discernable
only in the 90th and 98th percentiles. The data for the Richmond site exhibit no discernable
pattern. These figures show that seasonal peaks do not occur at the same time for all sites
and, indeed, there are some locations with no prominent seasonal pattern. Therefore, no
simple generalization would seem warranted in assessing the extent to which these urban,
fixed-location, ambient monitors reflect a component of an individual's collective exposure to
N02.
August 1991
7-6
DRAFT-DO NOT QUOTE OR CITE
-------
CONCENTRATION, PPM
0.03
.025
n n9
.015
.01
.005
n
^
0...
_Q *"~^.»--i*"®'!.T .T£ .".&..".
13
'
MSA Size |
> 1000 000 > 500,000 > 250,000 1
** •ii«ii^Ji
i < i i i i i i
*>* ~ — *
•©---.-.©
°
i i
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Wore: 1937MSA population estimates
Figure 7-3. Metropolitan area trends in the composite annual average nitrogen dioxide
concentration, 1980-1989 (U.S. Environmental Protection Agency, 1991).
7.2.4.2 Diurnal Patterns
In Chapters 15 and 16, 0.2 ppm NO2 is mentioned as a possible benchmark
concentration above which physiologic responses may be detected. This Section discusses
stations reporting 1-h values above that level in 1988. '
In 1988, 216 NO2 monitoring stations reported to EPA's data bank a "valid" year's
data, that is, at least 75 %. of possible hourly values. Figure 7-5 compares their annual
averages with their second-high 1-h values. Four stations, all in California, reported annual
averages equalling or exceeding the annual standard: 0.053 ppm.
The eighteen stations reporting a second-high 1-h value greater than 0.2 ppm are
identified by state in Figure 7-5. The high values for the Oklahoma station, in Muskogee; •
and the New York City station have been discounted. At the Muskogee station, a string of
August 1991
7-7
DRAFT-DO NOT QUOTE OR CITE
-------
A. LONG BEACH, CALIFORNIA '86-'89
B. DENVER, COLORADO '86-'89
C. CLEVELAND, OHIO '86-'S9
t-
D. RICHMOND, VIRGINIA '8e-'89
Figure 7-4. Monthly 50th, 90th and 98th percentiles of 1-h NO2 concentrations at
selected stations, 1986 - 1989. (Annual averages are shown in parentheses.)
August 1991
7-8 DRAFT-DO NOT QUOTE OR CITE
-------
1
aoi
1-HOUR 8BCOND HIGH ppm
Figure 7-5. Annual average NO2 vs second high 1-h concentration at 216 stations-
1988. (Second high 1-h values > 0.2 ppm are identified by state.)
1 "0.3" values beginning one midnight are deemed spurious readings; at the New York City
2 station, 10 and 11 PM values four times the preceding values, followed by a 9-h gap, are
3 also suspect. The diurnal incidences of credible 1-h NO2 values greater than 0.2 ppm for the
4 other 16 stations are listed in Table 7-2. At most stations, late morning is the period when
5 these events are most likely to occur.
6 The 1-h data from the remaining 16 stations have been examined in detail, hour by
7 hour, to serve two purposes: to place these high values in perspective with the general
8 distributions of 1-h values, and to show the changing shape of the general distribution of all
9 1-h values through the 24 h in response to the cumulative influences of local emissions and
0 meteorology.
August 1991
7-9
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-2. HOURLY INCIDENCE OF NO2 CONCENTRATIONS GREATER THAN
0.2 PPM FOR STATIONS WITH MORE THAN ONE OCCURRENCE, 1988
Clock Hour
1988 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Anaheim, CA
Azusa, CA
Burbank, CA
Hawthorne, CA
Ln Habra, CA
Long Beach, CA
Los Angeles (01 13), CA
Los Angeles (1103), CA
Lymvood, CA
Pico Rivera, CA
San Diego, CA
Whittier, CA
Worcester, MA
Minneapolis, MN
Manchester, NH
Bayonne, NJ
2211
1 1 2
12 1133211
12311
1 1
2342 1
2 3
67663 13
132
1 1
22 11111
1 2
1 1
1 1 1
11 1
1 1
1 Figure 7-6 presents the hour-by-hour percent frequency distributions of 1-h NO2 values
2 for four of the stations selected from the group of sixteen in Table 7-2. The San Diego, CA,
3 location evidently often receives midday ventilation from sea breezes, causing the
4 distributions of 1-h values to contract and shift toward lower concentrations through the early
5 afternoon. Toward sunset, the upper tails of the distributions begin to extend toward higher
6 concentrations but the peaks remain around 0.01 ppm. After midnight, a subset of higher
7 concentrations emerges and, around sunrise, has shifted the distribution peaks to around
8 0.04 ppm. (The 3 AM or 4 AM hour is used for calibration checks at California stations; no
9 data are reported.)
August 1991 7-10 DRAFT-DO NOT QUOTE OR CITE
-------
I—I—I—I—I—I—i i i i1 i—I—i—I—r T"l "I" "I
0.00 0.02 0.04 O.OB 0.08 0.10 0.12 0.14 0.16 0.18 O20
1-hour HO2 concentration, ppm
0.00 0.02 0.04 0.03 0.08 0.10 0.12 0.14 0.16 0.18 0.20
1 -ho ur HO2 concentration, ppm
WOHCESTEH, MASSACHUSETTS; W
1800 MRS
1200 HHS
0800 MRS
0000 MRS
0.00 0.02 0,04 0.08 0.08 0.10 0.12 0.14 0,16 0,18 020
1-hour NO2 ccnoemratton, ppm
BAYONNE, NEW JERSEY; <83
1800 HHS
12OOHRS
0000 MRS
~i—i—i—i—i—i—i i i i i—i—i—i—i—i—i—i—i—r
0.00 0.02 0.04 0.08 0.08 0.10 0.12 0.14 0.16 0.18 0.20
1-hour NO2conceRMior>, ppm
Figure 7-6. Hourly relative frequency distributions of 1-h N(>2 values at four selected
stations for 1988, with numbers of values greater than 0.2 ppm.
August 1991
7-11
DRAFT-DO NOT QUOTE OR CITE
-------
1 Distributions for the Burbank, CA station, situated along the northern edge of the valley
2 near the mountains, are notably broader than those for the San Diego station. They exhibit
3 their narrowest spread between midnight and sunrise. Morning rush-hour emissions extend
4 the upper tail and shift the peak higher, briefly. Through midday the peaks shift somewhat
5 lower but the distributions remain broad. From late afternoon to midnight the peaks shift
6 higher and the distributions broaden further.
7 The Bayonne, NJ and Worcester, MA stations have similar patterns: their distributions
8 broaden during morning rush-hour, shift toward lower peaks through early afternoon when
9 ventilation usually improves, then broaden and shift toward higher concentrations through the
10 evening hours when winds characteristically subside and inversions form.
11 From this survey of NO2 data, it is concluded that 1-h concentrations greater than
12 0.2 ppm are infrequent events that lie well above the general distribution of 1-h values.
13 Begging the question of data validity, these excursions are presumably produced by the rare
14 coincidence of emissions and meteorological conditions, a distal rather than a proximal
15 feature on the very attenuated tail of a station's main data distribution. By some criteria,
16 they would be judged outliers rather than composing the tail of the main distribution and,
17 indeed, closer scrutiny might reveal additional instances of suspect validity.
18
19 7.2.4.3 Distributional Patterns
20 From the group of 216 stations with valid data for 1988, discussed in the previous
21 section, the subset of 43 stations with annual averages > 0.03 ppm are examined here.
22 Twenty-three stations are located in California, the other 20 are located in 13 other states.
23 For this group of stations with annual averages above this concentration of potential interest,
24 this question is posed: What is the relationship between the annual average and the incidence
25 of 1-h values above selected thresholds? In Figure 7-7, the percent of 1-h values greater than
26 0.03 ppm and 0.05 ppm are plotted vs. the annual averages for these 43 stations. For this
27 subset of stations in the upper portion of the annual average distribution, the percentages of
28 1-h values above the chosen thresholds bear a reasonably linear relationship to the stations'
29 annual averages: for the 0.03 group, R2 = 0.805; for the 0.05 group, R2 = 0.924.
August 1991 7-12 DRAFT-DO NOT QUOTE OR CITE
-------
s-
100%
00% -
80% -
70% -
60% -
60% -
40% -
30% -
20% -
10% -
0%
vi.*
n n
DD
+ + H-
0.025
0.035 0.045
Annual Average NO2, ppm
0.055
0.065
Figure 7-7. Percent of 1-h values above 0.03 and 0.05 ppm vs. annual
averages > 0.03 ppm, 1988.
1 Complete distributions of a year's 1-h values for four stations are compared in
2 Figure 7-8. Two stations were selected from the California basin area, two from the east
3 coast, with annual averages approximately in the middle of the annual average range depicted
4 in Figure 7-7. While the Los Angeles and Baltimore stations have similar annual averages,
5 the Baltimore station has a higher percentage of values around 0.04 ppm, and its distribution
6 slides under the Los Angeles distribution at around 0.07 ppm. Likewise, the Anaheim and
7 New York City annual averages are similar, but the New York City station has a higher
8 percentage of 0.05 ppm values, then drops under the Anaheim distribution at 0.07 ppm.
9 This very limited comparison suggests that, at fixed-site ambient monitoring locations
10 with annual averages above the national average (see Figure 7-2), percentages of values in the
11 middle of the 1-h distribution are rather consistently related to the annual average.
August 1991
7-13
DRAFT-DO NOT QUOTE OR CITE
-------
85*
80% -
25* -
20* -
1
S
Q.
Ann.X,ppm
X Los Angelas, CA (0113) OJ085
A Anaheim, CA (0001) 0.046
O Baltimore, MD (0040) OJ034
D Mow York City, MY (00105 0.041
16S
10% -
O05 0.1 0.15
1-hour NO2 concentration, ppm
•a-a -m & naonanca—
0,2 OJS 0.3
Figure 7-8. Relative distributions of 1-h NO2 values at selected stations, 1988.
1
2
3
4
5
6
I
2
3
4
5
6
7
However, as the discussions in Sections 7.2.4.1 and 7.2.4.2 have already shown, patterns of
more elevated N(>2 occurrence in the California basin area stand discernibly above those in
most of the rest of the country, the result of a unique combination of sources and
meteorology.
7.3 INDOOR AIR CONCENTRATIONS OF NITROGEN OXIDES
7.3.1 Background
In recent years there has been a growing realization that exposures to air contaminants
occur across a number of microenvironments (residences, industrial and non-industrial
workplaces, community air, automobiles, public access buildings, etc.) in which people spend
their time indoors and a major portion of their time at home (Szalai, 1972). Ideally, the total
exposure to a given air contaminant or category of air contaminants should be assessed over
August 1991
7-14
DRAFT-DO NOT QUOTE OR CITE
-------
1 all microenvironments in evaluating adverse health or comfort effects and in formulating cost
2 effective mitigation efforts to reduce or minimize the risks associated with exposure.
3 The indoor residential environment is particularly important in assessing total air
1 contaminant exposure because: (1) individuals spend the major portion of their time indoors
5 (Szalai, 1972); (2) the highest concentrations of several important air contaminants occur
5 indoors (National Research Council, 1981); and (3) the most susceptible segments of the
7 population (the old, the young and the infirm) are exposed for long periods of time. In
3 addition, efforts to reduce energy consumption through weatherization programs and the use
? of supplemental space heaters, have further increased potentially hazardous air contaminant
D levels indoors and thus have increased the relative importance of the nonindustrial indoor
1 settings.
I Nitrogen oxides are introduced to indoor environments through emissions from a variety
3 of combustion sources and in the infiltration or ventilation air from outdoors. The resulting
1 indoor concentration, both average and peak, is dependent on a complex interaction of several
5 interrelated factors affecting the introduction, dispersion and removal of the nitrogen oxides.
5 These factors include, for example, such variables as: (1) the type, nature (factors affecting
7 the generating rate of NOX) and number of sources; (2) source use characteristics;
3 (3) building characteristics; (4) infiltration or ventilation rates; (5) air mixing between and
? . within compartments in an indoor space; (6) removal rates and potential remission or
D generation by indoor surfaces and chemical transformations; (7) existence and effectiveness of
L air contaminant removal systems; (8) outdoor concentrations; etc. The interaction of these
I factors to produce the resulting indoor concentrations is usually considered within the
3 framework of the mass balance principal.
\ In its simplest form, where equilibrium conditions are assumed for a single
5 compartment with complete mixing and no air cleaner, the mass balance model can be
5 represented by the following equation:
\ •c.-c.'C, (?_i}
?
3 PAC0 . . ,
1 where: Cj = = Outdoor Air Contribution
I . ' A+K, . (7_2)
3
August 1991 7-15 DRAFT-DO NOT QUOTE OR CITE
-------
s/v
C9 = - = Indoor Source Contribution
1 2 A + K (7.3)
2
3
4 . ........
A
5 and where: Cj = steady-state indoor concentration of NOX (/*g/nr)
6 Cj = contribution to indoor NOX from outdoor air (/*g/m3)
7 C2 = contribution to indoor NOX from indoor sources (^tg/m3)
8 C0 = outdoor NOX (/*g/m3)
9 P = fraction of outdoor NOX that penetrates the building shell
10 A = air exchange rate in air changes per hour— ACH (IT1)
11 S = generation rate or source strength of NOX
12 V = volume of the indoor space (m)
13 K = removal rate of NOT by indoor chemical
1
14 transformations— equivalent ACH (h"1)
15
16
17 This simplified form of the model is generally used to evaluate NOX levels indoors. In
18 actuality, however, indoor spaces are often multi-compartments with incomplete mixing
19 where the source generation and contaminant removal rates and air contaminant
20 concentrations vary considerably in time. Equation 7-1 is particularily useful for determining
21 the impact on indoor air contaminant concentrations from sources that are used over relatively
22 long periods of time (e.g., unvented kerosene or gas space heaters) where steady-state or
!
23 equilibrium conditions are reached. When applied to sources which are intermittent in their
24 use (e.g., gas range or tobacco combustion), Equation 7-1 averages over the off/on periods of
25 the sources to determine average input parameters for the model. Short term indoor
26 concentrations of air contaminants associated with sources whose use varies considerably with
27 time can be modeled with the differential version of Equation 7-1 when detailed information
28 on the time variability of the source use, mixing and removal terms are available. Field data
29 on short term variability of contaminant concentrations and associated variables are, however,
30 not available.
31 This chapter summarizes the available data on the levels of nitrogen dioxide (NOo)
32 indoors largely within the framework of the simplified mass balance model shown in
33 Equation 7-1. Nitrogen dioxide concentrations measured indoors in homes with no known
34 sources are compared to outdoor levels as a function of season of the year, housing type and
35 region. Data on average and peak NO2 levels indoors as a function of individual and
36 combinations of indoor sources are reviewed with an emphasis on attempting to assess the
37 average contribution to indoor levels as a function source type, housing type and region.
August 1991 , 7-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 Data on the spatial distribution of NO2 between and within rooms indoors as a function of
2 source type are reviewed as are recent data on the removal of NO2 by indoor surfaces.
3 Available results on efforts to model (both empirical and physical/chemical) indoor NO2
4 levels are presented and discussed as are efforts to use chamber generated emission factors for
5 major indoor sources to predict levels in homes under actual use conditions.
6 Nitrogen dioxide is the major oxide of nitrogen considered in this chapter since a
7 considerable amount of indoor sampling data exists for it. Little air sampling data exists on
8 indoor concentrations of other oxides of nitrogen. The residential indoor environment has
9 been the major indoor microenvironment for that NO2 levels have been measured. There are
0 few data available on NO2 levels in other indoor microenvironments. This chapter focuses
1 on the findings of the major field studies that have evaluated NO2 levels in the residential
2 indoor microenvironment.
.3
4 7.3.2 Residences Without Indoor Sources
5 In the absence of any indoor source of NO2, indoor NO2 concentrations are a function
6 of the building envelope penetration factor (P), the air exchange rate (A), the reactivity rate
7 (K) and the. outdoor concentration (C0). This condition is represented in Equation 7-1 for
.8 steady-state conditions when S/V is set to zero:
.9
•0 PAC0
Cj = = Outdoor Air Contribution
\l A+K (7-2)
>2
13 There are no chamber or field studies that have measured the penetration factor (P) for NO2
i4 or field studies that have separate measures of both ventilation and the reactivity rate for
>5 NO2. Limited field data for ventilation rates in residences exists. There have, however,
16 been several field studies that have investigated levels of NO2 in residences. As part of these
i7 study designs, indoor and outdoor levels of NO2 were monitored in subsamples of homes that
18 had no known indoor sources of NO2. The indoor/outdoor ratios of NO2 measured in these
J9 studies provide general information on the role of the penetration factor, the air exchange rate
50 and the reactivity rate (PA/(A+K)) in impacting indoor NO2 concentrations in residences
51 without known NO2 sources.
. August 1991 7-17 DRAFT-DO NOT QUOTE OR CITE
-------
1 Table 7-3 presents the average outdoor NO2 concentrations (^g/m3) measured in several
2 large field studies and the corresponding average NO2 indoor/outdoor ratios by location in the
3 residences without indoor NO2 sources. The table also presents a breakdown of the data by
4 geographical location, housing type and season of the year. A number of observations can be
5 made from the data presented in Table 7-3 regarding the average indoor/outdoor NO2 ratio
6 for residences without known indoor sources:
8 (1) average ratios are always less than one except for the living room location
9 in the Melia et al. (1982).
10
11 (2) average ratios are highest in the summer and lowest in the winter with the
12 fall and spring periods falling between the winter and summer values.
13
14 (3) the average ratios among studies by season and location in the residence,
15 except for the British studies (Goldstein et al., 1979 and Melia et aL,
16 1982), the mobile home study (Petreas et al., 1988) and the Onondaga
17 County N.Y. study (Hartwell et al., 1988), is consistent. For example,
18 summer kitchen values range from 0.78 to 0.91, while winter values range
19 from 0.39 to 0.65.
20
21 (4) generally the ratios show a slight trend by location in a residence
22 decreasing from kitchen to the living area to the bedroom.
23
24 (5) ratios do not vary appreciably by outdoor concentration.
25
26 (6) ratios determined in the British studies appear to be very different than
27 those found in the U.S. studies with the British studies being higher.
28
29 (7) except for mobile homes, geographic location and housing type have little
30 effect on the ratios.
31
32
33 The above observations are based only upon the average indoor and outdoor
34 concentrations reported for each study listed in Table 7-3. These studies typically did not
35 report report the standard deviations or standard errors of the average indoor/outdoor ratios
36 from which the above observations are drawn, thus not allowing a test for statistical
37 significance of the trends. Such an analysis is needed, but beyond the scope of this
38 document.
August 1991 7-18 DRAFT-DO NOT QUOTE OR CITE
-------
c
era
c
v°+
\^5
v^5
t— fc
«-J
vo
O
j> •
TJ
7^
O
55
O
H
O
a
o
w.
o.
o
H
W
TABLE 7-3. AVERAGE OUTDOOR CONCENTRATIONS OF NO2 0*g/m3) AND AVERAGE INDOOR/OUTDOOR
RATIOS IN HOMES WITHOUT GAS APPLIANCES OR UNVENTED SPACE
HEATERS FROM FIELD STUDIES OF PRIVATE RESIDENCES
Reference
Southern California
Gas (1986)
Leaderer et al.
(1986)
Marbury et al.
(1988)
Petreas et al.
(1988)
Quackenboss
et al. (1986)
Quackenboss
et al. (1987) .
Ryan et al.
(1988)
Koontz et al.
(1986)
Hartwell et al.
(1988)
Spengler et al.
(1983)
Clausing et al.
(1984)
Location
Southern
California
New Haven,
Connecticut
Albuquerque,
New Mexico
California
Portage,
Wisconsin
Tucson,
Arizona
Boston,
Massachusetts
Northern
Central Texas
Suffolk County,
New York
Onondago County,
New York
Portage,
Wisconsin
Watertown,
Massachusetts
Housing Type
MIXED8
Single family
unattached
MIXED
Mobile homes
MIXED
MIXED
MIXED
Single family
unattached
Single family
unattached
Single family
unattached
Single family
unattached
Not given
Averaging
Time
7 days
14 days
14 days
7 days
7 days
14 days
14 days
5 days
7 days
7 days
7 days
3-4 days
Seasons
Summer
Spring
Winter
Winter
Winter 1
Winter 2
Summer
Winter
Summer
Winter
Summer
Spring/Fall
Winter
Summer
Fall
Winter/Spring
Winter
Winter
Winter
Average over
all seasons
November
December
Number of
Homes
70
100
69
60
60
56
46
23
47
47
56
41
23
117
117
124
9
49
66
25
18
10
Average NO2
Outdoors
71.9
43.5
91.2
13.2
14.1
19.6
25.9
44.6
15.2
17.2
19.9
25.6
36.8
31.7
37.8
33.5
53.8
35.5
21.7
12.8
37.0
46.0
Indoor/Outdoor Ratios
Kitchen
0.80
0.72
0.56
0.56
—
00
0.61
0.27
0.91
0.65
0.86
0.71
0.64
0.76
0.43
0.53
0.47
0.70
0.65
0.65
0.39
Bedroom
0.75
0.60
0.47
0.55
0.50
0.32
0.54
0.26
0.72
0.45
0.76
0.55
0.52
0.75
0.40
0.47
—
—
0.51
0.51
0.30
-------
> TABLE 7-3 (cont'd). AVERAGE OUTDOOR CONCENTRATIONS OF NO2 fcg/m3) AND AVERAGE
-------
The similarity in the indoor/outdoor ratios among studies is best highlighted by a
comparison of studies with the largest sample sizes: Southern California Gas (SoCalGas)
(1986), Ryan et al. (1988) and Leaderer et al. (1986). Despite differences in geographic
location, housing types and outdoor concentrations the average ratios by season are very
close. The Southern California Gas (1986) and Ryan et al. (1988) residences were selected
by random from clusters of homes while the Leaderer et al. (1986), was a convience sample.
Table 7-3 presents the average ratios determined from each study, while there is in fact
a distribution of ratios around each of the means reported. No study reported the distribution
of ratios, but Leaderer et al. (1986) did report an overall house NO2 indoor/outdoor ratio of
0.58 + 0.31 (n=123) for the winter sample period, demonstrating the variability of the ratio.
In the absence of frequency distributions of NO2 indoor/outdoor ratios for each field
study, the distributions of indoor and outdoor concentrations by season for data reported for
two large field studies can be evaluated for indoor and outdoor concentrations for residences
without known indoor NO2 sources. The cumulative frequency distributions of
concentrations of NO2 for residences without indoor sources by location in the residence and
for outdoors for two different geographic areas (Southern California and New Haven, CT.)
during the winter season is shown in Figures 7-9 and 7-10. A similar cumulative frequency
distribution for the summer period for homes monitored in the Southern California Gas study
is shown in Figure 7-11. The winter distributions for the two studies (Figures 7-9 and 7-10)
are similar despite the large differences in outdoor concentrations. Figures 7-9 and 7-10
highlight the substantial differences in indoor and outdoor concentrations during the winter
period. Figures 7-9 and 7-11, for the same population of homes in the Southern California
Gas study, highlight the differences in the distributions of NO2 concentrations indoors relative
to the outdoor levels as a function of season (winter verses summer). The distributions are
different with indoor levels much closer in concentration to outdoor levels in the summer
period.
The variability of the indoor/outdoor ratio by season observed in Table 7-3 is clearly
demonstrated in a study by Atkins and Law (1987). In this study two week NO2
concentrations indoors and outdoors over a 12-mo period were made in three houses without
any indoor source of NO2. The results, Figure 7-12, show the pronounced temporal trend.
August 1991 , 7-21 : DRAFT-DO NOT QUOTE OR CITE
-------
100
80
60
40
20
0
JANUARY 1985
- SOUTHERN CALIFORNIA
- GAS STUDY
ELECTRIC RANGE/
NO GAS APPLIANCES
D BEDROOMS
A KITCHEN
• OUTDOORS
I
20
40
60
80
100
120
Figure 7-9. Cumulative frequency distribution of NO2 concentrations (one-week
sampling period) by location for homes with no gas appliances for a winter
period in Southern California.
Source: Southern California Gas Co. (1986).
1 The seasonal variability in the ratio is in part due to seasonal variability in the air tightness of
2 the residences. In the winter months outside doors, windows, etc. are closed and air entering
3 the residences infiltrates through the building envelope with more effective removal of
4 outdoor NO2. During the summer period window, doors, etc. are more typically open
5 minimizing removal by the building envelope. No information is available on the impact on
6 the summer ratio for homes with air conditioning. Seasonal differences in the ratio and
7 variability in the ratio within a season from residence is also, no doubt, due to variations in
8 the penetation factor and reactivity rate for NO2. Variations in the penetration factor and
9 reactivity rate, like ventilation, can have a substantial impact on the indoor/outdoor ratios of
10 NO2. There is, however, little information on the variability of these two additional factors.
11 Indoor concentrations of NO2 in residences without indoor sources of NO2 will be
12 below outdoor levels, thus providing some degree of protection from outdoor concentrations.
13 Indoor/outdoor NO2 ratios in these homes will vary considerably as a function of season with
14 the ratio being lowest in the winter and highest in the summer. The indoor/outdoor ratio
15 does not appear to vary substantially as a function of geographic area, housing type or
August 1991
7-22
DRAFT-DO NOT QUOTE OR CITE
-------
100-
15
20
W
* OUTDOORS 13.2 (144)
A BEDROOMS 7.3 (146)
• KITCHEN 7.6 (147)
Q UVIHG ROOMS 7.3 (146)
I I I
SO 35
25
NQgCQNCEMTRATIQN,
Mgure 7-10. Cumulative frequency distribution and arithmetic means of NO2
concentrations (two week sampling period) by location for homes with no
kerosene heater and no gas range for a winter period in New Haven, CT,
area.
Source: Leaderer et al. (1984).
outdoor concentration. Only limited data are available on indoor/outdoor ratios for mobile
homes and for European residences.
7.3.3 Residences With Gas Appliances
It is estimated that gas (natural gas and liquid propane) is used for cooking, heating
water, or drying clothes in approximately 45.1% of all homes in the U.S. (U.S. Bureau of
the Census, 1982) and near 100% of the homes in other countries (e.g., The Netherlands).
Unvented, partially vented and improperly vented gas appliances, particularly the gas cooking
range and oven represent an important source category of NO2 emissions into the indoor
residential environment. Emissions of NO2 from these gas appliances (the source term, S, in
Equation 7-1) are a function of a number of variables relating the source type (range top or
oven, water heater, dryer, number of pilot lights, burner design, etc.), source condition (age,
maintenance, combustion efficiency, etc.), source use (number of burners used, frequency of
August 1991
7-23
DRAFT-DO NOT QUOTE OR CITE
-------
LU
O
DC
01
EL
1
s
O
100
80
60-
20
JULY 1984
SOUTHERN CALIFORNIA GAS STUDY
ELECTRIC RANGE/NO
GAS APPLIANCES
n-70
U BEDROOM
A KITCHEN
O OUTDOORS
20 40 60 80
NO2 CONCENTRATION,
100
120
Figure 7-11. Cumulative frequency distribution of NC>2 concentrations (one week
sampling period) by location for homes with no gas appliances for a
summer period in Southern California.
Source: Southern California Gas Co. (1986).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
use, fuel consumption rate, length of use, improper use, etc.) and venting of emissions
(existence and use of outside vents over ranges, efficiency of vents, venting of gas dryers,
etc.).
The factors that affect NO2 emissions from gas appliances into residences in
combination with the residence factors (house volume, number of rooms, infiltration rate,
room and whole house mixing rate, pollutant decay rate, etc.) result in indoor residential
NO2 concentrations associated with gas appliance use. The contribution of NO2 emissions
from gas appliances to indoor concentrations in simple terms can be represented by the source
contribution term of Equation 7-1:
C2 = S/V = Indoor Source Contribution
A + K
(7-3)
This assumes complete mixing in and between rooms and the representation of a highly
time varying source as an equivalent constant source. When the contribution of the outdoor
August 1991
7-24
DRAFT-DO NOT QUOTE OR CITE
-------
EC
OZ
00
INDOOR/OUTD
CONCENTRATI
°rf
ciZ
I*W
0.9
0.8
0.7
0.6
O.S
0.4
0.3
0.2
0.1
, .
""" ' m. ^9 * •
t * •
- * * • * »
*...••"**.*
..•*.* * :
"***..***.
. • * «. ••••*:..• •
"* • »»«*
• •." *"-. •• • * v*
A V V V
•
- ^ •
m
I I I I I I I ! I I I
JAN FEB MAR APR MAY JUN JUL AUQ SEP OCT NOV DEC
MONTH
Figure 7-12. Indoor/outdoor NO2 concentration ratios (2-week sampling periods) as a
function of time for three homes without indoor NO2 sources.
Source: Atkins and Law (1987).
1 NO2 levels are added (Equation 7-2), the resultant indoor concentration is determined
2 (Equation 7-1).
3 .
4 7.3.3.1 Average Indoor Concentrations and Estimated Contributions
5 There have been a large number of field studies, both in the U.S. and Europe, which
6 have sought to determine the levels (averaged over several days or more) of NO2 in
7 residences associated with the use of gas appliances and more specifically, gas ranges and
8 ovens. These field studies have been directed toward both assessing exposures to compliment
3 epidemiologic studies and to determine the range and distribution of indoor NC^ levels in
3 homes with gas cooking. A summary of the findings of the major field studies (those with
1 large sample sizes) directed toward assessing residential indoor NO2 levels in homes with gas
2 appliances is shown in Table 7-4. The results of 15 such studies (12 U.S., 2 British, and a
3 summary of several studies conducted in The Netherlands) are presented as the average NO2
4 concentrations measured outdoors and at various locations indoors by geographic location,
August 1991
7-25
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-4. INDOOR AND OUTDOOR CONCENTRATIONS OF NO
IN WITH GAS
eg
c
CO
1— »
VO
VO
-jj
o\
a
>
Tl
1
0
o
2
O
H
-------
c
(TQ
C
in
NO
SO
H
6
o
TABLE 7-4 (cont'd). INDOOR AND OUTDOOR CONCENTRATIONS OF NO2 0»g/m3) IN HOMES WITH GAS
APPLIANCES, AND THE CALCULATED AVERAGE CONTRIBUTION OF THOSE APPLIANCES
TO INDOOR RESIDENTIAL NO2 LEVELS
Average measured tlO^
Gig/m3)
Reference
HartweU et al.
(1988)
Goldstein et al.
(1985)
Spengler et al.
(1983)
*
Clausing et al.
(1984)
Goldstein et al.
(1979)
Melia et al.
(1982)
Location
Suffolk Co.,
NY
Onondago Co.,
NY
New York, NY
Portage,
WI
Watertown,
MA-
Middlesbrough,
UK
Middlesbrough,
UK
Housing
Type
Single family
unattached
Single family
unattached
Apartments
Single family
unattached
Not given
Not given
Not given
Averaging
Time
7 days
7 days
2 days
7 days
3-4 days
7 days
7 days
Gas Appliances Number of
Furnace Season Homes
oven/range
w/wo pilot
oven/range
w/wo pilot
oven/range
w/wo pilot
natural gas
oven/range
w/o pilots
LPgas
oven/range
w/o pilots
gas cooking
gas cooking
w/o pilots
gas cooking
Winter
Winter
Summer
Fall 1
Fall 2
Winter 1
Winter 2
Spring
AH seasons
All seasons
November
December
Winter
Winter
42
56
14
15
9
8
18
13
36
76
60
51
428
183
Outdoors
37.6
30.6
109
61
73
100
75
95
15.8
11.6
37
46
35
34.7
Kitchen Bedroom
77.5
62.6
122
96
108
121
126
121
65.5
65.6
74
86
213
—
—
0
98
65
66
76
63
82
36.7
37.6
45
46
58
60
Other
52.4
50
106
71
76
95
82
99
_
—
51
60
—
82.7
Indoor NO2 due to Source
Otg/m3)
Kitchen
60
41
30
53
45
61
81
55
55
58
50
68
179
—
Bedroom
—
—
6
22
15
16
18
16
29
31
'26
32
24
39
Other
37
27
13
28
25
35
37
33
_
—
33
44
—
61
Comment
1,9
9,11,12
1,13
1,9,14
1,15
1,16
o
c
W
O
o
-------
TO
c
en
VO
to
oo
73
a
o
TABLE 7-4 (coEt'd). INDOOR AND OUTDOOR CONCENTRATIONS OF N02 Oig/m3) IN HOMES WITH GAS
APPLIANCES, AND THE CALCULATED AVERAGE CONTRIBUTION OF THOSE APPLIANCES
TO INDOOR RESIDENTIAL NO2 LEVELS
Average measured NO2
(fig/m3)
Reference
Noy et al.
(1984)
Studies in the
Netherlands
Location
Arnet
Enschede
Edc
Vlagttwedde
Rotterdam I
Rotterdam 1
Housing
Type
Not given
Not given
Rural area
toner city
Inner city
Averaging
Time
7 days
7 days
7 days
7 days
7 days
Gas Appliances
Furntce Season
gas cooking Fall/Wint.
w/o pilots
water heaters
Number of
Homes
294
173
162
228
102
Outdoors
35
44
28
45
45
Kitchen
118
113
107
144
143
Bedroom
43
24
51
64
Other
97
54
51
80
73
Indoor JK>2 due to Source
(fig/nr)
Kitchen
97
89
90
117
117
Bedroom
17
7
24
37
Oilier
37
28
34
53
46
Comment
9,17
1. Background correction determined by multiplying the indoor/outdoor ratio for homes in the study with no indoor NOo sources for a given season times the outdoor NOo concentration measured
for the home with sources and subtracting the product from the indoor level measured in the house.
2, Homes contain forced air gas furnace. These homes are thought not to contribute significantly to indoor levels for this sample.
3. Homes with electric range/oven, forced air gas furnace, and gas water heater in home. Comparison is made with electric range/oven, forced air gas furnace, and gas water heater located outside
home.
4. Homes have gas range/oven with source contribution calculated after correction of a gas range/oven. Values are background corrected with gas stove.
5. Living room or activity room.
6. Sampling was done over 2 different periods for the same houses within the same winter period.
7. Outdoor values were obtained from 5 locations, housing type, mobile home.
8. Other location in home average, bedroom refers to average of levels in one or more bedrooms in house,
9. Other location in the main living room.
10. Other location is point nearest center of home.
11, 48-h samples over 30 consecutive days.
12. Indoor/outdoor (I/O) ratio is assessed to be 0.6, 0,7, and 0.85 for the Winter, Spring/Fall, and Summer periods respectively, for all locations, since no control home (no gas appliances)
measurements were available. Using these I/O ratios, the impact of sources was calculated as in footnote fl.
13. Each home was sampled 6 times over a one-year period.
14. Outdoor levels are average for homes with and without gas appliances.
15. Outdoor levels were recorded at 75 locations in the general sampling area and not home-specific. Bedroom levels were obtained for 107 of the 428 homes.
16. Outdoor levels were recorded at 82 locations in the general sampling areas and not home-specific. Outdoor levels were recorded at the beginning and end of the study.
17. Indoor/outdoor (I/O) ratio is assumed to be 0.6 for all locations, since no control home (no gas appliances, measurements were available. Using I/O ratio of 0.6, the impact of sources was
calculated as in footnote #1.
o
»
o
-------
1 housing type, sampling time and the type of cooking present. Many of the homes may have
2 had other gas appliances which may or may not have been explicitly described or analyzed,
3 It should be emphasized that only the average NO2 concentrations measured in each study are
4 presented and that there was a broad variation of concentrations associated with each mean.
5 All measurements employed passive NO2 monitors (Palmes et ah, 1976). Table 7-4 also
6 presents an estimate of the contribution of the apparent indoor source for all gas appliance to
7 the average indoor NO2 levels measured, in Equation 7-3. This was done by applying a
8 background correction factor (subtracting the contribution of outdoor concentrations) to the
9 measured indoor levels. The correction factor was determined by multiplying the mean
0 indoor/outdoor ratio (Equation 7-2) for homes in the study with no indoor NO2 sources
1 (Table 7-3) for a given season, times the mean outdoor NO2 concentration measured for the
2 homes with sources and then subtracting the product from the mean indoor levels.
3 Indoor/outdoor ratios of 0.6 for the winter, 0.85 for the summer, and 0.7 for the fall and
4 spring were assumed for those studies which did not have a sample of homes without indoor
5 sources.
5 A number of observations can be made from the data presented in Table 7-4 regarding
7 both the measured levels of NO2 in homes with gas appliances and the calculated contribution
B of the gas appliances to those levels:
) (1) measured indoor concentrations:
L
I (a) there is a very large variation in average indoor concentrations among
? U.S. studies with no apparent geographic pattern or association with
I housing type.
5
5 (b) the indoor concentrations recorded in the European studies are
1 considerably higher than those recorded in the U.S. studies, probably
I due to the geysers (demand gas water heaters) commonly used in many
) European countries.
)
(c) both indoor and outdoor concentrations are generally higher in the
! winter than the summer.
\
1- (d) a sharp concentrations gradient within a residence exists with the
J concentration highest in the kitchen and lowest in the bedroom.
August 1991 7-29 DRAFT-DO NOT QUOTE OR CITE
-------
1 (e) indoor concentrations are higher in homes that have gas cooking ranges
2 with pilot lights than those without pilot lights.
3
4 (f) the presence of wall or floor furnaces with gas appliances results in
5 higher concentrations than just gas appliances. Indoor NO2 levels
6 associated with wall or floor furnaces were thought to be due to leaky
7 flues (Southern California Gas, 1986). In a follow-up study (Southern
8 California Gas, 1987) the homes with high levels of NO2 were
9 investigated for the source of high levels and to determine the role of
10 leaky flues. The results of the follow-up study indicated that leaky flues
11 or maladjusted gas appliances were not, on average, a major cause of
12 elevated NG2 levels in the residences. Rather, it was postulated that ••
13 appliance use patterns, especially use of the gas range for space heating,
14 are the major source of elevated NO2 levels. The study found that 20%
15 of all residences with gas ranges use the ranges as supplemental heat
16 sources with 28% of those residences with wall or floor furnaces
17 reported occasional use of the gas range for space heating. This finding
18 is site specific and may be higher than would be expected at other
19 locations.
20
21 (2) calculated indoor source contributions:
22
23 (a) the contribution of gas cooking ranges and ovens to the average
24 concentrations of NO2 in the kitchen, bedroom and other rooms among
25 the U.S. studies by season is very consistent, varying by less than a
26 factor of two. Winter contributions to the kitchen, bedroom and other
27 locations across all studies averaged 50, 26, and 36 ^g/m3 respectively,
28 while summer contributions averaged 32, 16, and 19 pig/m3
29 respectively. The seasonal differences may be related to seasonal
30 differences in source use, infiltration and decay.
31
32 (b) the contribution of NO2 sources during the winter season is higher than
33 during the summer or fall/spring periods.
34
35 (c) the contribution of NO2 sources is highest in the kitchen and lowest in
36 the bedroom. •
37
38 (d) the contribution of NO2 to kitchen levels in the European studies is
39 higher than that in the U.S. studies (probably due to the European
40 geysers) but contribution to other locations in homes are similar in both
41 the European and U.S. studies.
42
43 (e) the extensive Southern California Gas study (1986) data indicate that
44 pilot lights in the gas cooking range add about 20 jig/m3 to the kitchen
45 levels in the winter and less to the other rooms. The water heater was
46 found on average to add approximately 12 ptg/m3 to the kitchen during
47 the winter and less to the rest of the house. This result, however, was
August 1991 7-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 . not found to be statistically significant due to high variability in the
2 measurements. The Southern California Gas data also provide
3 information on the associated concentration of indoor NO2 in homes
4 with gas fired wall and floor gas fired furnaces.
5
6 '....'
7 It is remarkable that the contribution of gas cooking to indoor NO2 levels is as
8 consistent as it is among studies for locations in the residences and by season, given the great
9 variability of the factors which govern the emissions (source type, source condition, source
10 use and source venting) and dilution and removal of NO2 indoors (house Volume, infiltration,
11 etc.). This consistency is not observed until the impact of outdoor concentrations is corrected
12 for. These background levels can vary considerably over time and geographic area. The
13 impact of gas cooking and possibly other unvented or improperly vented combustion sources
14 on indoor NO2 levels is superimposed upon the indoor background level resulting from
15 outdoor levels. In areas where outdoor levels are low, concentrations indoors from gas ,
16 appliances will be higher than and in many cases much higher than outdoor levels (e.g.,
17 Marbury et al., 1988; Quackenboss et al., 1987 and 1988; Spengler et al., 1983; Leaderer
18 et al., 1986; Ryan et al., 1988 ). If outdoor concentrations are high then indoor levels in
19 homes with gas appliances will be closer to and even lower than the outdoor levels (Southern
10 California Gas, 1986).
II The contribution of gas appliances to indoor levels of NO2 shown in Table 7-4 are
12 average concentrations calculated from the average levels reported by the investigators for
13 each field study. It is important to note that the variability around the calculated
14 contributions in Table 7-4 is generally large.
15
16 7.3.3.2 Spatial and Temporal Distributions
17 As demonstrated in Table 7-4 NO2 concentrations in residences with gas appliances
18 exhibit a pronounced variation by season and by location in a residence. The calculated
'.9 contribution of gas appliances to indoor levels of NO2 (corrected for outdoor contributions—
'0 Table 7-4) is highest in the winter and lowest in the summer, with the largest differences seen
>1 in the kitchen. The calculated total seasonal differences can be a factor of 1.5 to 2 (e.g.,
2 Quakenboss et al., 1986; 1987; Ryan et al., 1988; Southern California Gas, 1986; Spengler
August 1991 7-31 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al., 1983; Goldstein et al., 1985). The seasonal effect is related to variations in outdoor
2 NO2 levels, source use, infiltration and removal by interior surfaces.
3 Spatial distributions of NO2 within and among rooms in a house where a gas range or
4 oven is used is a function of mixing in the space. Goldstein et al. (1985) reported the
5 vertical distributions of NO2 levels in nine apartments in New York City where gas ranges
6 were used. The concentrations were 48-h average values (Palmes tubes) measured at five
7 elevations in the Mtchens and living rooms of each apartment. Figure 7-13 shows the result
8 of that study. A pronounced vertical gradient was observed in the kitchen, with the highest
9 levels observed at the ceiling and lowest at the floor. A similar but less pronounced gradient
10 was observed for the living area. Southern California Gas (1986) investigated the vertical
11 distribution of NO2 levels in gas cooking homes. The results showed a vertical gradient in
12 the kitchen for some but not all of the homes. The potential for a strong spatial NO2
13 gradient in kitchens with gas ranges suggests that placement of monitors in the kitchen during
14 field studies could have a significant impact on the concentrations measured. In considering
15 the results from such studies the monitor placement issue has to be considered in interpreting
16 the results.
17 All studies investigating NO2 concentrations in homes with gas appliances have found a
18 concentration gradient between rooms (Table 7-4), with the kitchen being highest and the
19 bedroom being lowest. This gradient is highlighted in Figure 7-14. Figure 7-14 also
20 highlights the seasonal differences in indoor NO2 levels in homes with gas appliances. In this
21 study (Spengler et al., 1983) season and location in the house were found to be statistically
22 significant predictors of NO2 levels in homes with gas appliances. The within home spatial
23 variations are related to such variables as air exchange rates among rooms, air mixing within
24 a room, volume of a house, location of the air sampler and the frequency and length of gas
25 appliance use.
26
27 7.3.3.3 Peak Indoor Concentrations
28 Peak indoor NO2 levels (highest levels on a time frame of three hours or less) in homes
29 with gas appliances occur during source use. Peak levels are associated with length of source
30 use, number of sources (e.g.5 number of gas burners used) and location at which the
31 measurement is taken relative to the source (e.g., immediately over the source or several feet
August 1991 7-32 DRAFT-DO NOT QUOTE OR CITE
-------
lu
of
o
§
o
8
CJ
UJ
X
8
7
6
5
4
3
2
1
0
LIVING ROOM
I
I
I
10 20 80 90 100 110 120 130
NO
Figure 7-13. Vertiele distribution of average NO2 concentrations (48-h sampling
periods) measured in nine New York City apartments. Plotted from data
by Goldstein et al. (1985).
away). Few studies have measured short term or peak NO2 levels in residences with gas
appliances.
In a study of the incidence of respiratory illness in households using gas and electricity
for cooking (Keller et al., 1979) used continuous chemiluminescence monitoring over 3-day
periods in 46 homes in the Columbus, Ohio area to measure continuous variations in indoor
NO2 and NO levels in relation to cooking times. The study found that variations in peak
NO2 levels in gas-cooking households reached as high as eight times the 24-h average values.
"3
In several households peak NO2 concentrations exceeded 1,900 ptg/m . The study did not
report the location of the sampler relative to the source, the number of sources, length of
source use or sample averaging time. As part of a study of respiratory disease rates and
August 1991
7-33
DRAFT-DO NOT QUOTE OR CITE
-------
1
uu
O
I
i
100-
90-
80-
70-
60-
50-
40-
30-
20-
10-
SUMMER
FALL
WINTER
OUTDOORS
LP-KIT
NG-KIT
O
- LP-BED
— NG-BED
D E-BED
1 2345676
PERIOD (JULY, 1980 - JUNE, 1981)
Figure 7-14. Mean NC>2 concentrations (1-week sampling periods) for eight sampling
periods by location in the home and type of cooking fuel.
Source: Spengler et al. (1983),
1
2
3
4
5
6
7
8
9
pulmonary function in children associated with NO2 exposure, Speizer et al. (1980) reported
short-term peak NO2 exposures in excess of 1,100 /ig/m3 in a kitchen within 3 ft of a gas
range/oven (gas oven on) with 1-h average peak exposure at approximately 665 i^gln?. Peak
level associated with a range top gas burner was 428 /tg/m3. Concentrations were monitored
continuously by chemiluminescence in only one house. Hosein and Bouhuys (1979), using
chemiluminescence, reported peak 2-h NO2 levels in a kitchen during use of a gas range of
over 3,000 /ig/m3 3 ft from the source.
In one of the first studies of the impact of gas ranges on indoor air quality, Wade et al.
(1975) reported continuous NO2 measurements (chemiluminescence) in three locations in four
August 1991
7-34
DRAFT-DO NOT QUOTE OR CITE
-------
1 houses with gas ranges. Measurements were spread over all four seasons and outdoor levels
2 were recorded. NO2 levels in the kitchen responded rapidly to gas range use with less rapid
3 response in other locations in the house. Peak 5-min concentrations 1 m from the gas range
4 exceeded 100 /*g/m3 over 50% of the time in two of the houses and 20% of the time in one
5 house. The data were not sufficient to construct cumulative frequency distributions for the
6 fourth house measured. Peak levels were considerably lower in other locations in the houses.
7 The NO2 frequency distributions for longer averaging times (e.g., 1 h) for other locations in
8 the houses were not reported.
9 The most extensive data collected to date on peak indoor levels of NO2 associated with
.0 gas appliance use is reported by Harlos et al. (1987). In this study an electrochemical-cell
.1 NO2 monitor was used to record time-averaging NO2 levels of 5 s and longer during cooking
.2 for both a stationary location near the gas range and for personal exposure for over
.3 50 volunteers. Table 7-5 shows the summary statistics reported from the study. The
.4 considerable variability in peak levels is due to source, source use and air mixing
5 characteristics in the Mtchen.
6
TABLE 7-5. SUMMARY STATISTICS FOR GAS RANGE NO2 MAXIMA (/tg/m3)
OVER SEVERAL AVERAGING TIMES
Seconds
Gas Cooking (n=18)
Mean Maxima
Standard Error
Maximum
Minimum
5
977
450
2,258
371
15
828
435
2,126
312
60
713
407
1,879
251
180
648
412
1,843
211
1,800
410
239
1,138
- 131
3,600
346
173
796
116-
Source: Harlos et al. (1987).
1 Lebret et al. (1987), conducted real-time NO2 concentration measurements at three
2 locations (kitchen, living room and bedroom) and outdoors for 12 Dutch homes using a
3 chemiluminescence monitor. Measurements were carried out over periods of 135 to 273 h.
4 The homes sampled had gas ranges and geysers (demand gas water heaters). Maximum 1-m
August 1991 7-35 DRAFT-DO NOT QUOTE OR CITE
-------
*2
1 average concentrations in the kitchens ranged from 400 to 3,808 ^g/nr in the kitchen, while
2 the living room and bedroom levels were typically on the order of 30% and 18% respectively
3 of the kitchen levels. Maximum 1-h average concentrations in the kitchen ranged from
4 230 to 2,055 jug/m3, while living room levels were typically 50% of the kitchen levels and
5 bedroom levels were about 30% of the Mtchen levels. The relative contribution to peak
6 levels from gas range use verses geyser use was not determined.
7 There is only a very limited data base available on short-term (less than three hours)
8 indoor concentrations of NO2 associated with gas appliance use. In the absence of adequate
9 field study data it is difficult to assess the short time-average levels indoors associated with
10 gas appliances. The existing data, however, indicate that short-term indoor NO2
11 concentrations will be considerably higher than those recorded for outdoors.
12
13 7.3.4 Unvented Space Heaters
14 Unvented space heaters are used in the colder climates to supplement central heating
15 systems or in more moderate climates as the primary source of heat. During the heating
16 season space heaters will generally be used for a number of hours during the day resulting in
17 emissions over a relatively long period of time.
18 Over the last several years there has been a dramatic increase in the use of unvented
19 kerosene space heaters in residential and commercial establishments primarily as a
20 supplemental heat source. The U.S. Consumer Product Safety Commission (1987) estimates
21 that a total of 17 million such heaters have been sold through 1987, with current yearly sales
22 of 1 million. A residential energy survey conducted by the U.S. Department of Housing and
23 Urban Development (1980) estimated that 3 million residences use unvented gas space heaters
24 (fueled by natural gas or propane) with their use more prevalent in the South Census region
25 of the U.S. The large number of unvented space heaters sold in the U.S. and the potential
26 for their high use, particularly during periods when energy costs rise quickly, make them an
27 important source of NO2 indoors.
28 As discussed in Chapter 4, NO2 emissions from unvented kerosene and gas space
29 heaters can vary considerably and are a function of heater design (convective, radiant,
30 combination burner designs, etc.), condition of heater and manner of operation (e.g., flame
31 setting). Levels of NO2 indoors resulting from heater use are a function of the heater
August 1991 7-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 emission variables along with heater use variables (number of hours of use, volume of house
2 heated, etc.) and the variables governing the dispersal and elimination of the NO2 emissions
3 indoors (infiltration, mixing, etc). The contribution of NO2 emissions from unvented
4 kerosene and gas space heaters to indoor concentrations in simple terms is given in the source
5 contribution term in Equation 7-3. Since kerosene and gas space heaters, unlike other gas
6 appliances, are typically used for several hours at a time they approach being a continuous
7 source and hence their contribution to indoor concentrations is : probably reasonably
8 represented by the steady state model in Equation 7-3. ......
9
10 7.3.4.1 Unvented Kerosene Space Heaters
11 Average Indoor Concentrations
12 The most extensive data collected to date on residential levels of NO2 associated with
13 the use of unvented kerosene space heaters are reported by Leaderer et al. (1986). This field
14 study of 333 homes in the New Haven, Connecticut area was conducted during the
15 1982-1983 heating season to both assess the range and distribution of air contaminants
16 associated with residential unvented combustion sources, with particular emphasis on NO2
17 levels related to kerosene heaters, and to assess exposures to compliment an epidemiologic
18 study of the health impact of heater use. The study employed a nested design for exposure
19 assessment that utilized questionnaires and several levels of air monitoring. Two-week
20 average NO2 levels were recorded in three locations in each house (kitchen, living room and
21 bedroom) and outdoors using Palmes tubes (Palmes et al., 1977).
22 The measured 2-week NO2 concentrations by location in the homes for six general
23 source categories are shown in Table 7-6. Also shown in Table 7-6 are the percent of homes
24 in which NO2 levels exceeded the primary ambient air quality standard. The findings
25 indicate that the greater the number of sources the higher the average concentrations of NO2.
26 Homes with one kerosene heater and no gas range/oven had NO2 levels four to five times
27 higher than the levels in homes without a heater or a gas range/oven. NO2 levels in homes
28 with a kerosene heater but no gas range/oven were roughly comparable to homes with a gas
29 range/oven only. The study also showed that homes with convective heaters had higher NO2
30 levels than homes with radiant heaters. It was noted that the average concentrations measured
31 may have been lower than would have been expected due to the particularly mild winter
August 1991 .7-37 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-6. TWO-WEEK AVERAGE NO2 LEVELS BY LOCATION FOR HOMES
IN SIX PRINCIPLE SOURCE CATEGORIES.*
NEW HAVEN, CONNECTICUT, AREA STUDY, WINTER, 1983
N02 Qtg/m3)
Source Category;
Location n Mean SDf %>100fig/m3
No Kerosene Heater or Gas Stove
Outdoors
House Average
Kitchen
Living Room
Bedroom
One Kerosene Heater, No Gas Stove
Outdoors
House Average
Kitchen
Living Room
Bedroom
No Kerosene Heater, Gas Stove
Outdoors
House Average
Kitchen
Living Room
Bedroom
One Kerosene Heater, Gas Stove
Outdoors
House Average
Kitchen
Living Room
Bedroom
Two Kerosene Heaters, No Gas Stove
Outdoors
House Average
Kitchen
Living Room
Bedroom
Two Kerosene Heaters, Gas Stove
Outdoors
House Average
Kitchen
Living Room
Bedroom
144
145
147
146
145
95
95
96
96
95
42
42
42
42
42
18
18
18
18
18
13
13
13
13
13
3
3
3
3
3
13.2
7.4
7.6
7.3
7.3
12.9
36.8
39.1
38.5
31.9
14.8
34.3
44.7
30.4
27.8
14.5
66.8
74.5
57.4
68.5
16.5
69.5
73.0
73.6
67.8
22.1
85.8
94.0
77.6
85.8
5.3
4.2
3.7
3.4
8.6
4.6
32.8
35.5
36.6
30.8
4.2
26.2
31.4
24.8
25.1
5.2
43.9
52.0
38.6
56.5
9.4
38.0
31.7
44.3
44.9
6.2
24.4
22.7
38.4
19.5
0
0
0
0
0
0
2.1
4.2
5.2
5.3
0
4.8
4.8
4.8
4.8
0
16.7
22.2
11.1
16.7
0
23.0
23.0
38.5
23.1
0
33.3
66.6
33.3
33.3
Repeat monitoring data (n=19) are included. Samples were lost for two homes; in one home the monitors
were capped early by the residents, and in the second home repeated efforts by the interviewers to retrieve the
monitors failed.
t SD, standard deviation.
August 1991 7-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 encountered during the study—the median of daily hours of kerosene heater use was only six
2 hours per day. ... .
3 The data in Table 7-6 represent the average levels over a 2-week period and do not
4 reflect the actual concentrations in the homes during heater use. Using the measured
5 concentrations and questionnaire data on neater use during air sampling and correcting for
6 outdoor levels, the authors calculated the indoor NO2 levels that may have existed during
7 actual heater use. The resultant calculated cumulative frequency distribution of NO2 levels
8 by location in the homes during heater use for homes with one kerosene heater and no gas
9 range/oven is shown in Figure 7-15. The adjusted data show that over 49% of the residences
0 with one kerosene heater had average NO2 concentrations in the house in excess of
1 100 /ig/m3 (0.053 ppm), with 8.4% in excess of 480 jttg/m3 (0.254 ppm). The levels would
2 be higher in homes with more than one kerosene heater and/or a gas range.
3 In a study of two Vermont homes in which kerosene space heaters were used, Ryan '
4 et al. (1983) found average NO2 concentrations indoors (Palmes tubes) to range from 19 to
o
5 304 jug/m over two sampling periods from 81 to 174 h. NO2 levels were highest in the
6 house that used two kerosene heaters as a primary heat source. Traynor et al. (1984), in a
7 field study of indoor air pollutants in residences with suspected combustion related sources,
8 reported one week average NO2 levels in three homes with kerosene heaters and five homes
9 with a kerosene heater and gas range. NO2 levels in the homes with kerosene heaters ranged
0 from 48 to 222 /ig/m3. The differences in concentrations reflected differences in usage.
1
2 Spatial Distributions
3 As shown in Table 7-6, NO2 levels in homes with a kerosene heater only do not
4 exhibits a pronounced concentration gradient among rooms in a house. Leaderer et al. (1984)
5 found no strong spatial gradient among rooms, which contrasted with the strong gradient
6 observed for homes with gas ranges. The relatively long operating periods for the heater, on
7 the order of several hours, and the strong convective heat output evidently foster rapid mixing
8 within the homes where they are used.
August 1991 7.39 DRAFT-DO NOT QUOTE'OR CITE
-------
b
UJ
o
or
I
2
O
100r
80
40
20
n-98
A. BEDROOM-0.057 ppm
| KITCHEN=0.088 ppm
O LIVING ROOM-O.Q81 ppm
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.0
NO2 CONCENTRATION, ppm
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
10
11
Figure 7-15. Cumulative frequency distribution and arithmetic means by location, of
average NO2 levels (2-week sampling periods) during kerosene heater use
for residences with one kerosene heater and no gas range, New Haven,
CT, area study, winter 1983. ;,
Source: Leaderer et al. (1986).
Peak Indoor Concentrations
Peak indoor concentrations of NO2 in residences with kerosene space heaters will occur
during source use. Since the heaters are used over several hours at a time, the peak levels
will, unlike levels associated with gas appliances, occur over periods of several hours. The
variability in peak levels from heater use period to heater use period for a given heater will
be driven by variations in infiltration rates, NO2 deposition rates and the volume of the house
heated (e.g., one room or several rooms).
As part of the nested air sampling design protocol used by Leaderer et al. (1987),
14 residences were monitored for periods of 43 to 209 h for NO2 levels in two locations in
the home (room with heater and a bedroom) and outdoors using a continuous
chemiluminescence monitor (Leaderer et al., 1984). Thirteen had kerosene heaters of which
August 1991
7-40
DRAFT-DO NOT QUOTE OR CITE
-------
1 four had a gas range while one house had a gas range but no kerosene heater. Peak levels in
2 the homes associated with kerosene heater use ranged from 19.3jo 847 /*g/m3 while peak
/5 ^ ^
3 levels typically exceeded 100 ^tg/m , In this set of houses levels were generally higher in the
I room where the heater was used. A sustained nature for peak levels over periods of several
5 hours was seen in the time trend for gases (outdoor levels subtracted) for a 24-hour period in
5 one of the houses monitored (Leaderer et al., 1986). Koontz.et'ai. (1986) conducted a study
7 in Texas and monitored two locations, in all homes (central location and remote location)
3 using Palmes tubes and by continuous chemiluminescence in a subset of homes.' The
? cumulative per cent distributions of NO2 levels measured in all hdfnes in both locations
) showed that 35% of the homes exceeded 100 /*g/m3 and 20% exceeded 480 ^g/m3.
I Table 7-7 presents a breakdown of the average levels recorded as a function of source
I catagories. The use of unvented gas space heaters (UVGSH) in homes either as a primary or
5 secondary heat source results in high levels of NO2 in those homes. The single most
1- important variable accounting for variations in indoor NO2 levels in homes using unvented
5 gas space heaters was the difference between indoor and outdoor temperatures. In homes
5 using the heaters as a primary heat source variations in indoor/outdoor temperature
7 differences accounted for 64% of the variation in NO2 levels and 33% of the variation f6r
J homes where the heaters are used as a supplemental heat source.
J In a study of 14 homes with one or more unvented gas space heaters (primary source of
) heat) in the Atlanta, Georgia area, McCarthy et al. (1987), measured NO2 levels by both
I chemiluminescence and passive monitors in two locations in the homes (room with the heater
I and a remote room in the house) and outdoors. Chemiluminescence measurements were
5 taken over 5-min periods in turn from each of the three sampling points for each house over a
!> 96-h sampling periods. The authors reported only the summary statistics for NO2 based on
j the continuously collected data. Eleven out of the fourteen UVGSH homes exceeded
5 100 fig/m3 during the sampling period. Mean values ranged from 40 to 1,460 ftg/m3 and
7 varied as a function of the use pattern of the heater. Only one of the homes used more than
I one heater during air sampling.
) The highest residential average NO2 concentrations observed for homes without leaky
) flues are associated with unvented gas space heaters. Unvented gas space heaters are a major
^ source of NO2 in residences.
August 1991 7-41 'DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-7. ONE-WEEK AVERAGE NO2 LEVELS IN HOMES IN NORTH
CENTRAL TEXAS BY SOURCE CATEGORY, WITH AND WITHOUT UNVENTED
-------
1 7.3.5 Other Sources
2 The major sources of NO2 in residences are unvented gas and kerosene space heaters,
3 gas appliances and outdoor NO2 levels. Improper use of gas appliances (e.g., using a gas
4 stove to heat living space) and improper operation of vented gas appliances (e.g., improper
5 use or malfunctioning gas appliances) can be important contributors to NO2 concentrations
6 measured indoors. Southern California Gas Company (1987) provides some data on the
7 contribution to indoor NO2 levels from improper use of gas appliances. The highest NO2
8 concentrations in the homes studies were associated with the use of a gas range/oven as a
9 supplemental heat source. There are no data that would allow estimation of the impact of
10 improperly operating wall or floor furnaces (spilling a portion of the exhaust fumes into the
11 home) on indoor NO2 levels.
12 Field data do not exist that would allow assessing the contribution of wood or coal
13 burning stoves or fireplaces to indoor NO2 levels. To the extent that there is leakage of the
14 exhaust gas into the living space during stoking the fire or through spilling of a portion of the
15 exhaust gas into the living space, these sources will contribute to indoor levels of NO2.
16 Using Palmes tubes, Good et al. (1982), compared seven day average NO2 levels in
17 homes with and without smokers and no gas ranges. Concentrations were measured in three
18 locations in each home (living room, bedroom and kitchen) and outdoors. There were a total
19 of 79 homes monitored (no gas range) over two seasons (winter and summer). Analysis of
20 the data indicates that the contribution to residential NO2 levels from cigarette smoking was
21 4 and 3 jttg/m3 in the summer and winter respectively. In another field study (Leaderer
22 et al., 1986) 14 day average NO2 levels were monitored using Palmes tubes in homes with no
23 known sources of NO2 (n=69) and homes having smokers only (n=61). Homes with
24 smokers showed significantly higher levels of NO2 only in the living room (t=2.09,
25 p=0.038). The mean NO2 level in the smokers' living room was 1.1 jug/m3 higher than in
26 the unexposed homes. Other studies, however, have found no impact on indoor residential
27 NO2 levels from cigarette smoking (e.g., Southern California Gas Company, 1986).
28
29 7.3.6 Modeling of Indoor Concentrations
30 Predictive or exploratory models for indoor concentrations of indoor air contaminants
31 generally take two basic forms. The first is a physical/chemical model that usually follows
August 1991 7-43 DRAFT-DO NOT QUOTE OR CITE
-------
1 some form of the general mass balance equation (e.g., Equation 7-1). The physical/chemical
2 modeling approach requires detailed information on the input parameters (source strengths,
3 infiltration rates, mixing, reaction rates, etc.) to predict the indoor concentrations. The input
4 parameter are either measured in chamber studies and in homes or estimated. The second
**
5 modelling approach is statistical in nature based upon empirical measurements. These models
6 make simple assumptions with little or no transformations of the independent variables input
7 to the model. The statistical models utilize as input parameters data obtained in large field
8 studies through both measurement and estimation (questionnaires). The statistical models are
9 typically simple linear models where the independent variables are used as they are recorded
10 ' from the questionnaires to explain variations in the concentrations of the air contaminants
11 measured. Both modeling approaches have been utilized in evaluating indoor concentrations
12 of NO2.
13
14 7.3.6.1 Physical/Chemical Models
15 The physical/chemical modeling approach has been used by a number of investigators in
16 chamber, test house and small field studies (involving a small number of homes) to estimate
17 emission rates of NO2 from combustion sources (e.g., Traynor et ah, 1982; Moschandreas
18 et ah, 1984; Leaderer, 1982), to estimate reactive decay rates (e.g., Yamanaka, 1984;
19 Borrazzo et ah, 1987; Leaderer et ah, 1986; Spicer et ah, 1986; Oezkaynak et ah, 1982), to
20 estimate the impact of ventilation and mixing on the spatial and temporal distribution of NO2
21 (e.g., Borrazzo et ah, 1987; Oezkaynak et ah, 1982; Traynor et ah, 1982), and to evaluate
22 the applicability of emission rates determined under controlled conditions in estimating indoor
23 concentrations of NO2 (e.g., Traynor et ah, 1982; Borrazzo et ah, 1987). More recently
24 three studies have been reported that utilize distributions of the input variables to (he mass
25 balance equation (emission rates, source use, decay rates, ventilation rates, etc.) determined
26 from the published literature, to estimate the distributions of NO2 levels indoors for specific
27 sources and combinations of sources (Traynor et ah, 1987; Hemphill et ah, 1987).- Use of
28 the physical/chemical models to evaluate model input parameters (e.g., source strength) in
29 explaining indoor levels of NO2 and efforts to use physical/chemical models to estimate
30 concentration distributions will be touched upon here.
August 1991 7-44 DRAFT-DO NOT QUOTE OR CITE
-------
1 . Borrazzo et al. (1987) applied a mass^balance model to NO2 levels measured in a town
2 house with gas appliances. NO2 emission rates were determined from a portable sampling
3 hood, reactive decay rates from a comparison of NO^ and sulfur hexafluoride (SF6) and
4 infiltration rates from SF6 decay rates. Comparing model predictions with measured
5 concentrations yielded a difference of 28 % for NO2. Differences in NO2 emission rates over
6 time of use of gas appliances and breakdown of the well-mixed single compartment model
7 assumption were thought to account for the discrepancies in the predicted versus measured
8 concentrations. In a study of the effects of ventilation on residential air pollution from a gas-.
9 fired range (Traynor et al., 1982), a gas range tested in a series of chamber studies was used
.0. in a test house and measured NOX levels were compared to those predicted by,a mass-balance
.1 model. In this study, infiltration, gas consumption and NO2 reactivity rates were measured.
.2 The results indicated good overall agreement between the measured and predicted NOX levels
.3 over the full test periods, although discrepancies in predicited and measured concentration
.4 were observed in the startup phase of the sources. The authors note that the main deficiency
.5 in the model is the assumption that the house is a single cell and as such does not address the
.6 spatial variation in concentrations.
.7 Traynpr et al. (1987) have reported on efforts to develop a macromodel for assessing
.8 indoor concentrations of combustion-generated pollutants. The model is a single-chamber,
.9 well-mixed mass-balance model (Equation 7-1) that utilizes experimentally derived estimates
JO of emission factors, building penetration factors and reactivity rates in combination with
II existing regional and national data (e.g., house volumes, market penetration of unvented
12 combustion sources) and source usage and infiltration models to estimate indoor pollutant
J3 concentration distributions. Deterministic and Monte Carlo simulation techniques are. used to
14 combine all of the inputs to yield the concentration distributions. The macro model will be
15 used to estimate the distribution of NO2 concentrations indoors. In a parallel development of
16 a statistical and physical/chemical model, Hemphill et al. (1987) developed a stochastic model
17 based upon the physical model to predict indoor NO2 concentration distributions in homes
IS using unvented gas space heaters as the primary source of heat. The physical/chemical model
19 utilizes experimental results to estimate input parameters and the field study data from Koontz
$0 et al. (1986) to relate indoor NO2 concentrations in homes with unvented gas space heaters
August 1991 7-45 DRAFT-DO NOT QUOTE OR CITE
-------
1 and indoor/outdoor temperature differences. The model predicts a broad distribution of
2 indoor NO2 concentrations indoors associated with use of unvented gas space heaters.
3 The use of physical/chemical (mass-balance) models (single compartment) to predict
4 indoor concentrations of NO2 indoors or distributions of concentrations in homes with
5 combustion sources requires accurate information on the input parameters. While data are
6 available for some of the input parameters under controlled experimental conditions (e.g.,
7 emission rates), there are very limited data available on the variability of the input parameters
8 in actual homes or the factors that control the variability of those inputs (e.g., variability of
9 emission or decay rates).. Obtaining field measurements or estimates of the inputs in large
10 numbers of homes would be expensive and time consuming. Such modeling efforts,
11 however, do help to identify the potential range of indoor NO2 concentrations and factors that
12 may result in high levels and the potential effectiveness of mitigation efforts.
13
14 7.3.6.2 Statistical/Empirical Models
15 Field studies that have measured NO2 concentrations in residences and associated
16 outdoor levels for time periods of a week or more have typically obtained questionnaire
i
17 information on sources in the residences, source use, building characteristics (house volume,
18 number of rooms, etc.), building use and meteorological conditions. In some cases additional
19 measurements such as temperatures have been recorded. Several investigators have attempted
20 to fit simple regression models to their field study data bases in an effort to determine if the
21 variations in NO2 levels seen among houses can be explained by variations in the
22 questionnaire responses and any additional measurements that may have been taken. The goal
23 has been to see how well questionnaire information or easily available information
24 (meteorological data) can predict indoor NO2 levels. In most cases a linear model has been
25 used but several investigators have used log transformations of variables. Table 7-8 presents
26 a summary of the regression models that have been fitted to large field study data bases. The
27 independent variables entered into the analysis (p <0.05), the R2, and standard error of the
28 estimate reported by the investigators are shown in the table. No standard errors were
29 reported for a number of the models. Linear regression models, with the exception of the
30 Petreas et al. (1988) model, explain from 40 to 70% of the variations in residential NO2
31 levels and typically have large standard errors associated with their estimates. While the log
August 1991 7-46 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-8. EMPIRICAL STATISTICAL MODELS (REGRESSION) FOR RESIDENTIAL NO CONCENTRATIONS
c
c
^o
^o
•~J
•t-
-J
O
5
^d
H
D
O
,_,
O
H
O
C!
O
H
O
REPORTED FROM FIELD STUDIES OF INDOOR
Reference
Southern
California
Gas Company
(1986)
Leaderer et al.
(1986)
• Marbury et al.
•(1988)
Petreas et al.
(1988)
Koonlz et al. .
(1986)
Noy et al.
(1984)
House
Location
bedroom
kitchen
bedroom
kitchen
living room
activity room
kitchen
bedroom
center of
house
kitchen
living 'room
Number of
^
Observations R
400 0.45
to to
" 578 0.63
.318 0.57
to
0.66
114 0.69
215 0.31
to to
262 0.39
29 0.40
to to
82 0.69
173 0.59
to
S.E. (fig/or) Sources
23.2 outdoor NO2
to gas range/oven
40.88 gas floor furnace
range pilots
oven pilots
gas water heater
age of gas oven
18.4 outdoor NO2
to convect./radiant
21 .2 kerosene heater
gas range/oven
other gas appl.
cig. smoking
141.7 outdoor NO2
gas stove with pilot light
gas stove w/o pilot light
gas dryer
floor/wall furnace
use ,of toaster/microwave
in gas stove house
— gas/cooking fuel
• location variable .
cigarette smoking
— unvented gas space heater
condition of
number of pilot lights
outdoor NO2
— outdoor NO2
presence of gas geyser
cooking fuel
type of space heating
log of NOj in rooms
Source
Use .
oven use
number of
occupants
oven cleaning
max temp
outdoor temp
convect./tadiant
kero. heater use
gas/oven use
number of
cigarettes
income level
difference between
indoor/outdoor
temp.
exposure to high
wind
use of gas range
for heat
LEVELS
Removal/
Dilution Interactive Terms
total house volume pilot lights ventilation volume
air exchange rate outdoor NO2
open windows
bedroom area
house age
number of bedrooms
gas range hood use
number of fireplaces convective kerosene heater use
squared
house volume
bedroom window
open
gas range exhaust fan
kitchen volume
number of doors
O
w
-------
1 transformations of variables have always produced a higher percent of explained variation due
2 to the skewed distribution of the original variables, interpretation of the coefficients in a
3 nonlinear model can require special attention.
4 The independent variables reported as being significant in each model are broken down
5 into four general catagories in Table 7-8: sources; source use; removal/dilution; and,
6 interactive terms. The only independent variables that are common to all models are those
7 that deal with the identification of sources in the residences and outdoor concentrations. The
8 identification of the sources accounted for the major portion of the explained variation in
9 indoor NO2 levels for all models. Those models that incorporate source use information or
10 proxies for source use generally produce better fitting models. Only one model, developed
11 from the most extensive data base (Southern California Gas Company, 1986), found a
12 number of variables related to the removal or dilution of NO2 indoors. Three models found
13 independent variable interactive terms to be significant. It is important to note that efforts to
14 model the most extensive data base on NO2 levels indoors associated with gas appliances
15 (Southern California Gas Company, 1986) only explained approximately 60% of the variation
16 in indoor levels with standard errors in the range of 40 /*g/m3. There is little uniformity
17 among the models in the form of the source use and removal/dilution terms or their
18 significance in the models.
19 Regression models developed from field studies employing questionnaires to explain
20 variations in indoor levels of NO2 have met with only moderate success. Better information,
21 through additional measurements and better questionnaire design, is needed on source type
22 and condition, source use, contaminant removal (infiltration and reactive decay) and between
23 and among room mixing, if the statistical/empirical models are to be used to estimate indoor
24 concentrations of NO2 in homes without measurements.
25
26 7.3.7 Predictions of Indoor and Outdoor Annual Averages
27 Environmental monitoring is expensive and there are many situations in which it is
28 necessary to estimate the annual average concentration of a pollutant from less than complete
29 data. In particular, either one 2-week average or two 2-week averages are often used to
30 characterize a site. The following section discusses the implications of this for both outdoor
31 and indoor sites.
August 1991 7-48 DRAFT-DO NOT QUOTE OR-CITE
-------
1 Several of the epidemiological studies used for quantitative analysis (see Chapter 14)
2 used single 2-week NO2 averages or used two 2-week NO2 averages to characterize both the
3 annual indoor and outdoor exposure. The following discussion considers the
4- representativeness of such estimates for outdoor exposure. To roughly estimate the possible
5 divergence from an actual annual mean resulting from selecting one or two 2-week averages,
5 data from EPA's Aerometric Information Retrieval System data bank (AIRS, 1991) were
7 selected. Forty sites were chosen (for their availability of data) to try to roughly estimate the
8 uncertainty of these estimates. The sites were split into three categories depending on the
3 value of the annual average: (1) .001 - .020 ppm, (2) .020 - .040 ppm, and. (3) > .040 ppm
D NO2. ^or eac^1 s*te' UP to 52 2-week averages (starting at the beginning of each week) were
1 calculated, along with the overall annual average. Each 2-week average can be thought of as
2 an estimate of the annual average, and the standard error of this estimate was calculated
3 across all sites within a particular exposure range. The standard error was estimated from the
4 absolute deviations instead of the deviations squared in order to avoid possible problems from
5 non-normally distributed data. In a similar manner, standard errors were estimated from two
5 2-week averages, using the average of 2-week averages that were 26 weeks apart. The
7 number, annual means, and average standard errors of the one 2-week and two 2-week
3 estimates are given in Table 7-9,
3 Table 7-9 shows that the standard error of the estimate goes up with the annual average
D itself. In general, one 2-week estimate has a standard error or about 25% of the mean,
1 whereas the two 2-week estimate has a standard error of about 15 % of the mean.
2
3
TABLE 7-9. AVERAGE ANNUAL NO2 MEANS (ppm) AND STANDARD ERRORS
OF ESTIMATES FOR 40 SITES IN THE U.S. BY ANNUAL AVERAGE
Annual Average
.001-.Q20
.020-.040
> .040
Number of Sites
7
16
17
Average Mean
0.0159
0.0313
0.0495
Average Standard
One 2-week Av.
.0046
.0080
.0111
Error of Estimate
Two 2-week Av.
.0028
.0041
.0056
August 1991 7-49 DRAFT-DO NOT QUOTE OR CITE
-------
1 A similar analysis was made for indoor sites using data from homes in the Albequerque
2 area supplied by Lambert (1991). One hundred sites were selected (by Lambert) including
3 25 electric homes and 75 gas stove homes. The extensive data set consisted of twenty-six
4 2-week averages for each home. Standard errors were calculated for three different
5 estimates: (1) one 2-week average, (2) two 2-week averages, and (3) a fixed value depending
6 on whether the home used a gas vs. electric stove. When all the electric homes were
7 combined, the overall average NO2 for the electric homes was 6.51 jwg/m3 (0.003 ppm). The
8 gas stove homes averaged 32.67 jwg/m3 (0.017 ppm). These two overall averages were used
9 as the fixed estimate of each home's annual NO2 level. The estimated standard errors were
10 calculated in the same manner as was done for the outdoor data. The results are in
11 Table 7-10.
12
TABLE 7-10. AVERAGE ANNUAL NO2 MEANS (jig/m3) AND STANDARD ERRORS
OF ESTIMATES FOR 100 HOMES BASED ON DATA OF LAMBERT (1991)
Source of
Exposure
Electric Stove
Gas Stove
Number of
Homes
25
75
Overall
Average
6.51
32.67
Average
One 2-week
2.44
18.66
Standard
Two
1.42
9.21
Error of the Estimate
2-week Fixed
2.81
22'.61
1 The standard errors from the estimates of indoor exposure are somewhat higher than the
2 standard errors of the outdoor estimates. The standard error of the one 2-week average is
3 about 50% of the mean and the standard error of the two 2-week average is about 25% of the
4 mean. The fixed estimate (using an overall average for electric homes and a separate overall
5 average for gas stove homes) had a slightly larger standard error than did the one 2-week
6 average. This suggests that measured 2-week averages are better for characterizing exposure
7 than is a fixed average based only on the presence or absence of a gas stove. A single
8 2-week average is only slighlty better, however. Although the standard errors are quite large
9 (compared with the mean), the estimates are adequate to distinguish a high exposure (usually
10 with a gas stove present) from a low exposure (usually with an electric stove present)
11 household.
August 1991 7-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 7.3.8 Eeactive Decay Rate of NO2 Indoors
2 A number of field studies of NO2 levels in residences have reported that NO2 is
3 removed more rapidly than can be accounted for by infiltration alone (Wade et al., 1975;
4 Macriss and ElMns, 1977; Oezkaynak et al., 1982; Ryan et al., 1983; Traynor et al., 1982;
5 Leaderer et al., 1986). NO2 indoors, is removed by infiltration/ventilation and by interior
5 surfaces and furnishings. The removal of NO2 by interior surfaces and furnishings and
7 reactions occurring in air is often referred to as the reactive decay rate of NO2. Failure to
3 account for the reactive decay rate (K in Equations 7-2 and 7-3) can lead to a serious
) underestimation of emission rate measurements in chamber and test house studies and a
) serious overestimate of indoor concentrations when using emission rates to model indoor
I levels and be a significant factor in the actual NO2 levels measured in residences. NO2
I reactive decay rate is typically determined by subtracting the decay of NO2, after a source is
5 shut off, to that of a relatively non-reactive gas (e.g., CO, CO2, SF6, NO). The measured
I reactive decay rates in the above mentioned field studies ranged from 0.1 to 1.6 air changes
) per hour. All studies noted that the reactive decay of NO2 is as important and in some cases
5 more important than infiltration in removing NO2 indoors. Leaderer et al. (1986), in the
1 continuous monitoring of NO2, NO, CO and CO2 in seven houses over periods ranging from
§ two to eight days, reported that the NO2 decay rate was always greater than that due to
J infiltration alone and was highly variable among houses and among time periods within a
) house.
I In an effort to identify the factors that control the NO2 reactive decay rate, a number of
I small chamber (MiyazaM, 1984; Spicer et al., 1986), large chamber (Moschandreas et al.,
5 1985; Leaderer et al., 1986) and test house studies (Yamanaka, 1984; Borrazzo et al., 1987;
I- Fortmann et al., 1987) have been conducted. The most extensive small chamber work is
i reported by Spicer et al. (1986) in which 35 residential materials were screened for NO2
> reactivity in a 1.64 m3 chamber, and in which a limited number of the materials were tested
J for the impact of relative humidity on the reactivity rate. Figure 7-16 shows the relative rates
I of NO2 removal for the materials screened. The figure indicates that many of the materials
> used for building construction and furnishings are significant sinks for NO2 and that their
) removal rate is highly variable. Many of the materials were found to reduce a significant
August 1991 7-51 DRAFT-DO NOT QUOTE OR CITE
-------
01234567 8
WALLBOARD
CEMENT BLOCK
WOOL CARPET
BRICK (USED)
MASONITE
COTTON/POLYESTER BEDSPREAD
PAINTED (FLAT LATEX) WALLBOARD
PLYWOOD
ACRYLIC FIBER CARPET
NYLON CARPET
VINYL WALLCOVERING (PAPERBACKED)
CEILING TILE
POLYESTER CARPET
ACRYUC CARPET
FURNACE FILTERS (NEW)
DEHUMIDIFIER
OAK PANELING
VINYL-COATED WALLPAPER
PARTICLE BOARD
FURNACE FILTERS (USED)
CERAMIC TILE
WOOL(80y«) POLYESTER(20%) FABRIC
COTTON TERRYCLOTH
PLANTS (WFTH SOIL
WALLTEX COVERING
WAXED ASPHALT TILES
WINDOW GLASS
USED FURNACE HEAT EXCHANGER
FORMICA COUNTER TOP
POLYETHYLENE SHEET
ASPHALT FLOOR TILES
VINYL FLOOR TILE
GALVANIZED METAL DUCT
PLASTIC STORM WINDOWS
01234567 8 9
RATE CONSTANT FOR N02 REMOVAL, 1/hr
Figure 7-16. Bar graph of N<>2 removal rate for various materials evaluated in a
1.64 m^ test chamber at 50% relative humidity.
Source: Spicer et al. (1986).
August 1991
7-52
DRAFT-DO NOT QUOTE OR CITE
-------
fraction of the removed NO2 to NO. In no cases was NO2 reemitted, although some
materials emitted NO. The authors noted that the materials that removed NO2 most rapidly
fall in two catagories; porous mineral materials of high surface area and cellulosic material
derived from vegetable matter. Higher relative humidities were found to enhance the
removal rate for some materials (e.g., wool carpet), reduce the removal rate for some (e.g.,
cement block) and have little effect on others (e.g., wallboard). In a series of small chamber
**
studies (0.69 nr) (MiyazaM, 1984) reactive decay rates for NO2 were found to vary as a
function of material type and increase with increasing surface area of the material, degree of
stirring in the chamber, temperature and relative humidity. A saturation effect was noted on
some of the carpets tested.
/•»
In a series of large chamber studies (34 m chamber), Leaderer et al. (1986) evaluated
the reactive decay rate of NO2 as a function of material type, surface area of material,
relative humidity and air mixing. The reactive decay rate was found to vary as a function of
material surface roughness and surface area. Carpeting was found to be most effective in
removing NO2 while painted wallboard was least effective., Increases in relative humidity
were associated with increases in removal rates for all materials tested, but the slope was a
shallow one. Of particular interest is the finding in this study that the degree of air mixing
and turbulence was a dominant variable in determining the reactive decay rate for NO2.
Moschandreas et al. (1985), evaluated six materials in a 14.5 m3 chamber and found
variations in decay rates by material types and a positive impact on NO2 decay rates in an
empty chamber with relative humidity.
Yamanaka (1984) in assessing NO2 reactive decay rates in a Japanese living room found
the decay to be comprised of both homogeneous and heterogeneous processes. The rates
were found to vary as a function of surface property and sharply as a function of relative
humidity. NO production during the decay was noted. In a test house study Fortmann et al.
(1987) noted that the NO2 decay rate tends to decrease as the concentration increases. It is
not clear whether this is due to surface saturation or second order kinetics. This study also
noted a sharp increase in NO levels during the NO2 decay indicating NO production as a
result of the NO2 decay. In a test house study conducted over a seven month period
Borrazzo et al. (1987) found that reaction rates for NO2 in. the test house were sensitive to
August 1991 7-53 DRAFT-DO NOT QUOTE OR CITE
-------
1 the location in the house where they were measured. This indicates that reaction losses
2 during transport of NO2 from room to room in a house may be important.
3 The reactive decay of NO2 in residences associated with interior surface materials and
4 furnishings is an important mechanism for removing NO2 in residences. NO2 reactive decay
5 rates vary as a function of the type of material and surface area of the material. The impact
6 of relative humidity on the decay rate is unclear with some studies showing a pronounced
7 impact (Yamanaka, 1984) and others showing moderate or little impact (e.g., Spicer et al.,
8 1986 and Leaderer et al., 1986). The degree of air mixing or turbulence can have an
9 important effect on the reactive decay rate. A by-product of NO2 removal by materials may
10 be NO production and a saturation effect may occur for some materials. Reactive decay of
11 NO2 in residences is highly variable between residences, within rooms in a residence and on
12 a temporal basis within a residence. The large number of variables controlling the reactive
13 decay rate make it very difficult to assess in large field studies through questionnaire or
14 integrated air sampling.
15
16
17 7.4 NITRIC AND NITROUS ACIDS CONCENTRATIONS
18 Brauer et al. (1991) used annular denuder filter pack sampling systems to measure the
19 gaseous pollutants HNO3, HONO, NO2 and NH3 during summer and winter periods in
20 Boston, Massachusetts. Outdoor levels of HNO3 exceeded indoor concentrations during both
21 seasons. For HONO and NH3, indoor concentrations were significantly higher than ambient
22 levels. Geometric mean indoor/outdoor NH3 ratios of 3.5 and 23 were measured for the
23 summer and winter sampling periods, respectfully. Summary data are shown in Table 7-11
24 and Figure 7-17.
25 Levels of NO2 (Figure 7-17B) were substantially higher than nitrogen acid species. The
26 low indoor concentrations of HNO3 are due to the lack of measurable indoor HNO3
27 production and the high surface reactivity of this acidic gas (Figure 7-17D). For all of the
28 homes, indoor HONO concentrations were greater than outdoor values (Figure 7-17A). The
29 data suggested that higher indoor NO2 concentrations are related to elevated indoor HONO
August 1991 7-54 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-11. SUMMARY STATISTICS FOR INDOOR AND OUTDOOR
CONCENTRATIONS (in ppb) OF GAS SPECIES MEASURED DURING
SUMMER AND WINTER SAMPLING PERIODS IN BOSTON
Compound
HNO3
MONO
NH3
NO2
Season
Winter
Summer
Winter
Summer
Winter
Summer
Winter
Summer
Location
Indoor
Outdoor
Indoor
Outdoor
Indoor
Outdoor
Indoor
Outdoor
Indoor
Outdoor
Indoor
Outdoor
Indoor
Outdoor
Indoor
Outdoor
. , N
29
24
31
29
29
' 24
• .31
29
18
18
31 .
29
NA
NA
30
30
Mean
1
5
: 2
19
1
8
1
-
-
17
18
.029
,59
.84
.42.
.44
.65
.27
.85
.34
.13
.10
.87
.86
.71
S.D.
.057
.47
.58
.83
3.35
.57
1.16
.48
6.40
.92
6.47
2.05
-
_
7.68
8.11
Range
-------
12
6-
35
I
B 10
o
vv HONO 35 ;
S 30
f 25 -
o
C
i 15 -
ju *"
_ * o
""»- Z. -
1 i . O -
t^ ' LJL o "
i i i i
Outdoor Indoor Outdoor Indoor
N-29 N-31 N-24 N-29
SUMMER WINTER
I 90%
• mean
-r- 25%
I 10%
(C) NH3 3 -
3.2.5 -
&
• B 2-
IE* "i c
8
1 1 -
* OJ
OK
Z
BJa ^~~^ r-Jn X Q
Outdoor Indoor Outdoor Indoor
N-29 N-S1 N-18 N-18
SUMMER WINTER
' 1^^
m
" i
Outdoor Indoor
N-30 N-30
SUMMER
(D) HN03
•• ••' "
B
— f— •
Outdoor Indoor Outdoor Indoor
N-29 N-31 N-24 N-29
SUMMER WINTER
Figure 7-17. Concentration distributions (in ppb) for gas phase species in Boston
(A) HONO; (B) NO2; (C) NH3; (D) HNO3. (N = number of valid
observations)
1 7.5 SUMMARY
2 7.5.1 Ambient Nitrogen Dioxide Levels
3 In urban areas, hourly patterns at fixed-site, ambient air monitors often follow a
4 biomodal pattern of morning and evening peaks, related to motor vehicular traffic patterns.
August 1991
7-56
DRAFT-DO NOT QUOTE OR CITE
-------
1 Sites affected by large stationary sources of NO2 (or NO that rapidly converts to NO2) are
2 often characterized by short episodes at relatively high concentrations.
3 The highest hourly and annual NO2 levels are reported from air monitoring stations in
4 California. The seasonal patterns at California stations are usually quite marked and reach
5 their highest levels through the fall and winter months, while stations elsewhere in the U.S.
6 usually have less prominent seasonal patterns and may peak in the winter, in the summer, or
7 contain little discernable variation.
8 Since 1980, the average level among reporting NO2 stations has been below 0.03 ppm,
9 with no significant trend evident. For 103 Metropolitan Statistical Areas reporting a valid
0 year's data for at least one station in 1988 and/or 1989, peak annual averages ranged from
1 0.007 to 0.061 ppm. The only recently measured exceedances of the annual standard,
2 0.053 ppm, have occurred at stations in southern California.
3 One-hour NO2 values can exceed 0.2 ppm, but, in 1988, only 16 stations (12 of which
4 are in California) reported an apparently credible second high 1-h value greater than 0.2 ppm.
5 Since at least 98% of 1-h values at most stations are below 0.1 ppm, these values above
6 0.2 ppm are quite rare excursions whose validity should be verified.
7
8 7.5.2 Indoor Nitrogen Dioxide Levels
9 Indoor concentrations of NO2 are a function of outdoor concentrations, indoor sources
0 (source type, condition of source, source use, etc.), infiltration/ventilation, air mixing within
1 and between rooms, reactive decay by interior surfaces and air cleaning or source venting. In
I homes without indoor sources of NO2, concentrations are lower than outdoor levels due to
3 removal by the building envelope and interior surfaceSj thus providing some degree of
\ protection from,outdoor concentrations. Indoor/outdoor ratios for homes without sources
5 vary considerably by season of the year, with the lowest ratios occurring in the winter and
3 highest occurring during the summer. The differences are probably due to seasonal
1 differences in infiltration rates, NO2 reactivity rates, the penetration factor and outdoor
\ concentrations. The indoor/outdoor ratio for these homes does not appear to vary by
) geographic area, housing type or outdoor concentration.
) Gas appliances (gas range/oven, water heater, etc.) are the major indoor source category
for indoor residential NO2 by virtue of the number of homes with such sources
August 1991 7-57 DRAFT-DO NOT QUOTE OR CITE
-------
1 (approximately 45% of all homes in the U.S.). NO2 levels in homes with gas appliances are
2 higher than those without such appliances. Within this catagory, the gas range/oven is by far
3 the major contributor, especially when used as a supplemntal heat source. Average indoor
* «2
4 concentrations (a one to two week measurement period) in excess of 100 /xg/m have been
5 measured in some homes with gas ranges. Homes with gas ranges with pilot lights have
6 higher NO2 levels than homes that have gas ranges without pilot lights. Average NO2
7 concentrations in homes with gas ranges/ovens exhibit a spatial gradient within and between
8 rooms. Kitchen levels are higher than other rooms and a steep vertical concentration gradient
9 in the kitchen has been observed in some homes, with concentrations being highest nearest
10 the ceiling. Average NO2 concentrations are highest during the winter months and lowest
11 during the summer months. This seasonal temporal gradient is probably related to seasonal
12 differences in infiltration, source use, NO2 reactivity rates indoors and outdoor
13 concentrations. The impact of gas appliance use on indoor NO2 levels may be superimposed
14 upon the background level resulting from outdoor concentrations. The results of field studies
15 of the impact of gas ranges on indoor NO2 are fairly consistent. Once corrected for the
16 contribution of outdoor concentrations, the average contribution of gas ranges to NO2 levels
17 indoors is similar by locations within homes and by season across the studies. Only very
18 limited data exists on short-term average (three hours or less) indoor concentrations of NO2
19 associated with gas appliance use. The limited data suggest that short-term averages of NO2
20 are several times the longer-term average concentrations measured.
21 Unvented kerosene and gas space heaters are important sources of NO2 in homes both
22 because of the NO2 production rate of the heaters and the length of time they are used. Field
23 studies indicate that average residential concentrations (1-week or 2-week average levels)
24 exhibit a wide distribution varying primarily with the amount of heater use and type of
25 heater. Under similar operating conditions unvented gas space heaters appear to be associated
26 with higher indoor NO2 levels than kerosene heaters. Average concentrations in homes using
27 unvented kerosene heaters have been measured well in excess of 100 jug/m3. In one study
28 calculations of NO2 residential levels during heater use (in homes without gas appliances)
29 indicate that approximately 50% of the homes* have concentrations above 100 /*g/m3 and 8%
<*•> S}
30 above 480 #g/m . Peak NO2 levels of 847 /ig/nr over a 1-h period in a home during use of
31 a kerosene heater have been measured. A large field study of indoor NO2 concentrations in
August 1991 7-58 DRAFT-DO NOT QUOTE OR CITE
-------
1 homes using unvented gas space heaters (most also had gas ranges) found that approximately
*3
2 70% of the homes had average concentrations in excess of 100 Mg/m and 20% in excess of
3 480 ftg/m3. This study found that the indoor/outdoor temperature difference was the best
4 indicator of indoor NO2 levels during the colder winter periods when heating demands are
5 greatest. No data have yet been published that provide concentrations during heater use nor
6 are any short-term peak indoor concentrations yet published for homes with unvented gas
7 space heaters. No spatial gradient of NO2 was found in homes with unvented kerosene space
8 heaters, contrary to the strong spatial gradient noted for homes with gas appliances. This is
9 probably due to the strong convective heat output and the long operating hours of the heaters,
D which result in rapid mixing .within the homes. Published data are not yet available on
1 spatial concentrations during the use of unvented gas space heaters.
I Improper use of gas appliances (e.g., using a gas range to heat a living space) and
3 improperly operating gas appliances or vented heating systems (e.g., out of repair gas range
\ or improper operation of a gas wall or floor furnace) can be important contributors to indoor
5 NO2 concentrations; but little data are available to assess the extent of that contribution.
5 Field data do not exist that would allow for an assessment of the contributions of wood or
? coal burning stoves or fireplaces to indoor NO2 concentrations, but such a contribution would
> be expected to be small. Cigarette smoking is expected to add little NO2 to homes.
> Efforts to model indoor NO2 levels have employed both physical/chemical and
> empirical/statistical models. Physical/chemical models have largely been applied to test
house data or to small samples of homes where detailed data on the factor impacting the
; concentrations have been measured. Empirical/statistical models have been developed from
large field study data bases. These employ questionnaire responses and measured physical
data (house volume, etc.) as independent variables and have met with moderate success.
The removal of NO2 indoors by surfaces (reactive decay) is often equal to or greater
than infiltration in removing NO2. NO2 reactive decay rates varies sharply as a function of
the type of material and surface area of the material. The degree of mixing or turbulence in
a space is also important. The role of relative humidity is not clear and some materials can
become NO2 saturated.
August 1991 7-59 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 AIRS, Aerometric Information Retrieval System [database]. (1991) [Data on NOJ. Research Triangle Park, NC:
3 U. S. Environmental Protection Agency, Office of Air Quality Planning and Standards. Disc^ IBM 3090.
4
5 Atkins, D. H. F.; Law, D. V. (1987) Indoor-outdoor nitrogen dioxide concentration ratios for homes with gas
6 and electric cooking. In: Seifert, B.*, Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air
7 '87: proceedings of the 4th international conference on indoor air quality and climate, v, 1, volatile
8 organic compounds, combustion gases, particles and fibres, microbiological agents; August; Berlin,
9 Federal Republic of Germany. Berlin, Federal Republic of Germany: Institute for Water, Soil and Air
10 Hygiene; pp. 383-389.
11
12 Borrazzo, J. E.; Osbom, J. F.; Fortmann, R. C.; Keefer, R. L.; Davidson, C. I. (1987) Modeling and
13 monitoring of CO, NO and NO2 in a modern townhouse. Atmos. Environ. 21: 299-311.
14
15 Borrazzo, J. E.; Peters, C.; Peck, S.; Davidson, C. I. (1987) Determination of NO2 loss rates from concentration
16 measurements in an occupied urban residence. In: Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.;
17 Wegner, J., eds. Indoor air '87: proceedings of the 4th international conference on indoor air quality and
18 climate, v. 1, volatile organic compounds, combustion gases, particles and fibres, microbiological agents;
19 August; Berlin, Federal Republic of Germany. Berlin, Federal Republic of Germany: Institute for Water,
20 Soil and Air Hygiene; pp. 321-325.
21
22 Brauer, M.; Koutrakis, P.; Keeler, G. J.; Spengler, J. D, (1991) Indoor and outdoor concentrations of inorganic
23 acidic aerosols and gases. J. Air Waste Manage. Assoc. 41: 171-181.
24
25 Clausing, P.; Mak, J. K.; Spengler, J. D.; Letz, R; (1984) Personal NO2 exposures of high school students. In:
26 Berglund, B.; Lindvall, T.; Sundell, J.", eds. Indoor air: proceedings of the 3rd international conference
27 on indoor air quality and climate, v. 4, chemical characterization and personal exposure; August;
28 Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building Research; pp. 135-139.
29 Available from: NTIS, Springfield, VA; PB85-104214.
30
31 Federal Register. (1979) Ambient air quality, surveillance. F. R. (May 10) 44: 27558.
32
33 Fortmann, R. C.; Nagda, N. L.; Harper, J. P. (1987) Radon mitigation through residential pressurization control
34 strategy. In: Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air '87:
35 proceedings of the 4th international conference on indoor air quality and climate, v. 2, environmental
36 tobacco smoke, multicomponent studies, radon, sick buildings, odours and irritants, hyperreactivities and
37 allergies; August; Berlin, Federal Republic of Germany. Berlin, Federal Republic of Germany: Institute
38 for Water, Soil and Air Hygiene; pp. 300-304.
39
40 Goldstein, B. D.; Melia, R. J. W.; Chinn, S.; Florey, C. du V.; Clark, D.; John, H. H. (1979) The relation
41 between respiratory illness in primary schoolchildren and the use of gas for cooking: II - factors affecting
42 nitrogen dioxide levels in the home. Int. J. Epidemiol. 8: 339-345.
43
44 Goldstein, I, P.; Hartel, D.; Andrews, L. R. (1985) Monitoring personal exposure to nitrogen dioxide. Presented
45 at: 78th annual meeting of the Air Pollution Control Association; June; Detroit, MI. Pittsburgh, PA: Air
46 Pollution Control Association; paper no. 85-85.7.
47
48 Good, B. W.; Vilcins, G.; Harvey, W. R.; Clabo, D. A., Jr.; Lewis, A. L. (1982) Effect of cigarette smoking
49 on residential NO2 levels. Environ. Int. 8: 167-175.
50
51
August 1991 7-60 DRAFT-DO NOT QUOTE OR CITE
-------
1 Harlos, D. P.; Spongier, J. D.; Billick, I. (1987) Continuous nitrogen dioxide monitoring during cooking and
2 commuting: personal and stationary exposures. In: Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.;
3 Wegner, J., eds. Indoor air '87: proceedings of the 4th international conference on indoor air quality and
4 climate, v. 1, volatile organic compounds, combustion gases, particles and fibres, microbiological agents;
5 August; Berlin, West Germany. Berlin, West Germany: Institute for Water, Soil and Air Hygiene;
6 pp. 278-282,
7
8 Hartwell, T., et al. (1988) An investigation of infiltration and indoor air quality: determination of the impact of
9 indoor combustion sources on indoor air quality - final report. Submitted to the New York State Energy
10 Research and Development Authority.
11
12 Hemphill, C. P.; Ryan, P. B.; Billick, I. H.; Nagda, N. L.; Koontz, M. D.; Fortmann, R. C. (1987) Estimation
13 of nitrogen dioxide concentrations in homes equipped with unvented gas space heaters. In: Seifert, B.;
14 Esdom, H.; Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air '87: proceedings of the 4th
15 international conference on indoor air quality and climate, v. 1, volatile organic compounds, combustion
16 gases, particles and fibres, microbiological agents; August; Berlin, Federal Republic of Germany. Berlin,
17 Federal Republic of Germany: Institute for Water, Soil, and Air Hygiene; pp. 420-424.
18
19 Hosein, R. et al. (1979)
10
11 Keller, M. D.; Lanese, R. R.; Mitchell, R. L; Cote, R. W. (1979) Respiratory illness in households using gas
J2 and electricity for cooking: I. survey of incidence. Environ. Res, 19: 495-503,
13
14 Koontz, M. D.; Fortmann, R. C.; Nagda, N. L.; Billick, I. H. (1986) Protocol for an indoor air quality
15 monitoring survey conducted in Texas. Presented at: 79th annual meeting of the Air Pollution Control
'6 Association; June; Minneapolis, MN. Pittsburgh, PA: Air Pollution Control Association; paper no.
17 86-6.3.
58
'9 Lambert, W. E. (1991) [Letter to Dr. Dennis Kotchmar concerning nitrogen dioxide measurements for a sample
10 of 100 homes participating in the UNM study of infant respiratory illness]. Albuquerque, NM: The
> 1 University of New Mexico, New Mexico Tumor Registry Medical Center; July 10.
'.2 .
i3 Leaderer, B. P. (1982) Air pollutant emissions from kerosene space heaters. Science (Washington, DC)
4 218: 1113-1115.
,5
•6 Leaderer, B. P.; Stolwijk, J. A. J.; Zagraniski, R. T.; Qing-Shan, M. (1984) A field study of indoor air
7 contaminant levels associated with unvented combustion sources. Presented at: 77th annual meeting of the
8 Air Pollution Control Association; June; San Francisco, CA. Pittsburgh, PA: Air Pollution Control
9 Association; paper no. 84-33.3.
0
1 Leaderer, B. P.; Zagraniski, R. T.; Berwick, M.; Stolwijk, J. A. J.; Qing-Shan, M. (1984) Residential
2 exposures to NO2, SO2 and HCHO associated with unvented kerosene space heaters, gas appliances and
3_ sidestream tobacco smoke. In: Berglund, B,; Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of the
4 3rd international conference on indoor air quality and climate, v. 4, chemical characterization and
5 personal exposure; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building
6 Research; pp. 151-156. Available from: NTIS, Springfield, VA; PB85-104214.
7
8 Leaderer, B. P.; Zagraniski, R, T.; Berwick, M,; Stolwijk, J. A. J. (1986) Assessment of exposure to indoor air
9 contaminants from combustion sources: methodology and application. Am, J. EpidemioL 124: 275-289,
0
1 Leaderer, B. P.; Zagraniski, R. T.; Berwick, M.; Stolwijk, J. A. J. (1987) Predicting NO2 levels in residences
2 based upon sources and source use: a multivariate model. Atmos. Environ. 21: 361-368.
3
August 1991 7-61 DRAFT-DO NOT QUOTE OR CITE
-------
1 Lcbret, E.; Noy, D.; Boley, J.; Brunekreef, B. (1987) Real-time concentration measurements of CO and NO2 in
2 twelve homes. In: Seifert, B.; Esdorn, H,; Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air '87:
3 proceedings of the 4th international conference on indoor air quality and climate, v. 1, volatile organic
4 compounds, combustion gases, particles and fibres, microbiological agents; August; Berlin, Federal
5 Republic of Germany. Berlin, Federal Republic of Germany: Institute for Water, Soil, and Air Hygiene;
6 pp. 435-439.
7
8 Macriss, R. A.; Elkins, R. H. (1977) Control of the level of NOX in the indoor environment. In: Kasuga, S.;
9 Suzuki, N.; Yamada, T.; Kimura, G.; Inagaki, K.; Onoe, K., eds. Proceedings of the fourth international
10 clean air congress; May; Tokyo, Japan. Japanese Union of Air Pollution Prevention Associations;
11 pp. 510-514.
12
13 Matbury, M. C.; Harlos, D. P.; Samet, J. M.; Spengler, J. D. (1988) Indoor residential NO2 concentrations in .
14 Albuquerque, New Mexico. JAPCA 38: 392-398.
15 '
16 McCarthy, S. M.; Tarmac, R. F.; Yocom, J. E. (1987) Indoor nitrogen dioxide exposure: the contribution from
17 unvented gas space heaters. In: Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds.
18 Indoor air "87: proceedings of the 4th international conference on indoor air quality and climate, v. 1,
19 volatile organic compounds, combustion gases, particles and fibres, microbiological agents; August;
20 Berlin, Federal Republic of Germany. Berlin, Federal Republic of Germany: Institute for Water, Soil,
21 and Air Hygiene; pp. 478-482.
22
23 Melia, R. J. W.; Florey, C. du V.; Morris, R. W.; Goldstein, B. D,; Clark, D.; John, H. H. (1982) Childhood
24 respiratory illness and the home environment. I. Relations between nitrogen dioxide, temperature and
25 relative humidity. Int. J. Epidemiol. 11: 155-163.
26
27 Miyazaki, T. (1984) Adsorption characteristics of NOX by several kinds of ulterior materials. In: Berglund, B.;
28 Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of the 3rd international conference on indoor air
29 quality and climate, v. 4, chemical characterization and personal exposure; August; Stockholm, Sweden.
30 Stockholm, Sweden: Swedish Council for Building Research; pp. 103-110. Available from: NTIS,
31 Springfield, VA; PB85-104214.
32
33 Moschandreas, D. J.; Relwani, S. M.; Macriss, R. A.; Cole, J. T. (1984) Differences and similarities of two
34 techniques used to measure emission rates from unvented gas appliances. In: Berglund, B.; Lindvall, T,;
35 Sundell, J., eds. Indoor air: proceedings of the 3rd international conference on indoor air quality and
36 climate, v. 4, chemical characterization and personal exposure; August; Stockholm, Sweden. Stockholm,
37 Sweden: Swedish Council for Building Research; pp. 375-379. Available from: NTIS, Springfield, VA;
38 PB85-104214.
39
40 Moschandreas, D. J.; Relwani, S. M.; O'Neill, H. J.; Cole, J. T.; Elkins, R. H.; Macriss, R. A. (1985)
41 Characterization of emission rates from indoor combustion sources. Chicago, IL: Gas Research Institute;
42 report no. GRI 85/0075. Available from: NTIS; Springfield, VA; PB86-103900.
43
44 National Research Council. (1981) Indoor pollutants. Washington, DC: National Academy Press.
45
46 Noy, D.; Lebret, E.; Boleij, J.; Brunekreef, B. (1984) Integrated NO2 exposure estimates. In: Berglund, B.;
47 Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of the 3rd international conference on indoor air
48 quality and climate, v. 4, chemical characterization and personal exposure; August; Stockholm, Sweden.
49 Stockholm, Sweden: Swedish Council for Building Research; pp. 37-42. Available from: NTIS,
50 Springfield, VA; PB85-104214.
51
52 Noy, D.; Lebret, E.; Willers, H.; Winlces, A.; Boleij, J. S. M.; Brunekreef, B. (1986) Estimating human
53 exposure to nitrogen dioxide: results from a personal monitoring study among housewives. Environ. Int.
54 12:407-411.
August 1991 7-62 DRAFT-DO NOT QUOTE OR CITE
-------
1 Oezkaynak, H.; Ryan, P. B.; Allen, G. A.; Turner, W. A. (1982) Indoor air quality modeling: compartmental
2 approach with reactive chemistry. Environ. Int. 8; 461-471.
3 . . ' . "
4 Palmes, E, D.; Gunnisoa, A. F.; DiMattio, J.; Tomezyk, C. (1976) Personal sampler for nitrogen dioxide. Am.
5 Ind. Hyg. Assoc. J. 37: 570-577.
6
7 Palmes, E. D.; Tomczyk, C.; DiMattio, J. (1977) Average NO2 concentrations in dwellings with gas or electric
8 stoves. Atmos. Environ. 11: 869-872.
9
10 Petreas, M.; Liu, K.-S,; Chang, B,-H.; Hayward, S. B.; Sexton, K. (1988) A survey of nitrogen dioxide levels
1 1 measured inside mobile homes. JAPCA 38: 647-651.
[2 - . . . - .-.,-,c--...: -
13 Quackenboss, J. J.; Spengler, J, D.; Kanarek, M. S.; Letz, R.; Duffy, C. P. (1986) Personal exposure to
.4 nitrogen dioxide: relationship to indoor/outdoor air quality and activity patterns. Environ. Sci. Technol.
.5 20: 775-783.
.6
.7 Quackenboss, J. J.; Lebowitz, M. D.; Crutchfield, C. D.; Burtchin, D. (1987) Indoor-outdoor relationships for
.8 paniculate matter and verification of exposure classifications. In: Seifert, B.; Esdorn, H.; Fischer, M.;
.9 Rueden, H.; Wegner, J., eds. Indoor air "87: proceedings of the 4th international conference on indoor
10 air quality and climate, v. 1, volatile organic compounds, combustion gases, particles and fibres,
'1 microbiological agents; August; Berlin, Federal Republic of Germany. Berlin, Federal Republic of
'.2 Germany: Institute for Water, Soil, and Air Hygiene; pp. 534-538.
3
'A Quackenboss, J. J.; Lebowitz, M. D.; Hayes, C. (1987) Epidemiological study of respiratory responses to
,5 indoor-outdoor air quality. In: Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds.
6 Indoor air '87: proceedings of the 4th international conference on indoor air quality and climate, v. 2,
7 environmental tobacco smoke, multicomponent studies, radon, sick buildings, odours, and irritants,
8 hyperreactivities and allergies; August; Berlin, Federal Republic of Germany. Berlin, Federal Republic of
9 Germany: Institute for Water, Soil and Air Hygiene; pp. 198-202.
0
1 Quackenboss, J. J,; Bronnimann, D.; Camilli, A. E.; Lebowitz, M. D. (1988) Bronchial responsiveness in
2 children and adults in association with formaldehyde, particulate matter, and environmental tobacco
3 smoke exposures. Am. Rev. Respir. Dis. 137(suppl.): 253.
4
5 Ryan, P. B.; Spengler, J. D.; Letz, R. (1983) The effects of kerosene heaters on indoor pollutant concentrations:
6 a monitoring and modeling study. Atmos. Environ. 17: 1339-1345.
7
8 Ryan, P. B.; Soczek, M. L.; Treitman, R. D.; Spengler, J. D.; Billick, I. H. (1988) The Boston residential NO2
9 characterization study — II. survey methodology and population concentration estimates. Atmos. Environ.
0 22: 2115-2125.
1
2 Southern California Gas Company. (1986) Residential indoor air quality characterization study of nitrogen
3 dioxide. Phase I. Volumes 1, 2 and 3. Southern California Gas Company; October.
4
5 Southern California Gas Company. (1987) Residential indoor air quality characterization study of nitrogen
6 dioxide. Phase II. Volumes 1, 2, and 3. Southern California Gas Company; December.
7 . .
8 Speizer, F. E.; Ferris, B., Jr.; Bishop, Y. M. M.; Spengler, J. (1980) Respiratory disease rates and pulmonary
3 function in children associated with NO2 exposure. Am. Rev. Respir. Dis. 121: 3-10.
3
1 Spengler, J. D.; Ferris, B. G., Jr.; Dockery, D. W.; Speizer, F. E. (1979) Sulfur dioxide and nitrogen dioxide
2 levels inside and outside homes and the implications on health effects research. Environ. Sci. Technol.
3 13: 1276-1280.
August 1991 7-63 DRAFT-DO NOT QUOTE OR CITE
-------
1 Spengler, J. D.; Duffy, C. P.; Letz, R.; Tibbitts, T. W.; Ferris, B. G., Jr. (1983) Nitrogen dioxide inside and
2 outside 137 homes and implications for ambient air quality standards and health effects research. Environ.
3 Sci. Technol. 17: 164-168.
4
5 Spicer, C. W.; Coutant, R. W.; Ward, G. F. (1986) Investigation of nitrogen dioxide removal from inddor air
6 [final report (December 1984 - September 1986)]. Chicago, IL: Gas Research Institute; report no.
7 GRI-86/0303. Available from: NTIS, Springfield, VA; PB87-185971.
8
9 Spicer, C. W.; Coutant, R. W.; Ward, G. F.; Joseph, D. W,; Gaynor, A. J.; Billick, I. H. (1987) Rates and
10 mechanisms of NO2 removal from indoor air by residential materials. In: Seifert, B.; Esdorn, H.;
11 Fischer, M,; Rueden, H.; Wegner, J., eds. Indoor air '87: proceedings of the 4th international
12 conference on indoor air quality and climate, v. 1, volatile organic compounds, combustion gases,
13 particles and fibres, microbiological agents; August; Berlin, Federal Republic of Germany. Berlin,
14 Federal Republic of Germany: Institute for Water, Soil and Air Hygiene; pp. 371-375.
15
16 Szalai, A., ed. (1972) The use of time: daily activities of urban and suburban populations in 12 countries. The
17 Hague, The Netherlands: Mouton and Co.
18
19 Traynor, G. W.; Anthon, D. W.; Hollowell, C. D. (1982) Technique for determining pollutant emissions from a
20 gas-fired range. Atmos. Environ. 16: 2979-2987.
21
22 Traynor, G. W.; Apte, M. G.; Carruthers, A. R.; DOIworth, J. F.; Grimsrud, D. T.; Thompson, W. T. (1984)
23 Indoor air pollution and interroom transport due to unvented space heaters. Berkeley, CA: Lawrence
24 Berkeley Laboratory; report no. LBL-17600. Available from: NTIS, Springfield, VA; DE84015949.
25
26 Traynor, G. W.; Girman, J. R.; Apte, M. G.; Dillworth, J. F.; White, P. D. (1985) Indoor air pollution due to
27 emissions from unvented gas-fired space heaters. J. Air Pollut. Control Assoc. 35: 231-237.
28
29 Traynor, G. W.; Aceti, J. C.; Apte, M. G.; Smith, B. V.; Green, L. L.; Smith-Reiser, A.; Novak, K. M.;
30 Moses, D. O. (1987) Macromodel for assessing indoor exposures to combustion-generated pollutants. In:
31 Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air '87: proceedings of the
32 4th international conference on indoor air quality and climate, v. 1, volatile organic compounds,
33 combustion gases, particles and fibres, microbiological agents; August; Berlin, West Germany. Berlin,
34 West Germany: Institute for Water, Soil and Air Hygiene; pp. 273-277.
35
36 U. S, Bureau of the Census. (1982) 1980 census of population and housing: supplementary report: provisional
37 estimates of social, economic, and housing characteristics: states and selected standard metropolitan
38 statistical areas. Washington, DC: U. S. Department of Commerce; Bureau of the Census report no. PHC
39 80-S1-1. Available from: U. S. Department of Commerce, Bureau of the Census, Washington, DC.
40
41 U. S. Environmental Protection Agency. (1990) National air quality and emissions trends report, 1988.
42 Research Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no.
43 EPA/450/4-90/002. Available from: NTIS, Springfield, VA; PB90-200114/XAB.
44
45 U. S. Environmental Protection Agency. (1991) National air quality and emissions trends report, 1989. Research
46 Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA/450/4-91/003.
47 Available from: NTIS, Springfield, VA; PB91-172247/XAB.
48
49 Wade, W. A., HI; Cote, W. A.; Yocom, J. E. (1975) A study of indoor air quality. J. Air Pollut. Control
50 Assoc. 25: 933-939.
51
52 Yamanaka, S. (1984) Decay rates of nitrogen oxides hi a typical Japanese living room. Environ. Sci. Teehnol.
53 18: 566-570.
54
August 1991 7-64 DRAFT-DO NOT QUOTE OR CITE
-------
i 8. ASSESSING TOTAL HUMAN EXPOSURE
2 TO NITROGEN DIOXIDE
3
4
5 8.1 INTRODUCTION
6 In the course of their daily activities, humans are exposed to nitrogen dioxide (NO2) in
7 a number of settings or environments (residential, industrial, non-industrial occupational,
8 transportation, outdoors, etc.). Human exposure to NO2 consists of contact at the air
9 boundary layer between the human and the environment at a specific concentration for a
0 specified period of time. The integrated NO2 exposure is the sum of the individual NO2
1 exposures over all possible time intervals for all settings or environments. Thus, the units of
2 exposure are concentration multiplied by time.
3 The assessment of human exposures to NO2 can be represented by the following
4 simplified basic model proposed by various researchers (e.g., Duan, 1981; Sexton and Ryan,
5 1988):
5 E = Ei Ej = Ej f& (8-1)
7
3 where: E is the total or integrated exposure to NO2 for an individual expressed as an average
) concentration over a specified period of time across all microenvironments; Ej is the average
) NO2 exposure in the 1th microenvironment; fj is the fraction of time spent in the ith
I microenvironment; and, Cj is the NO2 concentration in the ith microenvironment. This
I model essentially represents exposures as a linear combination of an individual's average
5 concentration in each microenvironment, weighted by the time in a microenvironment type.
I- A microenvironment is defined as a three-dimensional space, having a volume in which the
> . pollutant concentration is considered or assumed to be spatially uniform. In theory there can
') be a large number of microenvironments; in practice, however, microenvironments are
' aggregated into a reduced number for air monitoring and modelling applications (outdoors,
! public transportation, indoor residential, non-industrial occupational, industrial, public access
1 buildings, etc.). The NO2 concentration in each microenvironment can show substantial
i spatial and temporal variability and is a complex function of the sources, source use, and
dispersal and removal mechanisms. A more detailed discussion of the general theoretical
August 1991 8-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 concepts in assessing exposure to air contaminants can be found in a recent report prepared
2 by the National Research Council (1991).
3 Until recently, efforts to assess adverse health and comfort effects associated with NO2
4 and subsequent efforts to mitigate or reduce exposures have focused on measurements of
5 ambient air quality obtained from fixed location monitoring sites. Numerous studies on
6 indoor concentrations of NO2 (see Chapter 7) combined with studies on human activity
7 patterns (e.g., Szalai, 1972; Chapin, 1974; Robinson, 1977; Ott, 1988; Schwab et al., 1990)
8 have clearly underscored the limited role of ambient air NO2 measurements in assessing
9 exposures. While outdoor NO2 levels clearly impact indoor concentrations, and in some
10 cases may dominate indoor levels, indoor concentrations are typically not well represented by
11 outdoor measurements when there are indoor sources (Chapter 7). In addition, direct
12 individual exposure to outdoor concentrations accounts on average for a small portion of a
13 person's total exposure because of the small amount of time typically spent outdoors. It is
14 important to note, however, that for a significant portion of the population, outdoor
15 exposures can account for a high fraction of their total exposure (e.g., police). Total
16 personal exposure to NO2 and a consideration of the microenvironments in which those
17 exposures take place is essential in assessing associated health effects and in developing
18 effective mitigation measures.
19 Central to the design of a human exposure assessment effort is the identification
20 of the health or comfort effect under study, the ascertainment of the individual contaminant
21 thought to be associated with that effect and specification of the contaminant exposure on a
22 time scale corresponding to the effect. It is in fact the biological response time that is central
23 to the development of the exposure assessment methodology. The principal biological
24 response of human exposures to NO2 known at this time is respiratory disease. Thus,
25 assessing long-term exposures to NO2 on the order of days or weeks, is of primary
26 importance. It is important to note, however, that short-term high episodes can elevate total
27 exposures (concentration x time) and may be of importance.
28 In developing and conducting an exposure assessment protocol, the application of the
29 acquired data must be considered. The three principal applications for NO2 exposure
30 assessment efforts are: environmental epidemiology; risk assessment; and risk management.
31 In environmental epidemiology, misclassifications and the influence of confounders can be
August 1991 g-2 DRAFT-DO NOT QUOTE OR CITE
-------
1 minimized when NO2 exposures of the study population are adequately assessed. In addition,
2 the application of an appropriate NO2 exposure assessment methodology will reduce the error
3 or uncertainty in calculating risks associated with NO exposure, aid in formulating cost
4 effective mitigation efforts to minimize risk and permit the monitoring of progress in
5 reducing the risk.
6 Adequate exposure assessment for NO2 is particularly critical in conducting and
7 evaluating epidemiological studies. Failure to measure or estimate exposures adequately and
8 address the uncertainty in the measured exposures can lead to misclassification errors (Shy
9 et al., 1978; Gladen et al., 1979; Ozkaynak et al., 1986; Wfflet, 1989; Dosemeci et al.,
0 1990; Lebret, 1990). Early studies comparing incidence of respiratory illness in children
1 living in homes with and without gas stoves used as a simple categorical variable of NO2
2 exposure the presence of a gas cooker. Such a dichotomous grouping can result in a severe
3 non-differential misclassificatioft error in assigning exposure categories. This type of error is
4 likely to underestimate the true relationship and could possibly result in a null finding.
5 Where catagories of exposure are assigned based upon measured or estimated concentrations,
6 the intermediate categories of exposure may be biased either away or towards the null. In
7 support of epidemiologic studies of NO2 it is important to employ an exposure assessment
8 protocol that will minimize the misclassification and characterize the uncertainty associated
9 with assessed exposures (Lambert et al., 1990).
0 Exposures to NO2 can be assessed by either direct or indirect methods. Direct methods
1 include biomarkers and personal monitoring (breathing zone measurements). Indirect
2 methods refer to the coupling of measurements of NO2 concentrations in various settings or
3 environments with models to estimate exposure. In this chapter the available data on human
4 exposure to NO2 will be reviewed within the context of the direct and indirect methods of
5 exposure assessment.
6
7
8 8.2 DIRECT METHODS
9 8.2.1 Biomarkers
0 Biological markers are cellular, biochemical or molecular measures that are obtained in
1 biological media such as human tissues, cells or fluids, and are indicative of exposure to
August 1991 8-3 DRAFT-DO NOT QUOTE OR CITE
-------
1 environmental chemicals (National Research Council, 1989). Biomarkers of exposure can
2 theoretically integrate total intake to the body from multiple sources of exposure to
3 environmental contaminants. If they are stable over time they can be used to integrate
4 exposure over time. They can be useful tools in elucidating mechanisms of disease or in
5 extrapolations between internal doses, routes of exposure, and species or tissues but do not
6 necessarily provide the direct link between environmental exposure and disease. Biomarkers
7 may be measures of the contaminant or its metabolites that are directly related to the specific
8 contaminant associated with the effect outcome (e.g., lead) or may only be a surrogate for
9 exposure to a complex source of environmental contaminants (e.g., cotinine).
10 The relationship between the biological marker and the air contaminant concentration
11 and length of exposure is typically poorly understood. Biomarkers are indicators that an
12 exposure has taken place but not necessarily a measure of exposure. Biomarkers by
13 themselves do not provide information on the environment or setting in which the exposure
14 takes place and hence on the factors which control the exposure.
15 Hydroxyproline excretion, an indicator of increased collagen catabolism or connective
16 tissue injury, has been proposed as a biomarker for exposure to NO2. Yanagisawa et al.
17 (1986, 1988) found an association between the hydroxyproline to creatinine ratio and daily
18 average personal NO2 levels in a sample of 800 women in two communities near Tokyo.
19 However, in chamber studies of normal males exposed to 0.6 ppm NO2 for three consecutive
20 days for four hours per day Muelenaer et al. (1988) saw no significant changes in
21 hydroxyproline excretion. More recently, Maples et al. (1991) have studied the potential of
22 the NO/heme protein complex as a useful biologic marker for NO2 exposure. At this time
23 there is no validated biomarker for NO2 exposure.
24
25 8.2.2 Personal Monitoring
26 Personal air monitoring provides a direct measurement of the concentrations of air
27 contaminants in the immediate breathing zone of an individual as a measure of personal
28 exposure. It provides a measure of exposure across the microenvironments or settings in
29 which individuals spend their time. Personal monitoring employs samplers (worn by
30 subjects) that record the concentration individuals are exposed to in the course of their normal
31 activity for periods of seconds to several days. These monitors can be passive dosimeters
August 1991 8-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 (e.g., Palmes tubes or Yanagisawa badges) that provide an integrated measure of exposure or
2 a portable monitor with data logger, which can provide a nearly continuous measure of
3 exposure. Little personal NO2 exposure data has been reported using portable monitors. The
4 vast majority of personal NO2 exposure data has been gathered using passive dosimeters.
5 Measures of integrated personal exposure (passive dosimetry) by themselves are not adequate
6 to determine the major sources of exposure and the settings in which the exposure take place.
7 Personal integrated measures of exposure need to be supplemented by personal activity diaries
8 and measures of the factors affecting those exposures if effective mitigation measures are to
9 be developed and instituted. A direct measure of personal exposure (integrated or
10 continuous) in epidemiologic studies of NO2 health effects, however, is highly desirable to
11 minimize misclassification and to uncover exposure response relationships. Personal
12 exposures on whole study populations or selected subsamples are not easy to obtain and can
13 be expensive. '
14 The ability to obtain measures of personal exposure is in large part controlled by the
15 availability,of accurate, small, easy to use and inexpensive personal monitors. The
16 development of the Palmes tube (Palmes et al., 1976) and more recently the NO2 badge
17 (Yanagisawa and Nishimura, 1982) has provided the sensitivity and specificity necessary to
18 conduct personal NO2 air monitoring at reasonable cost. These monitors are passive samplers
[9 that utilize diffusion to concentrate gases on a collection medium. The samples are then
JO returned to a. laboratory for subsequent laboratory analysis. These passive samplers have
II been used extensively in the monitoring of various microenvironments (Chapter 7) and to a
12 lesser extent in personal exposure monitoring.
!3 There are relatively few studies reported where subjects wore passive NO2 monitors in
14 the course of their daily activities to assess personal NO2 exposure. In these studies personal
15 exposures were typically compared to NO2 measurements made in the various
16 microenvironments in which they spent their time. Frequently these studies obtained
17 supplemental information, through time activity diaries, on the time subjects spent in various
18 microenvironments in which NO2 measurements are made. When microenvironmental
19 measurements and time activity diaries are obtained, the personal exposure measurements are
iO compared to the microenvironmental data, and personal exposure models are developed using
>1 both the microenvironmental measurements and time activity diaries. Many of the personal
August 1991 8-5 DRAFT-DO NOT QUOTE OR CITE
-------
1 monitoring studies have in fact resulted in the development and testing of indirect methods of
2 assessing personal NO2 exposures.
3 An extensive study on personal NO2 exposures as a function of outdoor and indoor
4 concentrations, indoor sources and time activity patterns was reported by Quackenboss et al.
5 (1986). In this study one-week NO2 samples were obtained during both a winter and summer
6 period using Palmes tubes for 350 volunteers residing in 82 homes in the' rural Portage, WI,
7 area. The personal samples were supplemented with household measurements made outside
8 and in the Mtchen and bedroom of each house, time spent in various microenvironments
9 (inside of home, outside, inside a motor vehicle, inside at work or school and at other indoor
10 locations) and information on household characteristics (i.e., existence of gas or electric
11 stove). Average NO2 personal exposures were poorly related to outdoor NO2 concentrations
12 (Figure 8-1) and strongly related to home average (average of the bedroom and kitchen)
13 concentrations for homes with and without gas cooking ranges (Figure 8-2). Outdoor
14 concentrations were considerably lower than personal exposures for individuals in homes with
15 gas cooMng ranges and higher than personal exposures for individuals in homes with electric
16 stoves. A comparison of the indoor and outdoor NO2 concentrations for this study as a
17 function of season and use of a gas or electric cooking range is discussed in Chapter 7.
18 The time activity diary results reported by Quackenboss et al. (1986) showed that more
19 than 65% of the time was spent at home by the subjects while about 15% of their time was
20 spent outdoors in the summer and 5% in the winter. Approximately 4% of the time was
21 spent in motor vehicles, up to 15% at work or school and approximately 8% in other indoor
22 environments. The study estimated personal exposures from home, average indoor and
23 outdoor levels weighted by the proportion of time spent there for three categories of people
24 (student, worker and other) by season and stove type. The estimated exposures were
25 compared to the measured personal exposures. The explained variance ranged from 3% to
26 71%. The measured exposures were generally not well predicted by the estimated exposures.
27 In an earlier study (a pilot for the one discussed above), Quackenboss et al. (1982)
28 measured personal NO2 exposures (from five to seven days) of 66 family members in
29 19 homes in Portage, WI, and NO2 concentrations in a bedroom, kitchen and outdoors for
30 each residence. Time budgets were obtained for each subject and information on cooking
31 fuel was obtained for each house. Personal exposures were found to be strongly associated
August 1991 8-6 DRAFT-DO NOT QUOTE OR CITE
-------
NITROGEN DIOXIDE LEVELS IN SUMMER
IT"
"1
•^
i
X
UJ
_J
<
z
Ul
D.
Ul
2
£
<
DU
40
SO
20
10
9
a
7
e
s
4
a
2
I
F • * * * *
3^- -|3 O .*.
- ** * * V** *°°
— d; "J* O
- ° * © 0 ctn 0 °° ° ° 0
O v *
— J,
— * D
~ * 0 ' '
—
E- LEGEND:
'^ o Electric
E_ * Gas
~i i i i I u 1 1 Inn I i ! i 1 1 1 1 III i i 1 1 1 1 ii 1 1 i i i i Ii mill 11
OUTDOOR CONCENTRATION Oag/m»)
NITROGEN DIOXIDE LEVELS IN WINTER
t
* .
"~* 50
Ul
DC
CO
i so
Ul
j£
^Sf
He!*
g «
cc
Ul
a.
Ul S
|
< 2
E
- * * *
=• ' *# ^ *
= ^ *** * **|"*
= "j," 0* *
- * * *
- * * * 0 o *
| °0Cs0o°o
- ° ^s °e®
~* O o ^^
r o
- O ° °
E a ° LEGEND:
H
1 o Electric
= * Gas
-i i i i Inn In ill i 1 1 1 1 hill i i i i liuil i i i i Ii in In ill i 1 i 1 ih
5 10 20
OUTDOOR CONCENTRATION
SO
Figure 8-1. Average personal NO2 exposure for each household compared with outdoor
concentrations for summer and winter.
Source: Quackenboss et al. (1986).
August 1991
8-7 DRAFT-DO NOT QUOTE OR CITE
-------
NITROGEN DIOXIDE LEVELS IN SUMMER
in
cc
o
i
a.
ui
I
I
70
60
60
40
SO
20
10
9
&
7
e
o
o
o* * £
o * **,,
°* y
Q>°
o
*
o <•
o *
LEGEND:
o Elect
* Gas
5 10 20 50
HOME AVERAGE CONCENTRATION
NITROGEN DIOXIDE LEVELS IN WINTER
•&• 100
10
I
a.
UJ
S; 2
o
_O
*
* *
** *
* *
° ' <5> O 0%
A^b00^
LEGEND;
0 Electric
* Gas
2 5 10 20 50
HOME AVERAGE CONCENTRATION
100
Figure 8-2. Average personal NO2 exposure for each home compared with average
indoor concentrations for summer and winter.
Source: Quackenboss et al. (1986),
August 1991
8-8
DRAFT-DO NOT QUOTE OR CITE
-------
1 with bedroom (r = 0.84—gas, r = 0.63—electric) and kitchen (r = 0.71—gas,
2 r = 0.6—electric) concentrations for both the gas and electric cooking homes but poorly
3 associated with outdoor concentrations (r = 0.4). Several models, using microenvironment
4 monitoring, time budgets and household characteristics, were applied to the data to relate
5 average personal NO2 exposures to indoor concentrations. The models explained as much as
6 90% of the variation in personal exposures.
7 As part of an epidemiological study of the impact of kerosene heaters on health,
8 Leaderer et al. (1986) compared personal one-week NO2 measurements on 23 adult subjects
9 with one-week NO2 measurements made in three locations in the subjects homes (kitchen,
,0 living room and bedroom) and outdoors. The homes monitored had a mix of NO2 sources
.1 (gas stove, kerosene heater, gas stove and kerosene heater and no known source). Eighty
.2 percent of the variation in total personal NO2 exposures were accounted for by variations in
3 indoor average NO2 concentrations for homes with the mix of sources (Figure 8-3). Total
4 personal NO2 exposures were found to be 90% of the house average concentration. When
5 compared to concentrations measured in different locations in the house, the bedroom
6 concentration was the best predictor (R2 = 0.88). Total personal exposures to NO2 were
7 not found to be related to outdoor concentrations of NO2 measured at each residence.
8 In a recent study, Harlos et al. (1987) obtained one-day personal NO2 measurements for
9 15 infants in homes with gas ranges. Corresponding NO2 measurements were made in the
0 infants bedroom and in the living room and kitchen of each house. As might be expected the
1 measured personal exposure correlated well with the bedroom concentrations (r = 0.78). An
2 effort to model the infants' personal exposure based upon the infants' activity patterns and
3 room concentrations resulted in a slightly stronger relationship with measured personal
4 exposure (r = 0.82).
5 In a study of 500 junior high students in Watertown, MA (Clausing et al., 1986),
6 personal NO2 concentrations were measured over a three to four day period between
7 November and December of 1982. NO2 measurements were made in the bedroom, living
8 room, kitchen, and outside for 200 .of the homes of the students on a time scale
9 corresponding to the personal monitoring. Time activity diaries were kept by the subjects
0 and a questionnaire was utilized to obtain information on home characteristics, particularly
1 sources of NO2 and their use. A variety of models, utilizing indoor and outdoor measured
August 1991 8-9 DRAFT-DO NOT QUOTE OR CITE
-------
e
01
12
Ld
o
Q.
X
LU
02
UJ
CL
130
120-
110-
100-
90-
80-
70 -
60-
50-
40-
30-
20 -
10-
0
* KEROSENE HEATERS & GAS STOVES
• NO SOURCE
A GAS STOVES
v KEROSENE HEATERS
T T
20 40 60
80 100 120 140
AVERAGE NCk/HOUSE, ug/m3
1
2
3
4
5
6
7
8
9
10
11
1
2
3
4
5
6
7
Figure 8-3. Comparison of the house average two-week NO2 concentrations with the
total personal NC>2 levels measured over the same time period for one adult
resident in each house, New Haven, CT, area, whiter 1983. Each house is
identified by the type of sources in the house and the fitted regression line
is presented.
Source: Leaderer et al. (1986).
NO2 concentrations, time activity information and home characteristics, were explored to
explain variations in measured personal NO2 exposure. The correlation between personal
exposure and outdoor concentrations was not significant at the 0.05 level. Models that
included indoor NO2 concentrations explained from 60% to 90% of the variation in personal
NO2 exposures. Excluding indoor concentrations and using cooking range characteristics (gas
range, presence of pilot lights, etc.) explained only 40% of the variation in personal NO2
exposures.
August 1991
8-10
DRAFT-DO NOT QUOTE OR CITE
-------
1 As part of a large study on indoor air pollution in the Netherlands (a sample of over
2 900 homes), a small study was conducted on a subsample of 14 families (11 mothers,
3 11 school children, and 8 pre-school children) to establish a relation between measured indoor
4 NO2 concentrations and personal exposure (Hoeket al., 1984). Weekly average personal
5 exposure of mothers, primary (mean age = 7.8 years) and pre-school (mean age =
6 3.4 years) children and weekly NO2 concentrations in several locations were measured.
7 A self-administered questionnaire was used to obtain time activity patterns. Measured and
8 calculated (from microenvironment measurements and time activity data) personal exposures
9 were highly correlated (R2 from 82% for the whole group to 92% for the preschool children)
10 but the calculated levels were on average 20% lower than the measured personal exposures.
11 The mothers' exposure was found to be a good predictor of the primary and preschool
12 children (R2 of 91 % and 93%, respectively) but overestimated the children's exposure by
13 approximately 20%.
14 One day personal exposures and corresponding microenvironment NO2 concentrations
15 (home, workplace and outdoors) were measured for 20 housewives and 44 office workers
16 (same office building) in Tokyo (Nitta and Maeda, .1982). Activity diaries were obtained
17 during the sampling period. Repeat sampling during a different season was conducted on a
18 subsample of the population. Outdoor concentrations were found to be very poor predictors
19 of personal exposures with correlation coefficients no higher than 0.33 for all comparisons by
20 group (housewife or worker) and season. Housewife personal exposures compared well with
21 both modelled personal exposures (from microenvironmental measurements and activity
22 diaries) for winter (r = 0.88 and 0.89) and moderately for the summer (r = 0.62 and 0.57).
23 Results for the office workers were similar.
24 Daily averages of personal NO2 exposures and indoor and outdoor NO2 concentrations
25 were measured for 40 housewives and their preschool children (subsample) living in and near
26 Tokyo for a winter, spring and summer period (Yanagisawa et al., 1984). An additional
27 9 housewives were monitored for 7 one-day periods each month for a year. Gas ranges were
28 used for cooking in almost all houses. Activity diaries and information on home
29 characteristics were obtained. Outdoor concentrations were not significantly related to
30 personal exposures. Indoor levels were above outdoor levels with the living room and
31 bedroom concentration averages close to the personal exposures of the housewives and
August 1991 8-11 DRAFT-DO NOT QUOTE OR CITE
-------
1 preschool children for the different locations and seasons. Kitchen concentrations were on
2 average higher than personal exposures. Measured and calculated personal exposures for the
3 9 housewives agreed well.
4 Twelve primary school children were monitored for a one-week period for NO2
5 exposures as part of a study of indoor air quality and respiratory symptoms (Hoek et al.,
6 1984). Time activity patterns and NO2 concentrations in the bedroom, living room and
7 kitchen of the homes of the subjects were also obtained. Outdoor NO2 concentrations were
8 obtained from a central monitoring site. The homes contained geisers, an important source of
9 NO2. Living room and bedroom concentrations best predicted personal exposures (R2 of
10 0.64 and 0.63) while the kitchen concentrations explained 35% of the variation of personal
11 exposures. Eighty-four percent of the variation in personal NO2 exposure was explained by
12 variations in the modelled exposure using time activity and microenvironment measurements.
13 Dockery et al. (1981) measured one-week personal NO2 exposures along with bedroom,
14 kitchen and outdoor concentrations for nine families in Topeka, Kansas, during a summer ';
15 period. The homes had a mix of electric and gas cooking ranges and activity diaries were
16 obtained. Mean personal exposures were similar to the bedroom means. Measured and
17 calculated personal exposures agreed well, with 86% of the variation explained.
18 As part of an epidemiologic study of an association between personal exposure to NO2
19 and respiratory illness in Hong Kong (Koo et al., 1990), 24-h personal measurements of NO2
20 using passive badges (Yanagisawa and Nishimura, 1982) were acquired on 362 children
21 age 13 and under and attending the same school, and 319 of their mothers. The passive
22 badges were also hung in the school classrooms and school playground. A questionnaire was
23 used to obtain data on indoor sources of air pollution (smoking habits of family members,
24 types of heating and cooking fuels, frequency of cooking, ventilation patterns, burning of
25 incense and mosquito coils, and mother's exposure to dust or fumes in the workplace. The
26 sampling was conducted during warm weather (ambient temperature of 27 °C). Ventilation
27 was supplied to the school through open windows. Variations in classroom NO2
28 concentrations explained 56% of the variation in personal exposure levels of the children.
29 Personal NO2 exposures of the mothers was not significantly correlated with the personal
30 exposures of their children (p >0.05). Neither the children's nor the mother's personal NO2
31 exposure related to number of cigarettes smoked in the home or the number of hours of
August 1991 8-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure to cigarette smoke. The children's exposure was not related to cooking and heating
2 habits while the mother's level was highest for LPG and kerosene users and lowest for piped .
3 gas. Increases in mother's personal NO2 levels were also seen when kitchens did not have
4 ventilating fans (11%), when incense was burned in the home (10%) and when the mother
5 reported dust exposure in the workplace (21%).
6 The largest field study of personel exposures reported to date was conducted.in the
7 Los Angles Basin (Schwab et al., .1990). In this study 700 individuals from 500 households
8 were monitored for two consecutive 24-h periods for their integrated NO2 exposure .using
9 Yanagisawa filter-badge monitors. In addition, NO2 concentrations were monitored in both
0 the bedroom and outside the home of each participant over the 48-h period they were
1 monitored for personal exposures. Data were collected on housing units (including indoor
2 sources), personal characteristics and time/activity patterns. The objective of the study was to
3 investigate the seasonal, spatial and demographic trends in personal NO2 exposure and the
4 influence of indoor and outdoor NO2 concentrations and activity patterns NO2 exposure.
5 Several analysis techniques (stepwise regression, analysis of variance and path analysis) were
6 applied to the collected data to determine the relative contribution to personal NO2 exposures
7 of factors hypothesized to influence those exposures. The factors considered .in the analysis
8 included indoor arid outdoor concentrations of NO2 activity patterns of the subjects, type of
9 cooking range (electric, gas without pilot lights, and gas with pilot lights), season (winter vs.
0 sununer), location (potential high, medium and low ambient NO2 levels) and six population
1 subgroups defined by age, sex, and employment status. In the analysis, the two consecutive
2 24-h. personal NO2 measurements were averaged, with the resultant variable used as the
3 dependent variable in the analysis.
4 The analysis indicated that the time spent in each of the microenyironments considered
O
5 was a poor predictor of personal exposure (Rz typically less than 0.10). The explained
6 variation in personal exposures increased to 55% when the bedroom concentration was
7 included in the model and decreased to 42% when only outdoor concentrations were
8 considered. Only small differences in the time/microenvironment relationships were observed
9 between groups defined by range type or population subgroups (explained variation changes
0 of 1 to 10% were observed). Analysis of variance showed that cooking range type, season
1 and geographic location explained 30% of the personal NO2 exposure and the addition of
August 1991 8-13 DRAFT-DO NOT QUOTE OR CITE
-------
1 outdoor NO2 levels increased the explained variation to 54%. Inclusion of bedroom
2 concentration increased the explained variation to 62%. The model with outdoor levels alone
f\ /•)
3 produced an Rz of 0.48 while using bedroom concentrations along resulted in Rz of 0.59.
4 Models developed for each of the population subgroups using bedroom and outdoor
5 NO2 concentrations and range type resulted in explained variations in personal NO2 levels
6 ranging from 60 to 78%. Approximately 15% of the variation in personal NO2 levels was
7 explained for non-working groups when compared to the population as a whole. The study
8 estimates that in the Los Angeles Basin 68% of the variation in personal NO2 exposures is
9 explained by measured outdoor and indoor concentrations and that 32% of the variation is
10 unexplained by the parameters measured. The authors noted that time/activity measures by
11 themselves are weak predictors of exposure, while surrogates of exposure such as cooking
12 range type and location are relativly good predictors. Outdoor concentrations in the Los
13 Angeles Basin are considerably higher than most other locations in the U.S. and as such it is
14 not obvious how the results of this study would be applicable to other regions.
15 One additional large field study of personal exposure to NO2 (conducted in the Boston
16 area) has recently been completed (Ryan et al., 1990). The results of this study, when
17 available, should contribute substantially to our understanding of the relationship between
18 personal exposure to NO2 and indoor and outdoor concentrations and time activity patterns.
19 The available measured data on personal NO2 exposures indicates that outdoor
20 measurements of NO2 (measured in the vicinity of residences) are generally not good
21 predictors of the personal exposure level, primarily because of the small amount of time
22 individuals spend outdoors. No studies have been reported which examine the relationship
23 between NO2 concentrations measured at central outdoor sites and personal NO2 exposures.
24 It is likely, however, given the outdoor spatial variability of NO2, that NO2 levels measured
25 at central sites would be poor predictors of personal NO2 exposures. Outdoor NO2
26 concentrations, however, dominate indoor levels when there are no indoor sources and when
27 outdoor concentrations are high (e.g., Los Angeles). NO2 concentrations measured in the
28 living room or bedroom of a house or a whole house average measurement is a better
29 predictor of personal exposure than outdoor measurements for the population as a whole. It
30 is important to note that there may be significant segments of the population for which indoor
31 NO2 levels may not be good predictors of their exposure (infants, police, etc.). Calculated
August 1991 8-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposures from microenvironmental measurements and activity diaries are good indicators of
2 personal exposure.
3
4
5 8.3 INDIRECT METHODS
6 Indirect methods employ various degrees of microenvironmental monitoring and
7 questionnaires to estimate and individual's or populations total NO2 exposure. They attempt
8 to measure and understand the underlying relationships between personal exposure and the
9 variables causing exposure so that NO2 exposures in other populations in other locations can
10 be estimated. Such models can provide exposure frequency distributions for the entire
11 population or segments of the population.
.2 Estimates of NO2 concentrations in various microenvironments (mieroenvironmental .
.3 monitoring) can provide information on the spatial and temporal distribution of concentrations
.4 in those environments and the factors (sources, emission rates, removal and dispersal
.5 mechanisms, etc.) affecting those concentrations. Estimates of microenvironmental NO2
.6 concentrations combined with time activity patterns can be used to estimate or model total
7 NO2 exposure (i.e., Equation 8-1). Questionnaires are used to gather information on
8 microenvironmental factors (sources, source use, volume, etc.), human activities or time
9 budgets or for the simple categorization of an individual's exposure status.
:0 Physical/chemical, empirical/statistical or hybrid models have been used to estimate
1 NO2 concentrations in various microenvironments. These models utilize inputs on sources
2 and emissions, air contaminant dispersal, contaminat reactions, removal mechanisms and
.3 responses to questionnaires. Several attempts to model indoor concentrations of NO2 have
.4 been reviewed in Chapter 7. Section 7.3 should be referred to for a detailed discussion of
5 efforts to model indoor NO2 concentrations. The indoor modelling approaches have been
6 very limited in applications because of lack of data on the variability of the input parameters
7 (emission rates, source use, mixing, reactive decay rates, etc.) in actual indoor environments
8 and lack of adequate questionnaires for obtaining information on building or household
9 characteristics. Efforts are underway to attempt to standardize questionnaires for use in
0 indoor air quality studies (i.e., Lebowitz et al., 1989). Such efforts will provide better input
1 data into models for predicting indoor NO2 concentrations. Ambient models do not provide
August 1991 8-15 DRAFT-DO NOT QUOTE OR CITE
-------
1 spatial and temporal NO2 concentrations on a scale needed to assess concentrations outside of
2 various microenvironments or at the interface between the outdoor air and an individual. No
3 models exist for transit microenvironments.
4 The primary method for determining exposures in epidemiologic studies of the effects of
5 NO2 has been the questionnaires (e.g., Melia et al., 1977, 1979). Typically these
6 questionnaires determined such information as the geographic location of a respondents
7 residence, type of occupation and whether a source exists in a house (gas cooking stove,
8 kerosene heater, etc.). Reliance on this type of data for exposure, as discussed earlier
9 (Section 8.2), can result in serious misclassification errors due to faulty exposure
10 specifications and failure to adequately account for confounding factors. These errors result
11 in the increased probability of masking or misrepresenting any exposure-effect relationships
12 which may exist. It is not enough, for example, to determine a respondent's exposure status
13 based upon their response to the question of existence of a gas cooking range in their home.
14 Chapter 7 indicates that there is a wide distribution of indoor NO2 levels in homes with gas,.
15 cooking ranges related to outdoor levels, indoor sources, indoor source use, condition of the
16 source, season of the year, indoor mixing, infiltration and indoor building materials and
17 furnishings.
18 Total personal exposures to NO2 are assessed indirectly by using a surrogate
19 micronenvironmental measurement or by combining modelled microenvironmental
20 concentrations with time-activity patterns. Section 8.2.2 indicates that a whole house NO2
21 measurement or a bedroom NO2 measurement is highly correlated with total personal NO2
22 exposure although there is a tendency for a underestimation of the actual level of exposure.
23 The good agreement is in large part due to the time spent in the residential environment-by
24 individuals. Presumably a model that predicts whole house NO2 levels could be used to
25 represent personal exposures. Combining microenvironmental monitoring data or modelled
26 concentrations with time activity patterns (Equation 8-1) for an individual has been used in a
27 number of the studies discussed in Section 8.2.2. The results indicated good agreement
28 between measured personal exposures and those calculated from time-activity patterns and
29 microenvironmental monitoring. Due to the potential for higher levels indoors and high level
30 of time spent indoors, outdoor levels contribute little to the total exposure while indoor
August 1991 8-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentrations dominate. In areas with high ambient NO2 levels such as the Los Angeles
2 Basin, the influence of outdoor levels would be greater. The time-activity portions of these
3 studies and the extensive review of time activity studies conducted by Ott (1989) clearly
4 demonstrate the importance of the home environment in terms of a time budget.
5 Figures 8-4a and 8-4b (Ott, 1989) highlight the portion of time spent in the home by
6 individuals employed outside the home and by full-time homemakers. It should be
7 emphasized again, however, Figures 8-4a and 8-4b represent average time budgets and that
8 segments of the population can deviate significantly from those averages.
9 .
10 8.3.1 Personal Exposure Models
11 Billick et al. (1991) report an effort to develop a model predictive of personal NO2
12 exposure based upon microenvironmental measurements and time activity patterns. Data used
13 to develop the model come from eight different sites in the U.S. that were obtained from
14 three different studies (Wilson et al., 1986; Ryan et al., 1988; Spengler et al., 1987). The
15 data base, reviewed all or in part in earlier publications (e.g., Butler et al., 1990; Drye
16 et al., 1989; Spengler et al., 1989), contains over 6,400 measurements of NO2 (Palmes
17 tubes) in over 1,700 households and represents a range of exposures levels over a diverse set
18 of geographic and climatic conditions in the U.S. The model was developed in two steps.
19 The first step was to develop a model predictive of indoor-outdoor and in-vehicle-outdoor
20 levels of NO2. The second step involved the integration of time-activity data with the models
21 developed in the first step.
22 The indoor-outdoor residential model was developed from a simplified version of the
23 general mass-balance equation (see Chapter 7, Equations 7-1, 7-2, and 7-3). The model
24 (Sexton etal., 1983; Drye etal., 1989) is given below:
25 . . •
26 Cin = mCout+b (8-2)
27
*2
28 where C^ is the indoor concentration (/ig/nr), m is the penetration coefficient for outdoor
29 NO2 and b is the concentration contribution by indoor sources (/zg/m3). In the above model
30 m and b are estimated from a multivariate analysis of data collected in the eight studies. The
August 1991 8-17 DRAFT-DO NOT QUOTE OR CITE
-------
INDOORS OTHER
5.4K
IK-TRANSIT
42V.
OUTDOORS
Figure 8-4a. Proportion of time spent by women who are full-time homemakers in
indoor, outdoor, and in-transit microenvironments.
Source: Data from time budget studies in 44 U.S. cities (Szalai [1972], Robinson [1977]); additional
interpretation and analyses appear in Ott (1989).
IN-TRANSIT
GK
OUTDOORS
Stk
INDOORS OTHER
Figure 8-4b. Proportion of time spent by employed persons in indoor, outdoor, and
in-transit microenvironments.
Source: Data from time budget studies in 44 U.S. cities (Szalai [1972], Robinson [1977]); additional
interpretation and analyses appear in Ott (1989).
August 1991
8-18
DRAFT-DO NOT QUOTE OR CITE
-------
1 model was estimated for NO2 concentrations separately for indoor location (e.g., bedroom,
2 kitchen, in-vehicle), season (winter, summer) and for cooking range type (gas vs. electric).
3 All the models used measured outdoor NO2 levels and the models for gas vs. electric ranges
4 used the inverse volume of the dwelling. The model for electric range contained a
5 dichotomous independent variable for the presence of a gas furnace while the model for a gas
6 cooking range dwellings were characterized by whether the appliance employs a continuously
7 burning pilot light and whether a microwave oven is present. A earlier version of this model
8 (Drye et al., 1989) based on a more limited data set is discussed in Chapter 7. The model
9 for in-vehicle-outdoor relationships uses data obtained in a study of NO2 concentration
10 measured inside and outside vehicles (Chan, 1990). The predictive models fitted from the
11 data for dwellings is shown in Tables 8-1 and 8-21. The in-vehicle-outdoor model uses a
'.2 range of indoor/outdoor ratios from 0.2 to 3.32. The fitted models resulted in R2 values
.3 ranging from 0.12 to 0.64 indicating that from 36 to 82% of the variation in indoor
.4 variations in NO2 levels remained unexplained.
.5 The model used to estimate the time weighted average NO2 exposure was presented
6 simply as the sum of the products of the time spent in each of the three environments (in
.7 dwellings, outdoors and in-vehicles) times the predicted concentration in each environment,
8 divided by the sum of the time spent in each environment (Equation 8-1). The concentrations
9 in each environment were determined from the above model (Equation 8-2, while the time in
0 each environment is developed from simulations formed by taking random draws from
1 distributions of observed amounts of time spent outdoor and in-vehicle (e.g., Schwab et al.,
2 1990). The utility of the model stems from its capacity to estimate personal NO2 exposure
3 profile for an area based on a limited set of relative easily obtainable information related to
4 microenvironmental concentrations and household characteristics. The model is a potentially
5 useful tool with which to develop and test the impact of mitigation measures. The utility of
6 the model was demonstrated by the authors in using it to estimating population-based
7 exposure profiles in Los Angeles (utilizing readily available data on outdoor concentrations,
8 housing characteristics and time-activity patterns) and demonstrating the impact on exposure
9 distributions of various exposure reduction scenarios. The overall model has not yet been
0 validated and the uncertainty associated with personal NO2 exposures has not been
August 1991 , 8-19 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 8-1. ELECTRIC-RANGE HOME LEAST SQUARES REGRESSION
COEFFICIENTS AND T-STATISTICS (IN PARENTHESES)*
Kitchen
Variable
Ambient NO2 (ppb)
l/(Home Volume)"
Furnace Fuel (gas= 1)
R2
Observations
Summer
0.61 (28.44)
-2.52 :|[|0:p|fl
4.50 (8.16)
0.58
646
Winter
0.52 (15.27)
42.22 (5.39)
-0.43
0.21
754
Bedroom
Summer
0.58 (32.15)
-2,87
3.64 (7.87)
0.64
641
Winter
0.43 (11.23)
41.81 (4.65)
-0.01 ' (0.01)
0,12
736
Regression Equation: Regression Equation: Indoor NO2 = /31 (ambient NO2 level) $2 (I/home volume) +
j83 (gas furnace present)
"Shaded t-statisties values are non-significant at the 95 % confidence level,
""Units: I/thousands of cubic feet.
Source: Billick et al. (1991).
TABLE 8-2. GAS-RANGE HOME LEAST SQUARES REGRESSION COEFFICIENTS
AND T-STATISTICS (IN PARENTHESES)*
Kitchen
Variable
Ambient NO2 (ppb)
No Pilot/Microwave
No Pilot/No Microwave
Pilot/Microwave
Pilot/No Microwave
l/(Home Volume)""
R2
Observations
Summer
0,85
2.51
4.05
14.30
16.66
28.21
(29.08)
(1.96)
(11
.10)
(10,17)
(3
0.58
876
.31)
Winter
0.70
7.48
7.39
26.11
28.96
37.85
(20.50)
(3.06)
(2.06)
(12.69)
(10.47)
(2.59)
0.41
952
Bedroom
Summer
0.70
1.22
2.56
6.85
7.60
28.04
(34.42)
-
(7.
(6.
(4.
0.63
868
55)
62)
55)
Winter
0.53
2.14
-0.34
11.92
13.06
56.20
(16.
(6.
(5.
(3.
0.29
947
60)
111
10)
04)
94)
Regression Equation: Indoor NO2 = j91 (ambient NO2 level) + /32 (no pilot/microwave)
+ /33 (no pilot/no microwave) + $4 (pilot/microwave)
-f /35 (pilot/no microwave) + j96 (I/home volume)
"Shaded t-statistics values are non-significant at the 95% confidence level.
"Units: 1/thousands of cubic feet.
Source: Billick et al. (1991).
August 1991
8-20
DRAFT-DO NOT QUOTE OR CITE
-------
1 determined. Such information will be needed before the overall utility of the model in NO2
2 risk management applications can be assessed.
3 - . .
4
5 8.4 SUMMARY
6 Exposure to NO2 occurs across a number of micorenvironments or settings. An
7 individual's integrated exposure is the sum of all of the individual NO2 exposures over all
8 time intervals for all microenvironments, weighted by the time in each microenvironment.
9 Accurate assessments of total NO2 exposure and the environments in which exposures take
0 place are essential to minimize misclassification errors in epidemiologic studies, in defining
1 population exposure distributions in risk assessment and in developing effective mitigation
2 measures in risk management. Personal NO2 exposures can be assessed by direct and indirect
3 measures. Direct measures include biomarkers and use of personal monitors. No validated
4 biomarkers for exposure are presently available for NO2. A limited number of studies have
5 been conducted in which personal exposures to NO2 were measured using passive monitors.
5 These studies generally indicate the outdoor levels of NO2, while'related to both personal
7 levels and indoor concentrations, are poor predictors of personal exposures for most
3 populations. Average indoor residential concentrations (e.g., whole house average or
) bedroom level) tend to be the best predictor of personal exposure, typically explaining 50 to
3 60% of the variation in personal exposures. In selected populations, the indoor residential
I environment may not be a good predictor of total exposure because of the higher percent of
I time spent in different environments and/or the potential for unusual NO2 concentration.
5 Indirect measures of personal exposure to NO2 employ various degrees of
!• microenvironmental monitoring and questionnaires to estimate an individual's or population's
5 total exposure. One such model, developed from an extensive monitoring and questionnaire
) data base, can estimate population exposure distributions from easily obtained data on outdoor
t NO2 concentrations, housing characteristics, and time-activity patterns. This model is
I proposed for use in evaluating the impact of various NO2 mitigation measures. The model is
> promising but it has not yet been validated and the uncertainty associated, with it has not been
) characterized.
August 1991 . 8-21 DRAFT-DO NOT QUOTE OR CITE
-------
1 REFERENCES
2 Billick, I. H,; Ozkaynak, H.; Butler, D. A.; Spengler, J. D, (1991) Predicting personal exposures to NO2 for
3 population-based exposure and risk evaluations. Presented at; 84th annual meeting of the Air and Waste
4 Management Association; June; Vancouver, BC, Canada. Pittsburgh, PA: Air and Waste Management
5 Association; paper no. 91-172.9.
6
7 Butler, D. A.; Ozkaynak, H.; Billick, I. H.; Spengler, J. D, (1990) Predicting indoor NO2 concentrations as a
8 function of home characteristics and ambient NO2 levels. In: Indoor air '90: precedings of the 5th
9 ' international conference on indoor air quality and climate, volume 2, characteristics of indoor air;
10 July-August; Toronto, ON, Canada. Ottawa, ON, Canada: International Conference on Indoor Air
11 Quality and Climate, Inc.; pp. 519-524.
12
13 Chan, C, (1990) Commuter exposures to volatile organic compounds [PhD. dissertation]. Boston, MA: Harvard
14 School of Public Health.
15
16 Chapin, F. S., Jr. (1974) Human activity patterns in the city. New York, NY: Wiley-Interscience Publishers.
17
18 Clausing, P.; Mak, J. K.; Spengler, J. D.; Letz, R. (1986) Personal NO2 exposures of high school students.
19 Environ. Int. 12: 413-417.
20
21 Doekery, D. W.; Spengler, J. D.; Reed, M. P.; Ware, J. (1981) Relationships among personal, indoor and
22 outdoor NO2 measurements. Environ. Int. 5: 101-107.
23
24 Dosemeci et al. (1990)
25
26 Drye, E. E.; Ozkaynak, H.; Burbank, B.; Billick, I. H.; Baker, P. E.; Spengler, J. D.; Ryan, P. B.; Colome,
27 S. D. (1989) Development of models for predicting the distribution of indoor nitrogen dioxide
28 concentrations. JAPCA 39: 1169-1177.
29
30 Duan, N. (1981) Microenvironment types: a model for human exposure to air pollution. Stanford, CA: Stanford
31 University, Dept. of Statistics; SIMS technical report no. 47.
32
33 Gladen, B.; Rogan, W. J. (1979) Misclassification and the design of environmental studies. Am. J. Epidemiol.
34 109: 607-616.
35
36 Harlos, D. P.; Marbury, M.; Samet, J.; Spengler, J. D. (1987) Relating indoor NO2 levels to infant personal
37 exposures. Atmos. Environ. 21: 369-376.
38
39 Hoek, G.; Brunekreef, B.; Meijer, R.; Scholten, A.; Boleij, J. (1984) Indoor nitrogen dioxide pollution and
40 respiratory symptoms of schoolchildren. Int. Arch. Occup. Environ. Health 55: 79-86.
41
42 Koo, L. C.; Ho, J. H-C.; Ho, C.-Y.; Matsuki, H.; Shimizu, H.; Mori, T.; Tominaga, S. (1990) Personal
43 " exposure to nitrogen dioxide and its association with respiratory illness in Hong Kong. Am. Rev. Respir.
44 Dis. 141: 1119-1126.
45
46 Lambert, W. E.; Samet, J. M.; Stidley, C. A.; Spengler, J. D. (1990) Classification of chronic exposure to
47 nitrogen dioxide. In: Indoor air '90: precedings of the 5th international conference on indoor air quality
48 and climate, volume 2, characteristics of indoor air; July-August; Toronto, ON, Canada. Ottawa, ON,
49 Canada: International Conference on Indoor Air Quality and Climate, Inc.; pp. 537-542.
50
51 Leaderer, B. P.; Zagraniski, R. T.; Berwick, M.; Stolwijk, J. A. J. (1986) Assessment of exposure to indoor air
52 contaminants from combustion sources: methodology and application. Am. J. Epidemiol. 124: 275-289.
53
August 1991 8-22 DRAFT-DO NOT QUOTE OR CITE
-------
1 Lebowitz, M, D.; Quackenboss, J. J,; Soczek, M. L,; Kollander, M.; Colome, S. (1989a) The new standard
2 environmental inventory questionnaire for estimation of indoor concentrations. JAPCA 39: 1411-1419,
3
4- Lebowitz, M. D.; Quackenboss, J. J.; Soczek, M. L.; Colome, S. D.; Lioy, P. J. (1989b) Workshop:
5 development of questionnaires and survey instruments. In: Nagda, N. L.; Harper, J. P., eds. Design and
S protocol for monitoring indoor air quality. Philadelphia, PA: American Society for Testing and
7 Materials; ASTM special technical publication 1002; pp. 203-216.
5
? Lebret, E. (1990) Errors in exposure measures. Toxicol. Ind. Health 6: 147-156.
)
t Maples, K. R.; Sandstrom, T.; Su, Y.-F.; Henderson, R. F. (1991) The nitric oxide/heme protein complex as a
1 biologic marker of exposure to nitrogen dioxide in humans, rats, and in vitro models. Am. J. Respir.
5 Cell Mol. Biol. 4: 538-543.
\
) Melia, R. J. W.; Florey, C. du V.; Altman, D. G.; Swan, A. V. (1977) Association between gas cooking and
) respiratory disease in children. Br. Med. J. 2: 149-152.
1
I Melia, R. J. W.; Florey, C. du V.; Chinn, S. (1979) The relation between respiratory illness in primary
) schoolchildren and the use of gas for cooking: I - results from a national survey. Int. J. Epidemiol.
) 8: 333-338,
\ Muelenaer, P.; Reid, H.; Morris, R.; Saltzman, L.; Horstman, D.; Collier, A.; Henderson, F. (1987) Urinary
» hydroxyproline excretion in young males exposed experimentally to nitrogen dioxide. In: Seifert, B.;
Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air '87: proceedings of the 4th
> international conference on indoor air quality and climate, v. 2, environmental tobacco smoke,
i multicomponent studies, radon, sick buildings, odours and irritants, hyperreactivities and allergies;
' August; Berlin, Federal Republic of Germany. Berlin, Federal Republic of Germany: Institute for Water,
1 Soil and Air Hygiene; pp. 97-103.
i • •
I National Research Council. (1987) Policies & procedures for control of indoor air quality. Washington, DC:
National Academy Press.
National Research Council. (1989) Biologic markers of pulmonary toxicology. Washington, DC: National
Academy Press.
Ott, W. (1989) Human activity patterns: a review of the literature for estimating time spent indoors, outdoors,
and in-transit. In: Starks, T. H., ed. Proceedings of the research planning conference on human activity
patterns. Las Vegas, NV: U. S. Environmental Protection Agency; EPA report no. EPA-600/4-89/004.
Ozkaynak, H.; Ryan, P. B.; Spengler, J. D.; Laird, N. M. (1986) Bias due to misclassification of personal
exposures in epidemiologic studies of indoor and outdoor air pollution. In: Berglund, B.;
Berglund, U.; Lindvall, T.; Spengler, J.; Sundell, J., eds. Indoor air quality: papers from the third
international conference on indoor air quality and climate; August 1984; Stockholm, Sweden. Environ.
Int. 12: 389-393.
%
Palmes, E. D.; Gunnison, A. F.; DiMattio, J.; Tomczyk, C. (1976) Personal sampler for nitrogen dioxide. Am.
Ind. Hyg. Assoc. J. 37: 570-577.
Quackenboss, J. J.; Kanarek, M, S.; Spengler, J. D.; Letz, R. (1982) Personal monitoring for nitrogen dioxide
exposure: methodological considerations for a community study. Environ. Int. 8: 249-258.
Quackenboss, J. J.; Spengler, J. D.; Kanarek, M. S.; Letz, R.; Duffy, C. P. (1986) Personal exposure to
nitrogen dioxide: relationship to indoor/outdoor air quality and activity patterns. Environ. Sci. Technol.
20: 775-783.
August 1991 8-23 DRAFT-DO NOT QUOTE OR CITE
-------
1 Robinson, J. P. (1977) How Americans use their time: a social psychological analysis of everyday behavior. New
2 York, NY: Praeger Publishers.
3
4 Ryan, P. B.; Koutrakis, P. (1990) Indoor air chemistry: an emerging field. In: Indoor air *90: precedings of the
5 5th international conference on indoor air quality and climate, volume 2, characteristics of indoor air;
6 July-August; Toronto, ON, Canada. Ottawa, ON, Canada: International Conference on Indoor Air
7 Quality and Climate, Inc.; pp. 489-494.
8
9 Ryan, P. B.; Soczek, M. L.; Treitman, R. D.; Spengler, J. D.; Billick, I. H. (1988) The Boston residential NO2
10 characterization study—Du survey methodology and population concentration estimates. Atmos. Environ.
11 22: 2115-2125. ;
12
13 Schwab, M.; Colome, S. D.; Spengler, J, D.; Ryan, P. B,; Billick, L H. (1990) Activity patterns applied to
14 pollutant exposure assessment: data from a personal monitoring study in Los Angeles. Toxicol. Ind.
15 Health 6: 517-532.
16
17 Sexton, K.; Ryan, P. B. (1988) Assessment of human exposure to air pollution: methods, measurements, and
18 models. In: Watson, A. Y.; Bates, R. R.; Kennedy, D., eds. Air pollution, the automobile, and public
19 health, Washington, DC: National Academy Press; pp. 207-238.
20
21 Sexton, K.; Letz, R.; Spengler, J. D. (1983) Estimating human exposure to nitrogen dioxide: an indoor/outdoor
22 modeling approach. Environ. Res. 32: 151-166.
23
24 Shy, C. M.; Kleinbaum, D. G.; Morgenstern, H. (1978) The effect of misclassification of exposure status in
25 epideroiological studies of air pollution health effects. Bull. N. Y. Acad. Med. 54: 1155-1165.
26
27 Spengler, J. D,; Ware, J,; Speizer, P.; Ferris, B.; Dockery, D.; Lebret, E.; Brunnekreef, B. (1987) Harvard's
28 indoor air quality respiratory health study. In: Seifert, B.; Esdom, H.; Fischer, M.; Rueden, H.;
29 Wegner, J., eds. Indoor air '87: proceedings of the 4th international conference on indoor air quality and
30 climate, v. 2, environmental tobacco smoke, multicomponent studies, radon, sick buildings, odours and
31 irritants, hyperreaetivities and allergies; August; Berlin, Federal Republic of Germany. Berlin, Federal
32 Republic of Germany: Institute for Water, Soil and Air Hygiene; pp. 218-223.
33
34 Spengler, J. D.; Ryan, P. B.; Schwab, M.; Billick, I. H.; Colome, S. D.; Becker, E. (1989) An overview of the
35 Los Angeles personal monitoring study. Presented at: First international symposium on total exposure
36 assessment methodology: a new horizon. Las Vegas, NV.
37
38 Szalai, A., ed. (1972) The use of time; daily activities of urban and suburban populations in 12 countries. The
39 Hague, The Netherlands: Mouton and Co.
40
41 Willet. (1989)
42
43 Wilson, A. G.; Seheuneman, E.; Tremayne, M. (1986) House depressurization and flue gas spillage. In:
44 Proceedings of the ASHRAE conference IAQ '86: managing indoor air for health and energy
45 conservation; April; Atlanta, GA. Atlanta, GA: American Society of Heating, Refrigerating and
46 Air-Conditioning Engineers, Inc.; pp. 417-429.
47
48 Yanagisawa, Y.; Nishimura, H. (1982) A badge-type personal sampler for measurement of personal exposure to
49 NO2 and NO in ambient air. Environ. Int. 8: 235-242.
50
51
August 1991 8-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 Yanagisawa, Y.; Matsuki, H.; Osaka, F.; Kasuga, H.; Nishimura, H. (1984) Annual variation of personal
2 exposure to nitrogen dioxide. In: Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of
3 the 3rd international conference on indoor air quality and climate, v. 4, chemical characterization and
4 personal exposure; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building
5 Research; pp. 33-36. Available from: NTIS, Springfield, VA; PB85-104214.
6
7 Yanagisawa, Y.; Nishimura, H.; Matsuki, H.; Osaka, F.; Kasuga, H. (1986) Personal exposure and health effect
8 relationship for NO2 with urinary hydroxyproline to creatinine ratio as indicator. Arch. Environ. Health
9 41: 41-48.
10
11 Yanagisawa, Y.; Nishimura, H.; Matsuki, H.; Osaka, F.; Kasuga, H. (1988) Urinary hydroxyproline to
12 creatinine ratio as a biological effect marker for exposure to NO2 and tobacco smoke. Atmos. Environ,
13 22: 2195-2203.
»U.S. GOVERNMENT PRINTING OFF1CI: 199! "S". I "0 0 3* 0 S » 7
August 1991 . 8-25 DRAFT-DO NOT QUOTE OR CITE
-------
United States Center for Environmental Research
Environmental Protection Information
A9ency Cincinnati OH 45268-1072
Official Business
Penalty for Private Use, S300
EPA/6QO/8-9l/Q49aA
------- |