Regulatory Impact Analysis for the Proposed
Revisions to the National Ambient Air Quality
Standards for Particulate Matter
-------
EPA-452/R-12-003
June 2012
Regulatory Impact Analysis for the Proposed Revisions to the National Ambient Air Quality
Standards for Particulate Matter
Contract No. EP-W-11-029
Work Assignment Number: 0-03
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Research Triangle Park, NC 27711
-------
CONTENTS
Section Page
Executive Summary ES-1
ES.l Overview ES-1
ES.2 Existing and Alternative PM Air Quality Standards ES-1
ES.2.1 Establishing the Baseline ES-3
ES.2.2 Emission Reduction Estimates by Alternative Standards (2020) ES-5
ES.2.3 Health and Welfare Benefits Analysis Approach ES-5
ES.2.4 Cost Analysis Approach ES-7
ES.2.5 Comparison of Benefits and Costs ES-7
ES.2.6 Conclusions of the Analysis ES-8
ES.3 Caveats and Limitations ES-11
ES.3.1 Benefits Caveats ES-11
ES.3.2 Control Strategy and Cost Analysis Caveats and Limitations ES-12
ES.3.3 Limitations of the Secondary Standard Analysis ES-13
ES.4 Discussion ES-13
ES.5 References ES-15
Chapter 1 Introduction and Background 1-1
1.1 Synopsis 1-1
1.2 Background 1-1
1.2.1 NAAQS 1-1
1.2.2 2006 PM NAAQS 1-2
1.3 Role of this RIA in the Process of Setting the NAAQS 1-2
1.3.1 Legislative Roles 1-2
1.3.2 Role of Statutory and Executive Orders 1-3
1.3.3 The Need for National Ambient Air Quality Standards 1-3
1.3.4 Illustrative Nature of the Analysis 1-4
-------
1.4 Overview and Design of the RIA 1-4
1.4.1 Modeling PM2.5 Levels in the Future (Analysis Year = 2020) 1-4
1.4.2 Existing and Alternative PM Air Quality Standards 1-5
1.4.3 Benefits Analysis Approach 1-6
1.4.4 Costs Analysis Approach 1-6
1.5 Organization of this Regulatory Impact Analysis 1-7
Chapter 2 Defining the PM Air Quality Problem 2-1
2.1 Synopsis 2-1
2.2 Particulate Matter (PM) Properties 2-1
2.2.1 PM2.5 2-4
2.2.2 Visibility 2-10
2.3 References 2-14
Chapter 3 Air Quality Modeling and Analysis 3-1
3.1 Synopsis 3-1
3.2 Modeling PM2.5 Levels in the Future 3-1
3.2.1 Air Quality Modeling Platform 3-1
3.2.2 Emissions Inventory 3-4
3.3 Modeling Results and Analyses 3-11
3.3.1 PM2.5 3-12
3.3.2 Visibility 3-19
3.4 References 3-22
Chapter 4 Control Strategies 4-1
4.1 Synopsis 4-1
4.2 PM2.5 Control Strategy Analysis 4-2
4.2.1 Establishing the Baseline 4-2
4.2.2 Alternative Standard Control Strategies 4-8
4.2.3 Emission Reductions Needed Beyond Identified Controls 4-14
4.3 Limitations and Uncertainties 4-16
4.4 References 4-18
IV
-------
4.A Additional Control Strategy Information 4.A-1
Chapter 5 Human Health Benefits Analysis Approach and Results 5-1
5.1 Synopsis 5-1
5.2 Overview 5-2
5.3 Updated Methodology Presented in this RIA 5-6
5.4 Human Health Benefits Analysis Methods 5-8
5.4.1 Health Impact Assessment 5-9
5.4.2 Economic Valuation of Health Impacts 5-10
5.5 Uncertainty Characterization 5-12
5.5.1 Monte Carlo Assessment 5-13
5.5.2 Concentration Benchmark Analysis for PM2.5 5-13
5.5.3 Alternative Concentration-Response Functions for PM2.5-
Related Mortality 5-14
5.5.4 Sensitivity Analyses 5-15
5.5.5 Distributional Assessment 5-16
5.5.6 Influence Analysis—Quantitative Assessment of Uncertainty 5-16
5.5.7 Qualitative Assessment of Uncertainty and Other Analysis
Limitations 5-17
5.6 Benefits Analysis Data Inputs 5-19
5.6.1 Demographic Data 5-19
5.6.2 Baseline Incidence and Prevalence Estimates 5-20
5.6.3 Effect Coefficients 5-24
5.6.4 Unquantified Human Health Benefits 5-42
5.6.5 Economic Valuation Estimates 5-44
5.6.6 Hospital Admissions and Emergency Department Valuation 5-57
5.6.7 Minor Restricted Activity Days Valuation 5-58
5.6.8 Growth in WTP Reflecting National Income Growth Over Time 5-59
5.7 Benefits Results 5-62
5.7.1 Benefits of Attaining Alternative Combinations of Primary PM2.5
Standards 5-62
5.7.2 Uncertainty in Benefits Results 5-63
5.7.3 Estimated Life Years Gained Attributable to Reduced PM2.5
Exposure and Percent of Total Mortality 5-74
-------
5.7.4 Analysis of Mortality Impacts at Various Concentration
Benchmarks 5-79
5.7.5 Additional Sensitivity Analyses 5-85
5.8 Discussion 5-86
5.9 References 5-87
5.A Distribution of the PM2.5-Related Benefits 5.A-1
5.B Additional Sensitivity Analyses Related to the Health Benefits Analysis 5.B-1
5.C Qualitative Assessment of Uncertainty 5.C-1
Chapter 6 Welfare Benefits Analysis Approach 6-1
6.1 Important Caveats Regarding this Chapter 6-1
6.2 Synopsis 6-1
6.3 Introduction to Welfare Benefits Analysis 6-1
6.4 Visibility Benefits 6-6
6.4.1 Visibility and Light Extinction 6-6
6.4.2 Visibility Valuation Overview 6-12
6.4.3 Recreational Visibility 6-14
6.4.4 Residential Visibility 6-22
6.4.5 Discussion of Visibility Benefits 6-29
6.5 Materials Damage Benefits 6-30
6.6 Climate Benefits 6-33
6.6.1 Climate Effects of Short Lived Climate Forcers 6-34
6.6.2 Climate Effects of Long-Lived Greenhouse Gases 6-37
6.7 Ecosystem Benefits and Services 6-37
6.7.1 Ecosystem Benefits for Metallic and Organic Constituents of PM 6-40
6.7.2 Ecosystem Benefits from Reductions in Nitrogen and Sulfur
Emissions 6-43
6.7.3 Ecosystem Benefits from Reductions in Mercury Emissions 6-61
6.7.4 Vegetation Benefits from Reductions in Ambient Ozone 6-64
6.8 References 6-71
6.A Additional Details Regarding the Visibility Benefits Methodology 6.A-1
VI
-------
Chapter 7 Engineering Cost Analysis 7-1
7.1 Synopsis 7-1
7.2 PM2.s Engineering Costs 7-2
7.2.1 Data and Methods—Identified Control Costs (non-EGU Point
and Area Sources) 7-2
7.2.2 Identified Control Costs 7-3
7.2.3 Extrapolated Costs 7-6
7.2.4 Total Cost Estimates 7-10
7.3 Changes in Regulatory Cost Estimates overTime 7-12
7.3.1 Examples of Technological Advances in Pollution Control 7-14
7.3.2 Influence on Regulatory Cost Estimates 7-16
7.3.3 Influence of Regulation on Technological Change 7-20
7.4 Uncertainties and Limitations 7-21
7.5 References 7-22
7.A Other Extrapolated Cost Approaches 7.A-1
Chapter 8 Comparison of Benefits and Costs 8-1
8.1 Synopsis 8-1
8.2 Analysis 8-1
8.3 Conclusions of the Analysis 8-2
8.4 Caveats and Limitations 8-3
8.4.1 Benefits Caveats 8-3
8.4.2 Control Strategy and Cost Analysis Caveats and Limitations 8-4
Chapter 9 Statutory and Executive Order Reviews 9-1
9.1 Synopsis 9-1
9.2 Executive Order 12866: Regulatory Planning and Review 9-1
9.3 Paperwork Reduction Act 9-1
9.4 Regulatory Flexibility Act 9-1
9.5 Unfunded Mandates Reform Act 9-2
9.6 Executive Order 13132: Federalism 9-2
VII
-------
9.7 Executive Order 13175: Consultation and Coordination with Indian
Tribal Governments 9-3
9.8 Executive Order 13045: Protection of Children from Environmental
Health and Safety Risks 9-3
9.9 Executive Order 13211: Actions that Significantly Affect Energy Supply,
Distribution or Use 9-4
9.10 National Technology Transfer and Advancement Act 9-4
9.11 Executive Order 12898: Federal Actions to Address Environmental
Justice in Minority Populations and Low-Income Populations 9-6
Chapter 10 Secondary Standards Analysis 10-1
10.1 Introduction 10-1
10.2 The Secondary NAAQS Standard 10-1
10.3 Visibility Benefits from PM Reduction 10-2
10.4 Baseline Modeling Projection Data (2020) 10-3
10.5 Impacts of Attaining a Distinct Secondary Standard 10-5
10.6 Limitations of Analysis 10-5
10.7 References 10-6
10.A 2017 Modeled Design Values by State, County, and Site 10.A-1
Chapter 11 Qualitative Discussion of Employment Impacts of Air Quality Regulations 11-1
11.1 Introduction 11-1
11.2 Influence of NAAQS Controls on Employment 11-1
11.3 The Current State of Knowledge Based on the Peer-Reviewed Literature .... 11-3
11.3.1 Immediate and Short-Run Employment Impacts 11-3
11.3.2 Long-Term Employment Impacts on the Regulated Industry 11-4
11.4 Conclusion 11-7
11.5 References 11-7
VIM
-------
LIST OF FIGURES
Number Page
2-1. Detailed Source Categorization of Anthropogenic Emissions of Primary PM2.5,
PMio and Gaseous Precursor Species S02, NOX, NH3 and VOCs for 2002 in Units of
Million Metric Tons (MMT). EGUs = Electricity Generating Units 2-3
2-2. Regional and Local Contributions to Annual Average PM2.5 by Particulate S042~,
Nitrate and Total Carbon (i.e., organic plus EC) for Select Urban Areas Based on
Paired 2000-2004 IMPROVE and CSN Monitoring Sites 2-5
2-3. Regional and Seasonal Trends in Annual PM2.5 Composition from 2002 to 2007
Derived Using the SANDWICH Method. Data from the 42 monitoring locations
shown on the map were stratified by region and season including cool months
(October-April) and warm months (May-September) 2-6
2-4. RCFM (left) versus SANDWICH (right) Pie Charts Comparing the Ambient and
PM2.5 FRM Reconstructed Mass Protocols on an Annual Average Basis 2-8
2-5. Maximum County-level PM2.5 Annual Design Values Calculated Using 2003-2007
FRM 24-hr Average PM2.5 Measurements 2-9
2-6. Maximum County-level PM2.5 24-hour Design Values Calculated Using 2003-2007
FRM 24-hr Average PM2.5 Measurements 2-10
2-7. Important Factors Involved in Seeing a Scenic Vista 2-11
2-8. Visibility in Selected National Parks and Wilderness Areas in the U.S., 1992-2008 2-13
2-9. Maximum County-level Visibility Design Values Calculated Using 2004-2006 24-
hr Average Speciated PM2.5 Measured Concentrations 2-14
3-1. Map of the CMAQ Modeling Domains Used for PM NAAQS RIA 3-2
3-2. Diagram of Rollback Method 3-19
4-1. Counties Projected to Exceed the Baseline and Analysis Levels of the PM2.5
Annual Standard Alternatives in 2020 4-5
4-2. Counties Projected to Exceed the Baseline and Analysis Levels of the PM2.5
24-hour Standard Alternatives in 2020 4-6
4-3. Counties Projected to Exceed the 12/35 ug/m3 Alternative Standard After
Meeting the Baseline (Current Standard) in 2020 4-9
4-4. Counties Projected to Exceed the 11/35 ug/m3 Alternative Standard After
Meeting the Baseline (Current Standard) in 2020 4-10
4-5. Counties Projected to Exceed the 11/30 ug/m3 Alternative Standard After
Meeting the Baseline (Current Standard) in 2020 4-11
5-1. Illustration of BenMAP Approach 5-10
5-2. Data Inputsand Outputs for the BenMAP Model 5-11
5-3. Estimated PM2.5-Related Premature Mortalities Avoided According to
Epidemiology or Expert-Derived PM2.5 Mortality Risk Estimate for 12/35 and
13/35 5-72
IX
-------
5-4. Total Monetized Benefits Using 2 Epidemiology-Derived and 12-Expert Derived
Relationships Between PM2.5 and Premature Mortality for 12/35 and 13/35 5-73
5-5a. Distribution of Estimated Avoided Premature Deaths by the Age at which these
Populations were Exposed in 2020 for 12/35 5-78
5-5b. Distribution of Estimated Life Years Gained by the Age at which these
Populations were Exposed in 2020 for 12/35 5-78
5-6. Number of Premature PM2.5-related Deaths Avoided for 12/35 and 13/35
According to the Baseline Level of PM2.5 and the Lowest Measured Air Quality
Levels of Each Mortality Study 5-84
5-7. Number of Premature PM2.5-related Deaths Avoided for 12/35 According to the
Baseline Level of PM2.5 and the Lowest Measured Air Quality Levels of Each
Mortality Study 5-85
5.A-1. PM2.5 Mortality Risk Modified by Educational Attainment in Counties Projected
to Exceed 12/35 5.A-5
5.A-2. PM2.5 Mortality Risk by Race in Counties Projected to Exceed 12/35 5.A-6
5.B-1. Alternate Lag Structures for PM2.5 Premature Mortality (Cumulative) 5.B-8
5.B-2. Exponential Lag Structures for PM2.5 Premature Mortality (Cumulative) 5.B-8
6-1. Important Factors Involved in Seeing a Scenic Vista (Malm, 1999) 6-7
6-2. Visibility in Selected National Parks and Wilderness Areas in the U.S., 1992-2008 6-9
6-3. Estimated Improvement in Annual Average Visibility Levels Associated with the
CAAA Provisions in 2020 6-10
6-4. Mandatory Class I Areas in the U.S 6-15
6-5. Visitation Rates and Park Regions for Class I Areas* 6-20
6-6. Residential Visibility Study City Assignment 6-27
6-7. Linkages between Categories of Ecosystem Services and Components of Human
Well-Being from Millennium Ecosystem Assessment (MEA, 2005) 6-38
6-8. Schematic of the Benefits Assessment Process (U.S. EPA, 2006c) 6-39
6-9. Schematics of Ecological Effects of Nitrogen and Sulfur Deposition 6-44
6-10. Nitrogen and Sulfur Cycling, and Interactions in the Environment 6-45
6-11. Areas Potentially Sensitive to Aquatic Acidification 6-48
6-12. Areas Potentially Sensitive to Terrestrial Acidification 6-50
6-13. Distribution of Red Spruce (pink) and Sugar Maple (green) in the Eastern U.S 6-51
6-14. Spatial and Biogeochemical Factors Influencing MeHg Production 6-59
6-15. Preliminary USGS Map of Mercury Methylation-Sensitive Watersheds 6-60
6-16. Visible Foliar Injury to Forest Plants from Ozone in U.S. by EPA Regions 6-68
6-17. Presence and Absence of Visible Foliar Injury, as Measured by U.S. Forest
Service, 2002 6-69
7-1. Technological Innovation Reflected by Marginal Cost Shift 7-13
10-1. Counties with Monitors Included in Analysis 10-4
10-2. Design Values in 2020, Prior to Full Attainment of a Primary Standard of 15/35 10-5
-------
LIST OF TABLES
Number Page
ES-1. Emission Reduction Estimates by Standard in 2020 (annual tons/year) ES-5
ES-2. Total Monetized Benefits, Total Costs, and Net Benefits in 2020 (millions of
2006$)-Full Attainment ES-9
ES-3. Regional Breakdown of Total Costs and Monetized Benefits Results ES-10
ES-4. Estimated Number of Avoided PM2.5 Health Impacts for Standard Alternatives-
Full Attainment ES-10
2-1. Annual Average FRM and CSN PM2.5 N03~ and NH4N03 Concentrations at Six Sites
during 2003 2-8
3-1. Geographic Specifications of Modeling Domains 3-3
3-2. Control Strategies and Growth Assumptions for Creating 2020 Base Case
Emissions Inventories from the 2005 Base Case 3-6
3-3. Area Definitions and PM2.5 Air Quality Ratios 3-16
4-1. Controls Applied in the Baseline forthe Current PM2.5 Standard 4-8
4-2. Number of Counties with Exceedances and Number of Additional Counties
Where Reductions were Applied 4-12
4-3. Emission Reductions from Known Controls for Alternative Standards 4-13
4-4. Emission Reductions Needed Beyond Known Control to Reach Alternative
Standards in 2020 (annual tons/year) 4-16
4-5. Summary of Qualitative Uncertainty for Elements of Control Strategies 4-17
4.A-1. Example PM Control Measures for NonEGU Point Source Categories 4.A-3
4.A-2. Example PM Control Measures for Area Sources 4.A-4
4.A-3. Example S02 Control Measures for NonEGU Point 4.A-5
4.A-4. Example NOX Control Measures for NonEGU Source Categories 4.A-7
4.A-5. Area County Definitions for S02 and NOX Emissions Reductions for Control
Strategy Analysis 4.A-7
4.A-6. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an
Annual Design Value (DV) Above 15 or 24-hr Design Value (DV) Above 35 in 2020
Base Case 4.A-14
4.A-7. Air Quality Ratios for Monitors in Counties With at Least One Monitor With an
Annual Design Value (DV) Above 13 in 2020 Baseline (15/35) 4.A-18
4.A-8. Air Quality Ratios for Monitors in Counties with at Least One Monitor With an
Annual Design Value (DV) Above 12 in 2020 Baseline (15/35) 4.A-19
4.A-9. Air Quality Ratios for Monitors in Counties With at Least One Monitor With an
Annual Design Value (DV) Above 11 in 2020 Baseline (15/35) 4.A-21
4.A-10. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an
Annual Design Value (DV) Above 11 or 24-hr Design Value (DV) Above 30 in 2020
Baseline (15/35) 4.A-23
XI
-------
4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35),
2020 12/35 and 2020 11/35 4.A-27
4.A-12.24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 4.A-71
5-1. Estimated Monetized Benefits of the Proposed and Alternative Combinations of
PM2.5 Standards in 2020, Incremental to Attainment of 15/35 (millions of 2006$) 5-1
5-2. Human Health Effects of Pollutants Potentially Affected by Attainment of the
Primary PM2.5 Standards 5-4
5-3. Baseline Incidence Rates and Population Prevalence Rates for Use in Impact
Functions, General Population 5-22
5-4. Asthma Prevalence Rates 5-23
5-5. Criteria Used When Selecting C-R Functions 5-26
5-6. Health Endpoints and Epidemiological Studies Used to Quantify Health Impacts
in the Main Analysis 5-27
5-7. Health Endpoints and Epidemiological Studies Used to Quantify Health Impacts
in the Sensitivity Analysis 5-28
5-8. Summary of Effect estimates from Associated With Change in Long-Term
Exposure to PM2.5 in Recent Cohort Studies in North America 5-34
5-9. Unit Values for Economic Valuation of Health Endpoints (2006$) 5-46
5-10. Influence of Applied VSL Attributes on the Size of the Economic Benefits of
Reductions in the Risk of Premature Death (U.S. EPA, 2006a) 5-52
5-11. Alternative Direct Medical Cost of Illness Estimates for Nonfatal Heart Attacks 5-56
5-12. Estimated Costs Over a 5-Year Period of a Nonfatal Myocardial Infarction (in
2006$) 5-57
5-13. Unit Values for Hospital Admissions 5-58
5-14. Elasticity Values Used to Account for Projected Real Income Growth 5-60
5-15. Adjustment Factors Used to Account for Projected Real Income Growth 5-62
5-16. Population-Weighted Air Quality Change for Adults (30+) for Alternative
Standards Relative to 15/35 5-63
5-17. Estimated number of Avoided PM2.5 Health Impacts for Alternative Combinations
of Primary PM2.5 Standards (Incremental to Attaining Current Suite of Primary
PM2.5 Standards) 5-65
5-18. Estimated Number of Avoided PM2.5-Related Deaths for Alternative
Combinations of Primary PM2.5 Standards (Incremental to Attaining Current Suite
of Primary PM2.5 Standards) 5-66
5-19. Estimated Monetized PM2.5 Health Impacts for Alternative Combinations of
Primary PM2.5 Standards (Incremental to Attaining Current Suite of Primary PM2.5
Standards) (Millions of 2006$, 3% discount rate) 5-67
5-20. Estimated Monetized PM2.5 Health Impacts for Alternative Combinations of
Primary PM2.5 Standards (Incremental to Attaining Current Suite of Primary PM2.5
Standards) (Millions of 2006$, 7% discount rate) 5-68
XII
-------
5-21. Estimated Monetized PM2.5-Related Deaths for Alternative Combinations of
Primary PM2.5 Standards (Incremental to Attaining Current Suite of Primary PM2.5
Standards)(Millionsof 2006$, 3% discount rate) 5-69
5-22. Estimated Monetized PM2.5-Related Deaths for Alternative Combinations of
Primary PM2.5 Standards (Incremental to Attaining Current Suite of Primary PM2.5
Standards) (Millions of 2006$, 7% discount rate) 5-70
5-23. Total Estimated Monetized Benefits of the for Alternative Combinations of
Primary PM2.5 Standards (Incremental to Attaining Current Suite of Primary PM2.5
Standards) (millions of 2006$) 5-71
5-24. Regional Breakdown of Monetized Benefits Results 5-71
5-25. Sum of Life Years Gained by Age Range from Changes in PM2.5 Exposure in 2020
for 12/35 5-77
5-26. Estimated Reduction the Percentage of All-Cause Mortality Attributable to PM2.5
for 12/35 from Changes in PM2.5 Exposure in 2020 5-79
5-27. Estimated Reduction in Incidence of Adult Premature Mortality Occurring Above
and Below the Lowest Measured Levels in the Underlying Epidemiology Studies
for 12/35 and 13/35 5-82
5-28. Percentage of Avoided Premature Deaths Occurring At or Above the Lowest
Measured Levels in the Underlying Epidemiology Studies for each Alternative
Combination of Primary PM2.5 Standards 5-83
5.A-1. Key Attributes of the Distributional Analyses in this Appendix 5.A-3
5.A-2. Numerical Values Used for Figures 5.A-1 and 5.A-2 Above 5.A-6
5.B-1. Values of the Time Constant (k) for the Exponential Decay Lag Function 5.B-6
5.B-2. Sensitivity of Monetized PM2.5-Related Premature Mortality Benefits to
Alternative Cessation Lag Structures, Using Effect Estimate from Krewski et al.
(2009) 5.B-7
5.B-3. Ranges of Elasticity Values Used to Account for Projected Real Income Growth 5.B-9
5.B-4. Ranges of Adjustment Factors Used to Account for Projected Real Income
Growth 5.B-9
5.B-5. Sensitivity of Monetized Benefits to Alternative Income Elasticities 5.B-9
5.B-6. Avoided Cases of Cardiovascular Emergency Department Visits, Stroke and
Chronic Bronchitis in 2020 (95th percentile confidence intervals) 5.B-11
5.B-7. Unit Values for Hospital Admissions in BenMAP 4.0.51 (Abt Associates, 2011) 5.B-12
5.B-8. Unit Values for Hospital Admissions in BenMAP 4.0.43 (Abt Associates, 2010) 5.B-12
5.B-9. Change in Monetized Hospitalization Benefits for 12/35 5.B-12
5.B-10.Summary of Effect Estimates From Associated With Change in Long-Term
Exposure to PM2.5 in Recent Cohort Studies in California 5.B-14
5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits.... 5.C-4
6-1. Welfare Effects by Pollutants Potentially Affected by Attainment of the PM
NAAQS 6-3
6-2. Quantified and Unquantified Welfare Benefits 6-4
6-3. Key Assumptions in the Light Extinction Estimates Affecting the Visibility Benefits
Analysis 6-12
6-4. WTP for Visibility Improvements in Class I Areas in Non-Studied Park Regions 6-20
XIII
-------
6-5. Summary of Key Assumptions in the Recreational Visibility Benefits 6-21
6-6. Summary of Residential Visibility Valuation Estimates 6-25
6-7. Summary of Key Assumptions in the Residential Visibility Benefits 6-28
6-8. Materials Damaged by Pollutants Affected by this Rule (U.S. EPA, 2011b) 6-33
6-9. Aquatic Status Categories 6-47
6.A-1. Available Information on WTP for Visibility Improvements in National Parks 6.A-8
6.A-2. Summary of Region-Specific Recreational Visibility Parameters to be Estimated in
Household Utility Functions 6.A-9
6.A-3. Mean Annual Household WTP for Changes in Visual Range for Recreational
Visibility (1990$) 6.A-21
6.A-4. Region-Specific Parameters for Recreational Visibility Benefits 6.A-21
6.A-5. Mean Annual Household WTP for Changes in Visual Range for Residential
Visibility 6.A-22
6.A-6. Parameters for Income Growth Adjustment for Visibility Benefits 6.A-23
7-1. Summary of Sectors, Control Costs, and Discount Rates for Known Control Costs
(millions of 2006$) 7-4
7-2. Partial Attainment Known Control Costs in 2020 for Alternative Standards
Analyzed (millions of 2006$) 7-5
7-3. Fixed Costs by Alternative Standard Analyzed (millions of 2006$) 7-10
7-4. Total Costs by Alternative Standard Analyzed (millions of 2006$) 7-12
7-5. Phase 2 Cost Estimates 7-18
7-6. Comparison of Inflation-Adjusted Estimated Costs and Actual Price Changes of
EPA Fuel Control Rules 7-19
7-7. Summary of Qualitative Uncertainty for Modeling Elements of PM Engineering
Costs 7-21
7.A-1. Extrapolated Costs by Alternative Standard Analyzed (millions of 2006$) 7.A-3
7.A-2. Sensitivity Analysis of Fixed-Cost Approach for Unknown Controls by Alternative
Standard Analyzed (millions of 2006$) 7.A-7
7.A-3. Sensitivity Analysis of Alternative Hybrid Approach for Unknown Controls by
Alternative Standard Analyzed (millions of 2006$) 7.A-8
8-1. Total Monetized Benefits, Total Costs, and Net Benefits in 2020 (millions of
2006$)—Full Attainment 8-2
8-2. Estimated Number of Avoided PM2.s Health Impacts for Standard Alternatives-
Full Attainment 8-3
10-1. Percentage of Monitors Projected to Exceed Alternative Secondary Standards in
2020, Prior to Attainment of Primary Standard of 15/35 10-4
10.A-1.2017 Modeled Design Values by State, County, and Site 10.A-1
XIV
-------
EXECUTIVE SUMMARY
ES.l Overview
Based on its review of the air quality criteria and the national ambient air quality
standards (NAAQS) for particulate matter (PM), the U.S. Environmental Protection Agency (EPA)
is proposing to revise the primary (health-based) and secondary (welfare-based) NAAQS for fine
particles (generally referring to particles less than or equal to 2.5 micrometers [u.m] in
diameter—PM2.5) to provide requisite protection of public health and welfare, respectively. As
has traditionally been done in NAAQS rulemakings, the EPA has conducted a Regulatory Impact
Analysis (RIA) to provide the public with illustrative estimates of the potential costs and health
and welfare benefits of attaining several alternative PM2.5 standards based on one possible set
of selected control strategies for reducing direct PM and PM precursor emissions.
In NAAQS rulemakings, the RIA is prepared for informational purposes only, and the
proposed decisions on the PM NAAQS discussed in the proposed rulemaking are not in any way
based on consideration of the information or analyses in the RIA. The RIA fulfills the
requirements of Executive Orders 12866 and 13563 and guidelines of the Office of
Management and Budget's (OMB) Circular A-4.1 Benefit and cost estimates provided in the RIA
are not additive to benefits and costs from other regulations, and the costs and benefits
identified in this RIA will not be realized until specific controls are mandated by State
Implementation Plans (SIPs) or other federal regulations.
ES.2 Existing and Alternative PM Air Quality Standards
Currently, two primary PM2.5 standards provide public health protection from effects
associated with fine particle exposures: the annual standard and the 24-hour standard. The
annual standard is set at a level of 15.0 u.g/m3, based on the 3-year average of annual
arithmetic mean PM2.5 concentrations. The 24-hour standard is set at a level of 35 u.g/m3, based
on the 3-year average of the 98th percentile of 24-hour PM2.5 concentration. In the RIA, the
current primary PM2.5 standard, including both annual and 24-hour averaging times is denoted
as 15/35.
In the current PM NAAQS review, the EPA is proposing to revise the level of the primary
annual PM2.5 standard within the range of 12 to 13 u.g/m3 in conjunction with retaining the level
of the 24-hour standard at 35 u.g/m3. In order to characterize the costs and benefits, it was
1 U.S. Office of Management and Budget. Circular A-4, September 17, 2003. Available at
http://www.whitehouse.gov/omb/circulars/a004/a-4.pdf.
ES-1
-------
necessary to identify discrete levels along this continuum. For purposes of this analysis, we
identified an annual standard of 12 u.g/m3 in conjunction with retaining the level of the 24-hour
standard at 35 u.g/m3 (denoted 12/35) and an annual standard of 13 u.g/m3 in conjunction with
retaining the level of the 24-hour standard at 35 u.g/m3 (denoted as and 13/35).
In addition to 12/35 and 13/35, the RIA also analyzes the benefits and costs of
incremental control strategies for two other alternative standards (11/35 and 11/30). The four
alternative standards analyzed are as follows:
• A revised annual standard level of 13 u.g/m3 in conjunction with retaining the
24-hour standard level at 35 u.g/m3 (13/35)
• A revised annual standard level of 12 u.g/m3 in conjunction with retaining the
24-hour standard level at 35 u.g/m3 (12/35)
• A revised annual standard level of 11 u.g/m3 in conjunction with retaining the
24-hour standard level at 35 u.g/m3 (11/35 )
• A revised annual standard level of 11 u.g/m3 in conjunction with a revised 24-hour
standard level at 30 u.g/m3 (11/30 )
In analyzing the current 15/35 standard (baseline), the EPA determined that all counties
would meet the 14/35 standard concurrently with meeting the existing 15/35 standard at no
additional cost. Consequently, no incremental costs or benefits are associated with 14/35 and
therefore, no need to present an analysis of 14/35 in this RIA.
Currently, the existing secondary PM2.5 standards are identical in all respects to the
primary standards. In the current PM NAAQS review, the EPA is proposing to add a distinct
standard for PM2.5 to provide protection from PM-related visibility impairment. Specifically, the
EPA is proposing to establish a separate secondary standard defined in terms of a PM2.5
visibility index, which would use speciated PM2.5 mass concentrations and relative humidity
data to calculate PM2.5 light extinction, similar to the Regional Haze Program; a 24-hour
averaging time; a 90th percentile form; and a level of either 30 deciviews (dv) or 28 dv. Based
on the air quality analysis conducted for the primary PM2.5 standard, all monitored areas are
estimated to be in attainment with both proposed secondary standard levels in 2020, assuming
full attainment of the primary PM2.5 standard. For the two optional levels proposed for the
secondary standard, no additional costs or benefits will be realized beyond those quantified for
meeting the primary PM2.5 standard in this RIA.
ES-2
-------
With regard to the primary and secondary standards for particles less than or equal to
10 u.m in diameter (PMi0), the EPA is proposing to retain the current primary and secondary
24-hour PMio standards. Both standards are the same. The current primary and secondary 24-
hour standards are set at a level of 150 u.g/m3, not to be exceeded more than once per year on
average over 3 years (EPA, 1997)2. Since the benefit-cost analysis of the alternative PMi0
standards was conducted when the standard was selected, this RIA does not repeat that
analysis here.
ES.2.1 Establishing the Baseline
The RIA is intended to evaluate the costs and benefits of reaching attainment with
potential alternative PM2.s standards. In order to develop and evaluate control strategies for
attaining a more stringent primary standard, it is important to first estimate PM2.s levels in 2020
given the current NAAQS standards (15/35) and air quality trends. Estimating the 2020 levels is
known as the baseline. Establishing this baseline allows us to estimate the incremental costs
and benefits of attaining any alternative primary standard.
The baseline includes reductions already achieved as a result of national regulations,
reductions expected prior to 2020 from recently promulgated national regulations3 (i.e.,
reductions that were not realized before 2005 but are expected prior to attainment of the
current PM standard), and reductions from additional controls which the EPA estimates need to
be included to attain the current standard (15/35). Reductions achieved as a result of state and
local agency regulations and voluntary programs are reflected to the extent that they are
represented in emission inventory information submitted to the EPA by state and local
agencies4. Two steps were used to develop the baseline. First, the reductions expected in
national PM2.s concentrations from national rules promulgated prior to this analysis were
considered (referred to as the base case). Below is a list of some of the major national rules
reflected in the base case. Refer to Chapter 3, Section 3.2.2 for a more detailed discussion of
the rules reflected in the base case emissions inventory.
• Light-Duty Vehicle Tier 2 Rule (U.S. EPA, 1999)
2 U.S. Environmental Protection Agency. 1997. Regulatory Impact Analyses for the Particulate Matter and Ozone
National Ambient Air Quality Standards and Proposed Regional Haze Rule. Available at:
http://www.epa.gov/ttn/oarpg/naaqsfin/ria.html.
3 The recently proposed Boiler MACT and CISWI reconsiderations are not included in the base case. These rules
were not yet proposed at the time of this analysis. It is not clear how the geographic scope of this rule will
match with the counties analyzed for this RIA—the costs may decrease but the magnitude is uncertain.
4 The amendments to the Low Emissions Vehicle Program (LEV-MI) in California are not included in the base case.
This program requires an approval of U.S. EPA via a waiver. At the time of this analysis the waiver had not been
submitted.
ES-3
-------
• Heavy Duty Diesel Rule (U.S. EPA, 2000)
• Clean Air Nonroad Diesel Rule (U.S. EPA, 2004)
• Regional Haze Regulations and Guidelines for Best Available Retrofit Technology
Determinations (U.S. EPA, 2005b)
• NOX Emissions Standard for New Commercial Aircraft Engines (U.S. EPA, 2005)
• Emissions Standards for Locomotives and Marine Compression-Ignition Engines (U.S.
EPA, 2008)
• Control of Emissions for Nonroad Spark Ignition Engines and Equipment (U.S. EPA,
2008)
• C3 Oceangoing Vessels (U.S. EPA, 2010)
• Hospital/Medical/lnfectious Waste Incinerators: New Source Performance Standards
and Emission Guidelines: Final Rule Amendments (U.S. EPA, 2009)
• Cross-State Air Pollution Rule (U.S. EPA, 2011a)
• Mercury and Air Toxics Standards (U.S. EPA, 2011b)
• Reciprocating Internal Combustion Engines (RICE) NESHAPs (U.S. EPA, 2010)
We did not conduct this analysis incremental to controls applied as part of previous
NAAQS analyses (e.g., 03, NOX, or S02) because the data and modeling on which these previous
analyses were based are now considered outdated and are not compatible with the current
PM2.s NAAQS analysis. In addition, all control strategies analyzed in NAAQS RIAs are
hypothetical. This analysis presents one scenario that states may employ but does not prescribe
how attainment must be achieved.
Second, because the base case reductions alone were not predicted to bring all areas
into attainment with the current standard (15/35), please see Figures 4-1 and 4-2 in Chapter 4
of this RIA, the EPA used a hypothetical control strategy to apply additional known controls to
illustrate attainment with that standard. To establish the baseline, additional control measures
were used in two sectors:5 Non-Electricity Generating Unit Point Sources (Non-EGUs) and Non-
Point Area Sources (Area).
5 In establishing the baseline, the EPA selected a set of cost-effective controls to simulate attainment of the current
PM2.5 standard. These control sets are hypothetical because states will ultimately determine controls as part of
the SIP process.
ES-4
-------
For addition details on the baseline, refer to Chapter 4 of this RIA.
ES.2.2 Emission Reduction Estimates by Alternative Standards (2020)
Emission reductions were calculated for the four alternative standards (13/35,12/35,
11/35, and 11/30) from a baseline of attaining the current standard of 15/35. Emission
reductions were calculated for the known control strategy analysis and the extrapolated cost
analysis for each alternative standard being analyzed. The EPA estimates the national-scale
emission reductions for each of the alternative standards as shown in Table ES-1.
Table ES-1. Emission Reduction Estimates by Standard in 2020 (annual tons/year)3
Alternative Standard
13/35
12/35
11/35
11/30
PM2.5
190
4,300
14,000
22,000
S02
0
970
19,000
23,000
NOX
0
0
1,500
8,200
3 Estimates are rounded to two significant figures.
ES.2.3 Health and Welfare Benefits Analysis Approach
The EPA estimated human health (e.g., mortality and morbidity effects) under full
attainment of the three alternative combinations of primary PM2.s standards. We considered an
array of health impacts attributable to changes in PM2.5 exposure. The EPA has incorporated an
array of policy and technical updates to the benefits analysis approach applied in this RIA,
including incorporation of the most recent follow-up to the American Cancer Society (ACS)
cohort (Krewski et al., 2009), updated health endpoints, new morbidity studies, updated
hospital cost-of-illness estimates, and an expanded uncertainty assessment. Each of these
updates is fully described in the benefits chapter. Even though the alternative primary
standards are designed to protect against adverse effects to human health, the emission
reductions have welfare co-benefits in addition to the direct human health benefits. The term
welfare co-benefits covers both environmental and societal benefits of reducing pollution, such
as reductions in visibility impairment, materials damage, and ecosystem damage. Despite our
attempts to quantify and monetize as many of the benefits as possible, welfare benefits are not
quantified or monetized in this analysis. Unquantified health benefits are discussed in Chapter
5, and unquantified welfare benefits are discussed in Chapter 6.
It is important to note that estimates of the health benefits from reduced PM2.5
exposure reported here contain uncertainty, including from the following key assumptions:
ES-5
-------
1. We assumed that all fine particles, regardless of their chemical composition, are
equally potent in causing premature mortality. This assumption is an important
assumption, because PM2.5 varies considerably in composition across sources, but
the scientific evidence is not yet sufficient to allow differentiation of effects
estimates by particle type.
2. We assumed that the health impact function for fine particles is linear within the
range of ambient concentrations under consideration. Thus, the estimates include
health benefits from reducing fine particles in areas with varied concentrations of
PM2.5, including both regions that are in attainment with the fine particle standard
and those that do not meet the standard down to the lowest modeled
concentrations.
As noted in the preamble to the proposed rule, the Policy Assessment (U.S. EPA, 2011c)
concludes that the range from the 25th to 10th percentiles of the air quality data used the
epidemiology studies is a reasonable range below which we have appreciably less confidence in
the associations observed in the epidemiological studies. In the RIA accompanying the
promulgated PM NAAQS, EPA will characterize the distribution of estimated PM-related health
benefits attributable to PM reductions occurring above and below the selected standard. For
12/35, we estimate that 51% and 92% of the estimated avoided premature deaths occur at or
above an annual mean PIVh.s level of 10 u.g/m3 (the lowest measured level (LML) of the Laden et
al. 2006 study) and 5.8 u.g/m3(the LML of the Krewskietal. 2009 study), respectively. For 13/35,
these estimates are 62% and 89%. These are the two source studies for the concentration-
response functions used to estimate mortality benefits. The EPA briefly describes the
uncertainties in the concentration-response functions below and in considerably more detail in
the benefits chapter of this RIA.
Although these concentration benchmark analyses (e.g., 25th percentile, 10th percentile,
and LML) provide some insight into the level of uncertainty in the estimated PM2.s mortality
benefits, EPA does not view these concentration benchmarks as a concentration threshold. The
best estimate of benefits includes estimates below and above these concentration benchmarks,
but uncertainty is higher in health benefits estimated at lower concentrations, with the lowest
confidence below the LML. Estimated health impacts reflecting air quality improvements both
below and above these concentration benchmarks are appropriately included in the total
benefits estimate. In other words, our increased confidence in the estimated benefits above
these concentration benchmarks should not imply an absence of confidence in the benefits
estimated below these concentration benchmarks.
ES-6
-------
It is important to note that these estimated benefits reflect specific control measures
and emission reductions that are needed to lower PM2.5 concentrations for monitors projected
to exceed the alternative standard analyzed. The result is that air quality will improve in
counties that exceed the alternative standards as well as surrounding areas that do not exceed
the alternative standards. It is not possible to apply controls that only reduce PM2.5 at the
monitor without affecting surrounding areas. In order to make a direct comparison between
the benefits and costs of these control strategies, it is appropriate to include all the benefits
occurring as a result of the control strategies applied.
We estimate benefits using modeled air quality data with 12km grid cells, which is
important because the grid cells are smaller than counties and PM2.5 concentrations vary
spatially within a county. Some grid cells in a county can be below the level of the alternative
standard even though the highest monitor value is above the alternative standard. Thus,
emission reductions lead to benefits in grid cells that are below the alternative standards even
within a county with a monitor that exceeds the alternative standard. We have not estimated
the fraction of benefits that occur only in counties that exceed the alternative standards.
ES.2.4 Cost Analysis Approach
The EPA estimated total costs under partial and full attainment of the alternative PM2.5
standards. The engineering costs generally include the costs of purchasing, installing, and
operating the referenced control technologies. The technologies and control strategies selected
for analysis are illustrative of one way in which nonattainment areas could meet a revised
standard. There are numerous ways to construct and evaluate potential control programs that
would bring areas into attainment with alternative standards, and the EPA anticipates that
state and local governments will consider programs that are best suited for local conditions.
The partial-attainment cost analysis reflects the costs associated with applying known
controls. Costs for full attainment include estimates for the engineering costs of the additional
tons of emissions reductions that are needed beyond identified controls, referred to as
extrapolated costs. The EPA recognizes that the extrapolated portion of the engineering cost
estimates reflects substantial uncertainty about which sectors and which technologies might
become available for cost-effective application in the future.
£5.2.5 Comparison of Benefits and Costs
In the analysis, we estimate the net benefits of the proposed range of annual PM2.5
standards of 12/35 to 13/35. For 12/35, net benefits are estimated to be $2.3 billion to $5.9
billion at a 3% discount rate and $2.0 billion to $5.3 billion at a 7% discount rate in 2020 (2006
ES-7
-------
dollars).6 For 13/35, net benefits are estimated to be $85 million to $220 million at the 3%
discount rate and $76 million to $200 million at the 7% discount rate.
The EPA estimated the net benefits of the alternative annual PM2.5 standard of 11/35 to
be $8.9 billion to $23 billion at a 3% discount rate and $8.0 billion to $21 billion at a 7%
discount rate in 2020. The EPA estimated the net benefits of the alternative annual PM2.s
standard of 11/30 to be $14 billion to $36 billion at a 3% discount rate and $13 billion to $33
billion at a 7% discount rate in 2020. All estimates are in 2006$.7
In analyzing the current 15/35 standard (baseline), the EPA determined that all counties
would meet the 14/35 standard concurrently with meeting the existing 15/35 standard at no
additional cost. No incremental costs or benefits are associated with 14/35 and consequently,
there is no analysis 14/35 in this RIA.
We provide these results in Table ES-2 and a regional percentage breakdown of costs
and benefits in Table ES-3. In Table ES-4, we provide the avoided health incidences associated
with these standard levels.
ES.2.6 Conclusions of the Analysis
The EPA's illustrative analysis has estimated the health and welfare benefits and costs
associated with the proposed revised PM NAAQS. The results for 2020 suggest there will be
significant health and welfare benefits and these benefits will outweigh the costs associated
with the illustrative control strategies in 2020.
6 Using a 2010$ year increases estimated costs and benefits by approximately 8%. Because of data limitations, we
were unable to discount compliance costs for all sectors at 3%. As a result, the net benefit calculations at 3%
were computed by subtracting the costs at 7% from the monetized benefits at 3%.
7 Using a 2010 $ year increases estimated costs and benefits by approximately 8%. Because of data limitations, we
were unable to discount compliance costs for all sectors at 3%. As a result, the net benefit calculations at 3%
were computed by subtracting the costs at 7% from monetized benefits at 3%.
ES-8
-------
Table ES-2. Total Monetized Benefits, Total Costs, and Net Benefits in 2020 (millions of
2006$a)—Full Attainment
Alternative
Standard
13/35
12/35
11/35
11/30
Total
3% Discount
Ratec
$2.9
$69
$270
$390
Costs
7% Discount
Rate
$2.9
$69
$270
$390
Monetized
3% Discount
Rate
$88 to $220
$2,300 to
$5,900
$9,200 to
$23,000
$14,000 to
$36,000
Benefits"
7% Discount
Rate
$79 to $200
$2, 100 to
$5,400
$8,300 to
$21,000
$13,000 to
$33,000
Net
3% Discount
Ratec
$85 to $220
$2,300 to
$5,900
$8,900 to
$23,000
$14,000 to
$36,000
Benefits"
7% Discount
Rate
$76 to $200
$2,000 to
$5,300
$8,000 to
$21,000
$13,000 to
$33,000
a Rounded to two significant figures. Using a 2010$ year increases estimated costs and benefits by approximately
8%.
b The reduction in premature deaths each year accounts for over 98% of total monetized benefits. Mortality risk
valuation assumes discounting over the Science Advisory Board-recommended 20-year segmented lag structure.
Not all possible benefits or disbenefits are quantified and monetized in this analysis. B is the sum of all
unquantified benefits. Data limitations prevented us from quantifying these endpoints, and as such, these
benefits are inherently more uncertain than those benefits that we were able to quantify.
c Due to data limitations, we were unable to discount compliance costs for all sectors at 3%. As a result, the net
benefit calculations at 3% were computed by subtracting the costs at 7% from monetized benefits at 3%.
For the lower end of the proposed standard range of 12/35, the EPA estimates that the
benefits of full attainment exceed the costs of full attainment by 34 to 86 times at a 3%
discount rate and 30 to 78 times at a 7% discount rate. For the upper end of the proposed
standard range of 13/35, the EPA estimates that the benefits of full attainment exceed the
costs of full attainment by 30 to 77 times at a 3% discount rate and 27 to 69 times at a 7%
discount rate. For the alternative standards, 11/35 and 11/30, the EPA estimates that the
benefits of full attainment exceed the costs of full attainment by 34 to 94 times at a 3%
discount rate and 30 to 85 times at a 7% discount rate.
ES-9
-------
Table ES-3. Regional Breakdown of Total Costs and Monetized Benefits Results
Alternative Combination of Standards
13 u.g/m3 annual
&
35 |ig/m3 24-hour
Region
East"
California
Rest of West
Total
Costs3
0%
100%
0%
Monetized
Benefits
0%
98%
2%
12 |ig/m3 Annual
&
35 ug/m3 24-hr
Total
Costs
<1%
94%
5%
Monetized
Benefits
27%
70%
3%
11 ug/m3 Annual
&
35 ug/m3 24-hr
Total
Costs
18%
67%
15%
Monetized
Benefits
53%
44%
3%
11 ug/m3 Annual
&
30 ug/m3 24-hr
Total
Costs
18%
54%
28%
Monetized
Benefits
43%
47%
10%
Costs are discounted at 7%.
b Includes Texas and those states to the north and east. Several recent rules such as MATS and CSAPR will have
substantially reduced PM2.5 levels by 2020 in the East, thus few additional controls would be needed to reach
12/35 or 13/35.
Table ES-4. Estimated Number of Avoided PM2.5 Health Impacts for Standard Alternatives-
Full Attainment3
Alternative Combination of Primary PM2.s Standards
Health Effect
Adult Mortality
Krewski et al. (2009)
Laden etal. (2006) (adult)
Woodruff etal. (1997) (infant)
Non-fatal heart attacks (age > 18)
Peters et al. (2001)
Pooled estimate of 4 studies
Hospital admissions— respiratory (all ages)
Hospital admissions— cardiovascular (age > 18)
Emergency department visits for asthma (age < 18)
Acute bronchitis (age 8-12)
Lower respiratory symptoms (age 7-14)
Upper respiratory symptoms (asthmatics age 9-11)
Asthma exacerbation (age 6-18)
Lost work days (age 18-65)
Minor restricted-activity days (age 18-65)
13/35
11
27
0
11
1
3
3
6
22
290
410
410
1,800
11,000
12/35
280
730
1
320
35
98
95
160
540
6,900
9,800
24,000
44,000
260,000
11/35
1,100
2,900
3
1,300
140
430
400
730
2,000
25,000
37,000
89,000
170,000
1,000,000
11/30
1,700
4,500
4
1,900
210
620
580
1,000
3,100
39,000
56,000
140,000
260,000
1,500,000
Incidence estimates are rounded to whole numbers with no more than two significant figures.
ES-10
-------
ES.3 Caveats and Limitations
EPA acknowledges several important limitations of the primary and secondary analysis.
These include:
ES.3.1 Benefits Caveats
* PM2.5 mortality co-benefits represent a substantial proportion of total monetized
benefits (over 98%). To characterize the uncertainty in the relationship between
PM2.5 and premature mortality, we include a set of twelve estimates of the
concentration-response function based on results of the PM2.5 mortality expert
elicitation study in addition to our core estimates. Even these multiple
characterizations omit the uncertainty in air quality estimates, baseline incidence
rates, populations exposed, and transferability of the effect estimate to diverse
locations. As a result, the reported confidence intervals and range of estimates give
an incomplete picture about the overall uncertainty in the PM2.5 estimates. This
information should be interpreted within the context of the larger uncertainty
surrounding the entire analysis.
• Most of the estimated avoided premature deaths occur at or above the lowest
measured PM2.5 concentration in the two studies used to estimate mortality
benefits. In general, we have greater confidence in risk estimates based on PM2.5
concentrations where the bulk of the data reside and somewhat less confidence
where data density is lower.
• We analyzed full attainment in 2020, and projecting key variables introduces
uncertainty. Inherent in any analysis of future regulatory programs are uncertainties
in projecting atmospheric conditions and source-level emissions, as well as
population, health baselines, incomes, technology, and other factors.
• There are uncertainties related to the health impact functions used in the analysis.
These include within-study variability; pooling across studies; the application of C-R
functions nationwide and for all particle species; extrapolation of impact functions
across populations; and various uncertainties in the C-R function, including causality
and shape of the function at low concentrations. Therefore, benefits may be under-
or over-estimates.
• This analysis omits certain unquantified effects due to lack of data, time, and
resources. These unquantified endpoints include other health and ecosystem
effects. The EPA will continue to evaluate new methods and models and select those
most appropriate for estimating the benefits of reductions in air pollution.
ES-11
-------
ES.3.2 Control Strategy and Cost Analysis Caveats and Limitations
Control Technology Data
* Technologies applied may not reflect emerging devices that may be available in
future years.
• Control efficiency data depend on equipment being well maintained.
• Area source controls assume a constant estimate of emission reductions, despite
variability in extent and scale of application.
Control Strategy Development
* States may develop different control strategies than the ones illustrated.
• Data on baseline controls from current SIPs are lacking.
• Timing of control strategies may be different than envisioned in the RIA.
• Controls are applied within the county with the violating monitor. It is possible that
additional known controls could be available in a wider geographical area.
• Unknown controls were needed to reach attainment in several counties. Costs
associated with these unknown controls were estimated using a fixed-cost per ton
methodology as well as an extrapolated cost methodology.
• Emissions reductions from mobile sources, EGUs, other PM2.5 precursors (i.e.,
ammonia and VOC), and voluntary programs are not reflected in the analyses.
Technological Change
* Emission reductions do not reflect potential effects of technological change that may
be available in future years.
• Effects of "learning by doing" are not accounted for in the emission reduction
estimates.
• Future technology developments in sectors not analyzed here (e.g., EGUs) may be
transferrable to non-EGU and area sources, making these sources more viable for
achieving future attainment at a lower cost than the cost presented in this analysis.
Engineering Cost Estimates
* Because of data limitations, we were unable to discount compliance costs for all
sectors at 3%.
• Estimates of private compliance cost are used as a proxy for social cost.
Unquantified Costs
* A number of costs remain unquantified, including administration costs of federal and
state SIP programs, and transactional costs.
ES-12
-------
ES.3.3 Limitations of the Secondary Standard Analysis
Visibility design values for 2020 were calculated using the CMAQ modeling information
and 2004-2006 ambient measurements. To determine the design values for meeting the
current primary PM2.5 standard and proposed alternative primary standards, we used a
methodology, described in Chapter 3, to estimate the small emissions reductions needed from
control measures to show attainment and to estimate the costs and benefits of attaining the
proposed alternative primary standards. It is not possible to apply this methodology to the
visibility design values.8 As a result, the only analysis available for the proposed alternative
secondary standards in 2020 is prior to full attainment of the current primary standard. All
monitors analyzed are projected to attain a secondary standard of 30 dv in the 2020 base case.
Given the 24-hr design value reductions that were included in simulating attainment of 15/35 in
the 2020 base case, we are confident in our conclusion that all monitors will also attain a
secondary standard of 28 dv when they attain the current primary standards.9
ES.4 Discussion
An extensive body of scientific evidence documented in the Integrated Science
Assessment for Particulate Matter (PM ISA) indicates that PM2.5 can penetrate deep into the
lungs and cause serious health effects, including premature death and other non-fatal illnesses
(U.S. EPA, 2009). As described in the preamble to the proposed regulation, the proposed
changes to the standards are based on an integrative assessment of an extensive body of new
scientific evidence (U.S. EPA, 2009). Health studies published since the PM ISA(e.g., Pope et al.
[2009]) confirm that recent levels of PM2.5 have had a significant impact on public health. Based
on the air quality analysis in this RIA, the EPA projects that nearly all counties with PM2.5
monitors in the U.S. would meet an annual standard of 12 u.g/m3 by 2020 without additional
As described in Chapter 3, we apply a methodology of air quality ratios to estimate the emissions reductions
needed to meet the current and proposed alternative levels for the primary standard. While this methodology
can estimate how these emissions reductions will affect changes in the future-year annual design value and the
corresponding response of the future-year 24-hr design value to changes in the annual design value, it is unable
to estimate how each of the PM2.5 species will change with these emission reductions. Given that estimating
changes in future-year visibility is dependent on the IMPROVE equation and how the PM2.5 species are
projected to change in time, we are unable to estimate visibility design values for meeting the current and
proposed alternative levels for the primary standard.
9 The projected 2020 base case design values for the secondary standard for the following monitors with id
numbers 60658001 (located in Riverside, CA), 60290014 (located in Kern, CA), and 60990005 (located in
Stanislaus, CA) are 29 dv, 30 dv, and 29 dv, respectively. The emissions reductions selected for simulating
attainment of 15/35 in the 2020 base case resulted in the following reductions in the 24-hr design values for
these three monitors: 11.1 u.g/m3, 21.9 u.g/m3 and 5.3 ug/m3, respectively. We believe that these emissions
reductions and 24-hr design value changes for simulating the current primary standard levels of 15/35 will be
enough to lower the projected 2020 secondary standard design values for these three monitors to 28 dv or
lower.
ES-13
-------
federal, state, or local PM control programs. This demonstrates the substantial progress that
the U.S. has made in reducing air pollution emissions over the last several decades. Regulations
such as the EPA's recent Mercury and Air Toxics Standards (MATS), the Cross-State Air Pollution
Rule (CSAPR), and other federal programs such as diesel standards will provide substantial
improvements in regional concentrations of PM2.5. Our analysis shows a few areas would still
need additional emissions reductions to address local sources of air pollution, including ports
and uncontrolled industrial emissions. For this reason, we have designed the RIA analysis to
focus on local controls in these few areas. We estimate that these additional local controls
would yield benefits well in excess of costs, by a ratio of at least 30 to 1.
The setting of a NAAQS does not compel specific pollution reductions, and as such does
not directly result in costs or benefits. For this reason, NAAQS RIAs are merely illustrative. The
NAAQS RIAs illustrate the potential costs and benefits of additional steps States could take to
attain a revised air quality standard nationwide beyond rules already on the books. We base
our illustrative estimates on an array of emission control strategies for different sources. The
costs and benefits identified in this RIA will not be realized until specific controls are mandated
by State Implementation Plans (SIPs) or other federal regulations. In short, NAAQS RIAs
hypothesize, but do not prescribe, the control strategies that States may choose to enact when
implementing a revised NAAQS.
It is important to emphasize that the EPA does not "double count" the costs or the
benefits of our rules. Emission reductions achieved under rules that require specific actions
from sources—such as MATS—are in the baseline of this NAAQS analysis, as are emission
reductions needed to meet the current NAAQS. For this reason, the cost and benefits estimates
provided in this RIA and all other NAAQS RIAs should not be added to the estimates for
implementation rules.
Furthermore, the monetized benefits estimates do not paint a complete picture of the
burden of PM to public health. For example, modeling by Fann et al. (2012) estimated that 2005
levels of air pollution were responsible for between 130,000 and 320,000 PM2.5-related deaths,
or between 6.1% and 15% of total deaths from all causes in the continental United States. The
monetized benefits associated with attaining the proposed range of standards appear modest
when viewed within the context of the potential overall public health burden of PM2.5 and
ozone air pollution estimated by Fann et al. (2012), but this is primarily because regulations
already on the books will make great strides toward reducing future levels of PM. One
important distinction between the total public health burden estimated for 2005 air pollution
levels and the estimated benefits in this RIA is that ambient levels of PM2.5 will have improved
ES-14
-------
substantially by 2020. For example, we estimate that S02 emissions in the U.S. would fall from
14 million tons in 2005 to less than 5 million tons by 2020 (a reduction of 66%). For this reason,
States will only need to achieve small air quality improvements to reach the proposed PM
standards. As shown in recent RIAs for the CSAPR (U.S. EPA, 2011a) and MATS (U.S. EPA,
2011b), implementing other federal and state air quality actions will address a substantial
fraction of the total public health burden of PM2.5and ozone air pollution.
The NAAQS are not set at levels that eliminate the risk of air pollution. Instead, the
Administrator sets the NAAQS at a level requisite to protect public health with an adequate
margin of safety, taking into consideration effects on susceptible populations based on the
scientific literature. The risk analysis prepared in support of this PM NAAQS reported risks
below these levels, while acknowledging that the confidence in those effect estimates is higher
at levels closer to the standard (U.S. EPA, 2010). While benefits occurring below the standard
are assumed to be more uncertain than those occurring above the standard, the EPA considers
these to be legitimate components of the total benefits estimate. Though there are greater
uncertainties at lower PM2.5 concentrations, there is no evidence of a population-level
threshold in PM2.5-related health effects in the epidemiology literature.
Lastly, the EPA was unable to monetize fully all of the benefits associated with reaching
these standards in this RIA, including other health effects of PM, visibility effects, ecosystem
effects, and climate effects. If the EPA were able to monetize all of the benefits, the benefits
would exceed the costs by an even greater margin. Even when considered in light of the
quantified and unquantified uncertainties identified in this RIA, we believe that implementing
the proposed range of standards would have substantial public health benefits that outweigh
the costs.
ES.5 References
Fann, N., A. Lamson, K. Wesson, D. Risley, S.C. Anenberg, and B.J. Hubbell. 2012. "Estimating
the National Public Health Burden Associated with Exposure to Ambient PM2.5 and
Ozone. Risk Analysis." Risk Analysis 32(1): 81-95.
Krewski D, et al. 2009. "Extended follow-up and spatial analysis of the American Cancer Society
study linking particulate air pollution and mortality." HEI Research Report, 140, Health
Effects Institute, Boston, MA. Pope, C.A., III, E. Majid, and D. Dockery. 2009. "Fine
Particle Air Pollution and Life Expectancy in the United States." New England Journal of
Medicine 360: 376-386.
ES-15
-------
U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for
Particulate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment—RTP Division. December. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010. Risk and Exposure Assessment for
Review of Particulate Matter. EPA-452/R-10-005. Office of Air Quality Planning and
Standards, Research Triangle Park, NC. September. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2011a. Regulatory Impact Analysis for the
Federal Implementation Plans to Reduce Interstate Transport of Fine Particulate Matter
and Ozone in 27 States; Correction of SIP Approvals for 22 States. June. Available on the
Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2011b. Regulatory Impact Analysis for the
Final Mercury and Air Toxics Standards. EPA-452/R-11-011. December. Available on the
Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2011c. Policy Assessment for the Review of
the Particulate Matter National Ambient Air Quality Standards. EPA-452/D-11-003. April.
Available on the Internet at
.
ES-16
-------
CHAPTER 1
INTRODUCTION AND BACKGROUND
1.1 Synopsis
This chapter summarizes the purpose and results of this Regulatory Impact Analysis
(RIA). This RIA estimates the human health and welfare benefits and costs of attaining several
particulate matter (PM) National Ambient Air Quality Standards (NAAQS) nationwide. According
to the Clean Air Act ("Act"), the EPA must use health-based criteria in setting the NAAQS and
cannot consider estimates of compliance cost. The EPA is producing this RIA both to provide the
public a sense of the benefits and costs of meeting a new NAAQS and to meet the requirements
of Executive Orders 12866 and 13563.
1.2 Background
1.2.1 NAAQS
Two sections of the Clean Air Act ("the Act") govern the establishment and revision of
NAAQS. Section 108 (42 U.S.C. 7408) directs the Administrator to identify pollutants that "may
reasonably be anticipated to endanger public health or welfare" and to issue air quality criteria
for them. These air quality criteria are intended to "accurately reflect the latest scientific
knowledge useful in indicating the kind and extent of all identifiable effects on public health or
welfare which may be expected from the presence of [a] pollutant in the ambient air." PM is
one of six pollutants for which the EPA has developed air quality criteria.
Section 109 (42 U.S.C. 7409) directs the Administrator to propose and promulgate
"primary" and "secondary" NAAQS for pollutants identified under section 108. Section
109(b)(l) defines a primary standard as "the attainment and maintenance of which in the
judgment of the Administrator, based on [the] criteria and allowing an adequate margin of
safety, [are] requisite to protect the public health." A secondary standard, as defined in section
109(b)(2), must "specify a level of air quality the attainment and maintenance of which in the
judgment of the Administrator, based on [the] criteria, [are] requisite to protect the public
welfare from any known or anticipated adverse effects associated with the presence of [the]
pollutant in the ambient air." Welfare effects as defined in section 302(h) [42 U.S.C. 7602(h)]
include but are not limited to "effects on soils, water, crops, vegetation, manmade materials,
animals, wildlife, weather, visibility and climate, damage to and deterioration of property, and
hazards to transportation, as well as effects on economic values and on personal comfort and
well-being."
1-1
-------
Section 109(d) of the Act directs the Administrator to review existing criteria and
standards at 5-year intervals. When warranted by such review, the Administrator is to retain or
revise the NAAQS. After promulgation or revision of the NAAQS, the standards are
implemented by the states.
1.2.2 2006PM NAAQS
In 2006, the EPA's final PM rule established a 24-hour standard of 35 u.g/m3 and
retained the annual standard of 15 u.g/m3. The EPA revised the secondary standards for fine
particles by making them identical in all respects to the primary standards. Following
promulgation of the final rule in 2006, several parties filed petitions for its review. On February
24, 2009, the U.S. Court of Appeals for the District of Columbia Circuit remanded the primary
annual PM2.s NAAQS to the EPA citing that the EPA failed to adequately explain why the
standard provided the requisite protection from both short- and long-term exposures to fine
particles, including protection for at-risk populations. The court remanded the secondary
standards to the EPA citing that the Agency failed to adequately explain why setting the
secondary PM standards identical to the primary standards provided the required protection
for public welfare, including protection from visibility impairment.
1.3 Role of this RIA in the Process of Setting the NAAQS
1.3.1 Legislative Roles
In setting primary ambient air quality standards, the EPA's responsibility under the law
is to establish standards that protect public health, regardless of the costs of implementing a
new standard. The Act requires the EPA, for each criteria pollutant, to set a standard that
protects public health with "an adequate margin of safety." As interpreted by the Agency and
the courts, the Act requires the EPA to create standards based on health considerations only.
The prohibition against the consideration of cost in the setting of the primary air quality
standard, however, does not mean that costs or other economic considerations are
unimportant or should be ignored. The Agency believes that consideration of costs and benefits
is essential to making efficient, cost-effective decisions for implementing these standards. The
impact of cost and efficiency is considered by states during this process, as they decide what
timelines, strategies, and policies make the most sense. This RIA is intended to inform the
public about the potential costs and benefits that may result when new standards are
implemented, but it is not relevant to establishing the standards themselves.
1-2
-------
1.3.2 Role of Statutory and Executive Orders
This RIA is separate from the NAAQS decision-making process, but several statutes and
executive orders still apply to any public documentation. The analysis required by these
statutes and executive orders is presented in Chapter 9.
The EPA presents this RIA pursuant to Executive Orders 12866 and 13563 and the
guidelines of Office of Management and Budget (OMB) Circular A-4.1 These documents present
guidelines for the EPA to assess the benefits and costs of the selected regulatory option as well
as more and less stringent options than those proposed or selected. In concordance with these
guidelines, the RIA also analyzes the benefits and costs of alternative combinations of primary
PM2.s standards, one combination that is more stringent than the existing standards, but less
stringent than the proposed standards and another combination that is more stringent than the
proposed standards (see Section 1.4.2).
In the current PM NAAQS review, the EPA is proposing to revise the level of the primary
annual PM2.s standard within the range of 12 to 13 u.g/m3 in conjunction with retaining the level
of the 24-hour standard at 35 u.g/m3 (denoted 12/35 and 13/35). In addition to the range of
12/35 to 13/35, the RIA also analyzes the benefits and costs of incremental control strategies
for two other alternative standards (11/35 and 11/30). In analyzing the current 15/35 standard
(baseline), the EPA determined that all counties would meet the 14/35 standard concurrently
with meeting the existing 15/35 standard at no additional cost. Consequently, no incremental
costs or benefits are associated with 14/35; thus, no analysis of 14/35 is presented.
Benefit and cost estimates provided in the RIA are not additive to benefits and costs
from other regulations, and, further, the costs and benefits identified in this RIA will not be
realized until specific controls are mandated by State Implementation Plans (SIPs) or other
federal regulations.
1.3.3 The Need for National Ambient Air Quality Standards
OMB Circular A-4 indicates that one of the reasons a regulation such as the NAAQS may
be issued is to address existing "externalities." An externality occurs when one party's actions
impose uncompensated benefits or costs on another party. Environmental problems are a
classic case of an externality. Setting primary and secondary air quality standards is one way the
1 U.S. Office of Management and Budget. Circular A-4, September 17, 2003, available at
.
1-3
-------
government can address an externality and thereby increase air quality and improve overall
public health and welfare.
1.3.4 Illustrative Nature of the Analysis
This NAAQS RIA is an illustrative analysis that provides useful insights into a limited
number of emissions control scenarios that states might implement to achieve revised NAAQS.
Because states are ultimately responsible for implementing strategies to meet any revised
standard, the control scenarios in this RIA are necessarily hypothetical in nature. Important
uncertainties and limitations are documented in the relevant portions of the analysis.
The illustrative goals of this RIA are somewhat different from other EPA analyses of
national rules, or the implementation plans states develop, and the distinctions are worth brief
mention. This RIA does not assess the regulatory impact of an EPA-prescribed national rule, nor
does it attempt to model the specific actions that any state would take to implement a revised
standard. This analysis attempts to estimate the costs and human and welfare benefits of cost-
effective implementation strategies that might be undertaken to achieve national attainment of
new standards. These hypothetical strategies represent a scenario where states use one set of
cost-effective controls to attain a revised NAAQS. Because states—not the EPA—will implement
any revised NAAQS, they will ultimately determine appropriate emissions control scenarios.
SIPs would likely vary from the EPA's estimates due to differences in the data and assumptions
that states use to develop these plans.
The illustrative attainment scenarios presented in this RIA were constructed with the
understanding that there are inherent uncertainties in projecting emissions and controls.
1.4 Overview and Design of the RIA
The RIA evaluates the costs and benefits of hypothetical national strategies to attain
several alternative PM standards.
1.4.1 Modeling PM2.5 Levels in the Future (Analysis Year = 2020)
A national-scale air quality modeling analysis was performed to estimate future-year
annual and 24-hour PM2.s concentrations and light extinction for the future year of 2020. Air
quality ratios were then developed using model responsiveness to emissions changes between
a recent year of air quality, 2005, and a future year of air quality, 2020. The air quality ratios
were used to determine potential control scenarios designed to attain the proposed alternative
NAAQS, as well as the costs of attaining these levels. These data were then used to estimate
1-4
-------
how air quality would change under each set of potential control scenarios, and as inputs to the
calculation of expected benefits from the alternative NAAQS considered in this assessment.
1.4.2 Existing and Alternative PM Air Quality Standards
Currently two primary PM2.5 standards provide public health protection from effects
associated with fine particle exposures. The annual standard is set at a level of 15.0 u.g/m3,
based on the 3-year average of annual arithmetic mean PM2.5 concentrations. The 24-hour
standard is set at a level of 35 u.g/m3, based on the 3-year average of the 98th percentile of
24-hour PM2.5 concentrations. In the RIA, the current suite of primary PM2.5 standards,
including both annual and 24-hour averaging times, is denoted as 15/35.
In the current PM NAAQS review, the EPA is proposing to revise the level of the primary
annual PM2.5 standard within the range of 12 to 13 u.g/m3 in conjunction with retaining the level
of the 24-hour standard at 35 u.g/m3 (denoted 12/35 and 13/35).
In addition to the range of 12/35 to 13/35, the RIA also analyzes the benefits and costs
of incremental control strategies for two other alternative standards (11/35 and 11/30). The
four alternative standards analyzed are as follows:
• A revised annual standard level of 13 u.g/m3 in conjunction with retaining the
24-hour standard level at 35 u.g/m3 (13/35)
• A revised annual standard level of 12 u.g/m3 in conjunction with retaining the
24-hour standard level at 35 u.g/m3 (12/35)
• A revised annual standard level of 11 u.g/m3 in conjunction with retaining the
24-hour standard level at 35 u.g/m3 (11/35 )
• A revised annual standard level of 11 u.g/m3 in conjunction with a revised 24-hour
standard level at 30 u.g/m3 (11/30 )
In analyzing the current 15/35 standard (baseline), the EPA determined that all counties
would meet the 14/35 standard concurrently with meeting the existing 15/35 standard at no
additional cost. Consequently, no incremental costs or benefits are associated with 14/35; thus,
no analysis of 14/35 is presented in this RIA.
Currently, the existing secondary PM2.5 standards are identical in all respects to the
primary standards. In the current PM NAAQS review, the EPA is proposing to add a distinct
standard for PM2.5 to provide protection from PM-related visibility impairment. Specifically, the
EPA is proposing to establish a separate secondary standard defined in terms of a PM2.5
1-5
-------
visibility index, which would use speciated PM2.5 mass concentrations and relative humidity
data to calculate PM2.5 light extinction, similar to the Regional Haze Program; a 24-hour
averaging time; a 90th percentile form; and a level of either 30 deciviews (dv) or 28 dv. Based
on the air quality analysis conducted for the primary PM2.5 standard, all monitored areas are
estimated to be in attainment with both proposed secondary standard levels in 2020, assuming
full attainment of the primary PM2.5 standard. For the two optional levels proposed for the
secondary standard, no additional costs or benefits will be realized beyond those quantified for
meeting the primary PM2.5 standard in this RIA.
With regard to the primary and secondary standards for particles less than or equal to
10 u.m in diameter (PMi0), the EPA is proposing to retain the current primary and secondary
24-hour PMio standards. Both standards are the same. The current primary and secondary
24-hour standards are set at a level of 150 u.g/m3, not to be exceeded more than once per year
on average over 3 years (EPA, 1997)2. Since the benefit cost analysis of the alternative PMi0
standards was conducted when the standard was selected, this RIA does not repeat that
analysis here.
1.4.3 Benefits Analysis Approach
The EPA estimated human health (e.g., mortality and morbidity effects) under full
attainment of several alternative PM standards. We considered an array of health impacts
attributable to changes in PM2.5. Even though the alternative primary standards are designed to
protect against adverse effects to human health, the emission reductions have welfare
co-benefits in addition to the direct human health benefits. The term welfare co-benefits covers
both environmental and societal benefits of reducing pollution, such as reductions in visibility
impairment, materials damage, and ecosystem damage. Despite our attempts to quantify and
monetize as many of the benefits as possible, many welfare benefits are not quantified or
monetized.
1.4.4 Costs Analysis Approach
The EPA estimated total costs under partial and full attainment of several alternative
PM standards. The engineering costs generally include the costs of purchasing, installing, and
operating the referenced control technologies. The technologies and control strategies selected
for analysis are illustrative of one way in which nonattainment areas could meet a revised
2 U.S. Environmental Protection Agency. 1997. Regulatory Impact Analyses for the Particulate Matter and Ozone
National Ambient Air Quality Standards and Proposed Regional Haze Rule. Available at:
http://www.epa.gov/ttn/oarpg/naaqsfin/ria.html.
1-6
-------
standard. There are numerous ways to construct and evaluate potential control programs that
would bring areas into attainment with alternative standards, and the EPA anticipates that
state and local governments will consider programs that are best suited for local conditions.
The partial attainment cost analysis reflects the costs associated with known controls. Costs for
full attainment include estimates for the engineering costs of the additional tons of emissions
reductions that are needed beyond identified controls, referred to as extrapolated costs. The
EPA recognizes that the extrapolated portion of the engineering cost estimates reflects
substantial uncertainty about which sectors, and which technologies, might become available
for cost-effective application in the future.
1.5 Organization of this Regulatory Impact Analysis
This RIA includes the following 11 chapters:
• Chapter 1: Introduction and Background. This chapter introduces the purpose of the
RIA.
• Chapter 2: Defining the PM2.s Air Quality Problem. This chapter characterizes the
nature, scope, and magnitude of the current-year PM2.s problem.
• Chapter 3: Air Quality Modeling and Analysis. The data, tools, and methodology used
for the air quality modeling are described in this chapter, as well as the post-
processing techniques used to produce a number of air quality metrics for input into
the analysis of costs and benefits.
• Chapter 4: Control Strategies. This chapter presents the hypothetical control
strategies, the geographic areas where controls were applied, and the results of the
modeling that predicted PM2.s concentrations in 2020 after applying the control
strategies.
• Chapter 5: Health Benefits Analysis Approach and Results. This chapter quantifies the
health-related benefits of the PM2.5-related air quality improvements associated
with several alternative standards.
• Chapter 6: Welfare Benefits Analysis Approach. This chapter quantifies and
monetizes selected other welfare effects, including changes in visibility, materials
damage, ecological effects from PM deposition, ecological effects from nitrogen and
sulfur emissions, vegetation effects from ozone exposure, ecological effects from
mercury deposition, and climate effects.
• Chapter 7: Engineering Cost Analysis. This chapter summarizes the data sources and
methodology used to estimate the engineering costs of partial and full attainment of
several alternative standards.
1-7
-------
Chapter 8: Comparison of Benefits and Costs. This chapter compares estimates of the
total benefits with total costs and summarizes the net benefits of several alternative
standards.
Chapter 9: Statutory and Executive Order Impact Analyses. This chapter summarizes
the Statutory and Executive Order impact analyses.
Chapter 10: Secondary Standards Analysis. This chapter contains an evaluation of the
regulatory impacts associated with a distinct secondary NAAQS for PM2.5.
Chapter 11: Economic Impacts—Employment. This chapter provides a qualitative
discussion of employment impacts of air quality regulations.
1-8
-------
CHAPTER 2
DEFINING THE PM AIR QUALITY PROBLEM
2.1 Synopsis
This chapter characterizes the nature, scope and magnitude of the current year PM
problem. It includes 1) a summary of the spatial and temporal distribution of PM2.5 and the
likely origin from direct emissions or atmospheric transformations of gaseous precursors;
2) discussion of what visibility is and how it is calculated from measured concentrations and
meteorological values; and 3) current year design values for PM2.5 and visibility.
2.2 Particulate Matter (PM) Properties
Particulate matter (PM) is a highly complex mixture of solid particles and liquid droplets
distributed among numerous atmospheric gases which interact with solid and liquid phases.
Particles range in size from those smaller than 1 nanometer (10~9 meter) to over 100
micrometer (u.m, or 10"6 meter) in diameter (for reference, a typical strand of human hair is 70
u.m in diameter and a grain of salt is about 100 u.m). Atmospheric particles can be grouped into
several classes according to their aerodynamic and physical sizes, including ultrafine particles
(<0.1 u.m), accumulation mode or 'fine' particles (0.1 to ~3 u.m), and coarse particles (>1 u.m).
For regulatory purposes, fine particles are measured as PM2.s and inhalable or thoracic coarse
particles are measured as PMi0-2.s, corresponding to their size (diameter) range in micrometers
and referring to total particle mass under 2.5 and between 2.5 and 10 micrometers,
respectively. The EPA currently has standards that measure PM2.s and PMi0.
Particles span many sizes and shapes and consist of hundreds of different chemicals.
Particles are emitted directly from sources and are also formed through atmospheric chemical
reactions; the former are often referred to as "primary" particles, and the latter as "secondary"
particles. Particle pollution also varies by time of year and location and is affected by several
weather-related factors, such as temperature, clouds, humidity, and wind. A further layer of
complexity comes from particles' ability to shift between solid/liquid and gaseous phases,
which is influenced by concentration and meteorology, especially temperature.
Particles are made up of different chemical components. The major chemical
components include carbonaceous materials (carbon soot and organic compounds), and
inorganic compounds including, sulfate and nitrate compounds that usually include ammonium,
and a mix of substances often apportioned to crustal materials such as soil and ash. As
mentioned above, particulate matter includes both "primary" PM, which is directly emitted into
the air, and "secondary" PM, which forms indirectly from emissions from fuel combustion and
2-1
-------
other sources. Primary PM consists of carbonaceous materials (soot and accompanying
organics) and includes:
• Elemental carbon, organic carbon, and crustal material directly emitted from cars,
trucks, heavy equipment, forest fires, some industrial processes and burning waste.
• Both combustion and process related fine metals and larger crustal material from
unpaved roads, stone crushing, construction sites, and metallurgical operations.
Secondary PM forms in the atmosphere from gases. Some of these reactions require
sunlight and/or water vapor. Secondary PM includes:
• Sulfates formed from sulfur dioxide (S02) emissions from power plants and industrial
facilities;
• Nitrates formed from nitrogen oxide (NOX) emissions from cars, trucks, industrial
facilities, and power plants; and
• Ammonium formed from ammonia (NH3) emissions from gas-powered vehicles and
fertilizer and animal feed operations. These contribute to the formation of sulfates
and nitrates that exist in the atmosphere as ammonium sulfate and ammonium
nitrate.1
• Organic carbon (OC) formed from reactive organic gas emissions, including volatile
organic compounds (VOCs), from cars, trucks, industrial facilities, forest fires, and
biogenic sources such as trees.1
As described above, organic carbon has both primary and secondary components. The
percentage contribution to total OC from directly emitted OC versus secondarily formed OC
varies based on location. In an urban area, near direct sources of OC such as cars, trucks, and
industrial sources, the percentage of primary OC may dominate, whereas, in a rural area with
more biogenic sources, OC may be mostly secondarily formed. In addition, emissions from
sources such as power plants and industrial facilities may have small amounts of directly
emitted PM2.5 speciated into sulfate. Figure 2-1 (EPA, 2006) shows, in detail, the sources
contributing to directly emitted PM2.5 and PMio, as well as PM precursors: S02, NOX, NH3, and
VOC.
1 Direct NH3 and VOC emissions are not controlled as part of the control strategy analysis. Emissions of PM2.5, NOX
and SO2 are controlled in the control strategies, for a complete discussion please refer to Chapter 4.
2-2
-------
n,
MK,
0*
CMC* W»d
•d Z* '• p*
\\
^-
0
NO,
MMT|
NH, |38 MMT)
VOC(tfl.6MMT»
Figure 2-1. Detailed Source Categorization of Anthropogenic Emissions of Primary PM2.s,
PMio and Gaseous Precursor Species SO2, NOX, NH3 and VOCs for 2002 in Units of Million
Metric Tons (MMT). EGUs = Electricity Generating Units
Source: U.S. EPA (2006)
2-3
-------
2.2.1 PM2.5
"Fine particles" or PM2.5 are particles with diameters that are less than 2.5 micrometers.
As discussed above, these particles are composed of both primary (derived directly from
emissions) and secondary (derived from atmospheric reactions involving gaseous precursors)
components.
2.2.1.1 Geographical Scale and Transport
Both local and regional sources contribute to particle pollution. Fine particles can be
transported long distances by wind and weather and can be found in the air thousands of miles
from where they formed. Nitrates and sulfates formed from NOX and S02 are generally
transported over wide areas leading to substantial background contributions in urban areas.
Organic carbon, which has both a primary and secondary component, can also be transported
but to a far lesser degree. In general, higher concentrations of elemental carbon and crustal
matter are found closest to the sources of these emissions.
Figure 2-2 shows how much of the PM2.5 mass can be attributed to local versus regional
sources for 13 selected urban areas (EPA, 2004).2 In each of these urban areas, monitoring sites
were paired with nearby rural sites. When the average rural concentration is subtracted from
the measured urban concentration, the estimated local and regional contributions become
apparent. We observe a large urban excess across the U.S. for most PM2.5 species but especially
for total carbon mass with Fresno, CA having the highest observed measure. Larger urban
excess of nitrates is seen in the western U.S. with Fresno, CA and Salt Lake City, UT significantly
higher than all other areas. These results indicate that local sources of these pollutants are
indeed contributing to the PM2.5 air quality problem in these areas. As expected for a
predominately regional pollutant, only a modest urban excess is observed for sulfates.
In the East, regional pollution contributes to more than half of total PM2.5
concentrations. Rural background PM2.5 concentrations are high in the East and are somewhat
uniform over large geographic areas. These regional concentrations come from emission
sources such as power plants, natural sources, and urban pollution and can be transported
hundreds of miles and reflect to some extent the denser clustering of urban areas in the East as
compared to the West. In the West, much of the measured PM2.5 concentrations tend to be
local in nature. These concentrations come from emission sources such as wood combustion
and mobile sources. In general, these data indicate that reducing regional S02 and local sources
2 The measured PM2.5 concentration is not necessarily the maximum for each urban area.
2-4
-------
of carbon in the East, and local sources of nitrate and carbon in the West will be most effective
in reducing PM2.5 concentrations.
N.Tl (
Amuai Average Concentration
at NrtrAtn. pgffi
Hwwttrk uv
:
Carbon
Figure 2-2. Regional and Local Contributions to Annual Average PM2.s by Particulate SO42 ,
Nitrate and Total Carbon (i.e., organic plus EC) for Select Urban Areas Based on Paired 2000-
2004 IMPROVE3 and CSNb Monitoring Sites
Interagency Monitoring of Protected Visual Environments (IMPROVE) http://vista.cira.colostate.edu/improve
Chemical Speciation Network (CSN)
2-5
-------
2.2.1.2 Regional and Seasonal Patterns
The chemical makeup of particles varies across the United States, as illustrated in
Figure 2-3. For example, the higher regional emissions of S02 in the East result in higher
absolute and relative amounts of sulfates as compared to the western U.S. Fine particles in
southern California generally contain more nitrates than other areas of the country. Carbon is a
substantial component of fine particles everywhere.
Cool
Warm
20-
I"
112
I *
* 4
Knrrr,v«.t
20
ta-
12-
B 4
02 03 04 OS Ofi 07 C2 03 04 M 06 Or
North Central
02 CG 04 03 06 37 02 03 M OS 06 07
Midwest
20-
10
02 03 04 OS 08 OT C7 03 04 Of
Cool
Warm
Northeast
CO 03 0* Q5 06 07 02 03 M 06 G* 07
Southeast
20-
020304090007 02 C3 M 06 06 07
Southern California
2Q
02 m o« as ofi 07 02 03 04 05 OB or
NOTM**
Figure 2-3. Regional and Seasonal Trends in Annual PM2.5 Composition from 2002 to 2007
Derived Using the SANDWICH Method. Data from the 42 monitoring locations shown on the
map were stratified by region and season including cool months (October-April) and warm
months (May-September)
Fine particles can also have a seasonal pattern. As shown in Figure 2-3, PM2.5 values in
the eastern half of the United States are typically higher in warmer weather when
2-6
-------
meteorological conditions are more favorable for the formation and build up of sulfates from
higher sulfur dioxide (S02) emissions from power plants in that region. Fine particle
concentrations tend to be higher in the cooler calendar months in urban areas in the West, in
part because fine particle nitrates and carbonaceous particles are more readily formed in cooler
weather, and wood stove and fireplace use increases direct emissions of carbon.
2.2.1.3 Composition of PM2.s as Measured by the Federal Reference Method
The speciation measurements in the preceding analyses represented data from EPA's
Speciation Trends Network, along with adjustments to reflect the fine particle mass associated
with these ambient measurements. In order to more accurately predict the change in PM2.5
design values for particular emission control scenarios, EPA characterizes the composition of
PM2.5 as measured by the Federal Reference Method (FRM). The current PM2.5 FRM does not
capture all ambient particles measured by speciation samplers as presented in the previous
sections. The FRM-measured fine particle mass reflects losses of ammonium nitrate (NH4N03)
and other semi-volatile organic compounds (SVOCs; negative artifacts). It also includes particle-
bound water (PBW) associated with hygroscopic species (positive artifacts) (Frank, 2006).
Comparison of FRM and collocated speciation sampler N03" values in Table 2-1 show that
annual average N03 retention in FRM samples for six cities varies from 15% in Birmingham to
76% in Chicago, with an annual average loss of 1 u.g/m3. The volatilization is a function of
temperature and relative humidity (RH), with more loss at higher temperatures and lower RH.
Accordingly, nitrate is mostly retained during the cold winter days, while little may be retained
during the hot summer days.
PM2.5 FRM measurements also include water associated with hygroscopic aerosol. This
is because the method derives fine particle concentrations from sampled mass equilibrated at
20-23 °C and 30-40% RH. At these conditions, the hygroscopic aerosol collected at more humid
environments will retain their particle-bound water. The water content is higher for more acidic
and sulfate-dominated aerosols. Combining the effects of reduced nitrate and hydrated aerosol
causes the estimated nitrate and sulfate FRM mass to differ from the measured ions simply
expressed as dry ammonium nitrate and ammonium sulfate. The composition of FRM mass is
denoted as SANDWICH based on the Sulfate, Adjusted Nitrate Derived Water and Inferred
Carbon approach from which they are derived. The PM2.5 mass estimated from speciated
measurements of fine particles is termed Reconstructed Fine Mass (RCFM). The application of
SANDWICH adjustments to speciation measurements at six sites is illustrated in Table 2-1 and
Figure 2-4. EPAs modeling incorporates these SANDWICH adjustments in the Model Attainment
Test Software (MATS) (Abt, 2010).
2-7
-------
Table 2-1. Annual Average FRM and CSN PM2.5 NO3 and NH4NO3 Concentrations at Six Sites
during 2003
Sampling Site
Location
Mayville, Wl
Chicago, IL
Indianapolis, IN
Cleveland, OH
Bronx, NY
Birmingham, AL
No. of
Observations
100
76
92
90
108
113
FRM
Mass
9.8
14.4
14.8
16.8
15.0
17.0
CSNa
2.5
2.8
2.5
2.9
2.4
1.1
N03-(u
FRMb
1.5
2.1
1.3
1.7
1.1
0.2
g/m3)
Difference
(CSN - FRM)
1.0
0.7
1.3
1.2
1.3
0.9
NH4NO3
CSN
3.2
3.7
3.2
3.7
3.1
1.4
(ug/m3)
FRM
1.9
2.8
1.6
2.2
1.4
0.2
Percent of
NH4NO3 in PM2.5
FRM Mass
CSN FRM
33% 19%
25% 19%
22% 11%
22% 13%
21% 9%
8% 1%
On denuded nylon-membrane filters for al sites except for Chicago, where denuded Teflon-membrane followed
by nylon filters were used.
On undenuded Teflon-membrane filters.
SANDWICH
(FRM)
SufotaMaw
NifaM MMD
TGM
CriiBlol
PflHiVfl
Figure 2-4. RCFM (left) versus SANDWICH (right) Pie Charts Comparing the Ambient and
PM2.5 FRM Reconstructed Mass Protocols on an Annual Average Basis3
a Estimated NH4* and PBW for SANDWICH are included with their respective sulfate and nitrate mass slices.
Circles are scaled in proportion to PM2.5 FRM mass.
2-8
-------
2.2.1.4 2004-2006 Design Values
The annual and 24-hour PM2.5 design values were calculated using 2003-2007 FRM
24-hour average PM2.5 concentration measurements and consistent with CFR Part 50.3
Figures 2-5 and 2-6 show the county-level maximum values for both the annual and 24-hour
standards, respectively. For the most part, counties in the center of the U.S. have PM2.s design
values that are above both 11 u.g/m3 for the annual standard and 30 u.g/m3 for the 24-hour
standard. In the East, the counties above the current NAAQS (i.e., 15 u.g/m3 annual and 35
u.g/m3 24-hour standards) are similar. In the West, there are fewer counties above the annual
level of 15 u.g/m3 than exceed the 24-hour standard of 35 u.g/m3.
• - ^<- ,1
I ' * ttfNft
IS iJ l'\ . *-_:::
Lag/end
98 wuitim CKcaad 1 5 uj'ttiS
"S .uhiinrri i Kill*" me- I H i»; :iv
|
I I JW tuirttas "
-------
E79 Limit; wftti nurtlws hM ?H 2 & 'IJJ« JKIJT valu«
Figure 2-6. Maximum County-level PM2.5 24-hour Design Values Calculated Using 2003-2007
FRM 24-hr Average PM2.s Measurements
2.2.2 Visibility
Air pollution can affect light extinction, a measure of how much the components of the
atmosphere scatter and absorb light. More light extinction means that the clarity of visual
images and visual range is reduced, all else held constant. Light extinction is the optical
characteristic of the atmosphere that occurs when light is either scattered or absorbed, which
converts the light to heat. Particulate matter and gases can both scatter and absorb light. Fine
particles with significant light-extinction efficiencies include sulfates, nitrates, organic carbon,
elemental carbon, and soil (Sisler, 1996). The extent to which any amount of light extinction
affects a person's ability to view a scene depends on both scene and light characteristics. For
example, the appearance of a nearby object (e.g., a building) is generally less sensitive to a
change in light extinction than the appearance of a similar object at a greater distance. See
Figure 2-7 for an illustration of the important factors affecting visibility.
2-10
-------
U0it front
cr 9tt«r*d into
wght
forming
•0*1 iott*r*0
out of Mflftt oatft
Figure 2-7. Important Factors Involved in Seeing a Scenic Vista
Source: Malm, 1999.
2.2.2.1 Calculating Visibility
Visibility degradation is often directly proportional to decreases in light transmittal in
the atmosphere. Scattering and absorption by both gases and particles decrease light
transmittance. To quantify changes in visibility, our analysis computes a light-extinction
coefficient, based on the work of Sisler (1996), which shows the total fraction of light that is
decreased per unit distance. This coefficient accounts for the scattering and absorption of light
by both particles and gases, and accounts for the higher extinction efficiency of fine particles
compared to coarse particles. Fine particles with significant light-extinction efficiencies include
sulfates, nitrates, organic carbon, elemental carbon (soot), and soil (Sisler, 1996).
As described in the Policy Assessment Document (EPA, 2011), the formula for total light
extinction (bext) in units of Mm"1 using the original IMPROVE equation is:
bext = 3 x /(RH) x [Sulfate] + 3 x /(RH) x [Nitrate] + 4 x [Organic Mass] +
10 x [Elemental Carbon] + Ix [Fine Soil] + 0.6 x [Coarse Mass] + 10 (2.1)
2-11
-------
where the mass concentrations of the components indicated in brackets are in units of u.g/m3,
and/(RH) is the unitless water growth term that depends on relative humidity. The final term in
the equation is known as the Rayleigh scattering term and accounts for light scattering by the
natural gases in unpolluted air. Since IMPROVE does not include ammonium ion monitoring, the
assumption is made that all sulfate is fully neutralized ammonium sulfate and all nitrate is
assumed to be ammonium nitrate.
Based upon the light-extinction coefficient, a unitless visibility index, called a "deciview,"
can also be calculated using Equation (2.2):
Deciviews = 10 * In (—} = 10 * In f^xt) (2.2)
\VR/ \ 10 J ^ '
where VR denotes visual range (in kilometers) and 3ext denotes light extinction (in Mm"1). The
deciview metric provides a scale for perceived visual changes over the entire range of
conditions, from clear to hazy. Under many scenic conditions, the average person can generally
perceive a change of one deciview. The higher the deciview value, the worse the visibility. Thus,
an improvement in visibility is a decrease in deciview value.
2.2.2.2 Geographical Scale and Variability
Annual average visibility conditions (reflecting light extinction due to both
anthropogenic and non-anthropogenic sources) vary regionally across the U.S. and by season
(U.S. EPA, 2009). Particulate sulfate is the dominant source of regional haze in the eastern U.S.
(>50% of the particulate light extinction) and an important contributor to haze elsewhere in the
country (>20% of particulate light extinction) (U.S. EPA, 2009). Particulate nitrate is an
important contributor to light extinction in California and the upper Midwestern U.S.,
particularly during winter (U.S. EPA, 2009). Smoke plumes from large wildfires dominate many
of the worst haze periods in the western U.S., while Asian dust only caused a few of the worst
haze episodes, primarily in the more northerly regions of the west (U.S. EPA, 2009). Higher
visibility impairment levels in the East are due to generally higher concentrations of fine
particles, particularly sulfates, and higher average relative humidity levels (U.S. EPA, 2009).
Humidity increases visibility impairment because some particles such as ammonium sulfate and
ammonium nitrate absorb water and form droplets that become larger when relative humidity
increases, thus resulting in increased light scattering (U.S. EPA, 2009).
Figure 2-8 shows the average trends in visual ranges at select monitors in the eastern
and western areas of the U.S. since 1992 using data from the IMPROVE monitoring network
(U.S. EPA (2008); IMPROVE (2010)). Because trends in haze are closely associated with trends in
2-12
-------
particulate sulfate and nitrate due to the simple relationship between their concentration and
light extinction, visibility trends have improved as emissions of S02 and NOX have decreased
overtime due to air pollution regulations such as the Acid Rain Program (U.S. EPA, 2009). For
example, Figure 2-8 shows that visual range increased nearly 50% in the eastern U.S. since
19924. While visibility trends have improved in most Class 1 areas5, the recent data show that
these areas continue to suffer from visibility impairment (U.S. EPA, 2009). Calculated from light
extinction efficiencies from Trijonis et al. (1987, 1988), annual average visual range under
natural conditions in the East is estimated to be 150 km ± 45 km (i.e., 65 to 120 miles) and 230
km ± 35 km (i.e., 120 to 165 miles) in the West (Irving, 1991).
I
A. Wislarn U S
B. fcaslern U.S.
Bwl viability rijrss
Wocst vlsiHilty days
IB Trt
'86 W TK) TJZ
W
: 3fl n-Kmfcflriig siles h Ete ««slem US Wd tl monitoring »(« in ihe ttespr US wlh
*
'V iui -anpt-s are calcutalid From Ifie measurM \BIK\S ol dirrerer: aroonir: j wthin aubnrna parties and V .
" light eifinciiai
Data stuirtt: JUMPHOUf. 20 W
^^B
^
Figure 2-8. Visibility in Selected National Parks and Wilderness Areas in the U.S.,
1992-2008a'b
Source: U.S. EPA (2008) updated, IMPROVE (2010).
2.2.2.3 2004-2006 Design Values
The secondary PM2.5lNAAQS standard consists of three parts: a level, averaging period,
and form. EPA proposes using a 3-year average, 90th percentile form for the standard,
calculated using 24-hr speciated PM2.5 measurements. EPA analyzed two proposed levels of 30
In Figure 2-8, the "best days" are defined as the best 20% of days, the "mid-range days" are defined as the middle
20%, and the "worst days" are defined as the worst 20% of days (IMPROVE, 2010).
' Class I areas are areas of special national or regional natural, scenic, recreational, or historic value for which the
Prevention of Significant Deterioration (PSD) regulations provide special protection.
2-13
-------
dv and 28 dv, as well as a more stringent standard of 25 dv. The ambient design values analyzed
in this RIA are based on measured 24-hour PM2.5 speciation data from 2004-20066. These data
were calculated as described in the Policy Assessment Document (EPA, 2011) and provided in
Chapter 13. Figure 2-9 shows the county-level maximum design values. 20 counties were above
30 dv and 90 counties were above 28 dv. For the more stringent proposed level, 77 additional
counties were above 25 dv. The large majority of these counties are located in the East.
Figure 2-9. Maximum County-level Visibility Design Values Calculated Using 2004-2006 24-hr
Average Speciated PM2.s Measured Concentrations
2.3 References
Abt Associates, 2010. User's Guide: Modeled Attainment Test Software.
http://www.epa.gov/scram001/modelingapps_mats.htm
' These years of ambient measurements were selected since they frame the air quality model year of 2005. As
discussed in Chapter 3, it is most appropriate to select ambient measurement years that include the model year
to allow for a more true projection of future year air quality using the air quality model.
2-14
-------
Frank, N.H., Retained Nitrate, Hydrated Sulfates, and Carbonaceous Mass in Federal Reference
Method Fine Particulate Matter for Six Eastern U.S. Cities, J. Air & Waste Manage. Assoc.
2006, 56, 500-511.
IMPROVE (Interagency Monitoring of Protected Visual Environments). 2010. 1992-2008 data
from the IMPROVE network based on the "New IMPROVE algorithm" (updated
November, 2008). Available on the Internet at
Irving, Patricia M., e.d., 1991. Acid Deposition: State of Science and Technology, Volume III,
Terrestrial, Materials, Health, and Visibility Effects, The U.S. National Acid Precipitation
Assessment Program, Chapter 24, page 24-76.
Malm, WC. 1999. Introduction to visibility. Colorado State University, Fort Collins, CO, USA,
1983, Revised edition, 1999. Available on the Internet at
.
Sisler, J.F. 1996. Spatial and Seasonal Patterns and Long Term Variability of the Composition of
the Haze in the United States: An Analysis of Data from the IMPROVE Network. Fort
Collins, CO: Cooperative Institute for Research in the Atmosphere, Colorado State
University.
Trijonis, J.C. et al. 1987. Preliminary extinction budget results from the RESOLVE program, pp.
872-883. In: P.J. Bhardwaja, et al. Visibility Protection Research and Policy Aspects. Air
Pollution Control Assoc., Pittsburgh, PA.
Trijonis, J.C. et al. 1988. RESOLVE Project Final Report: Visibility conditions and Causes of
Visibility Degradation in the Mojave Desert of California. NWC TP #6869. Naval Weapons
Center, China Lake, CA.
U.S. Environmental Protection Agency (EPA). 2004. Air quality criteria for particulate matter.
U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA/600/P-
99/002aF-bF.
U.S. Environmental Protection Agency (EPA). 2006. 2002 National Emissions Inventory Data and
Documentation. Located online at:
http://www.epa.gov/ttnchiel/net/2002inventory.html.
U.S. Environmental Protection Agency (EPA). 2008. National Air Quality Status and Trends
Through 2007. U.S. Environmental Protection Agency. Research Triangle Park. EPA-
454/R-08-006
U.S. Environmental Protection Agency (EPA). 2009. Integrated Science Assessment for
Particulate Matter. U.S. Environmental Protection Agency. Research Triangle Park.
EPA/600/R-08/139F
2-15
-------
U.S. Environmental Protection Agency (EPA). 2011. Policy Assessment for the Review of the
Particulate Matter National Ambient Air Quality Standards. Office of Air Quality Planning
and Standards, Research Triangle Park, NC.
2-16
-------
CHAPTER 3
AIR QUALITY MODELING AND ANALYSIS
3.1 Synopsis
In order to evaluate the health and environmental impacts of trying to reach the
alterative primary and secondary PM standards proposed in this RIA, it was necessary to use
models to predict concentrations in the future. The data, tools and methodology used for
projecting future-year air quality are described in this chapter, as well as the post-processing
techniques used to produce a number of air quality metrics for input into the analysis of costs
and benefits.
3.2 Modeling PM2.5 Levels in the Future
A national scale air quality modeling analysis was performed to estimate PM2.5
concentrations for the annual and 24-hour primary standards and light extinction for the future
year of 2020.1 Air quality ratios were then developed using model responsiveness to emissions
changes between a recent year of air quality, 2005, and a future year of air quality, 2020. The
air quality ratios were used to determine potential control scenarios designed to attain the
proposed alternative NAAQS, as well as the costs of attaining these levels. These data were
then used to estimate how air quality would change under each set of potential control
scenarios, and as inputs to the calculation of expected benefits from the alternative NAAQS
considered in this assessment.
3.2.1 Air Quality Modeling Platform
The 2005-based Community Multi-scale Air Quality (CMAQ) modeling platform was used
as the tool to project future-year air quality for 2020 and to estimate the costs and benefits for
attaining the current and proposed alternative NAAQS considered in this assessment. In
addition to the CMAQ model, the modeling platform includes the emissions, meteorology, and
initial and boundary condition data which are inputs to this model.
The CMAQ model is a three-dimensional grid-based Eulerian air quality model designed
to estimate the formation and fate of oxidant precursors, primary and secondary particulate
matter concentrations and deposition over regional and urban spatial scales (e.g., over the
contiguous U.S.) (Appel et al., 2008; Appel et al., 2007; Byun and Schere, 2006). Consideration
1 As described in more detail in this chapter, the future-year emissions inventory used in the air quality modeling
analysis is a combination of emissions sectors projected to 2017 and 2020. We have chosen to label the future-
year of modeling as "2020" because the EGU sector, which is projected to 2020, is of significant importance to
the concentrations of PM2.5 in the U.S.
3-1
-------
of the different processes (e.g., transport and deposition) that affect primary (directly emitted)
and secondary (formed by atmospheric processes) PM at the regional scale in different
locations is fundamental to understanding and assessing the effects of pollution control
measures that affect PM, ozone and deposition of pollutants to the surface. Because it accounts
for spatial and temporal variations as well as differences in the reactivity of emissions, CMAQ is
useful for evaluating the impacts of the control strategies on PM2.s concentrations. Version
4.7.1 of CMAQ was employed for this RIA modeling, as described in the Air Quality Modeling
Technical Support Document (EPA, 2011b).
3.2.1.1 Air Quality Modeling Domain
Figure 3-1 shows the modeling domains that were used as a part of this analysis. The
geographic specifications for these domains are provided in Table 3-1. All three modeling
domains contain 14 vertical layers with a top at about 16,200 meters, or 100 millibars (mb).
Two domains with 12 km horizontal resolution were used for modeling the 2005 base year and
2020 control strategy scenarios. These domains are labeled as the East and West 12 km
domains in Figure 3-1. Simulations for the 36 km domain were only used to provide initial and
boundary concentrations for the 12 km domains.
Figure 3-1. Map of the CMAQ Modeling Domains Used for PM NAAQS RIA
3-2
-------
Table 3-1. Geographic Specifications of Modeling Domains
36 km Domain
(148x112 Grid Cells)
Longitude Latitude
SW -121.77 18.17
NE -58.54 52.41
12 km East Domain
(279 x 240 Grid Cells)
Longitude Latitude
SW -106.79 24.99
NE -65.32 47.63
12 km West Domain
(213x192 Grid Cells)
Longitude Latitude
SW -121.65 28.29
NE -94.94 51.91
The model produces gridded air quality concentrations on an hourly basis for the entire
modeling domain. For this analysis, predictions from the East domain were used to provide
data for all areas that are east of approximately 104 degrees longitude. Model predictions from
the West domain were used for all areas west of this longitude.
3.2.1.2 Air Quality Model Inputs
CMAQ requires a variety of input files that contain information pertaining to the
modeling domain and simulation period. These include gridded, hourly emissions estimates and
meteorological data, and initial and boundary conditions. Separate emissions inventories were
prepared for the 2005 base year and the future year of 2020. All other inputs were specified for
the 2005 base year model application and remained unchanged for each future-year modeling
scenario.
CMAQ requires detailed emissions inventories containing temporally allocated (i.e.,
hourly) emissions for each grid-cell in the modeling domain for a large number of chemical
species that act as primary pollutants or precursors to secondary pollutants. The annual
emission inventories, described in Section 3.2.2, were preprocessed into CMAQ-ready inputs
using the SMOKE emissions preprocessing system. Meteorological inputs reflecting 2005
conditions across the contiguous U.S. were derived from Version 5 of the Mesoscale Model
(MM5). These inputs included hourly-varying horizontal wind components (i.e., speed and
direction), temperature, moisture, vertical diffusion rates, and rainfall rates for each grid cell in
each vertical layer. Details of the annual 2005 meteorological modeling are provided in the Air
Quality Modeling Technical Support Document: Final ECU NESHAP (EPA, 2011d).
The lateral boundary and initial species concentrations for the CMAQ simulations using
a 36 km domain are provided by a three dimensional global atmospheric chemistry and
transport model (GEOS-CHEM). The lateral boundary species concentrations varied with height
and time (every 3 hours). These data were used in CMAQ for the 36 km domain. Initial and
3-3
-------
boundary concentrations from the CMAQ 36 km domain were then used to provide initial and
boundary concentrations for CMAQ simulations using the East and West 12 km domains. The
development of model inputs is discussed in greater detail in the Air Quality Modeling Technical
Support Document: Final ECU NESHAP (EPA, 2011d).
3.2.1.3 Air Quality Model Evaluation
An operational model performance evaluation for PM2.s and its related speciated
components (e.g., sulfate, nitrate, elemental carbon, organic carbon) was performed to
estimate the ability of the CMAQ modeling system to replicate 2005 base year concentrations.
This evaluation principally comprises statistical assessments of model predictions versus
observations paired in time and space on an hourly, 24-hour, or weekly basis depending on the
sampling period of measured data. Details on the evaluation methodology and the calculation
of performance statistics are provided in the Air Quality Modeling Technical Support Document:
Final ECU NESHAP (EPA, 2011d). Overall, the model performance statistics for sulfate, nitrate,
organic carbon, and elemental carbon from the CMAQ 2005 simulation are within or close to
the ranges found in other recent applications. These model performance results give us
confidence that our applications of CMAQ using this 2005 modeling platform provide a
scientifically credible approach for assessing PM2.s concentrations for the purposes of the RIA.
3.2.2 Emissions Inventory
The future-year base-case inventory, projected from the 2005 Version 4.3 emissions
modeling platform, is the starting point for the baseline and control strategy for the Proposed
PM NAAQS emissions inventory. The Emissions Modeling for the Final Mercury and Air Toxics
Standard (MATS) TSD (EPA, 2011c) describes in detail the development of the 2005 base year
inventory, the projection methodology, and the controls applied to create the projected
inventory. Note that the referenced Emissions Modeling TSD describes the use of year 2015
emissions for EGUs and 2017 emissions for other sources, while this analysis used 2020
emissions for EGUs.
The ECU projected inventory represents demand growth, fuel resource availability,
generating technology cost and performance, and other economic factors affecting power
sector behavior. It also reflects environmental rules and regulations, consent decrees and
settlements, plant closures, and newly built units for the calendar year 2020. In this analysis,
the projected ECU emissions include the Final MATS policy case announced on December 21,
2011 and the Final Cross-State Air Pollution Rule (CSAPR) issued on July 6, 2011. The ECU
emissions were developed using version 4.10 Final MATS version of the Integrated Planning
3-4
-------
Model (IPM) and documented in detail at http://www.epa.gov/airmarkt/progsregs/epa-
ipm/toxics.html. The IPM is a multiregional, dynamic, deterministic linear programming model
of the U.S. electric power sector. Note that for this analysis, no further ECU control measures
were selected for illustrating attainment of the current and proposed alternative standard
levels, as discussed in Chapter 4, and the ECU emissions are unchanged between the future-
year base-case and control strategies.
The mobile source emissions were projected to 2017 using activity data. These
emissions represent the effects of the Clean Air Nonroad Diesel Rule, the Light-Duty Vehicle
Tier 2 Rule, the Heavy Duty Diesel Rule, and other finalized rules. Table 3-2 provides a
comprehensive list of the rules/control strategies and projection assumptions in the projected
base-case (i.e., reference case) inventory. A full discussion of the future year base inventory is
provided in the Emissions Modeling TSD. The 2017 onroad mobile source emissions were
developed by using the MOtor Vehicle Emission Simulator (MOVES)2 to create emission factors
that were then input to the Sparse Matrix Operator Kernel Emissions system (SMOKE). The
SMOKE-MOVES Integration Tools combined the county and temperature-specific emission
factors with the activity data to compute the actual emissions based on hourly gridded
temperature data.
The future year scenarios include the same year 2006 Canada and year 1999 Mexico
emissions as the 2005 base case. All 2005 and projected base case emissions inventories are
available on the EPA's Emissions Modeling Clearinghouse website at:
http://www.epa.gov/ttn/chief/emch/index.htmltftoxics. The inventories used to support this
analysis can be found under ftp://ftp.epa.gov/Emislnventory/2005v4 3/mats.
2More information is available online at: http://www.epa.gov/otaq/models/moves/index.htm
3-5
-------
Table 3-2. Control Strategies and Growth Assumptions for Creating 2020 Base Case
Emissions Inventories from the 2005 Base Case
Control Strategies and/or Growth Assumptions
(Grouped by Affected Pollutants or Standard and Approach Used to Apply
to the Inventory)
Non-EGU Point (ptnonipm) Controls
MACT rules, national, VOC: national applied by SCC, MACT
Boat Manufacturing
Wood Building Products Surface Coating
Generic MACT II: Spandex Production, Ethylene manufacture
Large Appliances
Miscellaneous Organic NESHAP (MON): Alkyd Resins, Chelating Agents,
Explosives, Phthalate Plasticizers, Polyester Resins, Polymerized Vinylidene
Chloride
Reinforced Plastics
Asphalt Processing & Roofing
Iron & Steel Foundries
Metal: Can, Coil
Metal Furniture
Miscellaneous Metal Parts & Products
Municipal Solid Waste Landfills
Paper and Other Web
Plastic Parts
Plywood and Composite Wood Products
Carbon Black Production
Cyanide Chemical Manufacturing
Friction Products Manufacturing
Leather Finishing Operations
Miscellaneous Coating Manufacturing
Organic Liquids Distribution (Non-Gasoline)
Refractory Products Manufacturing
Sites Remediation
Consent decrees on companies (based on information from the Office of
Enforcement and Compliance Assurance— OECA) apportioned to plants
owned/operated by the companies
DOJ Settlements: plant SCC controls for:
Alcoa, TX
Premcor (formerly Motiva), DE
Refinery Consent Decrees: plant/SCC controls
Pollutants
Affected
VOC
VOC, CO, NOX,
PM, SO2
All
NOX, PM, SO2
Approach
or
Reference:
EPA,
2007a
1
2
3
(continued)
3-6
-------
Table 3-2. Control Strategies and Growth Assumptions for Creating 2020 Base Case
Emissions Inventories from the 2005 Base Case (continued)
Control Strategies and/or Growth Assumptions
(Grouped by Affected Pollutants or Standard and Approach Used to Apply
to the Inventory)
Non-EGU Point (ptnonipm) Controls (continued)
Hazardous Waste Combustion
Municipal Waste Combustor Reductions— plant level
Hospital/Medical/lnfectious Waste Incinerator Regulations
Large Municipal Waste Combustors— growth applied to specific plants
MACT rules, plant-level, VOC: Auto Plants
MACT rules, plant-level, PM & SO2: Lime Manufacturing
MACT rules, plant-level, PM: Taconite Ore
Livestock Emissions Growth from year 2002 to year 2017 (some farms in the
point inventory)
NESHAP: Portland Cement (09/09/10)— plant level based on Industrial Sector
Integrated Solutions (ISIS) policy emissions in 2013. The ISIS results are from
the ISIS-Cement model runs for the NESHAP and NSPS analysis of July 28,
2010 and include closures.
New York ozone SIP controls
Additional plant and unit closures provided by state, regional, and the EPA
agencies and additional consent decrees. Includes updates from CSAPR
comments.
Emission reductions resulting from controls put on specific boiler units (not
due to MACT) after 2005, identified through analysis of the control data
gathered from the Information Collection Request (ICR) from the
Industrial/Commercial/lnstitutional Boiler NESHAP.
Reciprocating Internal Combustion Engines (RICE) NESHAP
Ethanol plants that account for increased ethanol production due to RFS2
mandate
State fuel sulfur content rules for fuel oil— effective only in Maine, New
Jersey, and New York
Nonpoint (nonpt sector) Projection Approaches
Municipal Waste Landfills: projection factor of 0.25 applied
Pollutants
Affected
PM
PM
NOX, PM, SO2
All (including Hg)
VOC
PM, SO2
PM
NH3, PM
Hg, NOX, S02, PM,
HCI
VOC, NOX, HAP
VOC
All
NOX, SO2, HCI
NOX, CO, PM, SO2
All
S02
All
Approach
or
Reference:
4
5
EPA, 2005
5
6
7
8
9
10; EPA,
2010
11
12
Section
4.2.13.2
13
14
15
EPA,
2007a
(continued)
3-7
-------
Table 3-2. Control Strategies and Growth Assumptions for Creating 2020 Base Case
Emissions Inventories from the 2005 Base Case (continued)
Control Strategies and/or Growth Assumptions
(Grouped by Affected Pollutants or Standard and Approach Used to Apply
to the Inventory)
Nonpoint (nonpt sector) Projection Approaches (continued)
Livestock Emissions Growth from year 2002 to 2017
New York, Connecticut, and Virginia ozone SIP controls
RICE NESHAP
State fuel sulfur content rules for fuel oil— effective only in Maine, New
Jersey, and New York
Residential Wood Combustion Growth and Change-outs from year 2005 to
2017
Gasoline and diesel fuel Stage II refueling via MOVES2010a month-specific
inventories for 2017 with assumed RFS2 and LDGHG fuels
Portable Fuel Container Mobile Source Air Toxics Rule 2 (MSAT2) inventory
growth and control from year 2005 to 2017
Phase II WRAP 2018 Oil and Gas
2008 Oklahoma and Texas Oil and Gas, and apply year 2017 projections for
TX, and RICE NESHAP controls to Oklahoma emissions.
Pollutants
Affected
NH3, PM
VOC
NOX, CO, VOC,
PM, SO2
S02
All
VOC, Benzene,
Ethanol
VOC
VOC, SO2, NOX,
CO
VOC, SO2, NOX,
CO, PM
Approach
or
Reference:
9
11,16
13
15
17
18
19
EMTSD
EMTSD
Approaches/References—Non-EGU Stationary Sources:
1. Appendix B in the MATS Proposal TSD:
http://www.epa.gov/ttn/chief/emch/toxics/proposed toxics rule appendices.pdf
2. For Alcoa consent decree, used http://cfpub.epa.gov/compliance/cases/index.cfm; for Motiva: used
information sent by State of Delaware
3. Used data provided by the EPA, OAQPS, Sector Policies and Programs Division (SPPD).
4. Obtained from Anne Pope, the US EPA—Hazardous Waste Incinerators criteria and hazardous air pollutant
controls carried over from 2002 Platform, v3.1.
5. Used data provided by the EPA, OAQPS SPPD expert.
6. Percent reductions and plants to receive reductions based on recommendations by rule lead engineer, and
are consistent with the reference: EPA, 2007a
7. Percent reductions recommended are determined from the existing plant estimated baselines and
estimated reductions as shown in the Federal Register Notice for the rule. SO2 percent reduction are
computed by 6,147/30,783 = 20% and PM10 and PM25 reductions are computed by 3,786/13,588 = 28%
8. Same approach as used in the 2006 Clean Air Interstate Rule (CAIR), which estimated reductions of "PM
emissions by 10,538 tpy, a reduction of about 62%." Used same list of plants as were identified based on
tonnage and SCC from CAIR: http://www.envinfo.com/caain/June04updates/tiop fr2.pdf
(continued)
3-8
-------
Table 3-2. Control Strategies and Growth Assumptions for Creating 2020 Base Case
Emissions Inventories from the 2005 Base Case (continued)
Control Strategies and/or Growth Assumptions
(Grouped by Affected Pollutants or Standard and Approach Used to Apply
to the Inventory)
Nonpoint (nonpt sector) Projection Approaches (continued)
Pollutants
Affected
Approach
or
Reference:
Approaches/References—Non-EGU Stationary Sources (continued):
9. Except for dairy cows and turkeys (no growth), based on animal population growth estimates from the US
Department of Agriculture (USDA) and the Food and Agriculture Policy and Research Institute. See
Section 4.2.10.
10. Data files for the cement sector provided by Elineth Torres, the EPA-SPPD, from the analysis done for the
Cement NESHAP: The ISIS documentation and analysis for the cement NESHAP/NSPS is in the docket of that
rulemaking-docket # EPA-HQ-OAR-2002-005. The Cement NESHAP is in the Federal Register: September 9,
2010 (Volume 75, Number 174, Page 54969-55066
11. New York NOX and VOC reductions obtained from Appendix J in NY Department of Environmental
Conservation Implementation Plan for Ozone (February 2008):
http://www.dec.nv.gov/docs/air pdf/NYMASIP7final.pdf.
12. Appendix D of Cross-State Air Pollution Rule:
ftp://ftp.epa.gov/Emislnventorv/2005v4 2/transportrulefinal eitsd appendices 28iun2011.pdf
13. Appendix F in the Proposed (Mercury and Air) Toxics Rule TSD:
http://www.epa.gov/ttn/chief/emch/toxics/proposed toxics rule appendices.pdf
14. The 2008 data used came from Illinois' submittal of 2008 emissions to the NEI.
15. Based on available, enforceable state sulfur rules as of November, 2010:
http://www.ilta.org/LegislativeandRegulatorv/MVNRLM/NEUSASulfur%20Rules 09.2010.pdf,
http://www.mainelegislature.org/legis/bills/bills 124th/billpdfs/SP062701.pdf,
http://switchboard.nrdc.org/blogs/rkassel/governor paterson signs new la.html,
http://green.blogs.nvtimes.com/2010/07/20/new-york-mandates-cleaner-heating-oil/
16. VOC reductions in Connecticut and Virginia obtained from CSAPR comments.
17. Growth and Decline in woodstove types based on industry trade group data, See Section 4.2.11.
18. MOVES (2010a) results for onroad refueling including activity growth from VMT, Stage II control programs
at gasoline stations, and phase in of newer vehicles with onboard Stage II vehicle controls.
http://www.epa.gov/otaq/models/moves/index.htm
19. VOC, benzene, and ethanol emissions for 2017 based on MSAT2 rule and ethanol fuel assumptions (EPA,
2007b)
Onroad Mobile and Nonroad Mobile Controls
(list includes all key mobile control strategies but is not exhaustive)
National Onroad Rules:
Tier 2 Rule: Signature date February, 2000
2007 Onroad Heavy-Duty Rule: February, 2009
Final Mobile Source Air Toxics Rule (MSAT2): February, 2007
Renewable Fuel Standard: March, 2010
Light Duty Greenhouse Gas Rule: May, 2010
Corporate Average Fuel Economy standards for 2008-2011
All
1
(continued)
3-9
-------
Table 3-2. Control Strategies and Growth Assumptions for Creating 2020 Base Case
Emissions Inventories from the 2005 Base Case (continued)
Control Strategies and/or Growth Assumptions
(Grouped by Affected Pollutants or Standard and Approach Used to Apply
to the Inventory)
Onroad Mobile and Nonroad Mobile Controls
(list includes all key mobile control strategies but is not exhaustive)
(continued)
Local Onroad Programs:
National Low Emission Vehicle Program (NLEV): March, 1998
Ozone Transport Commission (OTC) LEV Program: January, 1995
National Nonroad Controls:
Clean Air Nonroad Diesel Final Rule— Tier 4: June, 2004
Control of Emissions from Nonroad Large-Spark Ignition Engines and
Recreational Engines (Marine and Land Based): "Pentathalon Rule":
November, 2002
Clean Bus USA Program: October, 2007
Control of Emissions of Air Pollution from Locomotives and Marine
Compression-Ignition Engines Less than 30 Liters per Cylinder: October, 2008
Locomotive and marine rule (May 6, 2008)
Marine SI rule (October 4, 1996)
Nonroad large SI and recreational engine rule (November 8, 2002)
Nonroad SI rule (October 8, 2008)
Phase 1 nonroad SI rule (July 3, 1995)
Tier 1 nonroad diesel rule (June 17, 2004)
Aircraft (emissions are in the nonEGU point inventory):
Itinerant (ITN) operations at airports to 2017
Locomotives:
Energy Information Administration (EIA) fuel consumption projections for
freight rail
Clean Air Nonroad Diesel Final Rule— Tier 4: June 2004
Locomotive Emissions Final Rulemaking, December 17, 1997
Locomotive rule: April 16, 2008
Control of Emissions of Air Pollution from Locomotives and Marine: May
2008
Pollutants
Affected
voc
All
All
All
Approach
or
Reference:
2
3,4,5
6
EPA, 2009;
3; 4; 5
(continued)
3-10
-------
Table 3-2. Control Strategies and Growth Assumptions for Creating 2020 Base Case
Emissions Inventories from the 2005 Base Case (continued)
Control Strategies and/or Growth Assumptions
(Grouped by Affected Pollutants or Standard and Approach Used to Apply
to the Inventory)
Pollutants
Affected
Approach
or
Reference:
Onroad Mobile and Nonroad Mobile Controls
(list includes all key mobile control strategies but is not exhaustive)
(continued)
Commercial Marine:
Category 3 marine diesel engines Clean Air Act and International Maritime
Organization standards (April, 30, 2010)— also includes CSAPR comments.
EIA fuel consumption projections for diesel-fueled vessels
Clean Air Nonroad Diesel Final Rule—Tier 4
Emissions Standards for Commercial Marine Diesel Engines, December 29,
1999
Locomotive and marine rule (May 6, 2008)
Tier 1 Marine Diesel Engines, February 28, 2003
All
7, 3; EPA,
2009
Approaches/References—Mobile Sources
1. http://epa.gov/otaq/hwy.htm
2. Only for states submitting these inputs: http://www.epa.gov/otaq/lev-nlev.htm
3. http://www.epa.gov/nonroad-diesel/2004fr.htm
4. http://www.epa.gov/cleanschoolbus/
5. http://www.epa.gov/otaq/marinesi.htm
6. Federal Aviation Administration (FAA) Terminal Area Forecast (TAF) System, January 2010:
http://www.apo.data.faa.gov/main/taf.asp
7. http://www.epa.gov/otaq/oceanvessels.htm
3.3 Modeling Results and Analyses
The air quality modeling results were used in the RIA to estimate future-year PM2.5
concentrations for the 2020 base case and to calculate the air quality ratios that were used to
determine potential control scenarios designed to attain the current and proposed alternative
NAAQS. These data are then used to estimate the costs and benefits of attaining these current
and proposed NAAQS levels. Consistent with EPA guidance (EPA, 2007), the air quality modeling
results are applied in a relative sense to estimate 2020 future-year design values for PM2.s and
visibility for the base case as described in Sections 3.3.1.1 and 3.3.2.1. Air quality ratios are
calculated using the changes in the 2005 and 2020 base case design values and emissions as
described in Section 3.3.1.2. The data are then used to estimate the tons of emissions
reductions needed to show attainment of the current and alternative NAAQS levels as
3-11
-------
described in Section 3.3.1.3 and in Chapter 4. Based on the tons of emissions needed in each
county, annual standard design values are calculated for attaining the current and alternative
standard levels for input into the benefits assessment as described in Section 3.3.1.4.
Limitations of this approach are described in Section 3.3.1.5.
Additional data were also processed to calculate visibility design values for the 2020
base case. Details on this post-processing are discussed in Section 3.3.2.
3.3.1 PM2.5
As discussed in Chapter 1, this RIA evaluates the costs and benefits of attaining four
alterative combinations of standards relative to meeting the current primary PM2.5 standards
(15/35). The five alterative combinations of standards evaluated are: an annual standard level
of 14 u.g/m3 in conjunction with retaining the current 24-hour standard level at 35 u.g/m3
(14/35); an annual standard level of 13 u.g/m3 in conjunction with retaining the current 24-hour
standard level at 35 u.g/m3 (13/35); an annual standard level of 12 u.g/m3 in conjunction with
retaining the current 24-hour standard level at 35 u.g/m3 (12/35); an annual standard level of 11
u.g/m3 in conjunction with retaining the current 24-hour standard level at 35 u.g/m3 (11/35);
and an annual standard level of 11 u.g/m3 in conjunction with a 24-hour standard level of 30
u.g/m3 (11/30). We modeled to project future-year PM2.5 concentrations for a 2020 base case
using CMAQ and then estimated the air quality concentrations for meeting 15/35, 14/35,
13/35, 12/35, 11/35 and 11/30 using air quality ratios.
3.3.1.1 Calculating Future-year Design Values for 2020 Base Case
To estimate costs of attaining the alternative NAAQS, we use air quality modeling results
to predict the impact of the control strategies on future-year attainment. This is done by using
the air quality model results in a relative sense, as recommended by the EPA modeling guidance
(EPA, 2007), and estimating future-year PM2.5 relative reduction factors (RRFs). RRFs are ratios
that are calculated from the changes in PM2.5 species concentrations between recent-year and
future-year air quality modeling results. RRFs are calculated for each PM2.5 component. Future-
year estimates of the PM2.5 annual and 24-hour standard design values at monitor locations are
then calculated by applying the species-specific RRFs to ambient PM2.5 concentrations from the
IMPROVE Network, the Speciated Trends Network (STN), and the Federal Reference Method
(FRM) Network.
To more easily apply this methodology, EPA has created software, called Modeled
Attainment Test Software (MATS) (Abt, 2010), to calculate future-year PM2.5 annual and 24-
hour standard design values. For this RIA, the RRFs are based on the changes in modeled
3-12
-------
concentrations between the 2005 and 2020 base case. Ambient measurements used in MATS
are from IMPROVE and STN sites for 2004-2006 and FRM sites for 2003-2007. Output from
MATS includes the projected future-year annual and 24-hour standard design values, as well as
percentage sulfate, nitrate, ammonium, elemental carbon, organic carbon and crustal matter
contributing to the annual and 24-hour standard design values for each site. These data are
useful to better understand the PM species contributing to high PM2.5 concentrations and to
help determine what control measures might be most effective in reducing the future-year
design values to the proposed levels. Annual and 24-hour standard design values for 2005 and
2020 base case are discussed in Chapter 4 and shown in Appendix 4.
3.3.1.2 Calculating Future-year Design Values for Meeting the Current Standard and Proposed
Alternative Standard Levels
To estimate the tons of emissions reductions needed to reach attainment of the current
and proposed alternative standard levels, we calculated air quality ratios based on how
modeled concentrations changed with changes in emissions between a recent year of air
quality, 2005, and the future year of air quality, 2020. These air quality ratios represent an
estimate of how the annual standard design value at a monitor would change in response to
emissions reductions of S02, NOX, or direct PM2.5. Below are the details of how these air quality
ratios were estimated.
To calculate the air quality ratios for changes in response to emissions reductions of S02
and NOX we used the following methodology.
Step 1: The speciated changes in annual standard design values between 20053 and
2020 were obtained from the MATS (Abt, 2010) output files. For each monitor, we computed
the percent change in the NH4S04 and NH4N034 components of the annual standard design
value between 2005 and 2020, relative to the 2005 monitor annual standard design value.
Step 2: For NH4S041 and NH4N033 components, we computed the change in emissions of
S02 and NOX used in the air quality modeling for the 2005 and 2020 base case for groups of
3As described previously in this section, the "2005" annual design values are based on ambient measurements
used in MATS from IMPROVE and STN sites for 2004-2006 and FRM sites for 2003-2007. These years of
ambient measurements were selected since they frame the air quality model year of 2005. Because the air
quality model is used to predict the change in design values between recent and projected future year air
quality, with the modeled RRFs being applied to the recent year measured design values, it is important to
select ambient measurement years that include the model year to allow a more true prediction of the future
year air quality.
The NH4SO4 and NH4NO3 components are computed using the SO4, NO3, NH4 and water fraction from MATS as
described in EPA guidance (EPA, 2007).
3-13
-------
adjacent counties. This larger grouping of counties allowed us to better represent the more
regional nature of NH4S04 and NH4N03 formation and transport. These groupings of counties
were selected by including all counties within a state that bordered a county with a monitor
above the current or proposed alternative standard levels. For the state of California, where
these groupings could have expanded to include most of the counties within the state if we had
used the same selection criteria, we determined smaller groupings to more realistically
represent the area around a monitor from which emissions reductions would most influence
changes in the design value. This smaller grouping was also done for counties in Utah5. These
groups are listed in Appendix 4, Table 4.A-5.
Step 3: Using the data from Steps 1 and 2, we computed the percent change in the
NH4S04 component of the annual standard design value at each monitor per reduction of 1000
tons of S02 emissions in the surrounding counties between the 2005 and 2020 base case air
quality and emissions data. Similarly, we computed the percent change in the annual standard
NH4N03 component of the design value at each monitor per change in 1000 tons of NOx
emissions in the surrounding counties.
Step 4: The data from Step 3 are then used to compute the median value for all
monitors within the grouping of counties of the percent change in the NH4S04 and NH4N03
components of the annual standard design value per change in tons of S02 and N03 emissions,
respectively. This gives us an estimate for each grouping of monitors that indicates the
response of the sulfate and nitrate components of the annual standard design values to
changes in S02 and NOX emissions, relative to the PM2.5 speciation at the monitor.
Step 5: The percent change values from Step 4 are then multiplied by the NH4S04 and
NH4N03 speciation values at each monitor in the 2020 base case to produce the "air quality
ratios." These data give an estimate of how the annual standard design value (u.g/m3) at a
monitor would change if 1000 tons of S02 and/or NOX emissions were reduced in the county in
which the monitor is located.
To calculate the air quality ratios for changes in response to emissions reductions of
direct PM2.5 we follow the following methodology.
5 To determine the counties for the smaller groupings in California and Utah, we simply grouped counties together
geographically with a minimum of two counties allowed in each of the smaller groupings.
3-14
-------
Step 1: The speciated changes in annual standard design values between 20056 and
2020 were obtained from the MATS (Abt, 2010) output files. For each monitor, we computed
the percent change in the direct PM2.57 components of the annual standard design value
between 2005 and 2020, relative to 2005 monitor annual standard design value.
Step 2: We computed the change in emissions of direct PM2.5 in the emissions inventory
data used in the air quality modeling for the 2005 and 2020 base case for each county.
Step 3: Using the data from Steps 1 and 2, we computed the percent change in the
annual standard direct PM2.5 component of the design value at each monitor per change in tons
of direct PM2.5 emissions at the county level.
Step 4: The data from Step 3 are then used to compute the median value of the percent
change in the PM2.5 component of the annual standard design value per change in tons of direct
PM2.5 emissions for all monitors within the grouping of counties used to compute the S02 and
NOX air quality ratios (see Appendix 4, Table 4.A-5) within a state. We now have an estimate for
each grouping of monitors that indicates the response of the direct PM2.5 components of the
annual standard design values to changes in direct PM2.5 emissions, relative to the PM2.5
speciation at each monitor.
Step 5: The percent change values from Step 4 are then multiplied by the direct PM2.5
speciation values at each monitor in the 2020 base case to produce air quality ratios.
Step 6: The responsiveness of air quality at a specific monitor location to direct PM2.5
emission reductions will depend on several factors including the specific meteorology and
topography in an area and the nearness of the emissions source to the monitor. Because of the
more local influence of changes in directly emitted PM2.5 emissions on air quality, a monitor
where significant changes in direct PM2.5 emissions occurred between 2005 and 2020 due to
sources very close to a monitor can result in large non-representative values in Step 5. A large
change suggests the monitor is more responsive to PM2.5 emissions reductions than it actually
6As described previously in this section, the "2005" annual design values are based on ambient measurements
used in MATS from IMPROVE and STN sites for 2004-2006 and FRM sites for 2003-2007. These years of
ambient measurements were selected since they frame the air quality model year of 2005. Because the air
quality model is used to predict the change in design values between recent and projected future year air
quality, with the modeled RRFs being applied to the recent year measured design values, it is important to
select ambient measurement years that include the model year to allow a more true prediction of the future
year air quality.
7The direct PM2.s design value component is computed by summing the elemental carbon, organic carbon and
crustal portions of the design value.
3-15
-------
would be if those reductions were applied further away from the monitor. Given that the air
quality ratios must be applicable to multiple monitors across each county and must be
applicable for emissions reductions where we may not know the specific source location (e.g.,
extrapolated emissions reductions), the air quality ratios we employ should not be strongly
influenced by very local emissions changes which may have a much larger, non-representative
impact on air quality at the nearby monitor. To remedy this and obtain representative values
for air quality ratios, we separated all the counties for which PM2.5 air quality ratios were
computed (as shown in Table 4.A-10) into four areas of the country: East, West, Northern
California, and Southern California as shown in Table 3-38. We then computed a single
"trimmed" median PM2.5 ratio for each of the four areas after removing the highest ten percent
of the values in each area. That is, we calculated the median value of all the PM2.5 air quality
ratios in each area over counties for which air quality ratios had been calculated after the
highest ten percent of the values were removed. The resulting PM2.5 air quality ratios that are
used for all monitors in the four areas are shown in Table 3-3. These data give an estimate of
how the annual standard design value (u.g/m3) at a monitor would change if 1000 tons of direct
PM2.5 emissions were reduced in the county in which the monitor is located.
Table 3-3. Area Definitions and PM2.5 Air Quality Ratios
Area
East
West
Northern California
Southern California
States and Counties Included
Alabama, Georgia, Illinois, Michigan, New York, Ohio,
Pennsylvania, and Texas (all counties with air quality ratios)
Arizona, Idaho, Montana, Oregon, Utah, and Washington (all
counties with air quality ratios)
Butte, Colusa, Contra Costa, Fresno, Inyo, Kern, Kings, Merced,
Monterey, Placer, Plumas, Sacramento, San Joaquin, San Luis
Obispo, Santa Clara, Solano, Stanislaus, Sutter, Tulare, and Yolo,
California
Imperial, Los Angeles, Orange, Riverside, San Bernardino, San
Diego, and Ventura, California
PM2.5 Air Quality Ratio
(u.g/ms Change in Direct
PM2.5 per 1,000 Tons PM)
1.238
1.929
1.879
0.597
To be able to estimate how the 24-hour standard design value would change in
response to a lower annual design value, we computed the ratio of the change in the annual
California was separated into two areas because of the large number of counties analyzed and because of the
large differences seen in the PM2.s air quality ratios for the northern versus the southern counties of California.
3-16
-------
design value to the change in the 24-hour standard design value between 2005 and 2020 base
case for each monitor, and then use these data to compute the average value for all monitors
within a state that are included within a grouping shown in Appendix 4, Table 4.A-5. The end
result is an estimate for each state of the expected change in the 24-hour standard design value
per 1 u.g/m3 change in the annual standard design value. These values varied from state-to-
state but in general, a 1 u.g/m3 change in the annual standard design value corresponded to a
2-3 u.g/m3 change in the 24-hour standard design value at the same monitor. Tables of these
values are provided with the air quality ratios in Chapter 4.
3.3.1.3 Estimating Emissions Reductions and Costs of Attaining the Current and Proposed
Alternative PM2.s Standards
The air quality ratios described in Section 3.3.1.2 are used to determine the most
effective control measures for reducing the annual and 24-hour standard design values to
meeting the current and proposed alternative standard levels. The total amount of S02, NOX
and/or direct PM2.5 emissions reduced per county, based on the control measures selected for
each strategy, are then used in conjunction with the air quality ratios to estimate how the
annual and 24-hour standard design values would change in the counties with emissions
reductions. The details of control measure selections and their associated costs are described in
Chapter 4.
3.3.1.4 Estimating Changes in Annual Average PM2.sfor Benefits Inputs
MATS (Abt, 2010) can also provide gridded fields of changes in annual average PM2.5
concentrations for the entire CMAQ 12km domain. MATS does this by calculating RRFs at every
grid cell within the CMAQ domain for each future-year control scenario, and applying these
RRFs to ambient data that have been interpolated to cover all grid cells in the modeling
domain. The basic interpolation technique, called Voronoi Neighbor Averaging (VNA), identifies
the set of monitors that are nearest to the center of each CMAQ grid cell, and then takes an
inverse distance squared weighted average of the monitor concentrations. A "fused" spatial
field is then calculated by adjusting the interpolated ambient data (in each grid cell) up or down
by a multiplicative factor calculated as the ratio of the modeled concentration at the grid cell
divided by the modeled concentration at the nearest neighbor monitor locations (weighted by
distance). We use the 2005 and 2020 base case CMAQ modeling outputs, in conjunction with
the ambient measurements from the IMPROVE and STN sites for 2004-2006 and FRM sites for
2003-2007, to create a spatial surface for the 2020 base case.
3-17
-------
To generate spatial surfaces that represent annual average PM2.5 concentrations when
attaining the current and proposed alternative standard levels for the benefits analysis, we
need to adjust the 2020 base case spatial surface to reflect attainment of the current and
alternative standard levels. Given the MATS gridded annual average PM2.5 concentrations for
the CMAQ 12km domain and projected design values at monitors, the "monitor rollback"
approach is used to approximate the air quality change resulting from attaining alternative
NAAQS at each design value monitor. Figure 3-2 depicts the rollback process. This approach
aims to estimate the change in population exposure associated with attaining an alternate
NAAQS, relying on data from the existing monitoring network and the inverse distance variant
of the VNA interpolation method to adjust the CMAQ-modeled concentrations such that each
area attains the standard alternatives. Using the VNA spatial averaging technique, the annual
average PM2.5 spatial surface is smoothed to minimize sharp gradients in PM2.5 concentrations
in the spatial fields due to changes in the monitor concentrations.
3.3.1.5 Limitations of Using Air Quality Ratios
There are important limitations to the methodology of calculating and using air quality
ratios to predict the response of air quality to emissions changes. The air quality ratios are
calculated with results from only two CMAQ model runs and are based on the assumption that
the monitor design values would decrease with additional reductions in emissions of S02, NOX
and direct PM2.5 in the future similar to how the two CMAQ model runs predicted changes in air
quality concentrations. The uncertainty of this assumption will increase with increasing
emissions reductions needed to estimate attainment. In addition, the model response to
emissions changes are analyzed at a county-level or within a small group of counties, and we
assume that air quality concentrations at a monitor will decrease linearly with emissions
reductions in a county (e.g., direct PM2.5 emission reductions) or a group of counties (e.g., S02
and NOX emissions reductions). Because of the more local influence of changes in directly
emitted PM2.5 emissions on air quality, it is also particularly difficult for the air quality ratio
approach to estimate well how the design value at a monitor in a county would respond to
changes in direct PM2.5 emissions in a county without knowing the location of the source (e.g.,
extrapolated emissions reductions) relative to the location of the monitor.
The exact impact of using this methodology to estimate the emissions reductions
needed for attainment and the associated effect on the cost and benefits is uncertain and may
vary from monitor-to-monitor. We do not believe that this methodology tends towards any
general trend and does not always result in either an underestimation or overestimation of the
costs and benefits of attaining the proposed alternative standards.
3-18
-------
Figure 3-2. Diagram of Rollback Method
Use modeled air quality data
Step 1. Receive 12 km
to establish ratios between
CMAQ baseline air quality , .
design value and air quality
modeling . .
metric at each monitor.
Step 2. Rollback monitor design
values to attain each standard
alternative
Step 3. Interpolate
incremental reduction Convert interpolated Calculate
in design value change DV change to benefits for
to 12 km grid using equivalent change in each
VNA in BenMAP metric and adjust standard
model grid
3.3.2 Visibility
As described in the Policy Assessment Document (EPA, 2011a) and Chapter 2 of this RIA,
the formula for total light extinction (bext) in units of Mm"1 using the original IMPROVE equation
fcext = 3 x /(RH) x [Sulfate] + 3 x /(RH) x [Nitrate] + 4 x [Organic Mass] +
10 x [Elemental Carbon] + Ix [Fine Soil] + 0.6 x [Coarse Mass] + 10 (3.1)
where the mass concentrations of the components indicated in brackets are in units of u.g/m3,
and/(RH) is the unitless water growth term that depends on relative humidity. The final term in
the equation is known as the Rayleigh scattering term and accounts for light scattering by the
3-19
-------
natural gases in unpolluted air. Since IMPROVE does not include ammonium ion monitoring, the
assumption is made that all sulfate is fully neutralized ammonium sulfate and all nitrate is
assumed to be ammonium nitrate. Using equation (3.1), light extinction (bext) can then be
converted into units of deciviews, a scale frequently used in the scientific and regulatory
literature on visibility.
3.3.2.1 Calculating Future-year Visibility Design Values for 2020 Base Case
The visibility design value calculations are based on 24-hour averages and the 90th
percentile format. To estimate future-year visibility design values, we use the air quality
modeling results in a relative sense, as recommended by the EPA modeling guidance (EPA,
2007), and estimate future-year relative reduction factors (RRFs) for each speciated component
of the light extinction (bext) equation. To be consistent with visibility calculations described in
the Policy Assessment Document (EPA, 2011a), which focuses on PM2.5 visibility, we do not
include coarse mass (PMi0-2.5) in the calculation. The steps for projecting the future-year
visibility design values are described below.
Step 1: We extract 24-hour averages of sulfate, nitrate, organic mass, elemental carbon
and fine soil for the 2020 future-year modeled base case for each CMAQ grid cell in which a STN
monitor is located. For the assumption that sulfate is fully neutralized and all nitrate is assumed
to be ammonium nitrate, we multiply 24-hour average sulfate mass by 1.375 and nitrate mass
by 1.29.
Step 2: For the Regional Haze Program, there exists a gridded file of monthly averaged
/(RH) climatological mean values.9 Using these data, we assign a/(RH) values to each STN
monitor for each season10 by averaging the 3 monthly/(RH) values in each season using the
data from the closest available data point.
Step 3: Using the data from Step 1 &2, we calculate bext for every day in the 2005
modeled base case using equation (3.1) without the coarse mass component.
Step 4: For every season, we extract the top 10% worst modeled visibility days (i.e., top
9 days) in 2005 based on their bext values. Using the species concentrations for these nine
9U.S. EPA, Interpolating Relative Humidity Weighting Factors to Calculate Visibility Impairment and the Effects of
IMPROVE Monitor Outliers, prepared by Science Applications International Corporation, Raleigh, NC, EPA
Contract No. 68-D-98-113, August 30, 2001.
10Each season is defined as Winter ( Dec, Jan & Feb), Spring (Mar, Apr & May), Summer (Jun, Jul & Aug) and Fall
(Sep, Oct & Nov).
3-20
-------
maximum bext days, we calculate the average species specific bext value for each season for each
monitor location.
Step 5: We repeat step 3 for each of the future-year modeled scenario, and extract the
same calendar days that were selected in step 4 for 2005.
Step 6: Using the data from steps 4 and 5, we calculate the species specific Relative
Response Factors (RRFs) for each monitor in each season for each of the future-year modeled
scenario. This is done by dividing the average speciated bext value for the top 10% worse
visibility days for each season for every monitor in each future-year scenario by the average
species specific concentration for same the season for every monitor in 2005. In this way, we
will have an RRF for every monitor for each season for each future-year control scenario.
Step 7: The set of seasonal RRFs for each monitor for each future-year control scenario
are applied to the corresponding 2004-2006 ambient data11 and the 90th percentile value is
extracted. The end result is a set of visibility design values for each future-year scenario.
3.3.2.2 Calculating Future-year Visibility Design Values for Meeting the Current and Proposed
Alternative Standard Levels
It is important to understand how changes in the PM2.5 design values to simulate full
attainment for each proposed alternative NAAQS will affect visibility design values.
As described in Section 3.3.1.2, we apply a methodology of air quality ratios to estimate
the emissions reductions needed to meet the current and proposed alternative levels for the
primary standard for PM2.5. While this methodology can estimate how the emissions reductions
in each control scenario will affect changes in the future-year annual design values, and the
corresponding response of the future-year 24-hour design values to these changes in the
annual design value, it is unable to estimate how each of the PM2.5 species will change with
these emission reductions. Given that estimating changes in future-year visibility is dependent
on the IMPROVE equation (3.1) and how the PM2.5 species are projected to change in time, we
are unable to estimate visibility design values for meeting the current and proposed alternative
levels for the primary PM2.5 standard.
11 These years of ambient measurements were selected since they frame the air quality model year of 2005.
Because the air quality model is used to predict the change in design values between recent and projected
future year air quality, with the modeled RRFs being applied to the recent year measured design values, it is
important to select ambient measurement years that include the model year to allow a more true prediction of
the future year air quality.
3-21
-------
3.3.2.3 Estimating Changes in Visibility for Analyzing Welfare Benefits
The visibility calculations for the welfare benefits assessment are based on 24-hour
average light extinction (bext) values, averaged over the year and converted to units of
deciviews. As described in Sections 3.3.2.1 and 3.3.2.2, we calculated the visibility design values
for the 2020 base case but were unable to estimate how these visibility design values would
change for meeting the current and proposed alternative levels for the primary PM2.5 standard.
In this same way, we are unable to estimate the light extinction values for meeting the current
and proposed alternative levels for the primary PM2.5 standard, which are needed to assess
welfare benefits.
3.4 References
Abt Associates, 2010. User's Guide: Modeled Attainment Test Software.
http://www.epa.gov/scram001/modelingapps mats.htm.
Appel, K.W., Bhave, P.V., Gilliland, A.B., Sarwar, G., Roselle, S.J., 2008. Evaluation of the
community multiscale air quality (CMAQ) model version 4.5: Sensitivities impacting
model performance; Part II—particulate matter. Atmospheric Environment 42, 6057-
6066.
Appel, K.W., Gilliland, A.B., Sarwar, G., Gilliam, R.C., 2007. Evaluation of the Community
Multiscale Air Quality (CMAQ) model version 4.5: Sensitivities impacting model
performance Part I—Ozone. Atmospheric Environment 41, 9603-9615.
Byun, D.W., and K.L. Schere. 2006. "Review of the Governing Equations, Computational
Algorithms, and Other Components of the Models-3 Community Multiscale Air Quality
(CMAQ) Modeling System." J. Applied Mechanics Reviews 59(2):51-77.
Pitchford, M., and W. Malm. 1994. "Development and Application of a Standard Visual Index."
Atmospheric Environment 28(5): 1049-1054.
Sisler, J.F. July 1996. Spatial and Seasonal Patterns and Long Term Variability of the Composition
of the Haze in the United States: An Analysis of Data from the IMPROVE Network. Fort
Collins, CO: Cooperative Institute for Research in the Atmosphere, Colorado State
University.
U.S. Environmental Protection Agency (EPA). 2007. Guidance on the Use of Models and Other
Analyses for Demonstrating Attainment of Air Quality Goals for Ozone, PM2.s, and
Regional Haze. Office of Air Quality Planning and Standards, Research Triangle Park, NC.
U.S. Environmental Protection Agency (EPA). 2008. Regulatory Impact Analysis of the Proposed
Revisions to the National Ambient Air Quality Standards for Lead. Office of Air Quality
Planning and Standards, Research Triangle Park, NC.
3-22
-------
U.S. Environmental Projections Agency (EPA). 2010. Technical Support Document: Preparation
of Emissions Inventories for the Version 4, 2005-based Platform, USEPA, July, 2010. Also
available at: http://www.epa.gov/ttn/chief/emch/index.htmlW2005.
U.S. Environmental Protection Agency (EPA). 2011a. Policy Assessment for the Review of the
Particulate Matter National Ambient Air Quality Standards. Office of Air Quality Planning
and Standards, Research Triangle Park, NC.
U.S. Environmental Projections Agency (EPA). 2011b. Regulatory Impact Analysis for the Final
Mercury and Air Toxics Standards. Office of Air Quality Planning and Standards,
Research Triangle Park, NC. EPA-452/R-11-011.
U.S. Environmental Projections Agency (EPA). 2011c. Emissions Modeling for the Final Mercury
and Air Toxics Standards Technical Support Document. Also available at:
http://www.epa.gov/ttn/chief/emch/index.htmltftoxics.
U.S. Environmental Projections Agency (EPA). 2011d. Air Quality Modeling Technical Support
Document: Final ECU NESHAP. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. EPA-454/R-11-009.
3-23
-------
CHAPTER 4
CONTROL STRATEGIES
4.1 Synopsis
In order to estimate the costs and benefits of alternative PM2.5 standards, the U.S. EPA
has analyzed hypothetical control strategies that areas across the country might employ to
attain alternative more stringent annual standards of 13, 12, and 11 u.g/m3 in conjunction with
retaining the 24-hour standard of 35 u.g/m3, as well as an alternative more stringent annual
standard of 11 u.g/m3 in conjunction with an alternative more stringent 24-hour standard of 30
ug/m3 (referred to as 13/35, 12/35,11/35, and 11/30). The U.S. EPA also analyzed a 14/35
alternative standard and determined that all counties would meet such a standard concurrent
with meeting the existing 15/35 standard at no additional costs and with no additional benefits
because of significant air quality improvements from the Mercury and Air Toxics Standards
(MATS), the Cross-State Air Pollution Rule (CSAPR), and other Clean Air Act rules as described in
Chapter 3, Section 3.2.2. Thus, there is no need to present an analysis of 14/35.
For the purposes of this discussion, it will be helpful to define some terminology. These
definitions are specific to this analysis:
• Base Case—Emissions projected to the year 2020 reflecting current state and federal
programs, including the Cross-State Air Pollution Rule and the Mercury and Air
Toxics Standards. This does not include control programs specifically for the purpose
of attaining the current PM2.5 standard (15/35).
• Baseline—Emissions projections to the year 2020 reflecting the base case plus
additional emission reductions needed to reach attainment of the current PM2.5
Standard (15/35).
• Alternative Standard Analysis—Emission reductions and associated hypothetical
controls needed to reach attainment of the alternative standards. These reductions
and controls are incremental to the baseline.
• Design Value—A metric that is compared to the level of the National Ambient Air
Quality Standard (NAAQS) to determine compliance. Design values are typically used
to classify nonattainment areas, assess progress towards meeting the NAAQS, and
develop control strategies. The design value for the annual PM2.5 standard is
calculated as the 3-year average of annual means for a single monitoring site or a
group of monitoring sites. The design value for the 24-hour standard is calculated as
the 3-year average of annual 98th percentile 24-hour average values recorded at
each monitoring site.
4-1
-------
The U.S. EPA has analyzed the impact that additional emissions controls across
numerous sectors would have on predicted ambient PM2.5 concentrations incremental to a
baseline, which includes the current PM2.5 standard as well as other major rules such as CSAPR
and MATS. Thus, the analysis for a revised standard focuses specifically on incremental
improvements beyond the current standard and other existing major rules, and uses control
options that might be available to states for application by 2020. The hypothetical control
strategies presented in this RIA represent illustrative options for achieving emissions reductions
to move towards a national attainment of a tighter standard. It is not a recommendation for
how a tighter PM2.5 standard should be implemented, and states will make all final decisions
regarding implementation strategies once a final NAAQS has been established.
In order to analyze these hypothetical control strategies incremental to attainment of
the current standard and beyond other existing major rules, the U.S. EPA employed a multi-
stage approach. First, the U.S. EPA identified controls to be included in the base case (e.g.,
reflecting current standard of 15/35) to reflect current state and federal programs. Next
additional controls were applied to attain the current PM2.5 standard. The current state and
federal programs combined with the additional controls needed for attainment of the current
PM2.5 standard make up the baseline for this analysis. Once the baseline was established, we
applied additional known controls within counties containing a monitor predicted to exceed the
standard alternatives of 13/35, 12/35, 11/35, and 11/30 so as to bring them into attainment
with the various alternatives in 2020.1 This chapter presents the hypothetical control strategies
and the results in 2020 after their application. For most of these alternative standards,
application of known control measures did not achieve attainment. In such cases, additional
emission reductions beyond the capability of known controls were estimated in order to reach
full attainment.
4.2 PM2.5 Control Strategy Analysis
4.2.1 Establishing the Baseline
The RIA is intended to evaluate the costs and benefits of reaching attainment with
alternative PM2.5 standards. In order to develop and evaluate hypothetical control strategies for
attaining a more stringent primary standard, it is important to first estimate PM2.5 levels in
20202 given the current NAAQS standards (15/35) and trends. This scenario is known as the
1 Refer to Table 4-2 for details on the number of counties with exceedances and the number of additional counties
where reductions were applied.
4-2
-------
baseline. Establishing this baseline allows us to estimate the incremental costs and benefits of
attaining any alternative primary standard.
The baseline includes reductions already achieved as a result of national regulations,
reductions expected prior to 2020 from recently promulgated national regulations3 (i.e.,
reductions that were not realized before 2005 but are expected prior to attainment of the
current PM standard), and reductions from additional controls which the U.S. EPA estimates
need to be included to attain the current standard (15/35). Reductions achieved as a result of
state and local agency regulations and voluntary programs are reflected to the extent that they
are represented in emission inventory information submitted to the U.S. EPA by state and local
agencies4. Two steps were used to develop the baseline. First, the reductions expected in
national PM2.s concentrations from national rules promulgated prior to this analysis were
considered (referred to as the base case). Below is a list of some of the major national rules
reflected in the base case. Refer to Chapter 3, Section 3.2.2 for a more detailed discussion of
the rules reflected in the base case emissions inventory.
• Light-Duty Vehicle Tier 2 Rule (U.S. EPA, 1999)
• Heavy Duty Diesel Rule (U.S. EPA, 2000)
• Clean Air Nonroad Diesel Rule (U.S. EPA, 2004)
• Regional Haze Regulations and Guidelines for Best Available Retrofit Technology
Determinations (U.S. EPA, 2005b)
• NOX Emission Standard for New Commercial Aircraft Engines (U.S. EPA, 2005)
• Emissions Standards for Locomotives and Marine Compression-Ignition Engines (U.S.
EPA, 2008)
• Control of Emissions for Nonroad Spark Ignition Engines and Equipment (U.S. EPA,
2008)
• C3 Oceangoing Vessels (U.S. EPA, 2010)
3 The recently proposed Boiler MACT and CISWI reconsiderations are not included in the base case. These rules
were not yet proposed at the time of this analysis. It is not clear how the geographic scope of this rule will
match with the counties analyzed for this RIA—the costs may decrease but the magnitude is uncertain.
4 The amendments to the Low Emissions Vehicle Program (LEV-MI) in California are not included in the base case.
This program requires an approval of U.S. EPA via a waiver. At the time of this analysis the waiver had not been
submitted.
4-3
-------
• Hospital/Medical/lnfectious Waste Incinerators: New Source Performance Standards
and Emission Guidelines: Final Rule Amendments (U.S. EPA, 2009)
• Reciprocating Internal Combustion Engines (RICE) NESHAPs (U.S. EPA, 2010)
• Cross-State Air Pollution Rule (U.S. EPA, 2011)
• Mercury and Air Toxics Standards (U.S. EPA, 2011)
Note that we did not conduct this analysis incremental to controls applied as part of
previous NAAQS analyses (e.g., 03, NOX, or S02) because the data and modeling on which these
previous analyses were based are now considered outdated and are not compatible with the
current PM2.s NAAQS analysis. In addition, all control strategies analyzed in NAAQS RIAs are
hypothetical. Second, because the base case reductions alone were not predicted to bring all
areas into attainment with the current standard (2 counties are projected to exceed an
alternative standard of 13/35 and 18 counties are projected to exceed an alternative standard
of 12/35—see Section 4.2.2.1 for more details), the U.S. EPA used a hypothetical control
strategy to apply additional known controls to illustrate attainment with the current PM2.s
standard. Additional control measures were used in two sectors to establish the baseline:5 Non-
Electricity Generating Unit Point Sources (Non-EGUs) and Non-Point Area Sources (Area).
The 2020 baseline for this analysis presents one scenario of future year air quality based
upon specific control measures, including federal rules such as CSAPR and MATS, years of air
quality monitoring and emissions data. This analysis presents one illustrative strategy relying on
the identified federal measures and other strategies that states may employ. States may
ultimately employ other strategies and/or other federal rules may be adopted that would also
help in achieving attainment. The U.S. EPA plans to issue the final rule no later than December
14, 2012 and intends to complete designations two years following promulgation of the final
rule. Under the Clean Air Act, States are required to submit State implementation plans within 3
years of the effective date of the designations. The plans are required to show attainment as
expeditiously as practicable but no later than 5 years following the effective date of the
designations, with the possibility, in certain cases, of an attainment date up to 10 years from
the effective date of the designations, considering the severity of air quality concentrations in
the area and the availability and feasibility of emission control measures. Designations will
likely be based on air quality data from 2011-2013, but attainment will not occur until 2020 at
5 In establishing the baseline, the U.S. EPA selected a set of cost-effective controls to simulate attainment of the
current PM2.5 standard. These control sets are hypothetical as states will ultimately determine controls as part
of the SIP process.
4-4
-------
the earliest. Thus EPA's projections for control costs and benefits focus on the year 2020. The
number of counties that will be part of the designations process may be different than the
number of counties projected to exceed as part of this analysis. Refer to Section IX of the
proposed PM2.5 NAAQS for more details concerning implementation requirements for the
proposed NAAQS.
Two maps of the country are presented in Figures 4-1 and 4-2, which show the
predicted concentrations for year 2020 for the 575 counties with PM2.5 annual design values
and 569 counties with 24-hour design values prior to applying controls to meet the current
standard of 15/35. Control measures were applied to 14 counties in the baseline analysis to
meet the current PM2.5 standard.
&?!> tcuiww wBi timlDis hate PM 2 3 antna
M ug« art projetifo to Kteel ii 11511 ?
'Daddvona riuitBZR pr|KtHlta«w:K
-------
4.2.1.1 Controls Applied in the Baseline
The purpose of identifying and analyzing hypothetical baseline controls for PM2.5 and its
precursors is to establish a level of emissions associated with ambient concentrations that
would meet the current PM2.5 standard. The additional known controls included in the baseline
to simulate attainment with current PM2.5 NAAQS are listed in Table 4-1; details regarding the
individual controls are provided in Appendix 4.A. Controls were applied to directly emitted
PM2.5 and the PM2.5 precursors of NOX and S02 given that nitrate, sulfate, and primary PM2.5
species usually dominate measured PM2.5 based on speciation data measured at the Chemical
Speciation Network (CSN) sites. Control measures that directly reduced emissions of PM2.5 were
determined to be most effective close to the exceeding monitors with N0xand S02 controls
supplemented depending upon the monitor speciation data. PM2.5 control measures were
L*g*nd
5tS csuils mtn reonton na« PW J *.
••-, KU'una -junits J e piqeiea m ejceto -30 uotnS
Figure 4-2. Counties Projected to Exceed the Baseline and Analysis Levels of the PM2.s
24-hour Standard Alternatives in 2020
4-6
-------
applied in the county containing the exceeding monitor for the non-EGU point and area source
emissions. If additional emission control was needed, S02 and NOX control measures were
applied within the county exceeding and any contiguous county.6 Additional control measures
were not applied to electric generating units (EGUs) due to the extensive nature of controls
resulting from the inclusion of MATS and CSAPR, and additional controls were not applied to
mobile sources due to our inability to capture regional reductions using the air quality screening
methodology employed for this analysis.
6 Refer to Table 4-2 for details on the number of counties with exceedances and the number of additional counties
where reductions were applied.
4-7
-------
Table 4-1. Controls Applied in the Baseline for the Current PM2.5 Standard4
Pollutant
NOX
PM2.5
S02
NOX
PM2.5
S02
Control Measure
Selective Catalytic Reduction (SCR)
Non-selective Catalytic Reduction (NSCR)
Oxy-Firing
Bio-solid Injection
SCR+ Steam Injection
Low NOX Burners (LNB)
LNB + SCR
LNB + Selective Non-catalytic Reduction (SNCR)
Fabric Filters
Dry Electrostatic Precipitators (ESPs)
Wet ESPs
Venturi Scrubbers
Spray Dryer Absorber (SDA)
Flue Gas Desulfurization (FGD)
Wet FGDs
Water Heater + LNB Space Heaters
Low-NOx Burners for Residential Natural Gas
Fireplace Inserts for Home Heating
Basic Smoke Management Practices and
Establishment of Smoke Management Programs for
Prescribed Burning and other Open Burning**
Woodstove Advisory Program
ESPs for Commercial Cooking
Fuel Switching for Stationary Source Fuel Combustion
Low Sulfur Home Heating Fuel
15/35
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
12/35 11/35
X
X
X
X
X X
X
X
X
X
X
X
X X
X X
X X
X
X
X
11/30
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
* As discussed elsewhere in this chapter, no known controls were applied for 14/35 or 13/35.
**lncludes specific practices such as episodic bans on open burning, and substituting chipping for open burning.
4.2.2 Alternative Standard Control Strategies
After establishing the baseline of attaining the current standard of 15/35, additional
emission reductions needed to meet four alternative standards 13/35,12/35, 11/35, and 11/30
were calculated.
4.2.2.1 Counties Exceeding Alternative Standards
Only two counties are projected to exceed an alternative standard of 13/35 using the
results from the baseline analysis. These are Riverside County, CA and San Bernardino County,
CA. Figures 4-3 through 4-5 show the counties projected to exceed the alternative standards
12/35, 11/35, and 11/30, respectively. Six counties are projected to exceed an alternative
4-8
-------
standard of 12/35, eighteen counties are projected to exceed an alternative standard of 11/35,
and thirty-five counties are projected to exceed an alternative standard of 11/30. For a
complete list of monitor values see Appendix 4.A.
Legend
Ha ow-TOw -fir- n.-jnlm how PH E 3 2i-*inir
jrtyjji .Jtiqn vatjas o4 iinlch
~]
Figure 4-3. Counties Projected to Exceed the 12/35 ug/m3 Alternative Standard After
Meeting the Baseline (Current Standard) in 2020
4-9
-------
Figure 4-4. Counties Projected to Exceed the 11/35 ug/m3 Alternative Standard After
Meeting the Baseline (Current Standard) in 2020
4-10
-------
Legend
m in (ranvn ravt PM 2 £ 2i-nour ar.my
jmu?l ^5171 vgdlKS 01 whKI
' T caiiHHt Jffl pr^MCiiW t> •jiLiihC id-naur any
ij r L aunkiM us (y.3»ecMfl nai m AI^UD iflvw
Figure 4-5. Counties Projected to Exceed the 11/30 ug/m3 Alternative Standard After
Meeting the Baseline (Current Standard) in 2020
In developing the control strategies for this RIA, the U.S. EPA first applied known
controls to reach attainment. For these control strategies, controls for two sectors were used in
developing the control analysis, as discussed previously: non-EGU point and area sources. An
approach similar to that taken for the baseline analysis was used in the analysis for the control
strategies for the alternative standards. Due to the lack of air quality modeling for the control
strategies, county-specific ratios of air quality response to emission reductions were applied
based on recent air quality modeling results. A least cost framework was adapted to adjust for
the use of the air quality to emissions ratios.
In this analysis, PM2.5 controls were applied first because they were more cost-effective
and the air quality ratio approach is generally more accurate for PM2.5 emission changes than
for emission changes from the precursors, S02 and NOX. If additional control was needed S02
and NOX controls were added.
4-11
-------
It should be noted that while PM2.5 controls were applied only within the counties with
monitors projected to exceed the alternative standard being analyzed, S02 and NOX controls
were applied in the exceeding county as well as in the surrounding counties because of the
transport of NOX and S02 across counties. Table 4-2 shows the number of exceeding counties
and the number of surrounding counties to which controls were applied for the alternative
standards analyzed. For a complete list of geographic areas for all alternative standards see
Appendix 4.A.
Table 4-2. Number of Counties with Exceedances and Number of Additional Counties
Where Reductions were Applied
Alternative Number of Counties with Number of Additional Counties where
Level exceedances reductions were applied
13/3523~
12/35 6 25~
11/351886~
11/3035134
There were some areas where known controls did not achieve enough emission
reductions to attain the alternative annual standards in 2020. To complete the analysis, the U.S.
EPA then estimated the additional emission reductions required to reach attainment. The
methodology used to develop those estimates and those calculations are presented in Section
4.2.3.
4.2.2.2 Non-EGU and Area Controls Applied for Alternative Standards
Non-EGU point and area control measures were identified using the U.S. EPA's Control
Strategy Tool7 (CoST). Many of these controls are summarized in Appendix 4.A.
Area source emissions data are generated at the county level, and therefore controls for
this emission sector were applied to the county. Area source controls were applied to NOX, S02,
and PM2.5. Table 4-1 lists the major controls applied to each sector. The same controls that
were applied in the baseline analysis were applied to additional sources and additional counties
in the analyses for the alternative standards. Controls for area sources were applied to: home
heating, restaurant operations, prescribed burning, and other open burning.
7 See http://www.epa.gov/ttn/ecas/cost.htm for a description of CoST.
4-12
-------
The analysis for non-EGUs applied NOX, S02, and PM2.5 controls to the following source
categories: industrial boilers, commercial and institutional boilers, sulfuric acid plants (both
standalone and at other facilities such as copper and lead smelters), primary metal plants (iron
and steel mills, lead smelters), mineral products (primarily cement kilns), and petroleum
refineries. Among the control measures applied were: wet FGD scrubbers and spray dryer
absorbers (SDA) for S02 reductions, fabric filters for PM2.5 reductions, and SCR and low NOX
burners for NOX.
To more accurately depict available controls, the U.S. EPA employed a decision rule in
which controls were not applied to any non-EGU or area sources with 50 tons/year of emissions
or less. This decision rule is the same rule we employed for sources in the previous PM2.5
NAAQS RIA completed in 2006. The reason for applying this decision rule is based on a finding
that most point sources with emissions of this level or less had controls already in place. This
decision rule helps fill gaps in information regarding existing controls on non-EGU sources.
4.2.2.3 Emission Reductions
Table 4-3 shows the emission reductions from known controls for the alternative
standards analyzed.
Table 4-3. Emission Reductions from Known Controls for Alternative Standardsa
Emission Reductions in 2020 (annual tons/year)
Alternative
Standard
13/35°
12/35a
11/35
11/30
Region
East
West
CA
Total
East
West
CA
Total
East
West
CA
Total
East
West
CA
Total
PM
9 ^
—
—
—
—
670
60
730
5,000
90
800
5,900
4,700
1,900
3,700
10,000
so
—
—
—
—
—
—
—
—
12,000
550
—
13,000
12,000
3,900
—
16,000
NOX
—
—
—
—
—
—
—
—
850
620
—
1,500
850
7,400
—
8,200
a Estimates are rounded to two significant figures.
All known controls were applied in the baseline analysis. Thus, no additional known controls were available.
4-13
-------
4.2.3 Emission Reductions Needed Beyond Identified Controls
For each alternative standard and geographic area that cannot reach attainment with
known controls, we estimated the additional emission reductions needed beyond identified
known controls for PM2.5and for the two PM2.5 precursors (S02and NOX) to attain the standard.
In Appendix 4.A, we provide estimates of the relationship between additional emission
reductions for each pollutant and air quality improvement.
Because three different pollutants affect ambient levels of PM2.5 in this analysis there
are many different combinations of pollutant reductions that would result in the required air
quality improvements. To determine which pollutant reductions to include in our analysis, we
employ a least cost approach using what we call the hybrid cost methodology. A detailed
discussion of this methodology appears in Chapter 7. In Appendix 7.A, we show cost estimates
for each additional emission reduction for each pollutant and geographic area. As expected, the
unit costs increase under this methodology as more of a particular pollutant is controlled. The
mix of pollutants controlled, however, varies by area because each area has a different
combination of known controls applied, varying amounts of additional air quality improvement
required, and different amounts of uncontrolled emissions remaining.
The process used to determine the emission reductions needed for each pollutant is
described below. First, the U.S. EPA examined the emissions remaining for each geographic
area and pollutant (NOX, PM2.5, S02). Each pollutant has a marginal cost curve that increases
(e.g., the third ton removed costs more than the second ton removed, and so on). Each
pollutant has an estimated effectiveness at reducing the ambient concentration of PM2.5 per
ton of emissions controlled (see Chapter 3 for more details) that varies by geographic area. The
U.S. EPA used a least cost methodology to determine the optimum way to reach attainment.
The optimization methodology used to estimate the quantity of PM2.5 and PM2.5 precursors
needed for each geographic area is described in more detail in Chapter 7 and Appendix 7.A.
Because the marginal cost equation for each pollutant is expected to be less accurate
for the very last portion of a pollutant in an area, and it is unlikely an area would reduce all
anthropogenic emissions to zero on one pollutant prior to controlling others, we added the
constraint that no more than 90% of the remaining emissions in an area for a given pollutant
can be reduced from emission reductions beyond known control measures. This decision was
based upon the rationale that no geographic area would be able to eliminate 100% of the
emissions of a pollutant given current control measures. This methodology used the marginal
cost curves for each of the three pollutants along with the air quality to emissions response
4-14
-------
ratios to determine the most cost-effective way to achieve the necessary levels of air quality
improvement. A detailed discussion of the methodology, including formulas and description of
how parameters were estimated, can be found in Chapter 7.
The emission reductions needed beyond known controls are shown in Table 4-4. For a
listing of emission reductions needed by county for the unknown controls, see Appendix 4.A.
For the alternative standard 13/35 there are only two counties projected to exceed—Riverside
County, CA and San Bernardino County, CA. For Riverside County, all known controls were
applied in the analysis to illustrate attainment of the baseline (15/35). Thus, no known controls
remained for demonstrating attainment of more stringent standards. For the other alternative
standards (12/35, 11/35, and 11/30), known controls accounted for over 70% of the needed
emission reductions.
The emissions reductions estimated using the hybrid methodology together with
reductions associated with known controls form the basis of the cost and benefit estimates.
However, a different mix of reductions in S02 emissions and PM2.5 emissions may have been
identified as least cost using a fixed cost per ton approach rather than the hybrid approach.8
Using the hybrid methodology, the less expensive pollutant to reduce will be selected
until the marginal cost to reduce the next ton exceeds the marginal cost to reduce the next ton
of an alternate pollutant. At that point, the methodology chooses a mix of pollutants to achieve
the least-cost solution. Since the cost per ton is held constant in the fixed-cost methodology,
the least-cost solution would select all available direct PM2.5 emissions reductions before
selecting S02 emissions reductions.9 Therefore, the hybrid methodology estimates PM2.5
emissions reductions lower than or equal to the fixed-cost methodology and S02 emission
reductions higher than or equal to the fixed-cost methodology.
Even so, for the proposed 12/35 and 13/35 standards, direct PM2.5 reductions account
for approximately 75%-100% of the reductions, and thus using the fixed cost per ton approach
to select the combination of emissions reductions would not have substantially changed the
mix of emissions reductions or the outcome of the cost and benefit analyses.
NOx reduction were not selected under either approach as the least cost alternative to achieve the necessary
PM2.5 reductions.
9 Because the marginal cost equation for each pollutant is expected to be less accurate for the very last portion of
a pollutant in an area, and it is unlikely an area would reduce all anthropogenic emissions to zero on one
pollutant prior to controlling others, we included the constraint that no more than 90% of the remaining
emissions in an area for a given pollutant can be reduced from emission reductions beyond known control
measures.
4-15
-------
That said, the hybrid approach still has a number of important uncertainties, and the
reliability of the method for extrapolating costs in cases where emissions reductions required
go well beyond known controls has not been evaluated. The degree of extrapolation for
emissions reductions in California in particular has caused us to rethink the use of the hybrid
method in providing a range of cost estimates for the proposed standards, therefore we
provide the hybrid approach as a sensitivity analysis in Appendix 7.A. We would like to take
comment on analyzing an alternate compliance pathway for California.
Table 4-4. Emission Reductions Needed Beyond Known Control to Reach Alternative
Standards in 2020 (annual tons/year)3
Alternative
Standard
13/35
12/35a
11/35
11/30
Region
East
West
CA
Total
East
West
CA
Total
East
West
CA
Total
East
West
CA
Total
PM2.5
—
—
190
190
—
210
3,400
3,600
89
1,100
6,500
7,700
1,400
3,500
7,200
12,000
S02 NOX
— —
— —
— —
— —
— —
10 —
960 —
970 —
— —
1,400 -
5,500 -
6,900 —
— —
1,700 -
5,500 -
7,200 —
a Estimates are rounded to two significant figures.
4.3 Limitations and Uncertainties
The U.S. EPA's analysis is based on its best judgment for various input assumptions that
are uncertain. As a general matter, the Agency selects the best available information from
engineering studies of air pollution controls and has set up what it believes is the most
reasonable modeling framework for analyzing the cost, emission changes, and other impacts of
regulatory controls. However, the estimates of emission reductions associated with our control
strategies above are subject to important limitations and uncertainties. We outline, and
4-16
-------
qualitatively assess the impact of, those limitations and uncertainties that are most significant.
EPA requests comment on the likelihood that new technologies that control direct PM2.5 and its
precursors will become available between now and 2020.
A number of limitations and uncertainties are associated with the analysis of emission
control measures are listed in Table 4-5. For a complete discussion of the terminology used
below please see Chapter 5.5.7.
Table 4-5. Summary of Qualitative Uncertainty for Elements of Control Strategies
Potential Source of Uncertainty
Magnitude Degree of
Direction of Impact Confidence
of on in Our Ability to
Potential Monetized Analytical Assess
Bias Costs3 Approach Uncertainty0
Uncertainties Associated with PM Concentration Changes
Projections of future levels of emissions and emissions Both
reductions necessary to achieve the NAAQS
Responsiveness of air quality model to changes in Both
precursor emissions from control scenarios
Air quality model chemistry, particularly for formation Both
of ambient nitrate concentrations
Post-processing of air quality modeled concentrations Both
to estimate future-year PM2.5 design value and spatial
fields of PM2.5 concentrations
Post-processing of air quality modeled concentrations Both
to estimate future-year visibility design value
"Rollback" methodology for simulating full-attainment Both
Medium Medium
Medium- Medium
high
Medium
High
High
High
Medium Medium
emerging devices that may be available in future
years
Control efficiency data is dependent upon
equipment being well maintained.
Area source controls assume a constant estimate of
emission reductions, despite variability in extent
and scale of application.
Tierl
Tierl
Tierl
Tierl
High Medium Tier 1
Tierl
Uncertainties Associated with Control Strategy Development
Control Technology Data
• Technologies annlipd mav not rpflprt m
Both Medium- High
ostrurrpnt high
Tier 2
Control Strategy Development
• States may develop different control strategies
than the ones illustrated
• Lack of data on baseline controls from current SIPs
Both Medium- Medium-
high high
TierO
4-17
-------
• Timing of control strategies may be different than
envisioned in RIA
• Controls are applied within the county with the
exceeding monitor. In some cases, additional
known controls are also applied in adjacent
contributing counties.
• Emissions growth and control from new sources
locating in these analysis areas is not included.
Technological Change Likely over- Medium- Low TierO
Emission reductions do not reflect potential effects estimate high
of technological change that may be available in
future years
• Effects of "learning by doing" are no accounted for
in the emission reduction estimates
Emission Reductions from Unidentified Controls Both High Low Tier 1
• emission control cut points for each pollutant
a Magnitude of Impact
High—If error could influence the total costs by more than 25%
Medium—If error could influence the total costs by 5%-25%
Low—If error could influence the total costs by less than 5%
b Degree of Confidence in Our Analytic Approach
High—The current evidence is plentiful and strongly supports the selected approach
Medium—Some evidence exists to support the selected approach, but data gaps are present
Low—Limited data exists to support the selected approach
c Ability to Assess Uncertainty (using WHO Uncertainty Framework)
Tier 0—Screening level, generic qualitative characterization
Tier 1—Scenario-specific qualitative characterization
Tier 2—Scenario-specific sensitivity analysis
Tier 3—Scenario-specific probabilistic assessment of individual and combined uncertainty
Future expected emissions are difficult to predict because they depend on many independent factors. Emission
inventories are aggregated from many spatially and technically diverse sources of emissions, so simplifying
assumptions are necessary to make estimating the future tractable.
4.4 References
U.S. Environmental Protection Agency (U.S. EPA).1999. Control of Air Pollution from New Motor
Vehicles: Tier 2 Motor Vehicle Emissions Standards and Gasoline Sulfur Control
Requirements. Office of Transportation and Air Quality. Available at
http://www.epa.gov/tier2/frm/fr-t2reg.pdf.
U.S. Environmental Protection Agency (U.S. EPA).2000. Heavy-Duty Engine and Vehicle
Standards and Highway Diesel Fuel Sulfur Control Requirements. Office of
Transportation and Air Quality. Available at http://www.epa.gov/otaq/highway-
diesel/index.htm.
4-18
-------
U.S. Environmental Protection Agency (U.S. EPA). 2004. Control of Emissions of Air Pollution
from Nonroad Diesel Engines and Fuel. Office of Transportation and Air Quality.
Available at
http://www. regulations.gov/sea rch/Regs/contentStreamer?obiectld=09000064800be20
3&disposition=attachment&contentType=pdf.
U.S. Environmental Protection Agency (U.S. EPA).2005. New Emission Standards for New
Commercial Aircraft Engines. Office of Transportation and Air Quality. Available at
http://www.epa.gov/oms/regs/nonroad/aviation/420f05015.htm.
U.S. Environmental Protection Agency (U.S. EPA).2005b. Regional Haze regulations and
Guidelines for Best Available Retrofit Technology Determinations. Office of Air Quality
Planning and Standards. Available athttp://www.epa.gov/fedrgstr/EPA-
AIR/2005/Julv/Day-06/al2526.pdfand
http://www.epa.gov/visibility/fs 2005 6 15.html.
U.S. Environmental Protection Agency (U.S. EPA). 2005. Clean Air Interstate Rule. Office of
Atmospheric Programs. Washington, D.C. Available at
http://edocket.access.gpo.gov/2005/pdf/05-5723.pdf.
U.S. Environmental Protection Agency (U.S. EPA). 2007. Guidance on the Use of Models and
other Analyses for Demonstrating Attainment of Air Quality Goals for Ozone, PM2.5,
and Regional Haze. Office of Air Quality Planning and Standards, Research Triangle Park,
NC. Available at http://www.epa.gov/ttn/scram/guidance/guide/final-03-pm-rh-
guidance.pdf.
U.S. Environmental Protection Agency (U.S. EPA).2008. Control of Emissions for Nonroad Spark
Ignition Engines and Equipment. Office of Transportation and Air Quality. Available at
http://www.epa.gov/otaq/regs/nonroad/marinesi-equipld/bondfrm.pdf.
U.S. Environmental Protection Agency (U.S. EPA).2008. Emissions Standards for Locomotives
and Marine Compression-Ignition Engines. Office of Transportation and Air Quality.
Available at http://www.epa.gov/nonroad/420f08004.pdf.
U.S. Environmental Protection Agency (U.S. EPA). 2009. Standards of Performance for New
Stationary Sources and Emission Guidelines for Existing Sources:
Hospital/Medical/lnfectious Waste Incinerators. Office of Air Quality Planning and
Standards, Research Triangle Park, NC. Available at
http://www.epa.gov/ttn/atw/129/hmiwi/fr06oc09.pdf.
U.S. Environmental Protection Agency (U.S. EPA).2010. Control of Emissions from New Marine
Compression-Ignition Engines at or Above 30 Liters per Cylinder. Office of
Transportation and Air Quality. Available at
http://www. regulations.gov/sea rch/Regs/contentStreamer?obiectld=0900006480ae43a
6&disposition=attachment&contentType=pdf.
4-19
-------
U.S. Environmental Protection Agency (U.S. EPA). 2011. Control Strategy Tool (CoST)
Documentation Report. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. Available at http://www.epa.gov/ttnecasl/cost.htm.
U.S. Environmental Protection Agency (U.S. EPA). 2011. Cross-State Air Pollution Rule. Office of
Atmospheric Programs. Washington, D.C. Available at
http://www.gpo.gov/fdsvs/pkg/FR-2011-08-08/pdf/2011-17600.pdf.
U.S. Environmental Protection Agency (U.S. EPA). 2011. Mercury and Air Toxics Standards.
Office of Atmospheric Programs. Washington, D.C. Available at
www.gpo.gov/fdsys/pkg/FR-2012-02-16/pdf/2012-806.pdf.
4-20
-------
APPENDIX 4.A
ADDITIONAL CONTROL STRATEGY INFORMATION
4.A.1 Control Measures for Stationary Sources
This appendix describes measures that were employed in this analysis to illustrate a
hypothetical scenario for controlling emissions of PM and precursors from non-EGU point and
area source categories to attain alternative annual and 24-hr air quality standards for PM2.5.
Most of the control measures available are add-on technologies but some other technologies
and practices that are not add-on in nature can reduce emissions of PM and PM precursors.
4.A.1.1 PM Emissions Control Technologies1
This section summarizes control measures focused on reduction of PM2.s from non-EGU
point and area sources. However, it should be noted that PMi0 will also be reduced by these
measures. The amount of PMi0 reduction varies by the fraction of PMi0 in the inlet stream to
the control measure and the specific design of the measure.
4.A. 1.1.1 PM Control Measures for NonEGU Point Sources
Most control measures for non-EGU point sources are add-on technologies. These
technologies include: fabric filters (baghouses), ESPs, and wet PM scrubbers. Fabric filters
collect particles with sizes ranging from below 1 micrometer to several hundred micrometers in
diameter at efficiencies in excess of 99%, and this device is used where high-efficiency particle
collection is required. A fabric filter unit consists of one or more isolated compartments
containing rows of fabric bags in the form of round, flat, or shaped tubes, or pleated cartridges.
Particle-laden gas passes up (usually) along the surface of the bags then radially through the
fabric. Particles are retained on the upstream face of the bags, and the cleaned gas stream is
vented to the atmosphere. The filter is operated cyclically, alternating between relatively long
periods of filtering and short periods of cleaning. Dust that accumulates on the bags is removed
from the fabric surface when cleaning and deposited in a hopper for subsequent disposal.
ESPs use electrical forces to move particles out of a flowing gas stream and onto
collector plates. The particles are given an electrical charge by forcing them to pass through a
corona, a region in which gaseous ions flow. The electrical field that forces the charged particles
to the plates comes from electrodes maintained at high voltage in the center of the flow lane.
Once particles are on the collector plates, they must be removed without re-entraining them
1 The descriptions of add-on technologies throughout this section are taken from the EPA Air Pollution Control Cost
Manual, Sixth Edition. This is found on the Internet at http://epa.gov/ttn/catc/products.htmlffcccinfo.
4.A-1
-------
into the gas stream. This is usually accomplished by rapping the plates mechanically which
loosens the collected particles from the collector plates, allowing the particles to slide down
into a hopper from which they are evacuated. This removal of collected particles is typical of a
"dry" ESP. A "wet" ESP operates by having a water flow applied intermittently or continuously
to wash away the collected particles for disposal. The advantage of wet ESPs is that there are
no problems with rapping re-entrainment or with "back coronas" (unintended injection of
positively charged ions which reduces the charge on particles and lowers the collection
efficiency). The disadvantage is that the collected slurry must be handled more carefully than a
dry product, adding to the expense of disposal. ESPs capture particles with sizes ranging from
below 1 micrometer to several hundred micrometers in diameter at efficiencies from 95 to up
to 99% and higher.
Wet PM scrubbers remove PM and acid gases from waste gas streams of stationary
point sources. The pollutants are removed primarily through the impaction, diffusion,
interception and/or absorption of the pollutant onto droplets of liquid. The liquid containing
the pollutant is then collected for disposal. Collection efficiencies for wet scrubbers vary by
scrubber type, and with the PM size distribution of the waste gas stream. In general, collection
efficiency decreases as the PM size decreases. Collection efficiencies range from in excess of
99% for venturi scrubbers to 40% to 60% for simple spray towers. Wet scrubbers are generally
smaller and more compact than fabric filters or ESPs, and have lower capital cost and
comparable operation and maintenance (O&M) costs. Wet scrubbers, however, operate with a
higher pressure drop than either fabric filters or ESPs, thus leading to higher energy costs. In
addition, they are limited to lower waste gas flow rates and operating temperatures than fabric
filters or ESPs, and also generate sludge that requires additional treatment or disposal. This RIA
only applies wet scrubbers to fluid catalytic cracking units (FCCUs) at petroleum refineries.
In addition, we also examined additional add-on control measures specifically for steel
mills. Virtually all steel mills have some type of PM control measure, but there is additional
equipment that in many cases could be installed to further reduce emissions. Capture hoods
that route PM emissions from a blast furnace casthouse to a fabric filter can provide 80% to
90% additional emission reductions from a steel mill. Other capture and control systems at
blast oxygen furnaces (BOFs) can also provide 80% to 90% additional reductions.
Table 4.A-1 lists some of these technologies. For more information on these
technologies, refer to the EPA Air Pollution Control Cost Manual.1
4.A-2
-------
Table 4.A-1. Example PM Control Measures for NonEGU Point Source Categories
Control Measure
Sector(s) to which Control
Measure Can Apply
Control
Efficiency
(percent)
Average
Annualized
Cost/Ton
Fabric Filters
ESPs—wet or dry3
Industrial Boilers, Iron and Steel
Mills, Pulp and Paper Mills
Industrial Boilers, Iron and Steel
Mills, Pulp and Paper Mills
98 to 99.9 $2,000-$ 100,000
95 to 99.9 $1,000-$20,000
Wet Scrubbers
Secondary Capture and Control
Systems—Capture Hoods for Blast
Oxygen Furnaces
CEM Upgrade and Increased
Monitoring Frequency
Industrial Boilers, Iron and Steel 40 to 99 $750-$2,800
Mills
Coke Ovens 80 to 90 $5,000
NonEGUs with an ESP 5 to 7 $600-$5,000
a CoST contains equations to estimate capital and annualized costs for ESP and FF installation and operation. The
average annualized cost/ton estimates presented here for these control measures are outputs from our
modeling, not inputs. They also reflect applications of control where there is no PM control measure currently
operating except if the control measure is an upgrade (e.g., ESP upgrades).
4.A. 1.1.2 PM Control Measures for Area Sources
Specific controls exist for a number of stationary area sources. Area source PM controls
at stationary sources include:
• catalytic oxidizers on conveyorized charbroilers at restaurants (up to 80% reduction
ofPM),
• replacement of older woodstoves with ones compliant with the New Source
Performance Standard (NSPS) for residential wood combustion (up to 98% reduction
of PM2), and
• education and advisory programs to help users to operate woodstoves more
efficiently and with fewer emissions (up to 50% reduction of PM)
Another PM area source control measure, diesel particulate filters, can be applied to existing
diesel-fueled compression-ignition (C-l) engines to achieve up to a 90% reduction in fine PM.
This measure is being applied to new C-l engines as part of a NSPS that was implemented
beginning in 2006.
" This control measure is largely meant to simulate the effects of a woodstove changeout program as applied to
Libby, MT per the efforts of the U.S. EPA and several co-sponsors. For more information, refer to
http://www.epa.gov/woodstoves/how-to-guide.html.
4.A-3
-------
Table 4.A-2. Example PM Control Measures for Area Sources3
Control Measures
Sectors to which
These Control
Measures Can Apply
Control
Efficiency
(percent)
Average
Annualized
Cost/ton
Restaurants
Residential wood
combustion sources
Catalytic oxidizers for conveyorized charbroilers
Changeout of older woodstoves for new ones by a
woodstove changeout campaign or on sale of
property, or an education and advisory program
for woodstove users
Replace open burning of wood waste with Residential waste
chipping for landfill disposal sources
83
46 to near 100
Near 100
$1,300
$1,900
$3,500
a The estimates for these control measures reflect applications of control where there is no PM area source control
measure currently operating. Also, the control efficiency is for total PM, and thus accounts for PM10 and PM2.5.
Data for these measures is available in the CoST Control Measures Documentation Report at
http://www.epa.gov/ttn/ecas/models/CoST CMDB Document 2010-06-09.pdf.
4.A.1.2 SO2 Control Measures
4.A.1.2.1S02 Control Measures for NonEGU Point Sources
The S02 emission control measures used in this analysis are similar to those used in the
PM2.5 RIA prepared about four years ago. Flue gas desulfurization (FGD) scrubbers can achieve
95-98% control of S02for nonEGU point sources and for utility boilers. Spray dryer absorbers
(SDA) are another commonly employed technology, and SDA can achieve up to 90% or more
control of S02. For specific source categories, other types of control technologies are available
that are more specific to the sources controlled. Table 4.A-3 lists some of these technologies.
For more information on these technologies, please refer to the CoST control measures
documentation report.3
1 For a complete description of the control technologies used in CoST, please refer to the report at
http://www.epa.gov/ttn/ecas/models/CoST CMDB Document 2010-06-09.pdf.
4.A-4
-------
Table 4.A-3. Example SO2 Control Measures for NonEGU Point3
Control Measure
Wet and Dry FGD
scrubbers and SDA
Sectors to Which These Control
Measures Can Be Applied
ICI boilers— all fuel types, kraft pulp
mills, Mineral Products (e.g., Portland
cement plants (all fuel types), primary
metal plants, petroleum refineries
Control
Efficiency
(percent)
95-FGD
scrubbers,
90-forSDA
Average Annualized
Cost/Ton (2006$)
$800-$8,000-FGD
$900-$7,000-SDA
Increase percentage Sulfur recovery plants
sulfur conversion to meet
sulfuric acid NSPS (99.7%
reduction)
Sulfur recovery and/or tail Sulfuric Acid Plants
gas treatment
Cesium promoted catalyst Sulfuric Acid Plants with Double-
Absorption process
75-95
95-98
50%
$4,000
$1,000-$4,000
$1,000
Sources: CoST control measures documentation report, May 2008, NESCAUM Report on Applicability of NOX,
SO2, and PM Control Measures to Industrial Boilers, November 2008 available at
http://www.nescaum.org/documents/ici-boilers-20081118-final.pdf, and Comprehensive Industry
Document on Sulphuric Acid Plant, Govt. of India Central Pollution Control Board, May 2007. The estimates for
these control measures reflect applications of control where there is no SO2 control measure currently operating
except for the Cesium promoted catalyst.
4.A.1.2.2S02 Control Technology for Area Sources
Fuel switching from high to low-sulfur fuels is the predominant control measure
available for S02 area sources. For home heating oil users, our analyses included switching from
a high-sulfur oil (approximately 2,500 parts per million (ppm) sulfur content) to a low-sulfur oil
(approximately 500 ppm sulfur). A similar control measure is available for oil-fired industrial
boilers. For more information on these measures, please refer to the CoST control measures
documentation report.3
4.A.1.3 NOX Emissions Control Measures
4.A. 1.3.1 A/0X Control Measures for Non-EGU Point Sources
This section describes available measures for controlling emissions of NOX from non-EGU
point sources. In general, Iow-N0x burners (LNB) are often applied as a control technology for
industrial boilers and for some other non-EGU sources because of their wide applicability and
cost-effectiveness. While all controls presented in this analysis are considered generally
technically feasible for each class of sources, source-specific cases may exist where a control
technology is in fact not technically feasible.
4.A-5
-------
Several types of NOX control technologies exist for non-EGU sources: selective catalytic
reduction (SCR), selective noncatalytic reduction (SNCR), natural gas reburn (NCR), coal reburn,
and Iow-N0x burners. The two control measures chosen most often were LNB and SCR because
of their breadth of application. In some cases, LNB accompanied by flue gas recirculation (FGR)
is applicable, such as when fuel-borne NOX emissions are expected to be of greater importance
than thermal NOX emissions. When circumstances suggest that combustion controls are not
feasible as a control technology (e.g., sintering processes, coke oven batteries, sulfur recovery
plants), SNCR or SCR may be an appropriate choice. Finally, SCR can be applied along with a
combustion control such as LNB with overfire air (OFA) to further reduce NOX emissions. All of
these control measures are available for application on industrial boilers.
Besides industrial boilers, other non-EGU source categories covered in this RIA include
petroleum refineries, kraft pulp mills, cement kilns, stationary internal combustion engines,
glass manufacturing, combustion turbines, and incinerators. NOX control measures available for
petroleum refineries, particularly process heaters at these plants, include LNB, SNCR, FGR, and
SCR along with combinations of these technologies. NOX control measures available for kraft
pulp mills include those available to industrial boilers, namely LNB, SCR, SNCR, along with water
injection (Wl). NOX control measures available for cement kilns include those available to
industrial boilers, namely LNB, SCR, and SNCR. In addition, mid-kiln firing (MKF), ammonia-
based SNCR, and biosolids injection can be used on cement kilns where appropriate. Non-
selective catalytic reduction (NSCR) can be used on stationary internal combustion engines.
OXY-Firing, a technique to modify combustion at glass manufacturing plants, can be used to
reduce NOX emissions at such plants. LNB, SCR, and SCR combined with steam injection (SI) are
available measures for combustion turbines. Finally, SNCR is an available control technology at
incinerators. Table 4.A-4 lists typical examples of the control measures available for these
categories. For more information on these measures, please refer to the CoST control measures
documentation report.3
4.A-6
-------
Table 4.A-4. Example NOX Control Measures for NonEGU Source Categories3
Sectors to Which These Control Measures Control Efficiency Average Annualized
Control Measures Apply (percent) Cost/ton
LNB
LNB+FGR
SNCR (urea-based or
Industrial boilers— all fuel types, Petroleum
refineries, Cement manufacturing, Pulp
and Paper mills
Petroleum refineries
Industrial boilers— all fuel types, Petroleum
25 to 50%
55
45 to 75
$200 to $1,000
$4,000
$1,000 to $2,000
not)
SCR
refineries, Cement manufacturing, pulp
and paper mills, incinerators
Industrial boilers—all fuel types, Petroleum
refineries, Cement manufacturing, pulp
and paper mills, Combustion turbines
80 to 90
$2,000 to 7,000
OXY-Firing
NSCR
MKF
Biosolids Injection
SCR + SI
Glass manufacturing
Stationary internal combustion engines
Cement manufacturing— dry
Cement manufacturing— dry
Industrial boilers— all fuel types
85
90
25
23
95
$2,500 to 6,000
500
-$460 to 720
$300
$2,700
Source: CoST control measures documentation report (June 2010). Note: a negative sign indicates a cost savings
from application of a control measure. The estimates for these control measures reflect applications of control
where there is no NOX control measure currently operating except for post-combustion controls such as SCR and
SNCR. For these measures, the costs presume that a NOX combustion control (such as LNB) is already operating
on the unit to which the SCR or SNCR is applied.
4.A.2 Projected Monitor Design Values
Table 4.A-5. Area County Definitions for SO2 and NOX Emissions Reductions for Control
Strategy Analysis
FIPS Code
1007
1009
1073
1115
1117
1125
1127
State Name
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
County Name
Bibb
Blount
Jefferson
St. Clair
Shelby
Tuscaloosa
Walker
Area Label
AL
AL
AL
AL
AL
AL
AL
(continued)
4.A-7
-------
Table 4.A-5. Area County Definitions for SO2 and NOX Emissions Reductions for Control
Strategy Analysis (continued)
FIPS Code
42003
42005
42007
42019
42125
42129
4003
4019
4023
13021
13079
13153
13169
13207
13225
13289
6007
6011
6021
6063
6103
6115
17031
17043
17097
17197
16005
16007
State Name
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Arizona
Arizona
Arizona
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
California
California
California
California
California
California
Illinois
Illinois
Illinois
Illinois
Idaho
Idaho
County Name
Allegheny
Armstrong
Beaver
Butler
Washington
Westmoreland
Cochise
Pima
Santa Cruz
Bibb
Crawford
Houston
Jones
Monroe
Peach
Twiggs
Butte
Colusa
Glenn
Plumas
Tehama
Yuba
Cook
Du Page
Lake
Will
Bannock
Bear Lake
Area Label
all
all
all
all
all
all
AZ
AZ
AZ
bib
bib
bib
bib
bib
bib
bib
but
but
but
but
but
but
coo
coo
coo
coo
fra
fra
(continued)
4.A-8
-------
Table 4.A-5. Area County Definitions for SO2 and NOX Emissions Reductions for Control
Strategy Analysis (continued)
FIPS Code
16029
16041
16071
13045
13057
13063
13067
13077
13089
13097
13113
13121
13135
13117
6025
6073
42011
42029
42043
42071
42075
42133
16023
16033
16037
16049
16059
16085
State Name
Idaho
Idaho
Idaho
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
California
California
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
County Name
Caribou
Franklin
Oneida
Carroll
Cherokee
Clayton
Cobb
Coweta
De Kalb
Douglas
Fayette
Fulton
Gwinnett
Forsyth
Imperial
San Diego
Berks
Chester
Dauphin
Lancaster
Lebanon
York
Butte
Clark
Custer
Idaho
Lemhi
Valley
Area Label
fra
fra
fra
ful
ful
ful
ful
ful
ful
ful
ful
ful
ful
ful
imp
imp
Ian
Ian
Ian
Ian
Ian
Ian
lem
lem
lem
lem
lem
lem
(continued)
4.A-9
-------
Table 4.A-5. Area County Definitions for SO2 and NOX Emissions Reductions for Control
Strategy Analysis (continued)
FIPS Code
17005
17027
17083
17117
17119
17135
17163
26099
26115
26125
26161
26163
30001
30023
30029
30039
30047
30053
30061
30063
30077
30081
30089
36005
36047
36061
36081
36085
State Name
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Michigan
Michigan
Michigan
Michigan
Michigan
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
New York
New York
New York
New York
New York
County Name
Bond
Clinton
Jersey
Macoupin
Madison
Montgomery
StClair
Macomb
Monroe
Oakland
Washtenaw
Wayne
Beaverhead
Deer Lodge
Flathead
Granite
Lake
Lincoln
Mineral
Missoula
Powell
Ravalli
Sanders
Bronx
Kings
New York
Queens
Richmond
Area Label
mad
mad
mad
mad
mad
mad
mad
Ml
Ml
Ml
Ml
Ml
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
NY
NY
NY
NY
NY
(continued)
4.A-10
-------
Table 4.A-5. Area County Definitions for SO2 and NOX Emissions Reductions for Control
Strategy Analysis (continued)
FIPS Code
39035
39055
39085
39093
39103
39133
39153
41003
41019
41029
41035
41037
41039
41041
41043
41017
6003
6005
6013
6017
6061
6067
6095
6101
6113
6019
6027
6029
State Name
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
California
California
California
California
California
California
California
California
California
California
California
California
County Name
Cuyahoga
Geauga
Lake
Lorain
Medina
Portage
Summit
Benton
Douglas
Jackson
Klamath
Lake
Lane
Lincoln
Linn
Deschutes
Alpine
Amador
Contrasta
El Dorado
Placer
Sacramento
Solano
Sutter
Yolo
Fresno
Inyo
Kern
Area Label
OH
OH
OH
OH
OH
OH
OH
OR
OR
OR
OR
OR
OR
OR
OR
OR
sac
sac
sac
sac
sac
sac
sac
sac
sac
san
san
san
(continued)
4.A-11
-------
Table 4.A-5. Area County Definitions for SO2 and NOX Emissions Reductions for Control
Strategy Analysis (continued)
FIPS Code
6031
6039
6043
6047
6051
6053
6069
6077
6079
6085
6099
6107
6109
16009
16017
16035
16055
16057
16079
6037
6059
6065
6071
6111
48039
48071
48157
48167
State Name
California
California
California
California
California
California
California
California
California
California
California
California
California
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
California
California
California
California
California
Texas
Texas
Texas
Texas
County Name
Kings
Madera
Mariposa
Merced
Mono
Monterey
San Benito
San Joaquin
San Luis Obispo
Santa Clara
Stanislaus
Tula re
Tuolumne
Benewah
Bonner
Clearwater
Kootenai
Latah
Shoshone
Los Angeles
Orange
Riverside
San Bernardino
Ventura
Brazoria
Chambers
Fort Bend
Galveston
Area Label
san
san
san
san
san
san
san
san
san
san
san
san
san
sho
sho
sho
sho
sho
sho
sou
sou
sou
sou
sou
TX
TX
TX
TX
(continued)
4.A-12
-------
Table 4.A-5. Area County Definitions for SO2 and NOX Emissions Reductions for Control
Strategy Analysis (continued)
FIPS Code
48201
48291
48339
48473
49003
49005
49033
49057
49007
49011
49013
49023
49029
49035
49039
49043
49045
49049
49051
53033
53035
53037
53041
53045
53053
53067
53077
State Name
Texas
Texas
Texas
Texas
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Washington
Washington
Washington
Washington
Washington
Washington
Washington
Washington
County Name
Harris
Liberty
Montgomery
Waller
Box Elder
Cache
Rich
Weber
Carbon
Davis
Duchesne
Juab
Morgan
Salt Lake
Sanpete
Summit
Tooele
Utah
Wasatch
King
Kitsap
Kittitas
Lewis
Mason
Pierce
Thurston
Yakima
Area Label
TX
TX
TX
TX
Utahl
Utahl
Utahl
Utahl
Utah 2
Utah 2
Utah 2
Utah 2
Utah 2
Utah 2
Utah 2
Utah 2
Utah 2
Utah 2
Utah 2
WA
WA
WA
WA
WA
WA
WA
WA
4.A-13
-------
Table 4.A-6. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) Above 15 or
24-hr Design Value (DV) Above 35 in 2020 Base Case
FIPS
Code
42003
30047
30063
30081
6019
6029
6031
6099
6107
6037
6065
6071
49005
49035
High Monitor ID
420030064
300470028
300630031
300810007
60190008
60290010
60310004
60990005
61072002
60371301
60658001
60719004
490050004
490353007
State Name
Pennsylvania
Montana
Montana
Montana
California
California
California
California
California
California
California
California
Utah
Utah
County Name
Allegheny
Lake
Missoula
Ravalli
Fresno
Kern
Kings
Stanislaus
Tulare
Los Angeles
Riverside
San Bernardino
Cache
Salt Lake
2005 Annual
DV
20.31
9.00
10.52
9.01
16.99
18.94
17.28
14.21
18.51
17.66
20.95
19.01
11.56
12.02
2020 Annual
DV
12.91
7.71
9.13
7.89
12.71
14.34
13.24
10.85
14.10
13.14
16.30
14.96
9.78
9.72
2005 24-hr DV
64.2
43.6
44.6
45.1
60.2
64.5
58.0
51.4
56.6
48.7
59.1
55.5
56.9
45.3
2020 24-hr DV
41.0
38.3
37.4
37.3
41.0
45.9
42.4
37.0
40.1
39.7
46.5
41.5
42.6
36.1
24-hr to Annual
DV
Responsiveness
3.03
3.78
3.78
3.78
3.45
3.45
3.45
3.45
3.45
3.45
3.45
3.45
4.91
4.91
(continued)
-------
Un
Table 4.A-6. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) Above 15 or
24-hr Design Value (DV) Above 35 in 2020 Base Case (continued)
FIPS
Code
42003
30047
30063
30081
6019
6029
6031
6099
6107
6037
6065
6071
49005
49035
High
Monitor ID
420030064
300470028
300630031
300810007
60190008
60290010
60310004
60990005
61072002
60371301
60658001
60719004
490050004
490353007
State Name
Pennsylvania
Montana
Montana
Montana
California
California
California
California
California
California
California
California
Utah
Utah
County Name
Allegheny
Lake
Missoula
Ravalli
Fresno
Kern
Kings
Stanislaus
Tulare
Los Angeles
Riverside
San Bernardino
Cache
Salt Lake
NH4SO4
Component of
2005 Annual DV
9.1830
0.9752
1.2692
0.7614
2.3402
3.1194
2.7393
2.2879
2.9375
5.5145
4.5070
3.7956
1.4901
1.7974
NH4NO3
Component of
2005 Annual DV
0.7303
0.0936
1.0152
0.2929
4.1985
5.8850
4.7392
3.8313
5.3329
3.5811
5.2955
3.8907
2.1042
2.8766
Direct PM2.5
Component of
2005 Annual DV
10.4046
7.9383
8.2419
7.9577
10.4583
9.9388
9.8017
8.0978
10.2433
8.5679
11.1521
11.3297
7.9703
7.3465
NH4SO4
Component of
2020 Annual DV
4.1977
0.9007
1.1767
0.7034
1.9907
2.5838
2.2500
1.8519
2.4607
3.7217
3.3590
2.8912
1.2447
1.5642
NH4NO3
Component of
2020 Annual DV
0.7669
0.0730
0.7734
0.2380
2.7196
3.7690
2.8483
2.6903
3.2405
2.9601
3.9822
2.7994
1.5831
2.2080
(continued)
-------
Table 4.A-6. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) above 15 or
24-hr Design Value (DV) Above 35 in 2020 Base Case (continued)
FIPS
Code
42003
30047
30063
30081
6019
6029
> 6031
01 6099
6107
6037
6065
6071
49005
49035
High
Monitor ID
420030064
300470028
300630031
300810007
60190008
60290010
60310004
60990005
61072002
60371301
60658001
60719004
490050004
490353007
State Name
Pennsylvania
Montana
Montana
Montana
California
California
California
California
California
California
California
California
Utah
Utah
County Name
Allegheny
Lake
Missoula
Ravalli
Fresno
Kern
Kings
Stanislaus
Tulare
Los Angeles
Riverside
San Bernardino
Cache
Salt Lake
Direct PM2.5
Component
of 2020
Annual DV
7.9528
6.7370
7.1876
6.9574
8.0008
7.9904
8.1443
6.3118
8.4003
6.4628
8.9602
9.2726
6.9584
5.9562
Area
Label
all
MT
MT
MT
san
san
san
san
san
sou
sou
sou
Utahl
Utah 2
Change in Area SO2
Emissions Between
2005 and 2020
259568
853
853
853
5855
5855
5855
5855
5855
20834
20834
20834
639
8442
Change in Area NOX
Emissions Between
2005 and 2020
69304
9949
9949
9949
106130
106130
106130
106130
106130
230102
230102
230102
8192
33871
Change in County
PM2.s Emissions
Between 2005 and
2020
804
128
206
144
1151
1241
132
571
592
5223
1421
2026
124
799
(continued)
-------
Table 4.A-6. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) above 15 or
24-hr Design Value (DV) Above 35 in 2020 Base Case (continued)
FIPS
Code
42003
30047
30063
30081
6019
6029
6031
6099
6107
6037
6065
6071
49005
49035
High
Monitor ID
420030064
300470028
300630031
300810007
60190008
60290010
60310004
60990005
61072002
60371301
60658001
60719004
490050004
490353007
State Name
Pennsylvania
Montana
Montana
Montana
California
California
California
California
California
California
California
California
Utah
Utah
County Name
Allegheny
Lake
Missoula
Ravalli
Fresno
Kern
Kings
Stanislaus
Tulare
Los Angeles
Riverside
San Bernardino
Cache
Salt Lake
SO2 Air Quality
Ratio (u.g/m3
change in SO4
per 1,000 Tons
S02)
0.008
0.081
0.106
0.063
0.062
0.080
0.070
0.057
0.076
0.052
0.047
0.040
0.321
0.028
NOx Air Quality
Ratio (ug/m3
Change in NO3
per 1,000 Tons
NOX)
0.000
0.001
0.015
0.005
0.008
0.011
0.009
0.008
0.010
0.003
0.004
0.003
0.040
0.013
PM2.5 Air Quality
Ratio (|jg/m3
Change in Direct
PM2.5 per 1,000
Tons PM)
1.238
1.929
1.929
1.929
1.879
1.879
1.879
1.136
1.879
0.597
0.597
0.597
1.929
1.929
24-hr DV
Reduction
Needed for 35
5.543
2.878
1.990
1.815
5.548
10.442
6.939
1.578
4.634
4.236
11.081
6.041
7.168
1.240
Annual DV
Reduction
Associated With
the 24-hr DV
Meeting 35
1.829
0.761
0.526
0.480
1.608
3.027
2.011
0.457
1.343
1.228
3.212
1.751
1.460
0.253
-------
Table 4.A-7. Air Quality Ratios for Monitors in Counties With at Least One Monitor With an Annual Design Value (DV) Above 13 in
2020 Baseline (15/35)
FIPS
Code
6065
6071
High
Monitor ID
60658001
60710025
State
Name
California
California
County Name
Riverside
San Bernardino
2005 Annual
DV
20.95
19.67
2020
Annual DV
16.30
15.23
2005 24-hr
DV
59.1
51.9
2020 24-
hr DV
46.5
41.2
2020 15/35
Annual DV
13.08
13.12
NH4SO4 Component of
2005 Annual DV
4.5070
4.2309
(continued)
Table 4.A-7. Air Quality Ratios for Monitors in Counties With at Least One Monitor With an Annual Design Value (DV) Above 13 in
2020 Baseline (15/35) (continued)
FIPS
Code
High
Monitor
ID
State
Name
NH4NO3
Component
of 2005
County Name Annual DV
Direct PM2.5
Component
of 2005
Annual DV
NH4SO4
Component
of 2020
Annual DV
NH4NO3
Component
of 2020
Annual DV
Direct PM2.5
Component
of 2020
Annual DV
Area
Label
Change in Area
SO2 Emissions
Between 2005
and 2020
co
6065 60658001 California Riverside 5.2955 11.1521 3.3590 3.9822 8.9602 sou 20834
6071 60710025 California San Bernardino 4.0013 11.4396 3.0564 3.0890 9.0859 sou 20834
(continued)
Table 4.A-7. Air Quality Ratios for Monitors in Counties With at Least One Monitor With an Annual Design Value (DV) Above 13 in
2020 Baseline (15/35) (continued)
FIPS
Code
6065
6071
High
Monitor
ID
60658001
60710025
State
Name
California
California
County Name
Riverside
San Bernardino
Change in
Area NOX
Emissions
Between 2005
and 2020
230102
230102
Change in
County PM2.s
Emissions
Between 2005
and 2020
1421
2026
SO2 Air Quality
Ratio (u.g/m3
Change in SO4
per 1,000 Tons
S02)
0.047
0.043
NOx Air Quality
Ratio (ug/m3
Change in NO3
per 1,000 Tons
NOX)
0.004
0.003
PM2.5 Air Quality
Ratio (ug/m3
Change in Direct
PM2.5 per 1,000
Tons PM)
0.597
0.597
Annual DV
Reduction
Needed for
12 (ug/m3)
0.039
0.077
-------
Table 4.A-8. Air Quality Ratios for Monitors in Counties with at Least One Monitor With an Annual Design Value (DV) Above 12 in
2020 Baseline (15/35)
FIPS
Code
30053
6065
6071
26163
1073
4023
High
Monitor ID
300530018
60658001
60710025
261630033
10730023
40230004
State
Name
Montana
California
California
Michigan
Alabama
Arizona
2005 2020 2005 2020 202015/35 NH4SO4 Component of
County Name Annual DV Annual DV 24-hr DV 24-hr DV Annual DV 2005 Annual DV
Lincoln
Riverside
San Bernardino
Wayne
Jefferson
Santa Cruz
14.93
20.95
19.67
17.50
18.57
12.94
** Table 4.A-8. Air Quality Ratios for Monitors in Counties with
5 2020 Baseline (15/35) (continued)
ID
FIPS
Code
30053
6065
6071
26163
1073
4023
High
Monitor ID
300530018
60658001
60710025
261630033
10730023
40230004
State
Name
Montana
California
California
Michigan
Alabama
Arizona
County Name
Lincoln
Riverside
San Bernardino
Wayne
Jefferson
Santa Cruz
NH4NO3
Component
of 2005
Annual DV
0.4047
5.2955
4.0013
2.2041
0.2657
0.0149
12.60
16.30
15.23
12.35
12.34
12.06
at Least One
Direct PM2.5
Component
of 2005
Annual DV
13.3936
11.1521
11.4396
8.0960
10.7334
11.2381
42.7 35.3 12.53
59.1 46.5 13.08
51.9 41.2 13.12
43.9 31.3 12.35
44.1 27.9 12.34
36.1 33.8 12.06
Monitor With an Annual Design Value
NH4SO4 NH4NO3 Direct PM2.5
Component Component Component
of 2020 of 2020 of 2020
Annual DV Annual DV Annual DV
1.0433 0.3556 11.2031
3.3590 3.9822 8.9602
3.0564 3.0890 9.0859
4.2754 1.8639 6.2139
3.5837 0.2508 8.5105
1.5034 0.0121 10.5517
1.1342
4.5070
4.2309
7.2067
7.5801
1.6891
(continued)
(DV) Above 12 in
Area
Label
MT
sou
sou
Ml
AL
AZ
Change in
Area SO2
Emissions
Between
2005 and
2020
853
20,834
20,834
142,340
243,497
5,178
(continued)
-------
Table 4.A-8. Air Quality Ratios for Monitors in Counties with at Least One Monitor With an Annual Design Value (DV) Above 12 in
2020 Baseline (15/35) (continued)
NJ
O
FIPS
Code
30053
6065
6071
26163
1073
4023
High Monitor
ID
300530018
60658001
60710025
261630033
10730023
40230004
State Name
Montana
California
California
Michigan
Alabama
Arizona
County Name
Lincoln
Riverside
San Bernardino
Wayne
Jefferson
Santa Cruz
Change in
Area NOX
Emissions
Between
2005 and
2020
9,949
230,102
230,102
142,522
61,210
25,105
Change in
County
PM2.5
Emissions
Between
2005 and
2020
107
1,421
2,026
2,846
2,902
48
SO2 Air
Quality
Ratio (ug/m3
Change in
SO4 per
1,000 Tons
S02)
0.094
0.047
0.043
0.013
0.007
0.038
NOxAir
Quality
Ratio (|jg/m3
Change in
NO3 per
1,000 Tons
NOx)
0.007
0.004
0.003
0.002
0.000
0.000
PM2.5Air
Quality
Ratio (u.g/m3
Change in
Direct PM2.5
per 1,000
Tons PM)
1.929
0.597
0.597
1.238
1.238
1.929
Annual DV
Reduction
Needed for
12(ug/m3)
0.494
1.039
0.857
0.301
0.291
0.011
-------
NJ
Table 4.A-9. Air Quality Ratios for Monitors in Counties With at Least One Monitor With an Annual Design Value (DV) Above 11 in
2020 Baseline (15/35)
FIPS
Code
1073
4023
13021
17031
13121
6025
17119
26163
30053
36061
39035
6031
6107
6071
6059
6065
48201
High
Monitor ID
10730023
40230004
130210007
170310052
131210039
60250005
171191007
261630033
300530018
360610056
390350038
60310004
61072002
60710025
60590007
60658001
482011035
State
Name
Alabama
Arizona
Georgia
Illinois
Georgia
California
Illinois
Michigan
Montana
New York
Ohio
California
California
California
California
California
Texas
County Name
Jefferson
Santa Cruz
Bibb
Cook
Fulton
Imperial
Madison
Wayne
Lincoln
New York
Cuyahoga
Kings
Tulare
San Bernardino
Orange
Riverside
Harris
Change in Area SO2
Area Emissions Between 2005
Label and 2020
AL
AZ
bib
coo
ful
imp
mad
Ml
MT
NY
OH
san
san
sou
sou
sou
TX
243,497
5,178
57,968
100,781
91,890
5,629
40,890
142,340
853
32,091
115,275
5,855
5,855
20,834
20,834
20,834
81,104
Change in Area NOX
Emissions Between 2005
and 2020
61,210
25,105
23,656
153,366
96,664
51,787
26,245
142,522
9,949
47,190
61,836
106,130
106,130
230,102
230,102
230,102
121,308
Change in County PM2.5
Emissions Between 2005
and 2020
2,902
48
276
5,267
1,360
295
700
2,846
107
537
984
132
592
2,026
940
1,421
4,910
(continued)
-------
Table 4.A-9. Air Quality Ratios for Monitors in Counties With at Least One Monitor With an Annual Design Value (DV) Above 11 in
2020 Baseline (15/35) (continued)
NJ
FIPS
Code
1073
4023
13021
17031
13121
6025
17119
26163
30053
36061
39035
6031
6107
6071
6059
6065
48201
High
Monitor ID
10730023
40230004
130210007
170310052
131210039
60250005
171191007
261630033
300530018
360610056
390350038
60310004
61072002
60710025
60590007
60658001
482011035
State
Name
Alabama
Arizona
Georgia
Illinois
Georgia
California
Illinois
Michigan
Montana
New York
Ohio
California
California
California
California
California
Texas
SO2Air Quality Ratio
(ug/m3 Change in SO4
County Name per 1,000 Tons SO2)
Jefferson
Santa Cruz
Bibb
Cook
Fulton
Imperial
Madison
Wayne
Lincoln
New York
Cuyahoga
Kings
Tulare
San Bernardino
Orange
Riverside
Harris
0.0072
0.0376
0.0306
0.0119
0.0180
0.1164
0.0404
0.0128
0.0939
0.0529
0.0164
0.0698
0.0763
0.0428
0.0435
0.0470
0.0164
NOX Air Quality Ratio
(ug/m3 Change in NO3
per 1,000 Tons NOx)
0.0003
0.0001
0.0000
0.0000
0.0003
0.0023
0.0090
0.0019
0.0071
0.0022
0.0048
0.0085
0.0097
0.0031
0.0027
0.0040
0.0001
PM2.5 Air Quality
Ratio (ug/m3 Change
in Direct PM2.5 per
1,000 Tons PM)
1.238
1.929
1.238
1.238
1.238
0.597
1.238
1.238
1.929
1.238
1.238
1.879
1.879
0.597
0.597
0.597
1.238
Annual DV Reduction
Needed for 11
(ug/m3)
1.291
1.011
0.171
0.091
0.061
0.171
0.491
1.301
1.487
0.121
0.741
0.180
0.645
2.077
0.008
2.039
0.561
-------
Table 4.A-10. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) Above 11
or 24-hr Design Value (DV) Above 30 in 2020 Baseline (15/35)
FIPS
Code
1073
42003
4023
13021
6007
17031
16041
13121
6025
42071
16059
17119
26163
30047
30081
30063
30053
36061
High
Monitor ID
10730023
420030064
40230004
130210007
60070002
170310052
160410001
131210039
60250005
420710007
160590004
171191007
261630033
300470028
300810007
300630031
300530018
360610056
State Name
Alabama
Pennsylvania
Arizona
Georgia
California
Illinois
Idaho
Georgia
California
Pennsylvania
Idaho
Illinois
Michigan
Montana
Montana
Montana
Montana
New York
County Name
Jefferson
Allegheny
Santa Cruz
Bibb
Butte
Cook
Franklin
Fulton
Imperial
Lancaster
Lemhi
Madison
Wayne
Lake
Ravalli
Missoula
Lincoln
New York
2005
Annual DV
18.57
20.31
12.94
16.54
12.73
15.75
7.7
17.43
12.71
16.55
N/A
16.72
17.5
9
9.01
10.52
14.93
16.18
2020
Annual DV
12.34
12.91
12.06
11.22
9.56
11.14
6.68
11.11
11.22
10.73
N/A
11.54
12.35
7.71
7.89
9.13
12.6
11.17
Area Label
AL
all
AZ
bib
but
coo
fra
ful
imp
Ian
lem
mad
Ml
MT
MT
MT
MT
NY
Change in Area
SO2 Emissions
Between 2005
and 2020
243,497
259,568
5,178
57,968
2,320
100,781
228
91,890
5,629
119,209
169
40,890
142,340
853
853
853
853
32,091
Change in Area
NOX Emissions
Between 2005
and 2020
61,210
69,304
25,105
23,656
8,494
153,366
3,350
96,664
51,787
42,136
577
26,245
142,522
9,949
9,949
9,949
9,949
47,190
Change in
County PM2.s
Emissions
Between 2005
and 2020
2,902
804
48
276
693
5,267
40
1,360
295
8,866
44
700
2,846
128
144
206
107
537
(continued)
-------
Table 4.A-10. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) Above 11
or 24-hr Design Value (DV) Above 30 in 2020 Baseline (15/35) (continued)
FIPS
Code
39035
41035
41039
6067
6099
6019
6031
6107
16079
6059
6065
6071
48201
49005
49035
49011
49049
53053
High
Monitor ID
390350038
410350004
410392013
60670006
60990005
60190008
60310004
61072002
160790017
60590007
60658001
60710025
482011035
490050004
490350012
490110004
490494001
530530029
State Name
Ohio
Oregon
Oregon
California
California
California
California
California
Idaho
California
California
California
Texas
Utah
Utah
Utah
Utah
Washington
County Name
Cuyahoga
Klamath
Lane
Sacramento
Stanislaus
Fresno
Kings
Tulare
Shoshone
Orange
Riverside
San Bernardino
Harris
Cache
Salt Lake
Davis
Utah
Pierce
2005
Annual DV
17.37
11.2
11.93
11.88
14.21
16.99
17.28
18.51
12.08
15.75
20.95
19.67
15.42
11.56
N/A
10.31
10.52
10.55
2020
Annual DV
11.79
8.54
9.43
8.76
10.85
12.71
13.24
14.1
10.66
11.93
16.3
15.23
11.61
9.78
N/A
8.58
8.8
8.11
Area Label
OH
OR
OR
sac
san
san
san
san
sho
sou
sou
sou
TX
Utah 1
Utah 2
Utah 2
Utah 2
WA
Change in Area
SO2 Emissions
Between 2005
and 2020
115,275
1,489
1,489
9,448
5,855
5,855
5,855
5,855
555
20,834
20,834
20,834
81,104
639
8,442
8,442
8,442
11,269
Change in Area
NOX Emissions
Between 2005
and 2020
61,836
22,686
22,686
42,974
106,130
106,130
106,130
106,130
5,546
230,102
230,102
230,102
121,308
8,192
33,,871
33,871
33,871
91,530
Change in
County PM2.s
Emissions
Between 2005
and 2020
984
769
1,666
1,311
571
1,151
132
592
43
940
1,421
2,026
4,910
124
799
243
383
1,058
(continued)
-------
Table 4.A-10. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) Above 11
or 24-hr Design Value (DV) Above 30 in 2020 Baseline (15/35) (continued)
FIPS
Code
1073
42003
4023
13021
6007
17031
16041
13121
6025
42071
16059
17119
26163
30047
30081
30063
30053
36061
High
Monitor ID
10730023
420030064
40230004
130210007
60070002
170310052
160410001
131210039
60250005
420710007
160590004
171191007
261630033
300470028
300810007
300630031
300530018
360610056
State Name
Alabama
Pennsylvania
Arizona
Georgia
California
Illinois
Idaho
Georgia
California
Pennsylvania
Idaho
Illinois
Michigan
Montana
Montana
Montana
Montana
New York
County Name
Jefferson
Allegheny
Santa Cruz
Bibb
Butte
Cook
Franklin
Fulton
Imperial
Lancaster
Lemhi
Madison
Wayne
Lake
Ravalli
Missoula
Lincoln
New York
SO2Air
Quality Ratio
(ug/m3
Change in SO4
per 1,000
Tons SO2)
0.007
0.008
0.038
0.031
0.079
0.012
0.609
0.018
0.116
0.015
0.371
0.040
0.013
0.081
0.063
0.106
0.094
0.053
NOxAir
Quality Ratio
(ug/m3
Change in
NO3 per
1,000 Tons
NOX)
0.000
0.000
0.000
0.000
0.018
0.000
0.087
0.000
0.002
0.007
0.131
0.009
0.002
0.001
0.005
0.015
0.007
0.002
PM2.5 Air
Quality Ratio
(ug/m3
Change in
Direct PM2.5
per 1,000
Tons of PM2.5)
1.238
1.238
1.929
1.238
1.879
1.238
1.929
1.238
0.597
1.238
1.929
1.238
1.238
1.929
1.929
1.929
1.929
1.238
Annual DV
Reduction
Needed for
11 (ug/m3)
1.291
0.032
1.011
0.171
0.000
0.091
0.000
0.061
0.171
0.000
0.000
0.491
1.301
0.000
0.000
0.000
1.487
0.121
24-hr DV
Reduction
Needed for
30 (ug/m3)
0.000
5.000
3.332
0.000
1.674
0.000
0.010
0.000
1.331
0.029
0.837
0.000
0.888
5.000
5.000
3.993
4.637
0.000
Annual DV
Reduction
Corresponding
to the 24-hr DV
Meeting 30
(ug/m3)
0.000
1.650
1.262
0.000
0.369
0.000
0.002
0.000
0.386
0.010
0.192
0.000
0.299
1.323
1.323
1.056
1.227
0.000
(continued)
-------
Table 4.A-10. Air Quality Ratios for Monitors in Counties with at Least One Monitor with an Annual Design Value (DV) Above 11
or 24-hr Design Value (DV) Above 30 in 2020 Baseline (15/35) (continued)
FIPS
Code
39035
41035
41039
6067
6099
6019
6031
6107
16079
6059
6065
6071
48201
49005
49035
49011
49049
53053
High
Monitor ID
390350038
410350004
410392013
60670006
60990005
60190008
60310004
61072002
160790017
60590007
60658001
60710025
482011035
490050004
490350012
490110004
490494001
530530029
State Name
Ohio
Oregon
Oregon
California
California
California
California
California
Idaho
California
California
California
Texas
Utah
Utah
Utah
Utah
Washington
SO2Air
Quality Ratio
Change in SO4
per 1,000
County Name TonsSO2)
Cuyahoga
Klamath
Lane
Sacramento
Stanislaus
Fresno
Kings
Tulare
Shoshone
Orange
Riverside
San Bernardino
Harris
Cache
Salt Lake
Davis
Utah
Pierce
0.016
0.059
0.068
0.033
0.057
0.062
0.070
0.076
0.129
0.043
0.047
0.043
0.016
0.321
0.026
0.028
0.024
0.029
NOxAir
Quality Ratio
Change in
NO3 per 1,000
Tons NOX)
0.005
0.002
0.003
0.007
0.008
0.008
0.009
0.010
0.014
0.003
0.004
0.003
0.000
0.040
0.012
0.014
0.015
0.000
PM2.5 Air
Quality Ratio
(ug/m3
Change in
Direct PM2.5
per 1,000
Tons of PM2.5)
1.238
1.929
1.929
1.879
1.879
1.879
1.879
1.879
1.929
0.597
0.597
0.597
1.238
1.929
1.929
1.929
1.929
1.929
Annual DV
Reduction
Needed for
11 (ug/m3)
0.741
0.000
0.000
0.000
0.000
0.000
0.180
0.645
0.000
0.008
2.039
2.077
0.561
0.000
0.000
0.000
0.000
0.000
24-hr DV
Reduction
Needed for
30 (ug/m3)
0.000
0.363
3.520
2.438
1.300
0.402
5.000
1.333
1.475
0.525
5.000
3.537
0.000
5.000
4.821
0.864
3.461
0.515
Annual DV
Reduction
Corresponding
to the 24-hr DV
Meeting 30
(ug/m3)
0.000
0.082
0.793
0.707
0.377
0.117
1.449
0.386
0.337
0.152
1.449
1.025
0.000
1.018
0.982
0.176
0.705
0.146
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35
NJ
State Name
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
County Name
Baldwin
Clay
Colbert
DeKalb
Escambia
Etowah
Houston
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Madison
Mobile
Mobile
Mobile
Montgomery
2005 Annual
DV
11.44
13.27
12.75
14.13
13.19
14.87
13.22
18.57
15.46
13.52
15.89
17.15
15.1
14.42
14.53
13.83
12.9
12.36
11.51
14.24
2020 Annual
DV
7.45
8.42
8.21
8.74
9.29
9.38
9.32
12.34
10.33
8.77
10.03
11.72
9.98
9.02
9.3
8.54
8.8
8.33
7.51
9.77
2020 15/35
Annual DV
7.45
8.42
8.21
8.74
9.29
9.38
9.32
12.34
10.33
8.77
10.03
11.72
9.98
9.02
9.3
8.54
8.8
8.33
7.51
9.77
2020 13/35
Annual DV
7.45
8.42
8.21
8.74
9.29
9.38
9.32
12.34
10.33
8.77
10.03
11.72
9.98
9.02
9.3
8.54
8.8
8.33
7.51
9.77
2020 12/35
Annual DV
7.45
8.42
8.21
8.74
9.29
9.38
9.32
12.03
10.02
8.46
9.729
11.41
9.67
8.71
8.99
8.54
8.8
8.33
7.51
9.77
2020 11/35
Annual DV
7.45
8.42
8.21
8.74
9.29
9.38
9.32
11.00
8.99
7.43
8.69
10.38
8.64
7.68
7.96
8.54
8.8
8.33
7.51
9.77
2020 11/30
Annual DV
7.45
8.42
8.21
8.74
9.29
9.38
9.32
11.00
8.99
7.43
8.69
10.38
8.64
7.68
7.96
8.54
8.8
8.33
7.51
9.77
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
NJ
00
State Name
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arkansas
Arkansas
County Name
Morgan
Russell
Shelby
Sumter
Talladega
Tuscaloosa
Walker
Cochise
Coconino
Gila
Maricopa
Maricopa
Maricopa
Pima
Pima
Pinal
Pinal
Santa Cruz
Arkansas
Ashley
2005 Annual
DV
13.32
15.73
14.43
11.92
14.51
13.56
13.86
7
6.49
8.94
12.17
12.59
9.97
6.04
5.85
7.77
5.71
12.94
12.45
12.83
2020 Annual
DV
8.42
10.63
9.3
7.75
9.08
8.78
8.89
6.56
6.02
8.11
9.64
10.24
8.02
5.12
4.95
6.92
5.04
12.06
8.76
9.44
2020 15/35
Annual DV
8.42
10.63
9.3
7.75
9.08
8.78
8.89
6.56
6.02
8.11
9.64
10.24
8.02
5.12
4.95
6.92
5.04
12.06
8.76
9.44
2020 13/35
Annual DV
8.42
10.63
9.3
7.75
9.08
8.78
8.89
6.56
6.02
8.11
9.64
10.24
8.02
5.12
4.95
6.92
5.04
12.06
8.76
9.44
2020 12/35
Annual DV
8.42
10.63
9.3
7.75
9.08
8.78
8.89
6.56
6.02
8.11
9.64
10.24
8.02
5.12
4.95
6.92
5.04
12.02
8.76
9.44
2020 11/35
Annual DV
8.42
10.63
9.3
7.75
9.08
8.78
8.89
6.54
6.02
8.11
9.64
10.24
8.02
5.09
4.92
6.92
5.04
11.04
8.76
9.44
2020 11/30
Annual DV
8.42
10.63
9.3
7.75
9.08
8.78
8.89
6.54
6.02
8.11
9.64
10.24
8.02
5.10
4.93
6.92
5.04
10.79
8.76
9.44
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
NJ
ID
State Name
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
California
California
California
California
California
California
California
California
County Name
Crittenden
Faulkner
Garland
Mississippi
Phillips
Polk
Pope
Pulaski
Pulaski
Pulaski
Union
White
Alameda
Alameda
Butte
Calaveras
Colusa
Contra Costa
Fresno
Fresno
2005 Annual
DV
13.36
12.79
12.4
12.61
12.1
11.65
12.79
13.17
14.05
13.59
12.86
12.57
9.44
9.34
12.73
7.77
7.39
9.47
16.99
16.38
2020 Annual
DV
8.63
9.16
8.79
8.32
8.15
8.35
9.4
9.11
9.91
9.52
9.3
9.13
7.42
7.18
9.56
6.05
6.25
7.3
12.71
12.33
2020 15/35
Annual DV
8.63
9.16
8.79
8.32
8.15
8.35
9.4
9.11
9.91
9.52
9.3
9.13
7.42
7.18
9.56
6.05
6.25
7.3
9.76
9.42
2020 13/35
Annual DV
8.63
9.16
8.79
8.32
8.15
8.35
9.4
9.11
9.91
9.52
9.3
9.13
7.42
7.18
9.56
6.05
6.25
7.3
9.76
9.42
2020 12/35
Annual DV
8.63
9.16
8.79
8.32
8.15
8.35
9.4
9.11
9.91
9.52
9.3
9.13
7.42
7.18
9.56
6.05
6.25
7.3
9.76
9.42
2020 11/35
Annual DV
8.63
9.16
8.79
8.32
8.15
8.35
9.4
9.11
9.91
9.52
9.3
9.13
7.42
7.18
9.56
6.05
6.25
7.3
9.76
9.42
2020 11/30
Annual DV
8.63
9.16
8.79
8.32
8.15
8.35
9.4
9.11
9.91
9.52
9.3
9.13
7.42
7.18
9.18
6.05
6.25
7.3
8.00
7.67
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
o
State Name
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
County Name
Fresno
Imperial
Imperial
Imperial
Inyo
Kern
Kern
Kern
Kings
Lake
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
2005 Annual
DV
17.17
12.71
8.39
9.2
5.25
18.94
18.68
19.17
17.28
4.62
17.03
18.19
18
15.35
17.66
17.92
15.36
16.62
15.21
8.42
2020 Annual
DV
12.87
11.22
7.35
8.12
4.8
14.34
14.18
14.72
13.24
3.79
12.88
13.44
13.19
11.41
13.14
13.16
11.31
12.3
11.24
7.06
2020 15/35
Annual DV
9.96
11.22
7.35
8.12
4.12
7.98
7.84
8.36
11.22
3.79
9.87
10.34
10.03
8.57
10.07
10.07
8.28
9.35
8.34
4.58
2020 13/35
Annual DV
9.96
11.22
7.35
8.12
4.12
7.98
7.84
8.36
11.22
3.79
9.87
10.34
10.03
8.57
10.07
10.07
8.28
9.35
8.34
4.58
2020 12/35
Annual DV
9.96
11.22
7.35
8.12
4.12
7.98
7.84
8.36
11.22
3.79
9.82
10.29
9.984
8.53
10.02
10.02
8.23
9.30
8.29
4.55
2020 11/35
Annual DV
9.96
10.80
6.93
7.70
4.12
7.98
7.84
8.36
11.04
3.79
9.60
10.04
9.72
8.34
9.79
9.78
8.00
9.09
8.09
4.45
2020 11/30
Annual DV
8.21
10.80
6.93
7.70
3.87
7.29
7.16
7.68
9.33
3.79
9.60
10.04
9.72
8.34
9.79
9.78
8.00
9.09
8.09
4.45
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
State Name
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
County Name
Mendocino
Merced
Monterey
Nevada
Nevada
Orange
Orange
Placer
Plumas
Plumas
Riverside
Riverside
Riverside
Sacramento
Sacramento
Sacramento
San Bernardino
San Bernardino
San Bernardino
San Bernardino
2005 Annual
DV
6.46
14.78
6.96
5.16
6.71
15.75
11.33
9.8
9.75
11.46
18.91
10.31
20.95
11.88
11.44
10.53
19.67
10.29
19.14
10.77
2020 Annual
DV
5.14
11.24
5.39
3.83
5.31
11.93
9.43
7.24
7.8
8.91
14.84
8.66
16.3
8.76
8.75
8.01
15.23
8.2
14.86
9.16
2020 15/35
Annual DV
5.14
10.39
4.75
3.83
5.31
11.05
8.77
7.24
7.8
8.91
11.65
5.90
13.08
8.76
8.75
8.01
13.12
6.39
12.67
7.50
2020 13/35
Annual DV
5.14
10.39
4.75
3.83
5.31
11.05
8.77
7.24
7.8
8.91
11.61
5.86
13.04
8.76
8.75
8.01
13.04
6.32
12.60
7.42
2020 12/35
Annual DV
5.14
10.39
4.75
3.83
5.31
11.01
8.74
7.24
7.8
8.91
10.61
4.88
12.04
8.76
8.75
8.01
12.04
5.33
11.59
6.44
2020 11/35
Annual DV
5.14
10.39
4.75
3.83
5.31
10.75
8.53
7.24
7.8
8.91
9.62
3.97
11.04
8.76
8.75
8.01
11.04
4.39
10.58
5.54
2020 11/30
Annual DV
5.14
10.10
4.52
3.83
5.31
10.27
8.05
7.24
7.8
8.91
9.62
3.97
11.04
7.07
7.06
6.32
11.04
4.39
10.58
5.54
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
NJ
State Name
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
County Name
San Bernardino
San Diego
San Diego
San Diego
San Diego
San Diego
San Francisco
San Joaquin
San LuisObispo
San Luis Obispo
San Mateo
Santa Barbara
Santa Clara
Santa Clara
Shasta
Solano
Sonoma
Stanislaus
Sutter
Tulare
2005 Annual
DV
19.01
11.92
12.27
10.59
12.79
13.46
9.62
12.94
6.92
7.94
9.03
10.37
11.38
10.32
7.41
9.99
8.21
14.21
9.85
18.51
2020 Annual
DV
14.96
8.68
9.09
7.64
9.56
9.83
7.05
9.96
5.43
6.24
6.76
8.04
8.66
8.03
5.48
7.7
6.15
10.85
7.53
14.1
2020 15/35
Annual DV
12.90
8.68
9.09
7.64
9.56
9.83
7.05
9.07
4.76
5.47
6.76
8.04
7.75
7.13
5.48
7.7
6.15
9.32
7.53
11.69
2020 13/35
Annual DV
12.82
8.68
9.09
7.64
9.56
9.83
7.05
9.07
4.76
5.47
6.76
8.04
7.75
7.13
5.48
7.7
6.15
9.32
7.53
11.69
2020 12/35
Annual DV
11.83
8.68
9.09
7.64
9.56
9.83
7.05
9.07
4.76
5.47
6.76
8.04
7.75
7.13
5.48
7.7
6.15
9.32
7.53
11.69
2020 11/35
Annual DV
10.84
8.68
9.09
7.64
9.56
9.83
7.05
9.07
4.76
5.47
6.76
8.04
7.75
7.13
5.48
7.7
6.15
9.32
7.53
11.04
2020 11/30
Annual DV
10.84
8.68
9.09
7.64
9.56
9.83
7.05
8.78
4.52
5.21
6.76
8.04
7.42
6.81
5.48
7.7
6.15
8.43
7.53
10.54
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
uj
State Name
California
California
California
California
California
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
County Name
Ventura
Ventura
Ventura
Ventura
Yolo
Adams
Arapahoe
Boulder
Boulder
Delta
Denver
Denver
Elbert
El Paso
El Paso
Larimer
Mesa
Pueblo
San Miguel
Weld
2005 Annual
DV
10.68
9.74
11.68
10.69
9.03
10.06
7.96
8.32
6.96
7.44
9.37
9.76
4.4
6.73
7.94
7.33
9.28
7.45
4.65
8.19
2020 Annual
DV
7.84
7.69
8.79
7.64
7.15
7.58
6.02
6.68
5.6
5.74
7.07
7.37
3.48
4.85
5.67
5.95
7.34
5.73
4.09
6.61
2020 15/35
Annual DV
7.19
7.04
8.09
7.14
7.15
7.58
6.02
6.68
5.6
5.74
7.07
7.37
3.48
4.85
5.67
5.95
7.34
5.73
4.09
6.61
2020 13/35
Annual DV
7.19
7.04
8.09
7.14
7.15
7.58
6.02
6.68
5.6
5.74
7.07
7.37
3.48
4.85
5.67
5.95
7.34
5.73
4.09
6.61
2020 12/35
Annual DV
7.16
7.01
8.06
7.12
7.15
7.58
6.02
6.68
5.6
5.74
7.07
7.37
3.48
4.85
5.67
5.95
7.34
5.73
4.09
6.61
2020 11/35
Annual DV
7.01
6.86
7.89
7.00
7.15
7.58
6.02
6.68
5.6
5.74
7.07
7.37
3.48
4.85
5.67
5.95
7.34
5.73
4.09
6.61
2020 11/30
Annual DV
7.01
6.86
7.89
7.00
7.15
7.58
6.02
6.68
5.6
5.74
7.07
7.37
3.48
4.85
5.67
5.95
7.34
5.73
4.09
6.61
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Colorado
Connecticut
Connecticut
Connecticut
Connecticut
Connecticut
Connecticut
Connecticut
}> Connecticut
j^ Connecticut
Connecticut
Connecticut
Connecticut
Delaware
Delaware
Delaware
Delaware
Delaware
Delaware
Delaware
County Name
Weld
Fairfield
Fairfield
Fairfield
Fairfield
Hartford
Litchfield
New Haven
New Haven
New Haven
New Haven
New Haven
New London
Kent
Kent
New Castle
New Castle
New Castle
New Castle
Sussex
2005 Annual
DV
8.78
13.21
12.49
12.43
11.48
11.03
8.01
12.12
12.45
13.12
11.17
12.74
10.96
12.61
12.52
13.73
12.92
13.69
14.87
13.39
2020 Annual
DV
7.06
8.79
8.27
8.15
7.51
7.29
5.03
7.93
8.1
8.62
7.24
8.33
7.25
7.66
7.71
8.57
7.93
8.55
9.42
8.13
2020 15/35
Annual DV
7.06
8.79
8.27
8.15
7.51
7.29
5.03
7.93
8.1
8.62
7.24
8.33
7.25
7.66
7.71
8.57
7.93
8.55
9.42
8.13
2020 13/35
Annual DV
7.06
8.79
8.27
8.15
7.51
7.29
5.03
7.93
8.1
8.62
7.24
8.33
7.25
7.66
7.71
8.57
7.93
8.55
9.42
8.13
2020 12/35
Annual DV
7.06
8.79
8.27
8.15
7.51
7.29
5.03
7.93
8.1
8.62
7.24
8.33
7.25
7.66
7.71
8.57
7.93
8.55
9.42
8.13
2020 11/35
Annual DV
7.06
8.79
8.27
8.15
7.51
7.29
5.03
7.93
8.1
8.62
7.24
8.33
7.25
7.66
7.71
8.57
7.93
8.55
9.42
8.13
2020 11/30
Annual DV
7.06
8.79
8.27
8.15
7.51
7.29
5.03
7.93
8.1
8.62
7.24
8.33
7.25
7.66
7.71
8.57
7.93
8.55
9.42
8.13
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
Un
State Name
District of Columbia
District of Columbia
District of Columbia
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
County Name
District of Columbia
District of Columbia
District of Columbia
Alachua
Alachua
Bay
Brevard
Broward
Broward
Broward
Citrus
Duval
Duval
Escambia
Hillsborough
Hillsborough
Lee
Leon
Manatee
Marion
2005 Annual
DV
14.16
14.41
13.99
9.32
9.59
11.46
8.32
8.22
8.18
8.21
9
9.9
10.44
11.72
10.74
10.52
8.36
12.56
8.81
10.11
2020 Annual
DV
8.86
8.71
8.48
6.19
6.44
7.96
5.56
5.93
5.78
5.78
5.69
6.88
7.49
8.38
7.37
7.23
5.88
9
5.64
6.93
2020 15/35
Annual DV
8.86
8.71
8.48
6.19
6.44
7.96
5.56
5.93
5.78
5.78
5.69
6.88
7.49
8.38
7.37
7.23
5.88
9
5.64
6.93
2020 13/35
Annual DV
8.86
8.71
8.48
6.19
6.44
7.96
5.56
5.93
5.78
5.78
5.69
6.88
7.49
8.38
7.37
7.23
5.88
9
5.64
6.93
2020 12/35
Annual DV
8.86
8.71
8.48
6.19
6.44
7.96
5.56
5.93
5.78
5.78
5.69
6.88
7.49
8.38
7.37
7.23
5.88
9
5.64
6.93
2020 11/35
Annual DV
8.86
8.71
8.48
6.19
6.44
7.96
5.56
5.93
5.78
5.78
5.69
6.88
7.49
8.38
7.37
7.23
5.88
9
5.64
6.93
2020 11/30
Annual DV
8.86
8.71
8.48
6.19
6.44
7.96
5.56
5.93
5.78
5.78
5.69
6.88
7.49
8.38
7.37
7.23
5.88
9
5.64
6.93
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
en
State Name
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
County Name
Miami-Dade
Miami-Dade
Orange
Orange
Palm Beach
Palm Beach
Pinellas
Pinellas
Polk
St. Lucie
Sarasota
Seminole
Volusia
Bibb
Bibb
Chatham
Chatham
Clarke
Clayton
Cobb
2005 Annual
DV
9.45
8.14
9.61
9.5
7.84
7.7
9.82
9.52
9.53
8.34
8.77
9.51
9.27
16.54
13.94
13.74
13.93
14.9
16.5
16.15
2020 Annual
DV
6.72
6.42
6.43
6.28
5.83
5.69
6.64
6.41
6.55
5.77
5.77
6.33
6.08
11.22
9.1
9.33
9.56
9.69
10.65
10.6
2020 15/35
Annual DV
6.72
6.42
6.43
6.28
5.83
5.69
6.64
6.41
6.55
5.77
5.77
6.33
6.08
11.22
9.1
9.33
9.56
9.69
10.65
10.6
2020 13/35
Annual DV
6.72
6.42
6.43
6.28
5.83
5.69
6.64
6.41
6.55
5.77
5.77
6.33
6.08
11.22
9.1
9.33
9.56
9.69
10.65
10.6
2020 12/35
Annual DV
6.72
6.42
6.43
6.28
5.83
5.69
6.64
6.41
6.55
5.77
5.77
6.33
6.08
11.22
9.1
9.33
9.56
9.69
10.65
10.6
2020 11/35
Annual DV
6.72
6.42
6.43
6.28
5.83
5.69
6.64
6.41
6.55
5.77
5.77
6.33
6.08
11.04
8.92
9.33
9.56
9.69
10.65
10.6
2020 11/30
Annual DV
6.72
6.42
6.43
6.28
5.83
5.69
6.64
6.41
6.55
5.77
5.77
6.33
6.08
11.04
8.92
9.33
9.56
9.69
10.65
10.6
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
State Name
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
County Name
Cobb
DeKalb
DeKalb
Dougherty
Floyd
Fulton
Fulton
Glynn
Gwinnett
Hall
Houston
Lowndes
Muscogee
Muscogee
Muscogee
Paulding
Richmond
Richmond
Walker
Washington
2005 Annual
DV
15.42
15.48
15.37
14.46
16.13
15.84
17.43
12.25
16.07
14.16
14.19
12.58
14.94
15.39
14.16
14.12
15.61
15.68
15.49
15.14
2020 Annual
DV
9.88
9.54
9.55
10.24
10.63
9.82
11.11
8.74
10.45
9.16
9.34
9.26
9.94
10.4
9.58
8.76
10.75
10.84
9.86
10.42
2020 15/35
Annual DV
9.88
9.54
9.55
10.24
10.63
9.82
11.11
8.74
10.45
9.16
9.34
9.26
9.94
10.4
9.58
8.76
10.75
10.84
9.86
10.42
2020 13/35
Annual DV
9.88
9.54
9.55
10.24
10.63
9.82
11.11
8.74
10.45
9.16
9.34
9.26
9.94
10.4
9.58
8.76
10.75
10.84
9.86
10.42
2020 12/35
Annual DV
9.88
9.54
9.55
10.24
10.63
9.82
11.11
8.74
10.45
9.16
9.34
9.26
9.94
10.4
9.58
8.76
10.75
10.84
9.86
10.42
2020 11/35
Annual DV
9.88
9.54
9.55
10.24
10.63
9.82
10.98
8.74
10.45
9.16
9.32
9.26
9.94
10.4
9.58
8.76
10.75
10.84
9.86
10.42
2020 11/30
Annual DV
9.88
9.54
9.55
10.24
10.63
9.82
10.98
8.74
10.45
9.16
9.32
9.26
9.94
10.4
9.58
8.76
10.75
10.84
9.86
10.42
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
00
State Name
Georgia
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
County Name
Wilkinson
Ada
Bannock
Benewah
Canyon
Franklin
Idaho
Shoshone
Adams
Champaign
Champaign
Cook
Cook
Cook
Cook
Cook
Cook
Cook
Cook
Cook
2005 Annual
DV
15.27
8.41
7.66
9.59
8.46
7.7
9.58
12.08
12.5
12.5
12.53
15.21
14.81
15.75
15.03
14.89
14.77
15.24
12.78
12.76
2020 Annual
DV
10.3
7.6
6.97
8.59
7.46
6.68
8.8
10.66
8.91
8.39
8.4
11.11
10.69
11.14
10.54
10.5
10.77
10.78
8.9
8.87
2020 15/35
Annual DV
10.3
7.6
6.97
8.59
7.46
6.68
8.8
10.66
8.91
8.39
8.4
11.11
10.69
11.14
10.54
10.5
10.77
10.78
8.9
8.87
2020 13/35
Annual DV
10.3
7.6
6.97
8.59
7.46
6.68
8.8
10.66
8.91
8.39
8.4
11.11
10.69
11.14
10.54
10.5
10.77
10.78
8.9
8.87
2020 12/35
Annual DV
10.3
7.6
6.97
8.59
7.46
6.68
8.8
10.66
8.91
8.39
8.4
11.11
10.69
11.14
10.54
10.5
10.77
10.78
8.9
8.87
2020 11/35
Annual DV
10.3
7.6
6.97
8.59
7.46
6.68
8.8
10.66
8.91
8.39
8.4
10.97
10.55
11.00
10.40
10.36
10.63
10.64
8.76
8.73
2020 11/30
Annual DV
10.3
7.6
6.97
8.57
7.46
6.68
8.75
10.32
8.91
8.39
8.4
10.97
10.55
11.00
10.40
10.36
10.63
10.64
8.76
8.73
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
u>
ID
State Name
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
County Name
Cook
DuPage
Jersey
Kane
Kane
Lake
McHenry
McLean
Macon
Madison
Madison
Madison
Peoria
Randolph
Rock Island
Saint Clair
Saint Clair
Sangamon
Will
Will
2005 Annual
DV
15.48
13.82
12.89
13.32
14.34
11.81
12.4
12.39
13.24
16.72
14.01
14.32
13.34
13.11
12.01
15.58
14.29
13.13
13.63
11.52
2020 Annual
DV
10.86
9.84
8.78
9.51
10.25
8.34
8.86
8.58
9.21
11.54
9.7
9.97
9.41
8.7
8.61
10.63
9.66
9.39
9.69
7.96
2020 15/35
Annual DV
10.86
9.84
8.78
9.51
10.25
8.34
8.86
8.58
9.21
11.54
9.7
9.97
9.41
8.7
8.61
10.63
9.66
9.39
9.69
7.96
2020 13/35
Annual DV
10.86
9.84
8.78
9.51
10.25
8.34
8.86
8.58
9.21
11.54
9.7
9.97
9.41
8.7
8.61
10.63
9.66
9.39
9.69
7.96
2020 12/35
Annual DV
10.86
9.84
8.78
9.51
10.25
8.34
8.86
8.58
9.21
11.54
9.7
9.97
9.41
8.7
8.61
10.63
9.66
9.39
9.69
7.96
2020 11/35
Annual DV
10.72
9.84
8.51
9.51
10.25
8.34
8.86
8.58
9.21
10.82
8.98
9.25
9.41
8.7
8.61
10.34
9.40
9.39
9.69
7.96
2020 11/30
Annual DV
10.72
9.84
8.51
9.51
10.25
8.34
8.86
8.58
9.21
10.82
8.98
9.25
9.41
8.7
8.61
10.34
9.40
9.39
9.69
7.96
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
-p*
o
State Name
Illinois
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
County Name
Winnebago
Allen
Allen
Clark
Delaware
Dubois
Floyd
Henry
Howard
Knox
Lake
Lake
Lake
Lake
Lake
La Porte
La Porte
Madison
Marion
Marion
2005 Annual
DV
13.57
13.67
13.55
16.44
13.69
15.19
14.85
13.64
13.93
14.03
14.33
13.83
14.02
14.05
13.89
12.49
12.69
13.97
14.24
15.26
2020 Annual
DV
9.78
9.92
9.84
10.19
9.11
9.34
9
9.05
9.61
8.78
10.45
10.06
10.37
10.27
10.11
8.9
9.03
9.34
9.28
10.16
2020 15/35
Annual DV
9.78
9.92
9.84
10.19
9.11
9.34
9
9.05
9.61
8.78
10.45
10.06
10.37
10.27
10.11
8.9
9.03
9.34
9.28
10.16
2020 13/35
Annual DV
9.78
9.92
9.84
10.19
9.11
9.34
9
9.05
9.61
8.78
10.45
10.06
10.37
10.27
10.11
8.9
9.03
9.34
9.28
10.16
2020 12/35
Annual DV
9.78
9.92
9.84
10.19
9.11
9.34
9
9.05
9.61
8.78
10.45
10.06
10.37
10.27
10.11
8.9
9.03
9.34
9.28
10.16
2020 11/35
Annual DV
9.78
9.92
9.84
10.19
9.11
9.34
9
9.05
9.61
8.78
10.45
10.06
10.37
10.27
10.11
8.9
9.03
9.34
9.28
10.16
2020 11/30
Annual DV
9.78
9.92
9.84
10.19
9.11
9.34
9
9.05
9.61
8.78
10.45
10.06
10.37
10.27
10.11
8.9
9.03
9.34
9.28
10.16
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Iowa
Iowa
Iowa
Iowa
Iowa
County Name
Marion
Marion
Marion
Porter
Porter
St. Joseph
St. Joseph
St. Joseph
Spencer
Tippecanoe
Vanderburgh
Vanderburgh
Vanderburgh
Vigo
Vigo
Black Hawk
Clinton
Johnson
Linn
Montgomery
2005 Annual
DV
14.71
16.05
15.9
12.66
13.21
13.29
13.69
12.82
14.32
13.7
14.69
14.82
14.99
13.99
13.46
11.16
12.52
12.08
10.79
10.02
2020 Annual
DV
9.7
10.81
10.66
8.99
9.42
10.04
10.36
9.65
8.55
9.4
9.82
9.9
10.06
8.95
8.51
8.15
9.01
8.96
7.88
7.35
2020 15/35
Annual DV
9.7
10.81
10.66
8.99
9.42
10.04
10.36
9.65
8.55
9.4
9.82
9.9
10.06
8.95
8.51
8.15
9.01
8.96
7.88
7.35
2020 13/35
Annual DV
9.7
10.81
10.66
8.99
9.42
10.04
10.36
9.65
8.55
9.4
9.82
9.9
10.06
8.95
8.51
8.15
9.01
8.96
7.88
7.35
2020 12/35
Annual DV
9.7
10.81
10.66
8.99
9.42
10.04
10.36
9.65
8.55
9.4
9.82
9.9
10.06
8.95
8.51
8.15
9.01
8.96
7.88
7.35
2020 11/35
Annual DV
9.7
10.81
10.66
8.99
9.42
10.04
10.36
9.65
8.55
9.4
9.82
9.9
10.06
8.95
8.51
8.15
9.01
8.96
7.88
7.35
2020 11/30
Annual DV
9.7
10.81
10.66
8.99
9.42
10.04
10.36
9.65
8.55
9.4
9.82
9.9
10.06
8.95
8.51
8.15
9.01
8.96
7.88
7.35
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
County Name
Muscatine
Palo Alto
Polk
Polk
Polk
Pottawattamie
Scott
Scott
Scott
Van Buren
Woodbury
Wright
Johnson
Johnson
Johnson
Linn
Sedgwick
Sedgwick
Sedgwick
Shawnee
2005 Annual
DV
12.92
9.53
10.41
9.95
10.64
11.13
11.86
11.64
14.42
10.84
10.32
10.37
10.59
11.1
9.68
10.47
10.26
10.29
10.36
10.79
2020 Annual
DV
9.47
7.21
7.68
7.34
7.85
8.42
8.52
8.35
10.62
7.98
7.93
7.68
7.68
8.04
7
7.74
7.55
7.57
7.64
8.07
2020 15/35
Annual DV
9.47
7.21
7.68
7.34
7.85
8.42
8.52
8.35
10.62
7.98
7.93
7.68
7.68
8.04
7
7.74
7.55
7.57
7.64
8.07
2020 13/35
Annual DV
9.47
7.21
7.68
7.34
7.85
8.42
8.52
8.35
10.62
7.98
7.93
7.68
7.68
8.04
7
7.74
7.55
7.57
7.64
8.07
2020 12/35
Annual DV
9.47
7.21
7.68
7.34
7.85
8.42
8.52
8.35
10.62
7.98
7.93
7.68
7.68
8.04
7
7.74
7.55
7.57
7.64
8.07
2020 11/35
Annual DV
9.47
7.21
7.68
7.34
7.85
8.42
8.52
8.35
10.62
7.98
7.93
7.68
7.68
8.04
7
7.74
7.55
7.57
7.64
8.07
2020 11/30
Annual DV
9.47
7.21
7.68
7.34
7.85
8.42
8.52
8.35
10.62
7.98
7.93
7.68
7.68
8.04
7
7.74
7.55
7.57
7.64
8.07
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Kansas
Kansas
Kansas
Kansas
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
County Name
Shawnee
Sumner
Wyandotte
Wyandotte
Bell
Boyd
Bullitt
Campbell
Carter
Christian
Daviess
Fayette
Fayette
Franklin
Hardin
Henderson
Jefferson
Jefferson
Jefferson
Jefferson
2005 Annual
DV
10.93
9.89
12.73
10.93
14.1
14.49
14.92
13.67
12.22
13.2
14.1
14.36
14.87
13.37
13.58
13.93
15.55
15.35
15.31
14.74
2020 Annual
DV
8.26
7.4
9.34
7.92
8.59
8.77
9.15
8.13
7.08
8.02
8.25
8.64
9.05
7.93
8.04
8.89
9.43
9.29
9.26
8.84
2020 15/35
Annual DV
8.26
7.4
9.34
7.92
8.59
8.77
9.15
8.13
7.08
8.02
8.25
8.64
9.05
7.93
8.04
8.89
9.43
9.29
9.26
8.84
2020 13/35
Annual DV
8.26
7.4
9.34
7.92
8.59
8.77
9.15
8.13
7.08
8.02
8.25
8.64
9.05
7.93
8.04
8.89
9.43
9.29
9.26
8.84
2020 12/35
Annual DV
8.26
7.4
9.34
7.92
8.59
8.77
9.15
8.13
7.08
8.02
8.25
8.64
9.05
7.93
8.04
8.89
9.43
9.29
9.26
8.84
2020 11/35
Annual DV
8.26
7.4
9.34
7.92
8.59
8.77
9.15
8.13
7.08
8.02
8.25
8.64
9.05
7.93
8.04
8.89
9.43
9.29
9.26
8.84
2020 11/30
Annual DV
8.26
7.4
9.34
7.92
8.59
8.77
9.15
8.13
7.08
8.02
8.25
8.64
9.05
7.93
8.04
8.89
9.43
9.29
9.26
8.84
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Louisiana
}> Louisiana
£* Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
County Name
Kenton
Laurel
McCracken
Madison
Perry
Pike
Warren
Caddo
Calcasieu
Calcasieu
Concordia
East Baton Rouge
East Baton Rouge
Iberville
Iberville
Jefferson
Lafayette
Ouachita
Rapides
Tangipahoa
2005 Annual
DV
14.39
12.55
13.41
13.61
13.21
13.49
13.83
12.53
10.58
11.07
11.42
13.38
12.08
12.9
11.02
11.52
11.08
11.97
11.03
12.03
2020 Annual
DV
8.73
7.4
8.39
8.01
7.98
7.94
8.29
8.76
7.66
8.05
7.81
9.86
8.74
9.44
7.68
7.53
7.62
8.63
7.64
8.14
2020 15/35
Annual DV
8.73
7.4
8.39
8.01
7.98
7.94
8.29
8.76
7.66
8.05
7.81
9.86
8.74
9.44
7.68
7.53
7.62
8.63
7.64
8.14
2020 13/35
Annual DV
8.73
7.4
8.39
8.01
7.98
7.94
8.29
8.76
7.66
8.05
7.81
9.86
8.74
9.44
7.68
7.53
7.62
8.63
7.64
8.14
2020 12/35
Annual DV
8.73
7.4
8.39
8.01
7.98
7.94
8.29
8.76
7.66
8.05
7.81
9.86
8.74
9.44
7.68
7.53
7.62
8.63
7.64
8.14
2020 11/35
Annual DV
8.73
7.4
8.39
8.01
7.98
7.94
8.29
8.76
7.66
8.05
7.81
9.86
8.74
9.44
7.68
7.53
7.62
8.63
7.64
8.14
2020 11/30
Annual DV
8.73
7.4
8.39
8.01
7.98
7.94
8.29
8.76
7.66
8.05
7.81
9.86
8.74
9.44
7.68
7.53
7.62
8.63
7.64
8.14
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
-p*
Un
State Name
Louisiana
Louisiana
Maine
Maine
Maine
Maine
Maine
Maine
Maine
Maine
Maine
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
County Name
Terrebonne
West Baton Rouge
Androscoggin
Aroostook
Aroostook
Cumberland
Cumberland
Hancock
Kennebec
Oxford
Penobscot
Anne Arundel
Anne Arundel
Anne Arundel
Baltimore
Baltimore
Cecil
Harford
Montgomery
Prince George's
2005 Annual
DV
10.74
13.51
9.9
9.74
8.27
11.06
11.13
5.76
9.99
10.13
9.12
11.91
14.82
14.57
13.77
14.76
12.68
12.51
12.47
12.24
2020 Annual
DV
7.32
9.96
6.85
8.73
6.85
7.51
7.61
4.24
7
7.65
6.66
7.3
9.65
9.45
8.58
9.46
7.81
7.62
7.77
7.61
2020 15/35
Annual DV
7.32
9.96
6.85
8.73
6.85
7.51
7.61
4.24
7
7.65
6.66
7.3
9.65
9.45
8.58
9.46
7.81
7.62
7.77
7.61
2020 13/35
Annual DV
7.32
9.96
6.85
8.73
6.85
7.51
7.61
4.24
7
7.65
6.66
7.3
9.65
9.45
8.58
9.46
7.81
7.62
7.77
7.61
2020 12/35
Annual DV
7.32
9.96
6.85
8.73
6.85
7.51
7.61
4.24
7
7.65
6.66
7.3
9.65
9.45
8.58
9.46
7.81
7.62
7.77
7.61
2020 11/35
Annual DV
7.32
9.96
6.85
8.73
6.85
7.51
7.61
4.24
7
7.65
6.66
7.3
9.65
9.45
8.58
9.46
7.81
7.62
7.77
7.61
2020 11/30
Annual DV
7.32
9.96
6.85
8.73
6.85
7.51
7.61
4.24
7
7.65
6.66
7.3
9.65
9.45
8.58
9.46
7.81
7.62
7.77
7.61
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Massachusetts
Massachusetts
}> Massachusetts
en Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
County Name
Prince George's
Washington
Baltimore (City)
Baltimore (City)
Baltimore (City)
Baltimore (City)
Berkshire
Bristol
Essex
Essex
Essex
Hampden
Hampden
Hampden
Plymouth
Suffolk
Suffolk
Suffolk
Suffolk
Worcester
2005 Annual
DV
13.03
13.7
14.12
14.38
15.76
15.63
10.65
9.58
9.03
9.1
9.58
9.85
12.17
11.85
9.87
12.34
11.86
10.88
13.07
10.55
2020 Annual
DV
8.06
8.64
8.99
9.08
10.14
10.13
7.34
6.5
6.36
6.42
6.74
6.73
8.29
8.08
6.87
8.79
8.35
7.7
9.3
7.19
2020 15/35
Annual DV
8.06
8.64
8.99
9.08
10.14
10.13
7.34
6.5
6.36
6.42
6.74
6.73
8.29
8.08
6.87
8.79
8.35
7.7
9.3
7.19
2020 13/35
Annual DV
8.06
8.64
8.99
9.08
10.14
10.13
7.34
6.5
6.36
6.42
6.74
6.73
8.29
8.08
6.87
8.79
8.35
7.7
9.3
7.19
2020 12/35
Annual DV
8.06
8.64
8.99
9.08
10.14
10.13
7.34
6.5
6.36
6.42
6.74
6.73
8.29
8.08
6.87
8.79
8.35
7.7
9.3
7.19
2020 11/35
Annual DV
8.06
8.64
8.99
9.08
10.14
10.13
7.34
6.5
6.36
6.42
6.74
6.73
8.29
8.08
6.87
8.79
8.35
7.7
9.3
7.19
2020 11/30
Annual DV
8.06
8.64
8.99
9.08
10.14
10.13
7.34
6.5
6.36
6.42
6.74
6.73
8.29
8.08
6.87
8.79
8.35
7.7
9.3
7.19
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Massachusetts
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
}> Michigan
vj Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
County Name
Worcester
Allegan
Bay
Berrien
Genesee
Ingham
Kalamazoo
Kent
Macomb
Missaukee
Monroe
Muskegon
Oakland
Ottawa
Saginaw
St. Clair
Washtenaw
Washtenaw
Wayne
Wayne
2005 Annual
DM
11.29
11.84
10.93
11.72
11.61
12.23
12.84
12.89
12.7
8.26
13.92
11.61
13.78
12.55
10.61
13.34
12.3
13.88
14.52
15.88
2020 Annual
DM
7.69
8.31
7.94
8.34
8.28
8.64
9.19
9.17
9.13
6.33
9.46
8.39
9.62
8.87
7.74
9.87
8.58
9.82
10.33
11.17
2020 15/35
Annual DM
7.69
8.31
7.94
8.34
8.28
8.64
9.19
9.17
9.13
6.33
9.46
8.39
9.62
8.87
7.74
9.87
8.58
9.82
10.33
11.17
2020 13/35
Annual DM
7.69
8.31
7.94
8.34
8.28
8.64
9.19
9.17
9.13
6.33
9.46
8.39
9.62
8.87
7.74
9.87
8.58
9.82
10.33
11.17
2020 12/35
Annual DM
7.69
8.31
7.94
8.34
8.28
8.64
9.19
9.17
9.13
6.33
9.46
8.39
9.62
8.87
7.74
9.87
8.58
9.82
9.802
10.64
2020 11/35
Annual DM
7.69
8.31
7.94
8.34
8.28
8.64
9.19
9.17
9.13
6.33
9.46
8.39
9.62
8.87
7.74
9.87
8.58
9.82
8.92
9.76
2020 11/30
Annual DM
7.69
8.31
7.94
8.34
8.28
8.64
9.19
9.17
9.13
6.33
9.46
8.39
9.62
8.87
7.74
9.87
8.58
9.82
8.92
9.76
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
-p*
00
State Name
Michigan
Michigan
Michigan
Michigan
Michigan
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
County Name
Wayne
Wayne
Wayne
Wayne
Wayne
Cass
Dakota
Hennepin
Hennepin
Hennepin
Hennepin
Hennepin
Hennepin
Mille Lacs
Olmsted
Ramsey
Ramsey
Ramsey
Saint Louis
Saint Louis
2005 Annual
DV
14.57
14.32
13.39
17.5
14.67
5.7
9.3
9.76
9.14
9.59
9.54
9.56
9.33
6.54
10.13
11.32
11.02
9.63
6.1
6.19
2020 Annual
DV
10.34
10.3
9.28
12.35
10.44
4.79
7.09
7.39
6.93
7.25
7.23
7.26
7.07
5.3
7.6
8.68
8.34
7.36
5.04
4.97
2020 15/35
Annual DV
10.34
10.3
9.28
12.35
10.44
4.79
7.09
7.39
6.93
7.25
7.23
7.26
7.07
5.3
7.6
8.68
8.34
7.36
5.04
4.97
2020 13/35
Annual DV
10.34
10.3
9.28
12.35
10.44
4.79
7.09
7.39
6.93
7.25
7.23
7.26
7.07
5.3
7.6
8.68
8.34
7.36
5.04
4.97
2020 12/35
Annual DV
9.812
9.772
8.75
11.82
9.912
4.79
7.09
7.39
6.93
7.25
7.23
7.26
7.07
5.3
7.6
8.68
8.34
7.36
5.04
4.97
2020 11/35
Annual DV
8.93
8.89
7.87
10.94
9.03
4.79
7.09
7.39
6.93
7.25
7.23
7.26
7.07
5.3
7.6
8.68
8.34
7.36
5.04
4.97
2020 11/30
Annual DV
8.93
8.89
7.87
10.94
9.03
4.79
7.09
7.39
6.93
7.25
7.23
7.26
7.07
5.3
7.6
8.68
8.34
7.36
5.04
4.97
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
-p*
ID
State Name
Minnesota
Minnesota
Minnesota
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Missouri
Missouri
Missouri
Missouri
County Name
Saint Louis
Scott
Stearns
Adams
Bolivar
DeSoto
Forrest
Harrison
Hinds
Jackson
Jones
Lauderdale
Lee
Lowndes
Pearl River
Warren
Boone
Buchanan
Cass
Cedar
2005 Annual
DV
7.51
9
8.58
11.29
12.36
12.43
13.62
12.2
12.56
12.04
14.39
13.07
12.57
12.79
12.14
12.32
11.84
12.8
10.67
11.12
2020 Annual
DV
5.99
6.9
6.8
7.66
8.56
7.95
9.03
8.16
8.43
7.9
9.56
8.62
8.06
8.42
8.25
8.51
8.45
9.75
7.76
7.89
2020 15/35
Annual DV
5.99
6.9
6.8
7.66
8.56
7.95
9.03
8.16
8.43
7.9
9.56
8.62
8.06
8.42
8.25
8.51
8.45
9.75
7.76
7.89
2020 13/35
Annual DV
5.99
6.9
6.8
7.66
8.56
7.95
9.03
8.16
8.43
7.9
9.56
8.62
8.06
8.42
8.25
8.51
8.45
9.75
7.76
7.89
2020 12/35
Annual DV
5.99
6.9
6.8
7.66
8.56
7.95
9.03
8.16
8.43
7.9
9.56
8.62
8.06
8.42
8.25
8.51
8.45
9.75
7.76
7.89
2020 11/35
Annual DV
5.99
6.9
6.8
7.66
8.56
7.95
9.03
8.16
8.43
7.9
9.56
8.62
8.06
8.42
8.25
8.51
8.45
9.75
7.76
7.89
2020 11/30
Annual DV
5.99
6.9
6.8
7.66
8.56
7.95
9.03
8.16
8.43
7.9
9.56
8.62
8.06
8.42
8.25
8.51
8.45
9.75
7.76
7.89
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
}> Missouri
o Missouri
Missouri
Missouri
Missouri
Montana
Montana
Montana
Montana
Montana
Montana
Montana
County Name
Clay
Greene
Jackson
Jefferson
Monroe
Saint Charles
Sainte Genevieve
Saint Louis
Saint Louis
St. Louis City
St. Louis City
St. Louis City
St. Louis City
Cascade
Flathead
Flathead
Gallatin
Lake
Lake
Lewis and Clark
2005 Annual
DV
11.03
11.75
12.78
13.79
10.87
13.29
13.34
13.04
13.46
14.27
14.36
13.44
14.56
5.72
9.99
8.58
4.38
9.06
9
8.2
2020 Annual
DV
8.1
8.32
9.33
9.57
7.59
9.12
9.06
8.9
9.07
9.74
9.71
9.03
9.84
5.01
8.52
7.28
4.13
7.81
7.71
7.19
2020 15/35
Annual DV
8.1
8.32
9.33
9.57
7.59
9.12
9.06
8.9
9.07
9.74
9.71
9.03
9.84
5.01
8.46
7.23
4.13
7.05
6.94
7.19
2020 13/35
Annual DV
8.1
8.32
9.33
9.57
7.59
9.12
9.06
8.9
9.07
9.74
9.71
9.03
9.84
5.01
8.46
7.23
4.13
7.05
6.94
7.19
2020 12/35
Annual DV
8.1
8.32
9.33
9.57
7.59
9.12
9.06
8.9
9.07
9.74
9.71
9.03
9.84
5.01
8.46
7.23
4.13
7.05
6.94
7.19
2020 11/35
Annual DV
8.1
8.32
9.33
9.57
7.59
9.12
9.06
8.9
9.07
9.74
9.71
9.03
9.84
5.01
8.36
7.13
4.13
6.95
6.83
7.19
2020 11/30
Annual DV
8.1
8.32
9.33
9.57
7.59
9.12
9.06
8.9
9.07
9.74
9.71
9.03
9.84
5.01
8.33
7.11
4.13
5.74
5.62
7.19
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
01
State Name
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nevada
Nevada
Nevada
Nevada
County Name
Lincoln
Missoula
Ravalli
Rosebud
Sanders
Silver Bow
Yellowstone
Cass
Douglas
Douglas
Hall
Lancaster
Lincoln
Sarpy
Scotts Bluff
Washington
Clark
Clark
Clark
Clark
2005 Annual
DV
14.93
10.52
9.01
6.58
6.75
10.14
8.14
9.99
9.88
9.85
7.95
8.9
7.57
9.79
6.04
9.29
4.02
5.75
9.44
3.67
2020 Annual
DV
12.6
9.13
7.89
6.1
6.04
8.69
6.84
7.48
7.41
7.4
6.13
6.53
6.32
7.33
5.17
7.1
3.64
4.97
8.08
3.3
2020 15/35
Annual DV
12.53
8.33
7.40
6.1
6.00
8.69
6.84
7.48
7.41
7.4
6.13
6.53
6.32
7.33
5.17
7.1
3.64
4.97
8.08
3.3
2020 13/35
Annual DV
12.53
8.33
7.40
6.1
6.00
8.69
6.84
7.48
7.41
7.4
6.13
6.53
6.32
7.33
5.17
7.1
3.64
4.97
8.08
3.3
2020 12/35
Annual DV
12.04
8.33
7.40
6.1
6.00
8.69
6.84
7.48
7.41
7.4
6.13
6.53
6.32
7.33
5.17
7.1
3.64
4.97
8.08
3.3
2020 11/35
Annual DV
11.01
8.19
7.32
6.1
5.91
8.69
6.84
7.48
7.41
7.4
6.13
6.53
6.32
7.33
5.17
7.1
3.64
4.97
8.08
3.3
2020 11/30
Annual DV
11.04
7.28
6.08
6.1
5.89
8.69
6.84
7.48
7.41
7.4
6.13
6.53
6.32
7.33
5.17
7.1
3.64
4.97
8.08
3.3
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
01
NJ
State Name
Nevada
Nevada
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
County Name
Clark
Washoe
Belknap
Cheshire
Coos
Grafton
Hillsborough
Hillsborough
Hillsborough
Merrimack
Rockingham
Sullivan
Atlantic
Bergen
Camden
Camden
Essex
Gloucester
Hudson
Mercer
2005 Annual
DV
8.49
8.11
7.28
11.53
10.24
8.43
10.18
10.01
6.27
9.72
9
9.86
11.47
13.09
13.31
13.51
13.27
13.46
14.24
12.71
2020 Annual
DV
7.28
6.39
5.12
7.96
8.01
6.08
7.08
7.01
4.3
6.76
6.26
7.01
6.96
8.9
8.49
8.53
8.74
8.46
9.65
8.14
2020 15/35
Annual DV
7.28
6.39
5.12
7.96
8.01
6.08
7.08
7.01
4.3
6.76
6.26
7.01
6.96
8.9
8.49
8.53
8.74
8.46
9.65
8.14
2020 13/35
Annual DV
7.28
6.39
5.12
7.96
8.01
6.08
7.08
7.01
4.3
6.76
6.26
7.01
6.96
8.9
8.49
8.53
8.74
8.46
9.65
8.14
2020 12/35
Annual DV
7.28
6.39
5.12
7.96
8.01
6.08
7.08
7.01
4.3
6.76
6.26
7.01
6.96
8.9
8.49
8.53
8.74
8.46
9.65
8.14
2020 11/35
Annual DV
7.28
6.39
5.12
7.96
8.01
6.08
7.08
7.01
4.3
6.76
6.26
7.01
6.96
8.9
8.49
8.53
8.74
8.46
9.65
8.14
2020 11/30
Annual DV
7.28
6.39
5.12
7.96
8.01
6.08
7.08
7.01
4.3
6.76
6.26
7.01
6.96
8.9
8.49
8.53
8.74
8.46
9.65
8.14
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
}> New Jersey
u> New Jersey
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
County Name
Mercer
Middlesex
Morris
Morris
Ocean
Passaic
Union
Union
Union
Warren
Bernalillo
Bernalillo
Chaves
Dona Ana
Dona Ana
Grant
Sandoval
Sandoval
San Juan
Santa Fe
2005 Annual
DV
11.14
12.15
11.5
10.21
10.92
12.88
14.94
13.32
13.06
12.72
7.03
6.64
6.54
9.95
6.31
5.93
5
7.99
5.92
4.76
2020 Annual
DV
7
7.91
7.42
6.56
6.85
8.59
9.88
8.73
8.45
8.2
5.74
5.41
5.68
8.7
5.55
5.5
4.12
7.15
5.23
4.22
2020 15/35
Annual DV
7
7.91
7.42
6.56
6.85
8.59
9.88
8.73
8.45
8.2
5.74
5.41
5.68
8.7
5.55
5.5
4.12
7.15
5.23
4.22
2020 13/35
Annual DV
7
7.91
7.42
6.56
6.85
8.59
9.88
8.73
8.45
8.2
5.74
5.41
5.68
8.7
5.55
5.5
4.12
7.15
5.23
4.22
2020 12/35
Annual DV
7
7.91
7.42
6.56
6.85
8.59
9.88
8.73
8.45
8.2
5.74
5.41
5.68
8.7
5.55
5.5
4.12
7.15
5.23
4.22
2020 11/35
Annual DV
7
7.91
7.42
6.56
6.85
8.59
9.88
8.73
8.45
8.2
5.74
5.41
5.68
8.7
5.55
5.5
4.12
7.15
5.23
4.22
2020 11/30
Annual DV
7
7.91
7.42
6.56
6.85
8.59
9.88
8.73
8.45
8.2
5.74
5.41
5.68
8.7
5.55
5.5
4.12
7.15
5.23
4.22
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
01
State Name
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
County Name
Albany
Bronx
Bronx
Bronx
Chautauqua
Erie
Erie
Essex
Kings
Monroe
Nassau
New York
New York
New York
New York
Niagara
Onondaga
Orange
Queens
Richmond
2005 Annual
DV
11.83
15.43
13.09
13.45
9.8
12.62
12.64
5.94
14.2
10.64
11.66
16.18
14.8
13.61
15.41
11.96
10.08
10.99
12.18
13.31
2020 Annual
DV
8.43
10.87
8.68
9.41
6.32
8.69
8.65
4.43
9.67
7.63
7.69
11.17
10.06
9.48
10.6
8.55
6.89
7.28
8.15
8.72
2020 15/35
Annual DV
8.43
10.87
8.68
9.41
6.32
8.69
8.65
4.43
9.67
7.63
7.69
11.17
10.06
9.48
10.6
8.55
6.89
7.28
8.15
8.72
2020 13/35
Annual DV
8.43
10.87
8.68
9.41
6.32
8.69
8.65
4.43
9.67
7.63
7.69
11.17
10.06
9.48
10.6
8.55
6.89
7.28
8.15
8.72
2020 12/35
Annual DV
8.43
10.87
8.68
9.41
6.32
8.69
8.65
4.43
9.67
7.63
7.69
11.17
10.06
9.48
10.6
8.55
6.89
7.28
8.15
8.72
2020 11/35
Annual DV
8.43
10.87
8.68
9.41
6.32
8.69
8.65
4.43
9.67
7.63
7.69
9.88
8.77
8.19
9.31
8.55
6.89
7.28
8.15
8.72
2020 11/30
Annual DV
8.43
10.87
8.68
9.41
6.32
8.69
8.65
4.43
9.67
7.63
7.69
10.20
9.09
8.51
9.63
8.55
6.89
7.28
8.15
8.72
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
01
Un
State Name
New York
New York
New York
New York
New York
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
County Name
Richmond
St. Lawrence
Steuben
Suffolk
Westch ester
Alamance
Buncombe
Caswell
Catawba
Chatham
Cumberland
Davidson
Duplin
Durham
Edgecombe
Forsyth
Gaston
Guilford
Haywood
Jackson
2005 Annual
DV
11.59
7.29
9
11.52
11.73
13.94
12.6
13.19
15.31
11.99
13.73
15.17
11.3
13.57
12.37
14.28
14.26
13.79
12.98
12.09
2020 Annual
DV
7.58
5.6
5.83
7.51
7.7
8.38
7.71
7.74
9.17
7.07
8.77
8.89
6.97
8.31
7.76
8.28
8.4
8.17
8.54
7.42
2020 15/35
Annual DV
7.58
5.6
5.83
7.51
7.7
8.38
7.71
7.74
9.17
7.07
8.77
8.89
6.97
8.31
7.76
8.28
8.4
8.17
8.54
7.42
2020 13/35
Annual DV
7.58
5.6
5.83
7.51
7.7
8.38
7.71
7.74
9.17
7.07
8.77
8.89
6.97
8.31
7.76
8.28
8.4
8.17
8.54
7.42
2020 12/35
Annual DV
7.58
5.6
5.83
7.51
7.7
8.38
7.71
7.74
9.17
7.07
8.77
8.89
6.97
8.31
7.76
8.28
8.4
8.17
8.54
7.42
2020 11/35
Annual DV
7.58
5.6
5.83
7.51
7.7
8.38
7.71
7.74
9.17
7.07
8.77
8.89
6.97
8.31
7.76
8.28
8.4
8.17
8.54
7.42
2020 11/30
Annual DV
7.58
5.6
5.83
7.51
7.7
8.38
7.71
7.74
9.17
7.07
8.77
8.89
6.97
8.31
7.76
8.28
8.4
8.17
8.54
7.42
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
01
en
State Name
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Dakota
North Dakota
County Name
Lenoir
McDowell
Martin
Mecklenburg
Mecklenburg
Mecklenburg
Mitchell
Montgomery
New Hanover
Onslow
Orange
Pitt
Robeson
Rowan
Swain
Wake
Watauga
Wayne
Billings
Burke
2005 Annual
DV
11.12
14.24
10.86
15.31
14.74
14.8
12.75
12.35
9.96
10.98
13.12
11.59
12.78
14.02
12.65
13.54
12.05
12.96
4.61
5.9
2020 Annual
DV
6.88
8.92
6.71
9.25
8.77
8.82
7.61
7.33
6.14
6.77
7.86
7.26
8.03
8.35
7.76
8.3
6.85
8.31
4.18
5.51
2020 15/35
Annual DV
6.88
8.92
6.71
9.25
8.77
8.82
7.61
7.33
6.14
6.77
7.86
7.26
8.03
8.35
7.76
8.3
6.85
8.31
4.18
5.51
2020 13/35
Annual DV
6.88
8.92
6.71
9.25
8.77
8.82
7.61
7.33
6.14
6.77
7.86
7.26
8.03
8.35
7.76
8.3
6.85
8.31
4.18
5.51
2020 12/35
Annual DV
6.88
8.92
6.71
9.25
8.77
8.82
7.61
7.33
6.14
6.77
7.86
7.26
8.03
8.35
7.76
8.3
6.85
8.31
4.18
5.51
2020 11/35
Annual DV
6.88
8.92
6.71
9.25
8.77
8.82
7.61
7.33
6.14
6.77
7.86
7.26
8.03
8.35
7.76
8.3
6.85
8.31
4.18
5.51
2020 11/30
Annual DV
6.88
8.92
6.71
9.25
8.77
8.82
7.61
7.33
6.14
6.77
7.86
7.26
8.03
8.35
7.76
8.3
6.85
8.31
4.18
5.51
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
North Dakota
North Dakota
North Dakota
North Dakota
North Dakota
Ohio
Ohio
Ohio
j> Ohio
i Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
County Name
Burke
Burleigh
Cass
McKenzie
Mercer
Athens
Butler
Butler
Butler
Clark
Clermont
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Franklin
Franklin
2005 Annual
DV
5.78
6.61
7.72
5.01
6.04
12.39
15.74
15.36
14.9
14.64
14.15
15.46
13.76
17.37
16.47
17.11
15.97
14.14
15.27
15.08
2020 Annual
DV
5.35
5.59
6.44
4.59
5.38
7.12
9.9
10.08
9.63
9.55
8.58
10.28
9.08
11.79
11.04
11.51
10.64
9.45
9.82
9.7
2020 15/35
Annual DV
5.35
5.59
6.44
4.59
5.38
7.12
9.9
10.08
9.63
9.55
8.58
10.28
9.08
11.79
11.04
11.51
10.64
9.45
9.82
9.7
2020 13/35
Annual DV
5.35
5.59
6.44
4.59
5.38
7.12
9.9
10.08
9.63
9.55
8.58
10.28
9.08
11.79
11.04
11.51
10.64
9.45
9.82
9.7
2020 12/35
Annual DV
5.35
5.59
6.44
4.59
5.38
7.12
9.9
10.08
9.63
9.55
8.58
10.28
9.08
11.79
11.04
11.51
10.64
9.45
9.82
9.7
2020 11/35
Annual DV
5.35
5.59
6.44
4.59
5.38
7.12
9.9
10.08
9.63
9.55
8.58
9.53
8.34
11.04
10.29
10.76
9.89
8.71
9.82
9.7
2020 11/30
Annual DV
5.35
5.59
6.44
4.59
5.38
7.12
9.9
10.08
9.63
9.55
8.58
9.53
8.34
11.04
10.29
10.76
9.89
8.71
9.82
9.7
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
01
00
State Name
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
County Name
Franklin
Greene
Hamilton
Hamilton
Hamilton
Hamilton
Hamilton
Hamilton
Hamilton
Jefferson
Jefferson
Lake
Lawrence
Lorain
Lorain
Lucas
Lucas
Lucas
Mahoning
Mahoning
2005 Annual
DV
14.33
13.36
14.84
17.29
15.5
16.85
15.55
16.17
17.54
15.41
16.51
13.02
15.14
13.87
12.78
14.38
13.95
14.08
14.68
15.12
2020 Annual
DV
9.18
8.34
9.18
10.81
9.45
10.63
9.76
9.99
11.04
9.33
9.96
8.67
9.51
9.12
8.73
9.88
9.52
9.73
9.47
9.9
2020 15/35
Annual DV
9.18
8.34
9.18
10.81
9.45
10.63
9.76
9.99
11.04
9.33
9.96
8.67
9.51
9.12
8.73
9.88
9.52
9.73
9.47
9.9
2020 13/35
Annual DV
9.18
8.34
9.18
10.81
9.45
10.63
9.76
9.99
11.04
9.33
9.96
8.67
9.51
9.12
8.73
9.88
9.52
9.73
9.47
9.9
2020 12/35
Annual DV
9.18
8.34
9.18
10.81
9.45
10.63
9.76
9.99
11.04
9.33
9.96
8.67
9.51
9.12
8.73
9.88
9.52
9.73
9.47
9.9
2020 11/35
Annual DV
9.18
8.34
9.18
10.81
9.45
10.63
9.76
9.99
11.04
9.33
9.96
8.60
9.51
9.05
8.67
9.88
9.52
9.73
9.47
9.9
2020 11/30
Annual DV
9.18
8.34
9.18
10.81
9.45
10.63
9.76
9.99
11.04
9.33
9.96
8.60
9.51
9.05
8.67
9.88
9.52
9.73
9.47
9.9
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
01
ID
State Name
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
County Name
Montgomery
Montgomery
Portage
Preble
Scioto
Stark
Stark
Summit
Summit
Trumbull
Caddo
Cherokee
Kay
Lincoln
Mayes
Mayes
Muskogee
Oklahoma
Oklahoma
Ottawa
2005 Annual
DV
14.58
15.54
13.37
13.7
14.65
16.26
15.23
15.17
14.26
14.53
9.22
11.79
10.26
10.28
11.7
11.44
11.89
10.07
9.86
11.69
2020 Annual
DV
9.31
10
8.7
8.8
8.86
10.41
10.07
10.15
9.54
9.53
6.9
8.59
7.84
7.54
8.64
8.38
8.85
7.25
7.08
8.63
2020 15/35
Annual DV
9.31
10
8.7
8.8
8.86
10.41
10.07
10.15
9.54
9.53
6.9
8.59
7.84
7.54
8.64
8.38
8.85
7.25
7.08
8.63
2020 13/35
Annual DV
9.31
10
8.7
8.8
8.86
10.41
10.07
10.15
9.54
9.53
6.9
8.59
7.84
7.54
8.64
8.38
8.85
7.25
7.08
8.63
2020 12/35
Annual DV
9.31
10
8.7
8.8
8.86
10.41
10.07
10.15
9.54
9.53
6.9
8.59
7.84
7.54
8.64
8.38
8.85
7.25
7.08
8.63
2020 11/35
Annual DV
9.31
10
8.63
8.8
8.86
10.41
10.07
10.08
9.47
9.53
6.9
8.59
7.84
7.54
8.64
8.38
8.85
7.25
7.08
8.63
2020 11/30
Annual DV
9.31
10
8.63
8.8
8.86
10.41
10.07
10.08
9.47
9.53
6.9
8.59
7.84
7.54
8.64
8.38
8.85
7.25
7.08
8.63
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
o
State Name
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
County Name
Pittsburg
Sequoyah
Tulsa
Tulsa
Jackson
Jackson
Klamath
Lane
Lane
Lane
Lane
Multnomah
Multnomah
Union
Adams
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
2005 Annual
DV
11.09
12.99
11.52
11.37
10.32
5.41
11.2
8.64
6.35
7.56
11.93
9.13
8.35
8.35
13.05
15.24
14.66
20.31
13.07
13.84
2020 Annual
DV
8.09
9.68
8.45
8.41
7.68
4.24
8.54
6.39
4.89
5.75
9.43
6.23
5.81
6.74
8.16
9.75
9.26
12.91
7.8
8.49
2020 15/35
Annual DV
8.09
9.68
8.45
8.41
7.68
4.24
8.54
6.39
4.89
5.75
9.43
6.23
5.81
6.74
8.16
7.92
7.43
11.08
5.98
6.66
2020 13/35
Annual DV
8.09
9.68
8.45
8.41
7.68
4.24
8.54
6.39
4.89
5.75
9.43
6.23
5.81
6.74
8.16
7.92
7.43
11.08
5.98
6.66
2020 12/35
Annual DV
8.09
9.68
8.45
8.41
7.68
4.24
8.54
6.39
4.89
5.75
9.43
6.23
5.81
6.74
8.16
7.92
7.43
11.08
5.98
6.66
2020 11/35
Annual DV
8.09
9.68
8.45
8.41
7.68
4.24
8.54
6.39
4.89
5.75
9.43
6.23
5.81
6.74
8.16
7.89
7.40
11.04
5.95
6.63
2020 11/30
Annual DV
8.09
9.68
8.45
8.41
7.68
4.24
8.37
5.35
3.85
4.71
8.39
6.23
5.81
6.74
8.16
6.27
5.78
9.43
4.33
5.01
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
}> Pennsylvania
i-* Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
County Name
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Beaver
Berks
Bucks
Cambria
Centre
Chester
Cumberland
Dauphin
Delaware
Erie
Lackawanna
Lancaster
Lehigh
Luzerne
Mercer
2005 Annual
DV
15.36
15.25
16.26
15.3
14.44
16.38
15.82
13.42
15.4
12.78
15.22
14.45
15.13
15.23
12.54
11.73
16.55
14.5
12.76
13.28
2020 Annual
DV
9.76
9.33
10.06
9.43
8.84
10.51
10.58
8.45
9.45
8
9.64
9.32
9.54
9.74
8.39
7.53
10.73
9.67
8.39
8.47
2020 15/35
Annual DV
7.93
7.50
8.23
7.60
7.01
10.46
10.58
8.45
9.45
8
9.64
9.32
9.54
9.74
8.39
7.53
10.73
9.67
8.39
8.47
2020 13/35
Annual DV
7.93
7.50
8.23
7.60
7.01
10.46
10.58
8.45
9.45
8
9.64
9.32
9.54
9.74
8.39
7.53
10.73
9.67
8.39
8.47
2020 12/35
Annual DV
7.93
7.50
8.23
7.60
7.01
10.46
10.58
8.45
9.45
8
9.64
9.32
9.54
9.74
8.39
7.53
10.73
9.67
8.39
8.47
2020 11/35
Annual DV
7.90
7.47
8.20
7.57
6.98
10.46
10.58
8.45
9.45
8
9.64
9.32
9.54
9.74
8.39
7.53
10.73
9.67
8.39
8.47
2020 11/30
Annual DV
6.28
5.85
6.58
5.95
5.36
10.46
10.58
8.45
9.45
8
9.64
9.32
9.54
9.74
8.39
7.53
10.71
9.67
8.39
8.47
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
NJ
State Name
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Rhode Island
Rhode Island
Rhode Island
Rhode Island
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
County Name
Northampton
Perry
Philadelphia
Washington
Washington
Washington
Westmoreland
York
Providence
Providence
Providence
Providence
Beaufort
Charleston
Charleston
Chesterfield
Edgefield
Florence
Georgetown
Greenville
2005 Annual
DV
13.68
12.81
15.19
15.17
14.92
13.37
15.49
16.52
10.07
12.14
10.82
9.93
11.52
12.21
11.6
12.56
13.17
12.65
12.85
15.65
2020 Annual
DV
8.96
8.22
9.89
8.95
8.75
8.2
9.26
10.66
6.99
8.42
7.56
6.85
7.25
7.86
7.13
7.86
8.5
8.02
8.38
9.7
2020 15/35
Annual DV
8.96
8.22
9.89
8.90
8.70
8.15
9.21
10.66
6.99
8.42
7.56
6.85
7.25
7.86
7.13
7.86
8.5
8.02
8.38
9.7
2020 13/35
Annual DV
8.96
8.22
9.89
8.90
8.70
8.15
9.21
10.66
6.99
8.42
7.56
6.85
7.25
7.86
7.13
7.86
8.5
8.02
8.38
9.7
2020 12/35
Annual DV
8.96
8.22
9.89
8.90
8.70
8.15
9.21
10.66
6.99
8.42
7.56
6.85
7.25
7.86
7.13
7.86
8.5
8.02
8.38
9.7
2020 11/35
Annual DV
8.96
8.22
9.89
8.90
8.70
8.15
9.21
10.66
6.99
8.42
7.56
6.85
7.25
7.86
7.13
7.86
8.5
8.02
8.38
9.7
2020 11/30
Annual DV
8.96
8.22
9.89
8.90
8.70
8.15
9.21
10.66
6.99
8.42
7.56
6.85
7.25
7.86
7.13
7.86
8.5
8.02
8.38
9.7
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
uj
State Name
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
Tennessee
Tennessee
County Name
Greenville
Greenwood
Horry
Lexington
Oconee
Richland
Richland
Spartanburg
Brookings
Brown
Codington
Custer
Jackson
Minnehaha
Minnehaha
Pennington
Pennington
Pennington
Blount
Davidson
2005 Annual
DV
14.66
13.53
12.04
14.64
10.95
13.59
14.24
14.17
9.37
8.42
10.14
5.64
5.39
10.18
9.58
7.48
8.77
7.32
14.3
14.21
2020 Annual
DV
8.91
8.41
7.68
9.28
6.45
8.36
8.89
8.58
7.52
7.03
8.35
5.01
4.65
7.84
7.38
6.56
7.71
6.44
9.09
8.87
2020 15/35
Annual DV
8.91
8.41
7.68
9.28
6.45
8.36
8.89
8.58
7.52
7.03
8.35
5.01
4.65
7.84
7.38
6.56
7.71
6.44
9.09
8.87
2020 13/35
Annual DV
8.91
8.41
7.68
9.28
6.45
8.36
8.89
8.58
7.52
7.03
8.35
5.01
4.65
7.84
7.38
6.56
7.71
6.44
9.09
8.87
2020 12/35
Annual DV
8.91
8.41
7.68
9.28
6.45
8.36
8.89
8.58
7.52
7.03
8.35
5.01
4.65
7.84
7.38
6.56
7.71
6.44
9.09
8.87
2020 11/35
Annual DV
8.91
8.41
7.68
9.28
6.45
8.36
8.89
8.58
7.52
7.03
8.35
5.01
4.65
7.84
7.38
6.56
7.71
6.44
9.09
8.87
2020 11/30
Annual DV
8.91
8.41
7.68
9.28
6.45
8.36
8.89
8.58
7.52
7.03
8.35
5.01
4.65
7.84
7.38
6.56
7.71
6.44
9.09
8.87
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
}> Tennessee
j^ Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
County Name
Davidson
Davidson
Dyer
Hamilton
Hamilton
Hamilton
Knox
Knox
Knox
Lawrence
Loudon
McMinn
Maury
Montgomery
Putnam
Roane
Shelby
Shelby
Shelby
Shelby
2005 Annual
DV
13.99
12.97
12.28
15.67
13.73
15.16
15.47
15.64
15.18
11.69
15.49
14.29
13.21
13.8
13.37
14.49
13.71
13.43
13.68
12.04
2020 Annual
DV
8.63
7.85
7.8
9.91
8.18
9.38
9.66
9.77
9.24
7.39
10.01
8.93
8.39
8.65
8.04
8.94
8.73
8.47
8.6
7.53
2020 15/35
Annual DV
8.63
7.85
7.8
9.91
8.18
9.38
9.66
9.77
9.24
7.39
10.01
8.93
8.39
8.65
8.04
8.94
8.73
8.47
8.6
7.53
2020 13/35
Annual DV
8.63
7.85
7.8
9.91
8.18
9.38
9.66
9.77
9.24
7.39
10.01
8.93
8.39
8.65
8.04
8.94
8.73
8.47
8.6
7.53
2020 12/35
Annual DV
8.63
7.85
7.8
9.91
8.18
9.38
9.66
9.77
9.24
7.39
10.01
8.93
8.39
8.65
8.04
8.94
8.73
8.47
8.6
7.53
2020 11/35
Annual DV
8.63
7.85
7.8
9.91
8.18
9.38
9.66
9.77
9.24
7.39
10.01
8.93
8.39
8.65
8.04
8.94
8.73
8.47
8.6
7.53
2020 11/30
Annual DV
8.63
7.85
7.8
9.91
8.18
9.38
9.66
9.77
9.24
7.39
10.01
8.93
8.39
8.65
8.04
8.94
8.73
8.47
8.6
7.53
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
Un
State Name
Tennessee
Tennessee
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Utah
Utah
County Name
Sullivan
Sumner
Bowie
Dallas
Dallas
Dallas
Ector
El Paso
Harris
Harris
Harrison
Hidalgo
Jefferson
Nueces
Nueces
Orange
Tarrant
Tarrant
Box Elder
Cache
2005 Annual
DV
14.16
13.68
12.85
12.77
11.8
11.15
7.78
9.09
11.77
15.42
11.69
10.98
11.56
10.42
9.63
11.51
11.41
12.23
8.4
11.56
2020 Annual
DV
9.15
8.04
9.21
8.98
8.16
7.54
6.52
7.94
8.65
11.61
7.91
9.17
8.26
7.69
7.04
8.37
7.77
8.41
7.16
9.78
2020 15/35
Annual DV
9.15
8.04
9.21
8.98
8.16
7.54
6.52
7.94
8.65
11.61
7.91
9.17
8.26
7.69
7.04
8.37
7.77
8.41
7.10
8.32
2020 13/35
Annual DV
9.15
8.04
9.21
8.98
8.16
7.54
6.52
7.94
8.65
11.61
7.91
9.17
8.26
7.69
7.04
8.37
7.77
8.41
7.10
8.32
2020 12/35
Annual DV
9.15
8.04
9.21
8.98
8.16
7.54
6.52
7.94
8.65
11.61
7.91
9.17
8.26
7.69
7.04
8.37
7.77
8.41
7.10
8.32
2020 11/35
Annual DV
9.15
8.04
9.21
8.98
8.16
7.54
6.52
7.94
7.94
10.90
7.91
9.17
8.26
7.69
7.04
8.37
7.77
8.41
7.10
8.32
2020 11/30
Annual DV
9.15
8.04
9.21
8.98
8.16
7.54
6.52
7.94
7.94
10.90
7.91
9.17
8.26
7.69
7.04
8.37
7.77
8.41
7.09
7.30
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
en
State Name
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Vermont
Vermont
Vermont
Vermont
Vermont
Vermont
Virginia
County Name
Davis
Salt Lake
Salt Lake
Salt Lake
Salt Lake
Salt Lake
Utah
Utah
Utah
Utah
Weber
Weber
Weber
Addison
Addison
Bennington
Chittenden
Chittenden
Rutland
Arlington
2005 Annual
DV
10.31
11.68
9.21
11.3
12.02
8.33
10
10.52
8.88
8.78
11.16
9.28
9.36
8.94
8.91
8.52
9.27
10.02
11.08
14.27
2020 Annual
DV
8.58
9.29
7.75
9.05
9.72
6.91
8.37
8.8
7.44
7.37
9.23
7.71
7.8
6.75
6.71
6.02
7.16
7.76
8.15
8.87
2020 15/35
Annual DV
8.58
9.00
7.46
8.76
9.43
6.62
8.37
8.8
7.44
7.37
9.16
7.64
7.73
6.75
6.71
6.02
7.16
7.76
8.15
8.87
2020 13/35
Annual DV
8.58
9.00
7.46
8.76
9.43
6.62
8.37
8.8
7.44
7.37
9.16
7.64
7.73
6.75
6.71
6.02
7.16
7.76
8.15
8.87
2020 12/35
Annual DV
8.58
9.00
7.46
8.76
9.43
6.62
8.37
8.8
7.44
7.37
9.16
7.64
7.73
6.75
6.71
6.02
7.16
7.76
8.15
8.87
2020 11/35
Annual DV
8.58
9.00
7.46
8.76
9.43
6.62
8.37
8.8
7.44
7.37
9.16
7.64
7.73
6.75
6.71
6.02
7.16
7.76
8.15
8.87
2020 11/30
Annual DV
8.21
8.01
6.47
7.77
8.43
5.66
7.48
7.91
6.57
6.50
9.15
7.64
7.72
6.75
6.71
6.02
7.16
7.76
8.15
8.87
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
State Name
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Washington
Washington
Washington
Washington
County Name
Charles
Chesterfield
Fairfax
Fairfax
Fairfax
Henrico
Henrico
Loudoun
Page
Bristol City
Hampton City
Lynchburg City
Norfolk City
Roanoke City
Salem City
Virginia Beach City
King
King
King
Pierce
2005 Annual
DV
12.37
13.44
13.33
13.62
13.88
13.51
12.93
13.57
12.79
13.93
12.17
12.84
12.78
14.27
14.69
12.4
9.15
11.24
8.13
10.55
2020 Annual
DV
7.45
8.1
8.34
8.54
8.84
8.09
7.67
8.62
7.49
8.31
7.46
7.5
7.92
8.62
9.12
7.55
6.95
8.35
6.19
8.11
2020 15/35
Annual DV
7.45
8.1
8.34
8.54
8.84
8.09
7.67
8.62
7.49
8.31
7.46
7.5
7.92
8.62
9.12
7.55
6.95
8.35
6.19
8.11
2020 13/35
Annual DV
7.45
8.1
8.34
8.54
8.84
8.09
7.67
8.62
7.49
8.31
7.46
7.5
7.92
8.62
9.12
7.55
6.95
8.35
6.19
8.11
2020 12/35
Annual DV
7.45
8.1
8.34
8.54
8.84
8.09
7.67
8.62
7.49
8.31
7.46
7.5
7.92
8.62
9.12
7.55
6.95
8.35
6.19
8.11
2020 11/35
Annual DV
7.45
8.1
8.34
8.54
8.84
8.09
7.67
8.62
7.49
8.31
7.46
7.5
7.92
8.62
9.12
7.55
6.95
8.35
6.19
8.11
2020 11/30
Annual DV
7.45
8.1
8.34
8.54
8.84
8.09
7.67
8.62
7.49
8.31
7.46
7.5
7.92
8.62
9.12
7.55
6.95
8.35
6.19
7.94
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
oo
State Name
Washington
Washington
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
Wisconsin
Wisconsin
Wisconsin
County Name
Snohomish
Spokane
Berkeley
Brooke
Brooke
Cabell
Hancock
Harrison
Kanawha
Kanawha
Kanawha
Marion
Marshall
Monongalia
Ohio
Raleigh
Wood
Ashland
Brown
Dane
2005 Annual
DV
9.91
9.97
15.93
16.52
16.04
16.3
15.76
13.99
15.15
13.17
16.52
15.03
15.19
14.35
14.58
12.9
15.4
6.07
11.39
12.2
2020 Annual
DV
7.79
7.19
10.37
10.01
9.6
10.25
9.6
8.31
8.89
7.61
9.92
8.93
8.74
8.09
8.27
7.38
9.59
4.85
8.48
8.59
2020 15/35
Annual DV
7.79
7.19
10.37
10.01
9.6
10.25
9.6
8.31
8.89
7.61
9.92
8.93
8.74
8.09
8.27
7.38
9.59
4.85
8.48
8.59
2020 13/35
Annual DV
7.79
7.19
10.37
10.01
9.6
10.25
9.6
8.31
8.89
7.61
9.92
8.93
8.74
8.09
8.27
7.38
9.59
4.85
8.48
8.59
2020 12/35
Annual DV
7.79
7.19
10.37
10.01
9.6
10.25
9.6
8.31
8.89
7.61
9.92
8.93
8.74
8.09
8.27
7.38
9.59
4.85
8.48
8.59
2020 11/35
Annual DV
7.79
7.19
10.37
10.01
9.6
10.25
9.6
8.31
8.89
7.61
9.92
8.93
8.74
8.09
8.27
7.38
9.59
4.85
8.48
8.59
2020 11/30
Annual DV
7.79
7.19
10.37
10.01
9.6
10.25
9.6
8.31
8.89
7.61
9.92
8.93
8.74
8.09
8.27
7.38
9.59
4.85
8.48
8.59
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
en
ID
State Name
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wyoming
Wyoming
Wyoming
County Name
Dodge
Forest
Grant
Kenosha
Manitowoc
Milwaukee
Milwaukee
Milwaukee
Milwaukee
Milwaukee
Outagamie
Ozaukee
St. Croix
Sauk
Taylor
Vilas
Waukesha
Campbell
Campbell
Campbell
2005 Annual
DV
11.04
7.41
11.79
11.98
10.2
13.32
12.88
14.08
13.68
13.54
10.96
11.6
10.09
10.22
8.24
6.78
13.91
6.29
5.11
5.26
2020 Annual
DV
7.95
5.81
8.5
8.47
7.66
9.28
8.88
9.8
9.54
9.37
8.11
8.37
7.8
7.28
6.31
5.35
9.84
5.93
4.8
4.88
2020 15/35
Annual DV
7.95
5.81
8.5
8.47
7.66
9.28
8.88
9.8
9.54
9.37
8.11
8.37
7.8
7.28
6.31
5.35
9.84
5.93
4.8
4.88
2020 13/35
Annual DV
7.95
5.81
8.5
8.47
7.66
9.28
8.88
9.8
9.54
9.37
8.11
8.37
7.8
7.28
6.31
5.35
9.84
5.93
4.8
4.88
2020 12/35
Annual DV
7.95
5.81
8.5
8.47
7.66
9.28
8.88
9.8
9.54
9.37
8.11
8.37
7.8
7.28
6.31
5.35
9.84
5.93
4.8
4.88
2020 11/35
Annual DV
7.95
5.81
8.5
8.47
7.66
9.28
8.88
9.8
9.54
9.37
8.11
8.37
7.8
7.28
6.31
5.35
9.84
5.93
4.8
4.88
2020 11/30
Annual DV
7.95
5.81
8.5
8.47
7.66
9.28
8.88
9.8
9.54
9.37
8.11
8.37
7.8
7.28
6.31
5.35
9.84
5.93
4.8
4.88
(continued)
-------
Table 4.A-11. Annual Design Values (DVs) for the 2005 and 2020 Base Case and after Meeting the Current and Proposed
Alternative Standard Levels: 2020 Baseline (15/35), 2020 12/35 and 2020 11/35 (continued)
State Name
Wyoming
Wyoming
Wyoming
Wyoming
County Name
Converse
Fremont
La ramie
Sheridan
2005 Annual
DV
3.58
8.17
4.48
9.7
2020 Annual
DV
3.29
7.25
3.81
8.65
2020 15/35
Annual DV
3.29
7.25
3.81
8.65
2020 13/35
Annual DV
3.29
7.25
3.81
8.65
2020 12/35
Annual DV
3.29
7.25
3.81
8.65
2020 11/35
Annual DV
3.29
7.25
3.81
8.65
2020 11/30
Annual DV
3.29
7.25
3.81
8.65
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35) and
2020 11/30
FIPS
1003
1027
1033
1049
1053
1055
1069
1073
1073
1073
1073
1073
1073
1073
1073
1089
1097
1097
1101
1103
1113
1117
1119
1121
1125
1127
4003
4005
4007
Monitor ID
10030010
10270001
10331002
10491003
10530002
10550010
10690003
10730023
10731005
10731009
10731010
10732003
10732006
10735002
10735003
10890014
10970002
10970003
11010007
11030011
11130001
11170006
11190002
11210002
11250004
11270002
40031005
40051008
40070008
State Name
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Alabama
Arizona
Arizona
Arizona
County Name
Baldwin
Clay
Colbert
De Kalb
Escambia
Etowah
Houston
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Jefferson
Madison
Mobile
Mobile
Montgomery
Morgan
Russell
Shelby
Sumter
Talladega
Tuscaloosa
Walker
Cochise
Coconino
Gila
2005
24-hr DV
26.21
31.88
30.43
32.08
29.03
35.18
28.66
44.06
34.83
34.5
34.16
40.3
33.17
33.05
35.81
33.58
30.03
28.58
32.05
31.58
35.55
32.05
28.9
33.46
29.8
32.82
16.62
17.11
22.12
2020
24-hr DV
16.25
17.20
15.63
17.07
18.87
19.12
18.24
27.91
21.49
17.88
18.12
28.55
18.34
17.43
19.46
17.12
18.72
17.74
18.56
15.05
23.14
18.39
16.10
18.57
16.98
17.34
15.73
16.21
19.76
2020
15/35
24-hr DV
16.25
17.20
15.63
17.07
18.87
19.12
18.24
27.91
21.49
17.88
18.12
28.55
18.34
17.43
19.46
17.12
18.72
17.74
18.56
15.05
23.14
18.39
16.10
18.57
16.98
17.34
15.73
16.21
19.76
2020
11/30
24-hr DV
16.25
17.20
15.63
17.07
18.87
19.12
18.24
24.14
17.73
14.12
14.36
24.79
14.58
13.67
15.70
17.12
18.72
17.74
18.56
15.05
23.14
18.39
16.10
18.57
16.98
17.34
15.69
16.21
19.76
(continued)
4.A-71
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
4013
4013
4013
4019
4019
4021
4021
4023
5001
5003
5035
5045
5051
5107
5113
5115
5119
5119
5119
5139
5145
6001
6001
6007
6009
6011
6013
6019
6019
Monitor ID
40130019
40134003
40139997
40190011
40191028
40210001
40213002
40230004
50010011
50030005
50350005
50450002
50510003
51070001
51130002
51150003
51190007
51191004
51191005
51390006
51450001
60010007
60011001
60070002
60090001
60111002
60130002
60190008
60195001
State Name
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
Arkansas
California
California
California
California
California
California
California
California
County Name
Maricopa
Maricopa
Maricopa
Pima
Pima
Pinal
Pinal
Santa Cruz
Arkansas
Ashley
Crittenden
Faulkner
Garland
Phillips
Polk
Pope
Pulaski
Pulaski
Pulaski
Union
White
Alameda
Alameda
Butte
Calaveras
Colusa
Contra Costa
Fresno
Fresno
2005
24-hr DV
32.8
31.46
26.3
12.27
11.34
17.55
11.85
36.08
29.16
28.91
35.06
29.87
29.27
29.18
26.13
28.32
31.16
31.93
31.91
28.7
29.91
32.58
29.44
52.55
20.55
26.16
34.7
60.22
56.15
2020
24-hr DV
24.35
24.06
19.01
9.64
8.55
14.56
10.5
33.82
18.21
21.00
18.60
19.26
18.67
18.10
15.71
17.96
19.81
21.88
21.73
19.81
19.33
24.22
21.44
32.16
13.86
20.47
25.15
41.03
38.30
2020
15/35
24-hr DV
24.35
24.06
19.01
9.64
8.55
14.56
10.53
33.82
18.21
21.00
18.60
19.26
18.67
18.10
15.71
17.96
19.81
21.88
21.73
19.81
19.33
24.22
21.44
32.16
13.86
20.47
25.15
30.89
28.29
2020
11/30
24-hr DV
24.35
24.06
19.01
9.59
8.50
14.56
10.53
30.49
18.21
21.00
18.60
19.26
18.67
18.10
15.71
17.96
19.81
21.88
21.73
19.81
19.33
24.22
21.44
30.46
13.86
20.47
25.15
24.80
22.25
(continued)
4.A-72
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
6019
6025
6025
6025
6027
6029
6029
6029
6031
6033
6037
6037
6037
6037
6037
6037
6037
6037
6037
6037
6045
6047
6053
6057
6057
6059
6059
6061
6063
Monitor ID
60195025
60250005
60250007
60251003
60271003
60290010
60290014
60290016
60310004
60333001
60370002
60371002
60371103
60371201
60371301
60371602
60372005
60374002
60374004
60379033
60450006
60472510
60531003
60570005
60571001
60590007
60592022
60610006
60631006
State Name
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
County Name
Fresno
Imperial
Imperial
Imperial
Inyo
Kern
Kern
Kern
Kings
Lake
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Mendocino
Merced
Monterey
Nevada
Nevada
Orange
Orange
Placer
Plumas
2005
24-hr DV
58.83
40.21
21.63
23.32
20
64.54
61.65
60.38
58.06
12.94
49.85
49.7
50.97
42.4
48.71
50.2
42.2
41.42
39.38
17.11
15.3
46.15
14.35
13.93
16.55
43.76
33.85
29.88
29.33
2020
24-hr DV
40.35
31.82
17.11
18.84
17.87
45.93
44.08
44.51
42.42
11.83
35.54
37.38
38.39
31.36
39.72
39.16
28.18
34.83
33.49
13.71
9.415
31.30
10.98
11.13
12.65
34.02
29.68
20.31
22.60
2020
15/35
24-hr DV
30.32
31.82
17.11
18.84
15.53
24.00
22.21
22.60
35.49
11.83
25.18
26.69
27.52
21.57
29.16
28.51
17.75
24.66
23.49
5.166
9.41
28.37
8.783
11.13
12.65
31.01
27.43
20.31
22.60
2020
11/30
24-hr DV
24.28
30.40
15.70
17.42
14.69
21.62
19.86
20.23
28.94
11.83
24.24
25.66
26.43
20.77
28.17
27.50
16.78
23.78
22.66
4.73
9.41
27.38
7.98
11.13
12.65
28.31
24.94
20.31
22.60
(continued)
4.A-73
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
6063
6065
6065
6065
6067
6067
6067
6071
6071
6071
6071
6071
6073
6073
6073
6073
6073
6075
6077
6079
6079
6081
6083
6085
6085
6089
6095
6097
6099
Monitor ID
60631009
60651003
60652002
60658001
60670006
60670010
60674001
60710025
60710306
60712002
60718001
60719004
60730001
60730003
60730006
60731002
60731010
60750005
60771002
60792006
60798001
60811001
60830011
60850005
60852003
60890004
60950004
60970003
60990005
State Name
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
California
County Name
Plumas
Riverside
Riverside
Riverside
Sacramento
Sacramento
Sacramento
San Bernardino
San Bernardino
San Bernardino
San Bernardino
San Bernardino
San Diego
San Diego
San Diego
San Diego
San Diego
San Francisco
San Joaquin
San Luis Obispo
San Luis Obispo
San Mateo
Santa Barbara
Santa Clara
Santa Clara
Shasta
Solano
Sonoma
Stanislaus
2005
24-hr DV
32.44
48.88
24.22
59.13
49.22
41.55
39.55
51.9
23.11
52.85
37.56
55.5
30.58
35.55
24.11
33.28
33.17
30.91
41.88
15.03
22.58
29.41
24.07
38.61
35.9
20.42
34.76
29.1
51.48
2020
24-hr DV
24.05
39.75
18.47
46.57
32.92
30.10
28.08
41.28
16.64
39.80
33.65
41.53
21.96
25.35
17.52
23.94
23.49
22.10
29.94
11.58
17.17
21.72
16.45
27.62
25.34
12.79
25.26
18.67
37.06
2020
15/35
24-hr DV
24.05
28.75
8.969
35.49
32.92
30.10
28.08
34.02
10.43
32.26
27.93
34.44
21.96
25.35
17.52
23.94
23.49
22.10
26.89
9.280
14.54
21.72
16.45
24.48
22.26
12.79
25.26
18.67
31.79
2020
11/30
24-hr DV
24.05
21.74
2.31
28.45
27.12
24.30
22.27
26.86
3.53
25.03
21.16
27.32
21.96
25.35
17.52
23.94
23.49
22.10
25.87
8.46
13.63
21.72
16.45
23.36
21.16
12.79
25.26
18.67
28.73
(continued)
4.A-74
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
6101
6107
6111
6111
6111
6111
6113
8001
8005
8013
8013
8029
8031
8031
8039
8041
8041
8069
8077
8101
8113
8123
8123
9001
9001
9001
9001
9003
9005
Monitor ID
61010003
61072002
61110007
61110009
61112002
61113001
61131003
80010006
80050005
80130003
80130012
80290004
80310002
80310023
80390001
80410008
80410011
80690009
80770017
81010012
81130004
81230006
81230008
90010010
90011123
90013005
90019003
90031003
90050005
State Name
California
California
California
California
California
California
California
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Colorado
Connecticut
Connecticut
Connecticut
Connecticut
Connecticut
Connecticut
County Name
Sutter
Tulare
Ventura
Ventura
Ventura
Ventura
Yolo
Adams
Arapahoe
Boulder
Boulder
Delta
Denver
Denver
Elbert
El Paso
El Paso
Larimer
Mesa
Pueblo
San Miguel
Weld
Weld
Fairfield
Fairfield
Fairfield
Fairfield
Hartford
Litchfield
2005
24-hr DV
38.55
56.63
26.43
21.53
30.3
25.4
30.38
25.35
21.27
21.12
18.7
20.76
26.44
26.36
13.18
16.41
16.51
18.3
23.51
15.42
10.11
22.9
18.38
36.27
32.27
34.91
33.66
31.83
27.16
2020
24-hr DV
25.21
40.12
19.13
16.18
22.58
17.38
22.66
17.94
15.47
16.19
14.40
14.09
19.47
19.57
10.20
10.29
10.72
13.96
17.04
10.93
9.29
17.47
14.08
22.27
20.46
19.74
18.34
17.84
13.00
2020
15/35
24-hr DV
25.21
31.82
16.91
13.96
20.18
15.67
22.66
17.94
15.47
16.19
14.40
14.09
19.47
19.57
10.20
10.29
10.72
13.96
17.04
10.93
9.29
17.47
14.08
22.27
20.46
19.74
18.34
17.84
13.00
2020
11/30
24-hr DV
25.21
27.84
16.28
13.32
19.50
15.19
22.66
17.94
15.47
16.19
14.40
14.09
19.47
19.57
10.20
10.29
10.72
13.96
17.04
10.93
9.29
17.47
14.08
22.27
20.46
19.74
18.34
17.84
13.00
(continued)
4.A-75
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
9009
9009
9009
9009
9009
9011
10001
10001
10003
10003
10003
10003
10005
11001
11001
11001
12001
12001
12005
12009
12011
12011
12011
12017
12031
12031
12033
12057
12057
Monitor ID
90090026
90090027
90091123
90092008
90092123
90113002
100010002
100010003
100031003
100031007
100031012
100032004
100051002
110010041
110010042
110010043
120010023
120010024
120051004
120090007
120111002
120112004
120113002
120170005
120310098
120310099
120330004
120570030
120573002
State Name
Connecticut
Connecticut
Connecticut
Connecticut
Connecticut
Connecticut
Delaware
Delaware
Delaware
Delaware
Delaware
Delaware
Delaware
D.C.
D.C.
D.C.
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
County Name
New Haven
New Haven
New Haven
New Haven
New Haven
New London
Kent
Kent
New Castle
New Castle
New Castle
New Castle
Sussex
Washington
Washington
Washington
Alachua
Alachua
Bay
Brevard
Broward
Broward
Broward
Citrus
Duval
Duval
Escambia
Hillsborough
Hillsborough
2005
24-hr DV
35.65
35.58
38.37
33.68
34.45
32.03
32.14
31.5
34.36
32.65
33.5
36.66
33.78
36.35
34.95
34.16
21.35
20.98
28.08
20.73
18.34
18.63
15.96
21.22
23.72
24.35
28.8
23.44
22.25
2020
24-hr DV
20.29
20.10
21.78
19.01
20.00
16.66
18.35
17.57
21.64
17.11
22.01
22.47
19.48
20.93
20.29
20.14
12.18
13.56
18.31
12.88
13.22
13.16
10.92
11.83
16.13
17.77
20.86
15.31
13.16
2020
15/35
24-hr DV
20.29
20.10
21.78
19.01
20.00
16.66
18.35
17.57
21.64
17.11
22.01
22.47
19.48
20.93
20.29
20.14
12.18
13.56
18.31
12.88
13.22
13.16
10.92
11.83
16.13
17.77
20.86
15.31
13.16
2020
11/30
24-hr DV
20.29
20.10
21.78
19.01
20.00
16.66
18.35
17.57
21.64
17.11
22.01
22.47
19.48
20.93
20.29
20.14
12.18
13.56
18.31
12.88
13.22
13.16
10.92
11.83
16.13
17.77
20.86
15.31
13.16
(continued)
4.A-76
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
12071
12073
12081
12083
12086
12086
12095
12099
12099
12103
12103
12105
12111
12115
12117
12127
13021
13021
13051
13051
13063
13067
13067
13089
13089
13095
13115
13121
13121
Monitor ID
120710005
120730012
120814012
120830003
120861016
120866001
120952002
120990009
120992005
121030018
121031009
121056006
121111002
121150013
121171002
121275002
130210007
130210012
130510017
130510091
130630091
130670003
130670004
130890002
130892001
130950007
131150005
131210032
131210039
State Name
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
County Name
Lee
Leon
Manatee
Marion
Miami-Dade
Miami-Dade
Orange
Palm Beach
Palm Beach
Pinellas
Pinellas
Polk
St Lucie
Sarasota
Seminole
Volusia
Bibb
Bibb
Chatham
Chatham
Clayton
Cobb
Cobb
De Kalb
De Kalb
Dougherty
Floyd
Fulton
Fulton
2005
24-hr DV
17.7
27.03
19.57
22.56
19.13
18.6
21.83
17.73
18.22
21.73
20.8
19.3
18.18
19.22
22.08
22
33.56
30.74
28.45
27.9
35.88
35.04
34.12
33.44
33.92
34.15
35.12
34.13
37.66
2020
24-hr DV
12.27
18.52
11.54
13.49
11.54
12.83
12.99
13.63
12.75
14.58
13.91
12.88
11.46
12.19
12.45
12.81
21.70
17.76
18.78
18.29
20.95
19.81
19.35
18.19
19.54
23.51
21.22
19.56
22.76
2020
15/35
24-hr DV
12.27
18.52
11.54
13.49
11.54
12.83
12.99
13.63
12.75
14.58
13.91
12.88
11.46
12.19
12.45
12.81
21.70
17.76
18.78
18.29
20.95
19.81
19.35
18.19
19.54
23.51
21.22
19.56
22.76
2020
11/30
24-hr DV
12.27
18.52
11.54
13.49
11.54
12.83
12.99
13.63
12.75
14.58
13.91
12.88
11.46
12.19
12.45
12.81
21.27
17.33
18.78
18.29
20.95
19.81
19.35
18.19
19.54
23.51
21.22
19.56
22.45
(continued)
4.A-77
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
13127
13135
13139
13153
13185
13215
13215
13215
13223
13245
13245
13295
13303
13319
16001
16005
16009
16027
16041
16049
16059
16077
16079
17001
17019
17019
17031
17031
17031
Monitor ID
131270006
131350002
131390003
131530001
131850003
132150001
132150008
132150011
132230003
132450005
132450091
132950002
133030001
133190001
160010011
160050015
160090010
160270004
160410001
160490003
160590004
160770011
160790017
170010006
170190004
170191001
170310022
170310050
170310052
State Name
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
Idaho
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
County Name
Glynn
Gwinnett
Hall
Houston
Lowndes
Muscogee
Muscogee
Muscogee
Paulding
Richmond
Richmond
Walker
Washington
Wilkinson
Ada
Bannock
Benewah
Canyon
Franklin
Idaho
Lemhi
Power
Shoshone
Adams
Champaign
Champaign
Cook
Cook
Cook
2005
24-hr DV
26.13
32.81
30.11
29.63
25.68
31.38
34.58
30.25
33.02
32.7
31.97
30.98
30.83
33.16
28.36
27.08
32.94
31.8
36.76
28.43
36.53
33.36
38.16
31.41
31.32
30.04
36.61
36.11
40.26
2020
24-hr DV
18.04
18.28
18.95
17.62
16.90
21.83
22.33
20.09
18.66
23.41
21.68
18.36
18.99
20.66
25.01
24.18
28.46
26.29
30.49
26.37
31.32
29.75
31.96
18.25
18.99
19.02
28.61
25.08
26.51
2020
15/35
24-hr DV
18.04
18.28
18.95
17.62
16.90
21.83
22.33
20.09
18.66
23.41
21.68
18.36
18.99
20.66
25.01
24.18
28.46
26.29
30.49
26.37
31.32
29.75
31.96
18.25
18.99
19.02
28.61
25.08
26.51
2020
11/30
24-hr DV
18.04
18.28
18.95
17.59
16.90
21.83
22.33
20.09
18.66
23.41
21.68
18.36
18.99
20.66
25.01
24.18
28.38
26.29
30.43
26.18
30.48
29.75
30.48
18.25
18.99
19.02
28.25
24.72
26.15
(continued)
4.A-78
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
17031
17031
17031
17031
17031
17031
17031
17031
17031
17043
17065
17083
17089
17089
17097
17099
17111
17113
17115
17119
17119
17119
17119
17143
17157
17161
17163
17163
17167
Monitor ID
170310057
170310076
170311016
170312001
170313103
170313301
170314007
170314201
170316005
170434002
170650002
170831001
170890003
170890007
170971007
170990007
171110001
171132003
171150013
171190023
171191007
171192009
171193007
171430037
171570001
171613002
171630010
171634001
171670012
State Name
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
County Name
Cook
Cook
Cook
Cook
Cook
Cook
Cook
Cook
Cook
Du Page
Hamilton
Jersey
Kane
Kane
Lake
LaSalle
McHenry
McLean
Macon
Madison
Madison
Madison
Madison
Peoria
Randolph
Rock Island
StClair
StClair
Sangamon
2005
24-hr DV
37.37
38.05
43.03
37.7
39.65
40.22
34.31
32
39.17
34.64
31.6
32.18
33.85
34.83
33.08
28.92
31.58
33.43
33.25
37.31
39.16
34.97
34.03
32.76
28.96
30.9
33.7
31.91
33.41
2020
24-hr DV
25.25
25.52
28.93
26.75
27.06
26.35
22.60
21.74
29.02
25.32
16.98
19.67
23.48
25.12
21.19
19.62
20.88
20.92
18.70
24.71
25.29
22.05
19.78
21.05
20.19
22.57
22.13
22.79
21.67
2020
15/35
24-hr DV
25.25
25.52
28.93
26.75
27.06
26.35
22.60
21.74
29.02
25.32
16.98
19.67
23.48
25.12
21.19
19.62
20.88
20.92
18.70
24.71
25.29
22.05
19.78
21.05
20.19
22.57
22.13
22.79
21.67
2020
11/30
24-hr DV
24.89
25.16
28.56
26.38
26.70
25.98
22.23
21.38
28.65
25.32
16.98
18.93
23.48
25.12
21.19
19.62
20.88
20.92
18.70
22.72
23.30
20.06
17.80
21.05
20.19
22.57
21.35
22.07
21.67
(continued)
4.A-79
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
17197
17197
17201
18003
18003
18019
18035
18037
18039
18043
18065
18067
18083
18089
18089
18089
18089
18089
18089
18089
18089
18091
18091
18095
18097
18097
18097
18097
18097
Monitor ID
171971002
171971011
172010013
180030004
180030014
180190006
180350006
180372001
180390003
180431004
180650003
180670003
180830004
180890006
180890022
180890026
180890027
180890031
180891003
180892004
180892010
180910011
180910012
180950009
180970042
180970043
180970066
180970078
180970079
State Name
Illinois
Illinois
Illinois
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
County Name
Will
Will
Winnebago
Allen
Allen
Clark
Delaware
Dubois
Elkhart
Floyd
Henry
Howard
Knox
Lake
Lake
Lake
Lake
Lake
Lake
Lake
Lake
La Porte
La Porte
Madison
Marion
Marion
Marion
Marion
Marion
2005
24-hr DV
36.45
30.71
34.73
33.1
30.51
37.57
32.07
35.36
34.43
33.26
31.86
32.21
35.92
34.97
38.98
38.42
32.63
34
32.71
32.91
34.23
33
30.61
32.82
34.23
38.47
38.31
36.64
35.61
2020
24-hr DV
24.37
17.71
24.49
23.14
20.81
20.97
20.32
21.93
25.09
17.44
19.30
20.20
21.44
26.23
29.59
27.13
23.66
22.52
24.90
26.34
25.77
21.47
21.63
20.04
20.92
23.70
24.20
22.66
21.22
2020
15/35
24-hr DV
24.37
17.71
24.49
23.14
20.81
20.97
20.32
21.93
25.09
17.44
19.30
20.20
21.44
26.23
29.59
27.13
23.66
22.52
24.90
26.34
25.77
21.47
21.63
20.04
20.92
23.70
24.20
22.66
21.22
2020
11/30
24-hr DV
24.37
17.71
24.49
23.14
20.81
20.97
20.32
21.93
25.09
17.44
19.30
20.20
21.44
26.23
29.59
27.13
23.66
22.52
24.90
26.34
25.77
21.47
21.63
20.04
20.92
23.70
24.20
22.66
21.22
(continued)
4.A-80
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
18097
18097
18127
18127
18141
18141
18141
18147
18157
18163
18163
18163
18167
18167
19013
19045
19103
19113
19137
19139
19147
19153
19153
19153
19155
19163
19163
19163
19177
Monitor ID
180970081
180970083
181270020
181270024
181410014
181411008
181412004
181470009
181570008
181630006
181630012
181630016
181670018
181670023
190130008
190450021
191032001
191130037
191370002
191390015
191471002
191530030
191532510
191532520
191550009
191630015
191630018
191630019
191770006
State Name
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
County Name
Marion
Marion
Porter
Porter
St Joseph
St Joseph
St Joseph
Spencer
Tippecanoe
Vanderburgh
Vanderburgh
Vanderburgh
Vigo
Vigo
Black Hawk
Clinton
Johnson
Linn
Montgomery
Muscatine
Palo Alto
Polk
Polk
Polk
Pottawattamie
Scott
Scott
Scott
Van Buren
2005
24-hr DV
38.2
36.63
32.96
31.87
32.45
33.16
30.04
32.32
35.68
34.8
33.27
32.66
34.6
34.88
30.78
33.95
34.67
30.6
27.5
36.03
25.73
28.41
27.26
31.46
28.6
31.01
32.34
37.1
28.36
2020
24-hr DV
23.59
23.85
22.00
23.05
23.43
24.73
23.49
15.43
20.81
22.81
22.48
22.91
20.64
19.44
21.69
23.86
24.38
20.62
17.47
27.11
18.02
20.25
18.56
22.22
21.11
21.33
23.19
25.09
19.89
2020
15/35
24-hr DV
23.59
23.85
22.00
23.05
23.43
24.73
23.49
15.43
20.81
22.81
22.48
22.91
20.64
19.44
21.69
23.86
24.38
20.62
17.47
27.11
18.02
20.25
18.56
22.22
21.11
21.33
23.19
25.09
19.89
2020
11/30
24-hr DV
23.59
23.85
22.00
23.05
23.43
24.73
23.49
15.43
20.81
22.81
22.48
22.91
20.64
19.44
21.69
23.86
24.38
20.62
17.47
27.11
18.02
20.25
18.56
22.22
21.11
21.33
23.19
25.09
19.89
(continued)
4.A-81
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
19193
19197
20091
20091
20091
20107
20173
20173
20173
20177
20191
20209
20209
21013
21019
21029
21037
21043
21047
21059
21067
21067
21073
21093
21101
21111
21111
21111
21111
Monitor ID
191930017
191970004
200910007
200910009
200910010
201070002
201730008
201730009
201730010
201770010
201910002
202090021
202090022
210130002
210190017
210290006
210370003
210430500
210470006
210590005
210670012
210670014
210730006
210930006
211010014
211110043
211110044
211110048
211110051
State Name
Iowa
Iowa
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
County Name
Woodbury
Wright
Johnson
Johnson
Johnson
Linn
Sedgwick
Sedgwick
Sedgwick
Shawnee
Sumner
Wyandotte
Wyandotte
Bell
Boyd
Bullitt
Campbell
Carter
Christian
Daviess
Fayette
Fayette
Franklin
Hardin
Henderson
Jefferson
Jefferson
Jefferson
Jefferson
2005
24-hr DV
26.4
28.65
25.37
29.3
23.55
25.38
23.7
25.01
25.37
29.16
22.84
29.58
26.6
29.9
33.15
34.63
31.2
29.91
33.6
33.86
31.97
32.23
32.17
32.81
31.85
35.48
36.16
36.44
32.4
2020
24-hr DV
19.58
19.15
17.45
22.43
15.14
17.91
16.04
17.07
18.21
21.81
16.11
21.07
18.33
16.93
16.09
17.50
16.22
13.49
16.00
16.90
16.44
17.70
17.09
15.86
17.66
18.39
19.20
20.47
15.08
2020
15/35
24-hr DV
19.58
19.15
17.45
22.43
15.14
17.91
16.04
17.07
18.21
21.81
16.11
21.07
18.33
16.93
16.09
17.50
16.22
13.49
16.00
16.90
16.44
17.70
17.09
15.86
17.66
18.39
19.20
20.47
15.08
2020
11/30
24-hr DV
19.58
19.15
17.45
22.43
15.14
17.91
16.04
17.07
18.21
21.81
16.11
21.07
18.33
16.93
16.09
17.50
16.22
13.49
16.00
16.90
16.44
17.70
17.09
15.86
17.66
18.39
19.20
20.47
15.08
(continued)
4.A-82
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
21117
21125
21145
21151
21193
21195
21227
22017
22019
22019
22029
22033
22033
22047
22047
22051
22055
22073
22079
22105
22109
22121
23001
23003
23003
23005
23005
23009
23011
Monitor ID
211170007
211250004
211451004
211510003
211930003
211950002
212270007
220171002
220190009
220190010
220290003
220330009
220331001
220470005
220470009
220511001
220550006
220730004
220790002
221050001
221090001
221210001
230010011
230030013
230031011
230050015
230050027
230090103
230110016
State Name
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Louisiana
Maine
Maine
Maine
Maine
Maine
Maine
Maine
County Name
Kenton
Laurel
McCracken
Madison
Perry
Pike
Warren
Caddo
Calcasieu
Calcasieu
Concordia
East Baton Rouge
East Baton Rouge
Iberville
Iberville
Jefferson
Lafayette
Ouachita
Rapides
Tangipahoa
Terrebonne
West Baton Rouge
Androscoggin
Aroostook
Aroostook
Cumberland
Cumberland
Hancock
Kennebec
2005
24-hr DV
34.74
25.16
33.62
30.11
28.54
30.52
33.14
27.56
24.28
26.38
26.16
29.36
25.47
28.62
26.14
27.06
24.28
28.91
30.26
29.61
26.25
29.08
26.56
24.23
22.91
27.74
29.2
19.43
26.21
2020
24-hr DV
19.01
13.86
17.07
15.14
13.52
15.35
16.04
18.95
16.84
17.48
16.01
21.12
17.97
21.52
16.97
16.37
15.98
19.57
18.76
18.23
16.19
20.94
16.78
20.18
17.01
16.79
17.48
11.79
15.86
2020
15/35
24-hr DV
19.01
13.86
17.07
15.14
13.52
15.35
16.04
18.95
16.84
17.48
16.01
21.12
17.97
21.52
16.97
16.37
15.98
19.57
18.76
18.23
16.19
20.94
16.78
20.18
17.01
16.79
17.48
11.79
15.86
2020
11/30
24-hr DV
19.01
13.86
17.07
15.14
13.52
15.35
16.04
18.95
16.84
17.48
16.01
21.12
17.97
21.52
16.97
16.37
15.98
19.57
18.76
18.23
16.19
20.94
16.78
20.18
17.01
16.79
17.48
11.79
15.86
(continued)
4.A-83
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
23017
23019
24003
24003
24003
24005
24005
24015
24025
24031
24033
24033
24043
24510
24510
24510
24510
24510
24510
25003
25005
25009
25009
25009
25013
25013
25013
25023
25025
Monitor ID
230172011
230190002
240030014
240031003
240032002
240051007
240053001
240150003
240251001
240313001
240330030
240338003
240430009
245100006
245100007
245100008
245100035
245100040
245100049
250035001
250051004
250092006
250095005
250096001
250130008
250130016
250132009
250230004
250250002
State Name
Maine
Maine
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Maryland
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
County Name
Oxford
Penobscot
Anne Arundel
Anne Arundel
Anne Arundel
Baltimore
Baltimore
Cecil
Harford
Montgomery
Prince Georges
Prince Georges
Washington
Baltimore City
Baltimore City
Baltimore City
Baltimore City
Baltimore City
Baltimore City
Berkshire
Bristol
Essex
Essex
Essex
Hampden
Hampden
Hampden
Plymouth
Suffolk
2005
24-hr DV
28.36
22.03
33.23
35.55
36.16
33.33
35.84
30.82
31.21
30.93
31.73
33.46
33.43
33.38
34.74
37.21
37.75
39.01
38.16
31.06
25.07
28.72
26.85
27.8
27.26
32.3
33.13
28.48
29.45
2020
24-hr DV
18.96
14.30
17.56
22.09
23.69
20.06
21.66
19.07
17.19
17.19
17.45
18.03
20.15
21.16
22.38
24.33
24.85
25.23
26.21
19.54
15.30
18.00
14.98
17.48
17.08
20.35
20.80
16.46
19.75
2020
15/35
24-hr DV
18.96
14.30
17.56
22.09
23.69
20.06
21.66
19.07
17.19
17.19
17.45
18.03
20.15
21.16
22.38
24.33
24.85
25.23
26.21
19.54
15.30
18.00
14.98
17.48
17.08
20.35
20.80
16.46
19.75
2020
11/30
24-hr DV
18.96
14.30
17.56
22.09
23.69
20.06
21.66
19.07
17.19
17.19
17.45
18.03
20.15
21.16
22.38
24.33
24.85
25.23
26.21
19.54
15.30
18.00
14.98
17.48
17.08
20.35
20.80
16.46
19.75
(continued)
4.A-84
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
25025
25025
25025
25027
25027
26005
26017
26021
26049
26065
26077
26081
26099
26113
26115
26121
26125
26139
26145
26147
26161
26161
26163
26163
26163
26163
26163
26163
26163
Monitor ID
250250027
250250042
250250043
250270016
250270023
260050003
260170014
260210014
260490021
260650012
260770008
260810020
260990009
261130001
261150005
261210040
261250001
261390005
261450018
261470005
261610005
261610008
261630001
261630015
261630016
261630019
261630025
261630033
261630036
State Name
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Massachusetts
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
County Name
Suffolk
Suffolk
Suffolk
Worcester
Worcester
Allegan
Bay
Berrien
Genesee
Ingham
Kalamazoo
Kent
Macomb
Missaukee
Monroe
Muskegon
Oakland
Ottawa
Saginaw
StClair
Washtenaw
Washtenaw
Wayne
Wayne
Wayne
Wayne
Wayne
Wayne
Wayne
2005
24-hr DV
29.23
28.6
32.17
30.01
30.66
33.82
31.68
31.32
30.46
31.96
31.17
36.53
35.32
24.83
38.88
34.71
39.94
34.24
30.66
39.61
33.6
39.46
37.83
40.12
42.92
40.92
35.18
43.88
37.16
2020
24-hr DV
19.20
19.01
20.80
17.73
18.47
23.99
21.73
21.21
21.92
22.40
21.05
23.50
27.12
15.51
23.36
23.72
24.16
25.06
20.77
28.92
22.74
23.39
25.52
27.89
30.23
30.69
22.91
31.37
25.77
2020
15/35
24-hr DV
19.20
19.01
20.80
17.73
18.47
23.99
21.73
21.21
21.92
22.40
21.05
23.50
27.12
15.51
23.36
23.72
24.16
25.06
20.77
28.92
22.74
23.39
25.52
27.89
30.23
30.69
22.91
31.37
25.77
2020
11/30
24-hr DV
19.20
19.01
20.80
17.73
18.47
23.99
21.73
21.21
21.92
22.40
21.05
23.50
27.12
15.51
23.36
23.72
24.16
25.06
20.77
28.92
22.74
23.39
21.34
23.71
26.06
26.51
18.73
27.19
21.59
(continued)
4.A-85
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
26163
27021
27037
27053
27053
27053
27053
27053
27053
27095
27123
27123
27123
27137
27137
27137
27139
28001
28011
28033
28035
28047
28049
28059
28067
28081
28087
28149
29019
Monitor ID
261630039
270210001
270370470
270530050
270530961
270530963
270530965
270531007
270532006
270953051
271230866
271230868
271230871
271377001
271377550
271377551
271390505
280010004
280110001
280330002
280350004
280470008
280490010
280590006
280670002
280810005
280870001
281490004
290190004
State Name
Michigan
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Minnesota
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Mississippi
Missouri
County Name
Wayne
Cass
Dakota
Hennepin
Hennepin
Hennepin
Hennepin
Hennepin
Hennepin
Mille Lacs
Ramsey
Ramsey
Ramsey
St Louis
St Louis
St Louis
Scott
Adams
Bolivar
De Soto
Forrest
Harrison
Hinds
Jackson
Jones
Lee
Lowndes
Warren
Boone
2005
24-hr DV
37.03
18.02
25.42
27.25
25.52
26.07
24.71
25.44
26.76
22.03
28.04
28.38
26.36
20.31
19.51
23.53
24.98
27.48
28.98
30.82
30.48
29
28.83
26.96
31.21
32.18
32.44
30.26
30.23
2020
24-hr DV
26.93
13.84
18.28
19.26
17.99
18.83
18.35
17.85
18.04
16.93
20.57
20.59
19.72
15.63
14.20
16.56
17.93
16.79
19.20
15.85
20.78
18.33
17.03
16.48
20.85
16.69
17.28
19.28
19.03
2020
15/35
24-hr DV
26.93
13.84
18.28
19.26
17.99
18.83
18.35
17.85
18.04
16.93
20.57
20.59
19.72
15.63
14.20
16.56
17.93
16.79
19.20
15.85
20.78
18.33
17.03
16.48
20.85
16.69
17.28
19.28
19.03
2020
11/30
24-hr DV
22.75
13.84
18.28
19.26
17.99
18.83
18.35
17.85
18.04
16.93
20.57
20.59
19.72
15.63
14.20
16.56
17.93
16.79
19.20
15.85
20.78
18.33
17.03
16.48
20.85
16.69
17.28
19.28
19.03
(continued)
4.A-86
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
29021
29037
29039
29047
29077
29095
29099
29137
29183
29186
29189
29189
29510
29510
29510
29510
30013
30029
30029
30031
30031
30047
30047
30049
30053
30063
30081
30087
30089
Monitor ID
290210005
290370003
290390001
290470005
290770032
290950034
290990012
291370001
291831002
291860006
291890004
291892003
295100007
295100085
295100086
295100087
300131026
300290009
300290047
300310008
300310013
300470013
300470028
300490018
300530018
300630031
300810007
300870307
300890007
State Name
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
County Name
Buchanan
Cass
Cedar
Clay
Greene
Jackson
Jefferson
Monroe
St Charles
Ste Genevieve
St Louis
St Louis
St Louis City
St Louis City
St Louis City
St Louis City
Cascade
Flathead
Flathead
Gallatin
Gallatin
Lake
Lake
Lewis And Clark
Lincoln
Missoula
Ravalli
Rosebud
Sanders
2005
24-hr DV
30.1
25.61
28.7
28.04
28.27
27.88
33.43
27.83
33.16
31.44
32.03
33.21
33.16
33.24
32.5
34.35
20.15
27.14
27.17
29.55
12.2
27.03
43.66
33.53
42.71
44.64
45.11
19.73
20.42
2020
24-hr DV
21.61
16.89
18.77
20.00
18.75
20.33
21.12
17.98
20.01
18.83
20.46
23.50
20.79
21.03
22.13
22.07
17.08
22.76
24.28
26.24
11.3
23.75
38.36
28.27
35.36
37.47
37.30
18.34
18.25
2020
15/35
24-hr DV
21.61
16.89
18.77
20.00
18.75
20.33
21.12
17.98
20.01
18.83
20.46
23.50
20.79
21.03
22.13
22.07
17.08
22.56
24.11
26.24
11.39
20.89
35.49
28.27
35.12
34.48
35.48
18.34
18.10
2020
11/30
24-hr DV
21.61
16.89
18.77
20.00
18.75
20.33
21.12
17.98
20.01
18.83
20.46
23.50
20.79
21.03
22.13
22.07
17.08
22.05
23.65
26.24
11.39
15.95
30.49
28.27
29.50
30.48
30.48
18.34
17.71
(continued)
4.A-87
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
30093
30111
31025
31055
31055
31079
31109
31157
31177
32003
32003
32003
32003
32003
32031
33001
33005
33007
33009
33011
33011
33011
33013
33015
33019
34003
34007
34007
34013
Monitor ID
300930005
301111065
310250002
310550019
310550052
310790004
311090022
311570003
311770002
320030022
320030298
320030561
320031019
320032002
320310016
330012004
330050007
330070014
330090010
330110020
330111015
330115001
330131006
330150014
330190003
340030003
340070003
340071007
340130015
State Name
Montana
Montana
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nevada
Nevada
Nevada
Nevada
Nevada
Nevada
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Hampshire
New Jersey
New Jersey
New Jersey
New Jersey
County Name
Silver Bow
Yellowstone
Cass
Douglas
Douglas
Hall
Lancaster
Scotts Bluff
Washington
Clark
Clark
Clark
Clark
Clark
Washoe
Belknap
Cheshire
Coos
Grafton
Hillsborough
Hillsborough
Hillsborough
Merrimack
Rockingham
Sullivan
Bergen
Camden
Camden
Essex
2005
24-hr DV
35
19.38
28.3
25.7
25.76
19.16
24.77
16.66
24.01
9.13
12.43
25.26
8.6
20.93
30.78
20.55
30.23
26.5
23
28.66
27.33
25.9
25.65
26.35
28.92
37.03
36.5
37.37
38.38
2020
24-hr DV
28.15
15.76
20.73
19.25
18.99
14.31
17.87
13.90
17.68
8.18
10.10
19.31
7.53
16.35
20.85
11.31
18.74
17.08
14.67
18.87
19.03
12.93
15.12
15.67
16.49
22.46
20.93
20.89
22.59
2020
15/35
24-hr DV
28.15
15.76
20.73
19.25
18.99
14.31
17.87
13.90
17.68
8.18
10.10
19.31
7.53
16.35
20.85
11.31
18.74
17.08
14.67
18.87
19.03
12.93
15.12
15.67
16.49
22.46
20.93
20.89
22.59
2020
11/30
24-hr DV
28.15
15.76
20.73
19.25
18.99
14.31
17.87
13.90
17.68
8.18
10.10
19.31
7.53
16.35
20.85
11.31
18.74
17.08
14.67
18.87
19.03
12.93
15.12
15.67
16.49
22.46
20.93
20.89
22.59
(continued)
4.A-88
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
34017
34017
34021
34023
34027
34027
34029
34031
34039
34039
34039
34041
35001
35001
35005
35013
35013
35017
35043
35043
35045
35049
36001
36005
36005
36005
36013
36029
36029
Monitor ID
340171003
340172002
340210008
340230006
340270004
340273001
340292002
340310005
340390004
340390006
340392003
340410006
350010023
350010024
350050005
350130017
350130025
350171002
350431003
350439011
350450006
350490020
360010005
360050080
360050083
360050110
360130011
360290005
360291007
State Name
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New Mexico
New York
New York
New York
New York
New York
New York
New York
County Name
Hudson
Hudson
Mercer
Middlesex
Morris
Morris
Ocean
Passaic
Union
Union
Union
Warren
Bernalillo
Bernalillo
Chaves
Dona Ana
Dona Ana
Grant
Sandoval
Sandoval
San Juan
Santa Fe
Albany
Bronx
Bronx
Bronx
Chautauqua
Erie
Erie
2005
24-hr DV
39.08
41.43
34.75
34.82
32.32
31.5
31.56
36.3
40.47
37.35
36.82
34.06
18.6
16.43
15.68
32.95
13.8
13
10.3
15.68
12.4
9.78
34.26
38.87
34.74
36.11
29.15
35.35
33.61
2020
24-hr DV
25.73
29.62
18.67
19.67
18.32
16.03
16.14
21.13
24.64
21.11
21.32
20.46
14.61
13.12
12.43
26.90
11.66
12.21
8.01
13.73
10.91
8.57
22.51
26.00
21.32
25.41
15.82
25.54
23.32
2020
15/35
24-hr DV
25.73
29.62
18.67
19.67
18.32
16.03
16.14
21.13
24.64
21.11
21.32
20.46
14.61
13.12
12.43
26.90
11.66
12.21
8.01
13.73
10.91
8.57
22.51
26.00
21.32
25.41
15.82
25.54
23.32
2020
11/30
24-hr DV
25.73
29.62
18.67
19.67
18.32
16.03
16.14
21.13
24.64
21.11
21.32
20.46
14.61
13.12
12.43
26.90
11.66
12.21
8.01
13.73
10.91
8.57
22.51
26.00
21.32
25.41
15.82
25.54
23.32
(continued)
4.A-89
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
36031
36047
36055
36059
36061
36061
36061
36061
36063
36067
36071
36081
36085
36085
36089
36101
36103
36119
37001
37021
37033
37035
37037
37051
37057
37061
37063
37065
37067
Monitor ID
360310003
360470122
360551007
360590008
360610056
360610062
360610079
360610128
360632008
360671015
360710002
360810124
360850055
360850067
360893001
361010003
361030001
361191002
370010002
370210034
370330001
370350004
370370004
370510009
370570002
370610002
370630001
370650004
370670022
State Name
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
New York
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
County Name
Essex
Kings
Monroe
Nassau
New York
New York
New York
New York
Niagara
Onondaga
Orange
Queens
Richmond
Richmond
St Lawrence
Steuben
Suffolk
Westchester
Alamance
Buncombe
Caswell
Catawba
Chatham
Cumberland
Davidson
Duplin
Durham
Edgecombe
Forsyth
2005
24-hr DV
22.45
36.94
32.2
34.01
39.7
38.82
37.94
39.45
33.87
27.35
28.92
35.56
34.93
32.41
22.05
27.81
34.66
33.51
31.72
30.05
29.45
34.53
26.94
30.78
31.35
28.3
31.02
26.78
31.92
2020
24-hr DV
13.77
23.04
19.38
19.15
26.51
23.85
25.73
25.95
21.99
16.74
18.47
22.44
21.17
17.57
15.22
14.94
18.09
19.41
18.09
15.83
16.07
19.24
13.82
17.31
18.28
15.35
16.47
16.60
18.32
2020
15/35
24-hr DV
13.77
23.04
19.38
19.15
26.51
23.85
25.73
25.95
21.99
16.74
18.47
22.44
21.17
17.57
15.22
14.94
18.09
19.41
18.09
15.83
16.07
19.24
13.82
17.31
18.28
15.35
16.47
16.60
18.32
2020
11/30
24-hr DV
13.77
23.04
19.38
19.15
23.61
20.95
22.83
23.05
21.99
16.74
18.47
22.44
21.17
17.57
15.22
14.94
18.09
19.41
18.09
15.83
16.07
19.24
13.82
17.31
18.28
15.35
16.47
16.60
18.32
(continued)
4.A-90
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
37071
37081
37087
37099
37107
37111
37117
37119
37119
37119
37121
37123
37129
37133
37135
37147
37155
37159
37173
37183
37189
37191
38007
38013
38015
38017
38053
38057
39009
Monitor ID
370710016
370810013
370870010
370990006
371070004
371110004
371170001
371190010
371190041
371190042
371210001
371230001
371290002
371330005
371350007
371470005
371550005
371590021
371730002
371830014
371890003
371910005
380070002
380130003
380150003
380171004
380530002
380570004
390090003
State Name
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Dakota
North Dakota
North Dakota
North Dakota
North Dakota
North Dakota
Ohio
County Name
Gaston
Guilford
Haywood
Jackson
Lenoir
McDowell
Martin
Mecklenburg
Mecklenburg
Mecklenburg
Mitchell
Montgomery
New Hanover
Onslow
Orange
Pitt
Robeson
Rowan
Swain
Wake
Watauga
Wayne
Billings
Burke
Burleigh
Cass
McKenzie
Mercer
Athens
2005
24-hr DV
30.86
30.63
27.74
24.96
25.2
31.55
24.83
32.33
31.72
30.7
30.25
28.21
25.4
24.61
29.35
26.21
29.92
30.23
27.34
31.63
30.43
29.72
13.07
16.73
17.62
21.22
11.96
16.98
32.32
2020
24-hr DV
16.10
17.80
16.38
13.91
15.55
17.30
14.78
18.39
16.71
16.34
15.29
15.02
13.75
14.53
15.60
16.20
16.31
17.71
15.03
16.96
15.96
17.01
11.57
15.05
14.39
16.05
10.4
14.36
15.83
2020
15/35
24-hr DV
16.10
17.80
16.38
13.91
15.55
17.30
14.78
18.39
16.71
16.34
15.29
15.02
13.75
14.53
15.60
16.20
16.31
17.71
15.03
16.96
15.96
17.01
11.57
15.05
14.39
16.05
10.41
14.36
15.83
2020
11/30
24-hr DV
16.10
17.80
16.38
13.91
15.55
17.30
14.78
18.39
16.71
16.34
15.29
15.02
13.75
14.53
15.60
16.20
16.31
17.71
15.03
16.96
15.96
17.01
11.57
15.05
14.39
16.05
10.41
14.36
15.83
(continued)
4.A-91
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
39017
39017
39017
39017
39023
39025
39035
39035
39035
39035
39035
39035
39035
39049
39049
39049
39057
39061
39061
39061
39061
39061
39061
39061
39081
39081
39085
39087
39093
Monitor ID
390170003
390170016
390170017
390171004
390230005
390250022
390350027
390350034
390350038
390350045
390350060
390350065
390351002
390490024
390490025
390490081
390570005
390610006
390610014
390610040
390610042
390610043
390617001
390618001
390810017
390811001
390851001
390870010
390933002
State Name
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
County Name
Butler
Butler
Butler
Butler
Clark
Clermont
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Cuyahoga
Franklin
Franklin
Franklin
Greene
Hamilton
Hamilton
Hamilton
Hamilton
Hamilton
Hamilton
Hamilton
Jefferson
Jefferson
Lake
Lawrence
Lorain
2005
24-hr DV
39.23
37.14
37.93
37.13
35.37
34.46
36.6
36.58
44.2
38.57
42.12
38.67
34.25
38.51
38.46
34.16
32.21
37.66
38.24
36.73
37.3
35.95
38.81
40.6
40.7
41.96
37.16
33.77
31.56
2020
24-hr DV
23.53
19.79
20.19
19.29
19.40
17.07
24.82
21.66
29.73
23.28
26.89
22.85
20.89
21.09
20.11
18.98
16.98
17.85
19.50
18.87
20.82
18.63
20.11
22.10
24.05
22.59
21.08
18.28
19.08
2020
15/35
24-hr DV
23.53
19.79
20.19
19.29
19.40
17.07
24.82
21.66
29.73
23.28
26.89
22.85
20.89
21.09
20.11
18.98
16.98
17.85
19.50
18.87
20.82
18.63
20.11
22.10
24.05
22.59
21.08
18.28
19.08
2020
11/30
24-hr DV
23.53
19.79
20.19
19.29
19.40
17.07
22.59
19.46
27.50
21.05
24.66
20.62
18.67
21.09
20.11
18.98
16.98
17.85
19.50
18.87
20.82
18.63
20.11
22.10
24.05
22.59
20.88
18.28
18.91
(continued)
4.A-92
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
39095
39095
39095
39099
39099
39113
39113
39133
39135
39145
39151
39153
39153
39155
40015
40021
40071
40071
40081
40097
40097
40101
40109
40109
40115
40121
40135
40143
40143
Monitor ID
390950024
390950025
390950026
390990005
390990014
391130031
391130032
391330002
391351001
391450013
391510017
391530017
391530023
391550007
400159008
400219002
400710602
400719010
400819005
400970186
400979014
401010169
401090035
401091037
401159004
401210415
401359015
401430110
401431127
State Name
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
Oklahoma
County Name
Lucas
Lucas
Lucas
Mahoning
Mahoning
Montgomery
Montgomery
Portage
Preble
Scioto
Stark
Summit
Summit
Trumbull
Caddo
Cherokee
Kay
Kay
Lincoln
Mayes
Mayes
Muskogee
Oklahoma
Oklahoma
Ottawa
Pittsburg
Sequoyah
Tulsa
Tulsa
2005
24-hr DV
36.34
35.14
34.9
35.16
36.83
35.78
37.8
34.32
32.85
34.55
36.9
38.06
35.88
36.23
23.97
27.55
31.8
27.93
27.83
28.71
26.13
29.54
23.42
27.12
29.14
26.37
31.43
28.43
30.37
2020
24-hr DV
23.56
25.95
23.59
19.98
21.48
22.68
19.27
18.83
17.51
18.20
20.19
21.46
20.29
21.39
16.68
20.06
25.60
20.55
19.33
21.86
18.49
20.95
16.46
19.18
20.58
18.75
22.98
20.69
21.77
2020
15/35
24-hr DV
23.56
25.95
23.59
19.98
21.48
22.68
19.27
18.83
17.51
18.20
20.19
21.46
20.29
21.39
16.68
20.06
25.60
20.55
19.33
21.86
18.49
20.95
16.46
19.18
20.58
18.75
22.98
20.69
21.77
2020
11/30
24-hr DV
23.56
25.95
23.59
19.98
21.48
22.68
19.27
18.63
17.51
18.20
20.19
21.26
20.09
21.39
16.68
20.06
25.60
20.55
19.33
21.86
18.49
20.95
16.46
19.18
20.58
18.75
22.98
20.69
21.77
(continued)
4.A-93
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
41029
41029
41035
41039
41039
41039
41039
41051
41051
41061
42001
42003
42003
42003
42003
42003
42003
42003
42003
42003
42003
42003
42003
42007
42011
42017
42021
42027
42029
Monitor ID
410290133
410291001
410350004
410390060
410391007
410391009
410392013
410510080
410510246
410610119
420010001
420030008
420030021
420030064
420030067
420030093
420030095
420030116
420030133
420031008
420031301
420033007
420039002
420070014
420110011
420170012
420210011
420270100
420290100
State Name
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
County Name
Jackson
Jackson
Klamath
Lane
Lane
Lane
Lane
Multnomah
Multnomah
Union
Adams
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Allegheny
Beaver
Berks
Bucks
Cambria
Centre
Chester
2005
24-hr DV
33.72
14.51
44.08
32.55
15.63
23.96
48.95
29.88
23.22
27.38
34.93
39.44
35.16
64.27
36.48
45.6
38.77
42.56
39.23
41.34
40.3
37.52
37.86
43.42
37.71
34.01
39.04
36.28
36.7
2020
24-hr DV
23.50
10.43
30.85
21.42
10.16
16.59
34.01
19.10
15.24
22.64
20.05
22.23
19.37
41.03
17.26
24.65
21.02
22.61
24.73
21.40
20.93
21.63
20.14
23.46
27.18
20.66
19.60
21.01
22.40
2020
15/35
24-hr DV
23.50
10.43
30.85
21.42
10.16
16.59
34.01
19.10
15.24
22.64
20.05
16.72
13.86
35.49
11.76
24.65
15.51
17.10
24.73
15.88
15.40
16.12
14.62
23.32
27.18
20.66
19.60
21.01
22.40
2020
11/30
24-hr DV
23.50
10.43
30.14
16.81
5.56
11.98
29.40
19.10
15.24
22.64
20.05
11.72
8.86
30.48
6.76
24.65
10.51
12.10
24.73
10.88
10.40
11.12
9.62
23.32
27.18
20.66
19.60
21.01
22.40
(continued)
4.A-94
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
42041
42043
42045
42049
42069
42071
42077
42079
42085
42095
42099
42101
42101
42101
42125
42125
42125
42129
42133
44007
44007
44007
44007
45019
45025
45037
45041
45045
45045
Monitor ID
420410101
420430401
420450002
420490003
420692006
420710007
420770004
420791101
420850100
420950025
420990301
421010004
421010024
421010047
421250005
421250200
421255001
421290008
421330008
440070022
440070026
440070028
440071010
450190049
450250001
450370001
450410002
450450008
450450009
State Name
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Rhode Island
Rhode Island
Rhode Island
Rhode Island
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
County Name
Cumberland
Dauphin
Delaware
Erie
Lackawanna
Lancaster
Lehigh
Luzerne
Mercer
Northampton
Perry
Philadelphia
Philadelphia
Philadelphia
Washington
Washington
Washington
Westmoreland
York
Providence
Providence
Providence
Providence
Charleston
Chesterfield
Edgefield
Florence
Greenville
Greenville
2005
24-hr DV
38
38.04
35.24
34.46
31.55
40.83
36.4
32.46
36.3
36.72
30.46
36.53
35.96
37.3
35.52
33.5
38.14
37.12
38.24
29.46
30.62
28.1
28.8
27.93
28.77
32.23
28.81
31.86
32.55
2020
24-hr DV
25.33
26.65
21.08
20.16
17.61
30.51
24.04
20.14
20.84
22.79
20.22
21.31
19.61
21.84
19.95
18.59
17.33
18.80
28.16
17.18
18.87
17.47
17.32
15.52
16.09
17.45
16.50
18.71
18.18
2020
15/35
24-hr DV
25.33
26.65
21.08
20.16
17.61
30.51
24.04
20.14
20.84
22.79
20.22
21.31
19.61
21.84
19.81
18.44
17.20
18.66
28.16
17.18
18.87
17.47
17.32
15.52
16.09
17.45
16.50
18.71
18.18
2020
11/30
24-hr DV
25.33
26.65
21.08
20.16
17.61
30.47
24.04
20.14
20.84
22.79
20.22
21.31
19.61
21.84
19.81
18.44
17.20
18.66
28.16
17.18
18.87
17.47
17.32
15.52
16.09
17.45
16.50
18.71
18.18
(continued)
4.A-95
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
45047
45051
45063
45073
45079
45079
45083
46011
46013
46029
46033
46071
46099
46099
46103
46103
46103
47009
47037
47037
47037
47045
47065
47065
47065
47093
47093
47099
47105
Monitor ID
450470003
450510002
450630008
450730001
450790007
450790019
450830010
460110002
460130003
460290002
460330132
460710001
460990006
460990007
461030016
461030020
461031001
470090011
470370023
470370025
470370036
470450004
470650031
470651011
470654002
470930028
470931017
470990002
471050108
State Name
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Carolina
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
South Dakota
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
County Name
Greenwood
Horry
Lexington
Oconee
Richland
Richland
Spartanburg
Brookings
Brown
Codington
Custer
Jackson
Minnehaha
Minnehaha
Pennington
Pennington
Pennington
Blount
Davidson
Davidson
Davidson
Dyer
Hamilton
Hamilton
Hamilton
Knox
Knox
Lawrence
Loudon
2005
24-hr DV
30.01
28.3
32.86
27.98
31.38
33.2
32.46
23.54
18.73
23.67
14.36
12.73
24.17
23.98
17.2
18.58
15.95
32.54
33.5
30.93
32.71
31.92
33.25
29.74
33.53
36.66
33.46
28.48
32.2
2020
24-hr DV
16.33
16.66
19.02
14.56
17.08
19.00
17.99
17.21
14.31
17.81
12.03
10.27
17.48
16.91
14.48
16.21
13.29
18.75
18.05
16.82
16.15
17.63
20.49
14.88
18.24
20.46
19.45
14.91
19.93
2020
15/35
24-hr DV
16.33
16.66
19.02
14.56
17.08
19.00
17.99
17.21
14.31
17.81
12.03
10.27
17.48
16.91
14.48
16.21
13.29
18.75
18.05
16.82
16.15
17.63
20.49
14.88
18.24
20.46
19.45
14.91
19.93
2020
11/30
24-hr DV
16.33
16.66
19.02
14.56
17.08
19.00
17.99
17.21
14.31
17.81
12.03
10.27
17.48
16.91
14.48
16.21
13.29
18.75
18.05
16.82
16.15
17.63
20.49
14.88
18.24
20.46
19.45
14.91
19.93
(continued)
4.A-96
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
47107
47119
47125
47141
47145
47157
47157
47157
47157
47163
47165
48037
48113
48113
48113
48135
48141
48201
48203
48215
48355
48355
48361
48439
48439
49003
49005
49011
49035
Monitor ID
471071002
471192007
471251009
471410001
471450004
471570014
471570038
471570047
471571004
471631007
471650007
480370004
481130050
481130069
481130087
481350003
481410037
482011035
482030002
482150043
483550032
483550034
483611001
484391002
484391006
490030003
490050004
490110004
490350003
State Name
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Utah
Utah
Utah
Utah
County Name
Me Minn
Maury
Montgomery
Putnam
Roane
Shelby
Shelby
Shelby
Shelby
Sullivan
Sumner
Bowie
Dallas
Dallas
Dallas
Ector
El Paso
Harris
Harrison
Hidalgo
Nueces
Nueces
Orange
Tarrant
Tarrant
Box Elder
Cache
Davis
Salt Lake
2005
24-hr DV
32.73
30.96
36.3
32.66
30.24
32.25
32.52
33.5
29.88
31.13
33.66
29.42
27.44
25.7
24.21
17.81
22.93
30.81
25.95
26.42
27.55
20.74
27.78
25.34
25.76
33.2
56.95
38.95
47.36
2020
24-hr DV
17.70
16.91
17.88
16.31
15.93
16.94
16.25
16.90
15.72
18.99
15.20
19.29
17.73
16.73
15.08
13.75
19.47
21.23
17.31
22.24
18.66
12.40
18.57
16.26
16.82
27.74
42.65
31.35
33.48
2020
15/35
24-hr DV
17.70
16.91
17.88
16.31
15.93
16.94
16.25
16.90
15.72
18.99
15.20
19.29
17.73
16.73
15.08
13.75
19.47
21.23
17.31
22.24
18.66
12.40
18.57
16.26
16.82
27.46
35.48
31.35
32.06
2020
11/30
24-hr DV
17.70
16.91
17.88
16.31
15.93
16.94
16.25
16.90
15.72
18.99
15.20
19.29
17.73
16.73
15.08
13.75
19.47
19.46
17.31
22.24
18.66
12.40
18.57
16.26
16.82
27.44
30.48
29.56
27.20
(continued)
4.A-97
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
49035
49035
49035
49035
49035
49045
49049
49049
49049
49049
49057
49057
49057
50001
50001
50003
50007
50007
50021
51013
51036
51041
51059
51059
51059
51087
51087
51107
51139
Monitor ID
490350012
490351001
490353006
490353007
490353008
490450003
490490002
490494001
490495008
490495010
490570002
490570007
490571003
500010002
500010003
500030004
500070012
500070014
500210002
510130020
510360002
510410003
510590030
510591005
510595001
510870014
510870015
511071005
511390004
State Name
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Utah
Vermont
Vermont
Vermont
Vermont
Vermont
Vermont
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
County Name
Salt Lake
Salt Lake
Salt Lake
Salt Lake
Salt Lake
Tooele
Utah
Utah
Utah
Utah
Weber
Weber
Weber
Addison
Addison
Bennington
Chittenden
Chittenden
Rutland
Arlington
Charles City
Chesterfield
Fairfax
Fairfax
Fairfax
Henrico
Henrico
Loudoun
Page
2005
24-hr DV
50.14
37.73
47.84
45.38
30.07
30.53
38.18
44
35.9
35.93
38.58
33.6
36.16
28.2
31.73
26.47
29.84
30.13
30.6
34.18
31.76
31.25
34.47
33.72
33.31
31.95
29.18
34.45
30.06
2020
24-hr DV
36.73
30.29
34.97
36.10
25.11
26.09
29.25
33.95
27.60
28.13
30.01
26.35
28.34
17.10
18.53
15.89
18.90
21.58
22.73
18.75
16.76
15.30
18.50
18.14
19.36
16.57
14.29
19.24
16.65
2020
15/35
24-hr DV
35.31
28.87
33.55
34.68
23.69
26.09
29.25
33.95
27.60
28.13
29.68
26.03
28.02
17.10
18.53
15.89
18.90
21.58
22.73
18.75
16.76
15.30
18.50
18.14
19.36
16.57
14.29
19.24
16.65
2020
11/30
24-hr DV
30.48
24.04
28.70
29.80
19.01
25.32
24.92
29.61
23.36
23.86
29.66
26.01
27.99
17.10
18.53
15.89
18.90
21.58
22.73
18.75
16.76
15.30
18.50
18.14
19.36
16.57
14.29
19.24
16.65
(continued)
4.A-98
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
51520
51650
51680
51710
51770
51775
53033
53033
53033
53053
53061
53063
53063
54003
54009
54009
54011
54029
54033
54039
54039
54039
54049
54051
54061
54069
54081
54089
54107
Monitor ID
515200006
516500004
516800015
517100024
517700014
517750010
530330024
530330057
530330080
530530029
530611007
530630016
530630047
540030003
540090005
540090011
540110006
540291004
540330003
540390010
540390011
540391005
540490006
540511002
540610003
540690010
540810002
540890001
541071002
State Name
Virginia
Virginia
Virginia
Virginia
Virginia
Virginia
Washington
Washington
Washington
Washington
Washington
Washington
Washington
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
West Virginia
County Name
Bristol City
Hampton City
Lynchburg City
Norfolk City
Roanoke City
Salem City
King
King
King
Pierce
Snohomish
Spokane
Spokane
Berkeley
Brooke
Brooke
Cabell
Hancock
Harrison
Kanawha
Kanawha
Kanawha
Marion
Marshall
Monongalia
Ohio
Raleigh
Summers
Wood
2005
24-hr DV
30.24
29.01
30.71
29.66
32.7
34.06
28.78
29.16
22.03
41.82
34.36
29.7
29.86
34.51
39.43
43.9
35.1
40.64
33.53
34.73
33.1
36.98
33.68
33.98
35.65
32
30.67
31.26
35.44
2020
24-hr DV
16.20
15.90
15.77
16.97
18.00
19.37
20.60
20.74
16.10
31.00
26.99
19.15
18.60
23.43
22.36
25.30
18.06
20.54
15.75
16.61
16.24
18.25
15.56
16.98
14.68
16.50
14.42
14.30
17.75
2020
15/35
24-hr DV
16.20
15.90
15.77
16.97
18.00
19.37
20.60
20.74
16.10
31.00
26.99
19.15
18.60
23.43
22.36
25.30
18.06
20.54
15.75
16.61
16.24
18.25
15.56
16.98
14.68
16.50
14.42
14.30
17.75
2020
11/30
24-hr DV
16.20
15.90
15.77
16.97
18.00
19.37
20.60
20.74
16.10
30.41
26.99
19.15
18.60
23.43
22.36
25.30
18.06
20.54
15.75
16.61
16.24
18.25
15.56
16.98
14.68
16.50
14.42
14.30
17.75
(continued)
4.A-99
-------
Table 4.A-12. 24-hr Design Values (DVs) for the 2005 and 2020 Base Case and After Meeting
the Current and Proposed Alternative Standard Levels: 2020 Baseline (15/35)
and 2020 11/30 (continued)
FIPS
55003
55009
55009
55025
55027
55041
55043
55059
55071
55079
55079
55079
55079
55079
55087
55089
55109
55111
55119
55125
55133
56005
56005
56005
56009
56013
56021
56033
Monitor ID
550030010
550090005
550090009
550250047
550270007
550410007
550430009
550590019
550710007
550790010
550790026
550790043
550790059
550790099
550870009
550890009
551091002
551110007
551198001
551250001
551330027
560050877
560050892
560050899
560090819
560131003
560210001
560330002
State Name
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wyoming
Wyoming
Wyoming
Wyoming
Wyoming
Wyoming
Wyoming
County Name
Ashland
Brown
Brown
Dane
Dodge
Forest
Grant
Kenosha
Manitowoc
Milwaukee
Milwaukee
Milwaukee
Milwaukee
Milwaukee
Outagamie
Ozaukee
St Croix
Sauk
Taylor
Vilas
Waukesha
Campbell
Campbell
Campbell
Converse
Fremont
La ramie
Sheridan
2005
24-hr DV
18.61
36.56
35.86
35.57
31.82
25.26
34.35
32.78
29.7
38.67
37.38
39.92
35.56
37.78
32.87
32.53
26.66
28.63
25.38
22.61
35.48
18.63
12.55
12.66
10
29.8
11.93
30.86
2020
24-hr DV
12.48
24.89
25.53
24.20
21.63
17.13
24.95
22.88
21.11
26.12
24.90
26.08
24.18
25.75
23.37
22.69
19.57
21.31
18.12
16.14
24.68
17.12
12.19
12.19
9.33
23.81
10.0
27.17
2020
15/35
24-hr DV
12.48
24.89
25.53
24.20
21.63
17.13
24.95
22.88
21.11
26.12
24.90
26.08
24.18
25.75
23.37
22.69
19.57
21.31
18.12
16.14
24.68
17.12
12.19
12.19
9.33
23.81
10.07
27.17
2020
11/30
24-hr DV
12.48
24.89
25.53
24.20
21.63
17.13
24.95
22.88
21.11
26.12
24.90
26.08
24.18
25.75
23.37
22.69
19.57
21.31
18.12
16.14
24.68
17.12
12.19
12.19
9.33
23.81
10.07
27.17
4.A-100
-------
4.A.3 References
U.S. Environmental Protection Agency (EPA). Fugitive Dust Background Document and Technical
Information Document and Technical Information Document for Best Available Control
Measures (EPA-450/2-92-004). September 1992.
U.S. Environmental Protection Agency (EPA). "Chapter 13:2.1: Paved Roads," AP-42, Fifth
Edition, Volume I, 2003. Available at
http://www.epa.gov/ttn/chief/ap42/chl3/index.html.
U.S. Environmental Protection Agency (EPA). "Chapter 13:2.2: Unpaved Roads," AP-42, Fifth
Edition, Volume I, 2003. Available at
http://www.epa.gov/ttn/chief/ap42/chl3/index.html.
U.S. Environmental Protection Agency, Region 10 (EPA). Agricultural Burning, accessed
November 3, 2005 from
http://yosemite. epa.gov/R10/AIRPAGE. NSF/webpage/Agricultural+Burning.
U.S. Department of Agriculture (USDA), Agricultural Research Service. Air Quality Programs
Action Plan, October 6, 2000.
http://www.ars.usda.gov/research/programs/programs.htm7np code=203&docid=317
&page=3&pf=l&cg id=8.
Western Regional Air Partnership (WRAP). WRAP Fugitive Dust Handbook, prepared by
Countess Environmental, November 15, 2004. Available at
http://www.ndep.nv.gov/baqp/WRAP/INDEX.html.
4.A-101
-------
CHAPTER 5
HUMAN HEALTH BENEFITS ANALYSIS APPROACH AND RESULTS
5.1 Synopsis
This chapter presents the estimated human health benefits for the proposed National
Ambient Air Quality Standards (NAAQS) for particulate matter (PM). In this chapter, we quantify
the health-related benefits of the fine particulate matter (PM2.5)-related air quality
improvements resulting from the illustrative emission control scenarios that reduce emissions
of directly emitted particles and precursor pollutants including S02 and NOX to attain alternative
PM2.5 NAAQS levels in 2020.
These benefits are relative to a 2020 baseline reflecting attainment of the current
primary PM2.5 standards (i.e., annual standard at 15 u.g/m3 and 24-hour standard of 35 u.g/m3,
referred to as "15/35") that includes promulgated national regulations and illustrative emission
controls to simulate attainment with 15/35. We project PM2.5 levels in certain areas would
exceed 13/35, 12/35, 11/35, and 11/30 after illustrative controls to simulate attainment with
15/35. In analyzing the current 15/35 standard (baseline), EPA determined that all counties
would meet the 14/35 standard concurrently with meeting the existing 15/35 standard at no
additional cost. Consequently, there are no incremental costs or benefits for 14/35, and no
need to present an analysis of 14/35. Table 5-1 summarizes the total monetized benefits of
these alternative PM2.5 standards in 2020. These estimates reflect the sum of the estimated
PM2.5 mortality impacts identified and the value of all morbidity impacts.
Table 5-1. Estimated Monetized Benefits of the Proposed and Alternative Combinations of
PM2.5 Standards in 2020, Incremental to Attainment of 15/35 (millions of 2006$)a
13 ug/m3 Annual 12 ug/m3 Annual 11 ug/m3 Annual 11 ug/m3 Annual
Benefits Estimate & 35 ug/m3 24-Hour & 35 ug/m3 24-Hour & 35 ug/m3 24-Hour & 30 ug/m3 24-Hour
Economic value of avoided PM25-related morbidities and premature deaths using PM2.5 mortality estimate from
Krewski et al. (2009)
3% discount rate $88+B $2,300+B $9,2000+6 $14,000+6
7% discount rate $79 + 6 $2,100+6 $8,3000+6 $13,000+6
Economic value of avoided PM2.5-related morbidities and premature deaths using PM2.5 mortality estimate from
Laden et al. (2006)
3% discount rate $220+6 $5,900+6 $23,000+6 $36,000+6
7% discount rate $200+6 $5,400+6 $21,000+6 $33,000+6
a Rounded to two significant figures. Avoided premature deaths account for over 98% of monetized benefits here,
which are discounted over the SA6-recommended 20-year segmented lag. It was not all possible to quantify all
benefits due to data limitations in this analysis. "6" is the sum of all unquantified health and welfare benefits.
5-1
-------
For annual standards at 12 u.g/m3 and 13 u.g/m3, the majority of benefits (i.e., 70% and
98%, respectively) occur in California because this highly populated area is where the most air
quality improvement beyond 15/35 is needed to reach these levels. In addition, several recent
rules such as the Mercury and Air Toxics Standard (MATS) and the Cross-State Air Pollution Rule
(CSAPR) will have substantially reduced PM2.5 levels by 2020 in the East, thus few additional
controls would be needed to reach 12/35 or 13/35 in the East.
In general, we have greater confidence in risk estimates based on PM2.5 concentrations
where the bulk of the data reside and somewhat less confidence where data density is lower.
As noted in the preamble to the proposed rule, the range from the 25th to 10th percentiles of
the air quality data used the epidemiology studies is a reasonable range below which we have
appreciably less confidence in the associations observed in the epidemiological studies. Most of
the estimated avoided premature deaths occur at or above the lowest measured PM2.5
concentration in the two studies used to estimate mortality benefits.
In addition to PM2.5 benefits, implementation of emissions controls to attain the
alternative PM2.5 standards would reduce other ambient pollutants, such as S02, N02, and
ozone. However, because the method used in this analysis to simulate attainment does not
simulate changes in ambient concentrations of other pollutants, we were not able to quantify
the co-benefits of reduced exposure to other pollutants. In addition, due to data and
methodology limitations, we were unable to estimate additional health benefits associated
with exposure to PM2.5 or the additional benefits from improvements in welfare effects, such as
ecosystem effects and visibility. We describe the unquantified health benefits in this chapter
and the unquantified welfare benefits in Chapter 6.
5.2 Overview
This chapter contains a subset of the estimated health benefits of the proposed and
alternative PM2.5 standards in 2020 that EPA was able to quantify, given the available resources
and methods. The analysis in this chapter aims to characterize the benefits of the air quality
changes resulting from the implementation of new PM standards by answering two key
questions:
1. What are the health effects of changes in ambient particulate matter (PM2.5)
resulting from reductions in directly emitted PM2.5 and precursors due to the
attainment of a new PM2.5 standard?
2. What is the economic value of these effects?
5-2
-------
In this analysis, we consider an array of health impacts attributable to changes in PM2.5.
The Integrated Science Assessment for Particulate Matter (U.S. EPA, 2009b) identifies the
human health effects associated with ambient particles, which include premature mortality and
a variety of morbidity effects associated with acute and chronic exposures. Table 5-2
summarizes human health categories contained within the main benefits estimate as well as
those categories that were unquantified due to limited data or resources. It is important to
emphasize that the list of unquantified benefit categories is not exhaustive, nor is quantification
of each effect complete. In order to identify the most meaningful human health and
environmental co-benefits, we excluded effects not identified as having at least a causal, likely
causal, or suggestive relationship with the affected pollutants in the most recent
comprehensive scientific assessment, such as an Integrated Science Assessment (ISA). This does
not imply that additional relationships between these and other human health and
environmental co-benefits and the affected pollutants do not exist. Due to this decision
criterion, some effects that were identified in previous lists of unquantified benefits in other
RIAs have been dropped (e.g., UVb exposure).
The benefits analysis in this chapter relies on an array of data inputs—including air
quality modeling, health impact functions and valuation estimates among others—which are
themselves subject to uncertainty and may also in turn contribute to the overall uncertainty in
this analysis. We employ several techniques to characterize this uncertainty, which are
described in detail in section 5.4.
As described in Chapter 1, there are important differences worth noting in the design
and analytical objectives of NAAQS RIAs compared to RIAs for implementation rules, such as
the recent MATS rule (U.S. EPA, 2011d). The NAAQS RIAs illustrate the potential costs and
benefits of attaining a revised air quality standard nationwide based on an array of emission
control strategies for different sources, incremental to implementation of existing regulations
and controls needed to attain current standards. In short, NAAQS RIAs hypothesize, but do not
predict, the control strategies that States may choose to enact when implementing a revised
NAAQS. The setting of a NAAQS does not directly result in costs or benefits, and as such, NAAQS
RIAs are merely illustrative and are not intended to be added to the costs and benefits of other
regulations that result in specific costs of control and emission reductions. By contrast, the
emission reductions from implementation rules are generally for specific, well-characterized
sources, such as the recent MATS rule (U.S. EPA, 2011d). In general, EPA is more confident in
the magnitude and location of the emission reductions for implementation rules. As such,
emission reductions achieved under promulgated implementation rules such as MATS have
5-3
-------
been reflected in the baseline of this NAAQS analysis. Subsequent implementation rules will be
reflected in the baseline for the next PM NAAQS review. For this reason, the benefits estimated
provided in this RIA and all other NAAQS RIAs should not be added to the benefits estimated for
implementation rules.
Table 5-2. Human Health Effects of Pollutants Potentially Affected by Attainment of the
Primary PM2.s Standards
Benefits Category
Specific Effect
Effect Has Effect Has
Been Been
Quantified Monetized
More
Information
Improved Human
Reduced
incidence of
premature
mortality from
exposure to PM2.5
Health
Adult premature mortality based on cohort
study estimates and expert elicitation
estimates (age >25 or age >30)
Infant mortality (age <1)
Section 5.6
Section 5.6
Reduced
incidence of
morbidity from
exposure to PM2.5
Non-fatal heart attacks (age > 18)
Hospital admissions—respiratory (all ages)
Hospital admissions—cardiovascular (age
>20)
Emergency department visits for asthma (all
ages)
Acute bronchitis (age 8-12)
Lower respiratory symptoms (age 7-14)
Upper respiratory symptoms (asthmatics age
9-11)
Asthma exacerbation (asthmatics age 6-18)
Lost work days (age 18-65)
Minor restricted-activity days (age 18-65)
Chronic Bronchitis (age >26)
Emergency department visits for
cardiovascular effects (all ages)
Strokes and cerebrovascular disease (age 50-
79)
Other cardiovascular effects (e.g., other ages)
Other respiratory effects (e.g., pulmonary
function, non-asthma ER visits, non-bronchitis
chronic diseases, other ages and populations)
Reproductive and developmental effects
(e.g., low birth weight, pre-term births, etc)
Cancer, mutagenicity, and genotoxicity
effects
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
Section 5.6
PM ISA2
PM ISA2
PM ISA
2'3
PM ISA
2'3
(continued)
5-4
-------
Table 5-2. Human Health Effects of Pollutants Potentially Affected by Attainment of the
Primary PM2.s Standards (continued)
Benefits Category
Specific Effect
Effect Has Effect Has
Been Been
Quantified Monetized
More
Information
Reduced incidence Premature mortality based on short-term
of mortality from study estimates (all ages)
exposure to ozone
Premature mortality based on long-term
study estimates (age 30-99)
Hospital admissions—respiratory causes (age
>65)
Hospital admissions—respiratory causes (age
Emergency department visits for asthma (all
ages)
Minor restricted-activity days (age 18-65)
School absence days (age 5-17)
Decreased outdoor worker productivity (age
18-65)
Other respiratory effects (e.g., premature
aging of lungs)
Cardiovascular and nervous system effects
Reproductive and developmental effects
Ozone CD, Draft
Ozone ISA1
Ozone CD, Draft
Ozone ISA1
Ozone CD, Draft
ISA1
Ozone CD, Draft
ISA1
Ozone CD, Draft
ISA1
Ozone CD, Draft
ISA1
Ozone CD, Draft
ISA1
Ozone CD, Draft
ISA1
Ozone CD, Draft
ISA2
Ozone CD, Draft
ISA3
Ozone CD, Draft
ISA3
Reduced incidence Asthma hospital admissions (all ages)
of morbidity from
exposure to NO Chronic lung disease hospital admissions
2 (age > 65)
Respiratory emergency department visits (all
ages)
Asthma exacerbation (asthmatics age 4-18)
Acute respiratory symptoms (age 7-14)
Premature mortality
Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
NO2 ISA1
NO2 ISA1
N02 ISA1
NO2 ISA1
NO2 ISA
NO, ISA
2,3
NO, ISA
2,3
(continued)
5-5
-------
Table 5-2. Human Health Effects of Pollutants Potentially Affected by Attainment of the
Primary PM2.s Standards (continued)
Benefits Category
Specific Effect
Effect Has Effect Has
Been Been
Quantified Monetized
More
Information
Reduced incidence Respiratory hospital admissions (age > 65)
of morbidity from
exposure to SO2
Asthma emergency department visits (all
ages)
Asthma exacerbation (asthmatics age 4-12)
Acute respiratory symptoms (age 7-14)
Premature mortality
Other respiratory effects (e.g., airway
hyperresponsivenessand inflammation, lung
function, other ages and populations)
SO2 ISA1
SO2 ISA1
SO2 ISA1
SO2 ISA1
SO2 ISA
SO2 ISA2
2,3
Reduced incidence Neurologic effects—IQ loss
of morbidity from
exposure to
methylmercury
Other neurologic effects (e.g.,
developmental delays, memory, behavior)
IRIS; NRC, 2000
IRIS; NRC, 20002
(through role of
sulfate in
methylation)
Cardiovascular effects —
Genotoxic, immunologic, and other toxic —
effects
- IRIS; NRC,
20002'3
- IRIS; NRC,
20002'3
1 We assess these benefits qualitative due to time and resource limitations for this analysis.
We assess these benefits qualitatively because we do not have sufficient confidence in available data or
methods.
We assess these benefits qualitatively because current evidence is only suggestive of causality or there are other
significant concerns over the strength of the association.
We quantify these benefits in a sensitivity analysis, but not the main analysis.
5.3 Updated Methodology Presented in this RIA
The benefits analysis presented in this chapter incorporates an array of policy and
technical changes that the Agency has adopted since the previous review of the PM2.5 standards
in 2006, and since publication of the most recent major benefits analysis, documented in the
benefits chapter of the RIA accompanying the final MATS(U.S. EPA, 2011d). Below we note the
aspects of this analysis that differ from the MATS RIA:
1. Incorporation of the newest American Cancer Society (ACS) mortality study. I n 2009,
the Health Effects Institute published an extended analysis of the ACS cohort
(Krewski et al., 2009). Compared to the study it replaces (Pope et al., 2002), this new
analysis incorporates a number of methodological improvements that we describe in
detail below. The all-cause PM2.s mortality risk estimate drawn from Krewski et al.
5-6
-------
(2009) is identical to the Pope et al. (2002) risk estimate applied in recent EPA
analyses of long-term PM2.5 mortality but has narrower confidence intervals.
2. Updated health endpoints. We have moved the quantification of chronic bronchitis
from our main analysis to a sensitivity analysis. This change is consistent with the
findings of the Integrated Science Assessment (ISA) that the evidence for an
association between long-term exposure to PM2.5 and respiratory effects is more
tenuous (U.S. EPA, 2009). We also add two new sensitivity analyses, including an
assessment of PM-related cerebrovascular disease and cardiovascular emergency
department visits. The incorporation of these two new endpoints follows the
findings of the PM ISA that recent studies have strengthened the relationship
between PM2.5 exposure and cardiovascular outcomes (U.S. EPA, 2009b).
3. Incorporation of new morbidity studies. Since the publication of the 2004 Criteria
Document for Particulate Matter (U.S. EPA, 2004) the epidemiological literature has
produced a significant number of new studies examining the association between
short-term PM2.5 exposure and acute myocardial infarctions, respiratory and
cardiovascular hospitalizations, respiratory emergency department visits, acute
respiratory symptoms and exacerbation of asthma. Upon careful evaluation of this
new literature we have incorporated new studies into our health impact
assessment; in many cases we have replaced older single-city time-series studies
with newer multi-city time-series analyses.
4. Updated hospital cost-of-illness (COI), including median wage data. In previous
versions of BenMAP, estimates of hospital charges and lengths of hospital stays
were based on discharge statistics provided by the Agency for Healthcare Research
and Quality's Healthcare Utilization Project National Inpatient Sample (NIS) database
for 2000 (AHRQ, 2000). The newest version of BenMAP (version 4.0.51) used in this
analysis updated this information to use the 2007 database. The data source for the
updated median annual income is the 2007 American Community Survey (ACS,
2007).
5. Expanded uncertainty assessment. We have incorporated a more comprehensive
assessment of the various uncertain parameters and assumptions within the
benefits analysis.
While the list above identifies the major changes implemented since the MATS RIA, EPA
has updated several additional components of the benefits analysis since the 2006 PM NAAQS
RIA (U.S. EPA, 2006a). In the Portland Cement NESHAP proposal RIA (U.S. EPA, 2009a), the
Agency removed the threshold assumption in the concentration-response function for PM2.5-
related health effects and began using the benefits derived from the two major cohort studies
of PM2.s and mortality as the main benefits estimates, while still including a range of sensitivity
estimates based on EPA's PM2.5 mortality expert elicitation. In the N02 NAAQS proposal RIA
5-7
-------
(U.S. EPA, 2009a), we revised the estimate used for the value-of-a-statistical life to be
consistent with Agency guidance. In the proposed CSAPR (previously the "Transport Rule") (U.S.
EPA, 2010g), we incorporated the "lowest measured level" assessment to help characterize
uncertainty in estimates of benefits of reductions in PM2.s at lower baseline concentrations of
PM2.5. In the final CSAPR (U.S. EPA, 2011c), we updated the baseline incidence rates for hospital
admissions and emergency department visits and asthma prevalence rates. We direct the
reader to each of these RIAs for more information on these changes.
5.4 Human Health Benefits Analysis Methods
We follow a "damage-function" approach in calculating total benefits of the modeled
changes in environmental quality. This approach estimates changes in individual health and
welfare endpoints (specific effects that can be associated with changes in air quality) and
assigns values to those changes assuming independence of the values for those individual
endpoints. Total benefits are calculated simply as the sum of the values for all non-overlapping
health and welfare endpoints. The "damage-function" approach is the standard method for
assessing costs and benefits of environmental quality programs and has been used in several
recent published analyses (Levy et al., 2009; Fann et al., 2012a; Tagaris et al., 2009).
To assess economic value in a damage-function framework, the changes in
environmental quality must be translated into effects on people or on the things that people
value. In some cases, the changes in environmental quality can be directly valued, as is the case
for changes in visibility. In other cases, such as for changes in ozone and PM, a health and
welfare impact analysis must first be conducted to convert air quality changes into effects that
can be assigned dollar values. For the purposes of this RIA, the health impacts analysis (HIA) is
limited to those health effects that are directly linked to ambient levels of air pollution and
specifically to those linked to PM2.s.
We note at the outset that EPA rarely has the time or resources to perform extensive
new research to measure directly either the health outcomes or their values for regulatory
analyses. Thus, similar to Kunzli et al. (2000) and other, more recent health impact analyses, our
estimates are based on the best available methods of benefits transfer. Benefits transfer is the
science and art of adapting primary research from similar contexts to obtain the most accurate
measure of benefits for the environmental quality change under analysis. Adjustments are
made for the level of environmental quality change, the socio-demographic and economic
characteristics of the affected population, and other factors to improve the accuracy and
robustness of benefits estimates.
5-8
-------
5.4.1 Health Impact Assessment
The Health Impact Assessment (HIA) quantifies the changes in the incidence of adverse
health impacts resulting from changes in human exposure to PM2.5 and ozone air quality. HIAs
are a well-established approach for estimating the retrospective or prospective change in
adverse health impacts expected to result from population-level changes in exposure to
pollutants (Levy et al., 2009). PC-based tools such as the environmental Benefits Mapping and
Analysis Program (BenMAP) can systematize health impact analyses by applying a database of
key input parameters, including health impact functions and population projections—provided
that key input data are available, including air quality estimates and risk coefficients (Abt
Associates, 2010). Analysts have applied the HIA approach to estimate human health impacts
resulting from hypothetical changes in pollutant levels (Hubbell et al., 2005; Tagaris et al., 2009;
Fann et al., 2012a). EPA and others have relied upon this method to predict future changes in
health impacts expected to result from the implementation of regulations affecting air quality
(e.g., U.S. EPA, 2011d). For this assessment, the HIA is limited to those health effects that are
directly linked to ambient PM2.s concentrations. There may be other indirect health impacts
associated with implementing emissions controls, such as occupational health exposures.
The HIA approach used in this analysis involves three basic steps: (1) utilizing projections
of PM2.s air quality1 and estimating the change in the spatial distribution of the ambient air
quality; (2) determining the subsequent change in population-level exposure; (3) calculating
health impacts by applying concentration-response relationships drawn from the
epidemiological literature (Hubbell et al., 2009) to this change in population exposure.
A typical health impact function might look as follows:
Ay = y0 • (eP'^x - l) • Pop (5-1)
where y0 is the baseline incidence rate for the health endpoint being quantified (for example, a
health impact function quantifying changes in mortality would use the baseline, or background,
mortality rate for the given population of interest); Pop is the population affected by the
change in air quality; Ax is the change in air quality; and 3 is the effect coefficient drawn from
the epidemiological study. Figure 5-1 provides a simplified overview of this approach.
1 Projections of ambient PM2.5 concentrations for this analysis were generated using the Community Multiscale Air
Quality model (CMAQ). See Chapter 3 of this RIA for more information on the air quality modeling.
5-9
-------
:
toeremenal Air Quality
I m p rowsffic^c
>
LilMnit*
Figure 5-1. Illustration of BenMAP Approach
5.4.2 Economic Valuation of Health Impacts
After quantifying the change in adverse health impacts, the final step is to estimate the
economic value of these avoided impacts. The appropriate economic value for a change in a
health effect depends on whether the health effect is viewed ex ante (before the effect has
occurred) or ex post (after the effect has occurred). Reductions in ambient concentrations of air
pollution generally lower the risk of future adverse health effects by a small amount for a large
population. The appropriate economic measure is therefore ex ante Willingness to Pay (WTP)
for changes in risk. However, epidemiological studies generally provide estimates of the relative
risks of a particular health effect avoided due to a reduction in air pollution. A convenient way
to use these data in a consistent framework is to convert probabilities to units of avoided
statistical incidences. This measure is calculated by dividing individual WTP for a risk reduction
by the related observed change in risk. For example, suppose a measure is able to reduce the
risk of premature mortality from 2 in 10,000 to 1 in 10,000 (a reduction of 1 in 10,000). If
individual WTP for this risk reduction is $100, then the WTP for an avoided statistical premature
mortality amounts to $1 million ($100/0.0001 change in risk). Using this approach, the size of
5-10
-------
the affected population is automatically taken into account by the number of incidences
predicted by epidemiological studies applied to the relevant population. The same type of
calculation can produce values for statistical incidences of other health endpoints.
For some health effects, such as hospital admissions, WTP estimates are generally not
available. In these cases, we use the cost of treating or mitigating the effect. For example, for
the valuation of hospital admissions we use the avoided medical costs as an estimate of the
value of avoiding the health effects causing the admission. These cost-of-illness (COI) estimates
generally (although not in every case) understate the true value of reductions in risk of a health
effect. They tend to reflect the direct expenditures related to treatment but not the value of
avoided pain and suffering from the health effect.
We use the BenMAP model version 4.0.52 (Abt Associates, 2010) to estimate the health
impacts and monetized health benefits for the proposed standard. Figure 5-2 shows the data
inputs and outputs for the BenMAP model.
Census
Population Data
Modeled Baseline
and Post-Control
Ambient PM2.s
2020 Population
Projections
Woods & Poole
Population
Projections
PM2.5 Incremental Air
Quality Change
PM2.5 Health
Functions
Economic
Valuation
Functions
PM2.5-Related Health
Impacts
Background
Incidence and
Prevalence Rates
Monetized PM2.5-
related Benefits
Blue identifies a user-selected input within the BenMAP model
Green identifies a data input generated outside of the BenMAP
Figure 5-2. Data Inputs and Outputs for the BenMAP Model
5-11
-------
5.5 Uncertainty Characterization
In any complex analysis using estimated parameters and inputs from numerous models,
there are likely to be many sources of uncertainty. This analysis is no exception. As outlined
both in this and preceding chapters, this analysis includes many data sources as inputs,
including emission inventories, air quality data from models (with their associated parameters
and inputs), population data, population estimates, health effect estimates from epidemiology
studies, economic data for monetizing benefits, and assumptions regarding the future state of
the world (i.e., regulations, technology, and human behavior). Each of these inputs may be
uncertain and would affect the benefits estimate. When the uncertainties from each stage of
the analysis are compounded, even small uncertainties can have large effects on the total
quantified benefits.
After reviewing EPA's approach, the National Research Council (NRC) (2002, 2008),
which is part of the National Academies of Science, concluded that EPA's general methodology
for calculating the benefits of reducing air pollution is reasonable and informative in spite of
inherent uncertainties. The NRC also highlighted the need to conduct rigorous quantitative
analyses of uncertainty and to present benefits estimates to decision makers in ways that foster
an appropriate appreciation of their inherent uncertainty. Since the publication of these
reports, EPA has continued work to improve the characterization of uncertainty in both health
incidence and benefits estimates. In response to these recommendations, we have expanded
our previous analyses to incorporate additional quantitative and qualitative characterizations of
uncertainty.
To characterize uncertainty and variability, we follow an approach that combines
elements from two recent EPA analyses (U.S. EPA, 2010b; U.S. EPA, 2011a), and uses a tiered
approach developed by the World Health Organization (WHO) for characterizing uncertainty
(WHO, 2008). We present this tiered assessment as well as an assessment of the potential
impact and magnitude of each aspect of uncertainty In Appendix 5c. While data limitations
prevent us from treating each source of uncertainty quantitatively and from reaching a full-
probabilistic simulation of our results, we were able to consider the influence of uncertainty in
the risk coefficients and economic valuation functions by incorporating six quantitative analyses
described in more detail below:
1. A Monte Carlo assessment that accounts for random sampling error and between
study variability in the epidemiological and economic valuation studies;
5-12
-------
2. A concentration benchmark assessment that characterizes the distribution of
avoided PM2.5-related deaths relative to specific concentrations in the long-term
epidemiological studies used to estimate PM2.5-related mortality;
3. The quantification of PM-related mortality using alternative PM2.5 mortality effect
estimates drawn from two long-term cohort studies and an expert elicitation;
4. Sensitivity analyses of several aspects of PM-related benefits;
5. Distributional analyses of PM2.5-related benefits by location, race, income, and
education; and
6. An analysis of the influence of various parameters on total monetized benefits.
5.5.1 Monte Carlo Assessment
Similar to other recent RIAs, we used Monte Carlo methods for characterizing random
sampling error associated with the concentration response functions from epidemiological
studies and random effects modeling to characterize both sampling error and variability across
the economic valuation functions. The Monte Carlo simulation in the BenMAP software
randomly samples from a distribution of incidence and valuation estimates to characterize the
effects of uncertainty on output variables. Specifically, we used Monte Carlo methods to
generate confidence intervals around the estimated health impact and monetized benefits. The
reported standard errors in the epidemiological studies determined the distributions for
individual effect estimates for endpoints estimated using a single study. For endpoints
estimated using a pooled estimate of multiple studies, the confidence intervals reflect both the
standard errors and the variance across studies. The confidence intervals around the monetized
benefits incorporate the epidemiology standard errors as well as the distribution of the
valuation function. These confidence intervals do not reflect other sources of uncertainty
inherent within the estimates, such as baseline incidence rates, populations exposed and
transferability of the effect estimate to diverse locations. As a result, the reported confidence
intervals and range of estimates give an incomplete picture about the overall uncertainty in the
benefits estimates.
5.5.2 Concentration Benchmark Analysis for PM2.s
We include a concentration benchmark assessment, which identifies the baseline (i.e.,
pre-rule) annual mean PM2.5 levels at which populations are exposed and specific
concentrations in the two long-term cohort studies we use to quantify mortality impacts. This
analysis characterize avoided PM2.5-related deaths relative the the 10th and 25th percentiles of
5-13
-------
the the air quality data used the Krewski et al. (2009) study as well as the lowest measured level
(LML) of the Krewski et al. (2009) and Laden et al. (2006) studies.
5.5.3 Alternative Concentration-Response Functions for PM2.$ -Related Mortality
We assign the greatest economic value to the reduction in PM2.5 related mortality risk.
Therefore, it is particularly important to attempt to characterize the uncertainties associated
with reductions in premature mortality. To better understand the concentration-response
relationship between PM2.5 exposure and premature mortality, EPA conducted an expert
elicitation in 2006 (Roman et al., 2008; lEc, 2006).2 In general, the results of the expert
elicitation support the conclusion that the benefits of PM2.5 control are very likely to be
substantial.
Alternative concentration-response functions are useful for assessing uncertainty
beyond random statistical error, including uncertainty in the functional form of the model or
alternative study design. Thus, we include the expert elicitation results as well as standard
errors approaches to provide insights into the likelihood of different outcomes and about the
state of knowledge regarding the benefits estimates. In this analysis, we present the results
derived from the expert elicitation as indicative of the uncertainty associated with a major
component of the health impact functions, and we provide the independent estimates derived
from each of the twelve experts to better characterize the degree of variability in the expert
responses.
In previous RIAs, EPA presented benefits estimates using concentration response
functions derived from the PM2.5 Expert Elicitation (Roman et al., 2008) as a range from the
lowest expert value (Expert K) to the highest expert value (Expert E). However, this approach
did not indicate the agency's judgment on what the best estimate of PM2.5 benefits may be, and
EPA's independent Science Advisory Board raised concerns about this presentation (U.S. EPA-
SAB, 2008). Therefore, we began to present the cohort-based studies (Krewski et al., 2009;
Laden et al., 2006) as our core estimates in the proposal RIA for the Portland Cement NESHAP
(U.S. EPA, 2009a). Using alternate relationships between PM2.5and premature mortality
supplied by experts, higher and lower benefits estimates are plausible, but most of the expert-
based estimates of the mean PM2.5 effect on mortality fall between the two epidemiology-
based estimates (Roman et al., 2008). In addition to these studies, we have included a
discussion or other recent multi-state cohort studies conducted in North America, but we have
2 Expert elicitation is a formal, highly structured and well documented process whereby expert judgments, usually
of multiple experts, are obtained (Ayyub, 2002).
5-14
-------
not estimated benefits using the effect coefficients from these studies. Please note that the
benefits estimates results presented are not the direct results from the studies or expert
elicitation; rather, the estimates are based in part on the effect coefficients provided in those
studies or by experts.
Even these multiple characterizations with confidence intervals omit the contribution to
overall uncertainty from uncertainty in air quality changes, baseline incidence rates, and
populations exposed. Furthermore, the approach presented here does not yet include methods
for addressing correlation between input parameters and the identification of reasonable upper
and lower bounds for input distributions characterizing uncertainty in additional model
elements. As a result, the reported confidence intervals and range of estimates give an
incomplete picture about the overall uncertainty in the estimates. This information should be
interpreted within the context of the larger uncertainty surrounding the entire analysis.
5.5.4 Sensitivity Analyses
For some aspects of uncertainty, we have sufficient data to conduct sensitivity analyses.
In this analysis, we performed four such analyses for the proposed standard level. In particular,
we:
1. Assessed the sensitivity of the economic value of reductions in the risk of PM2.s-
related death according to differing assumptions regarding the lag between PM2.5
exposure and premature death. The timing of such premature deaths affects the
magnitude of the discounted PM2.5-related mortality benefits. In this sensitivity
assessment, we consider 6 alternative cessation lags.
2. Characterized the sensitivity of the economic value of the health endpoints valued
using willingness-to-pay estimates to a higher and a lower assumption regarding
income elasticity. As we discuss below, economic theory argues that individual
willingness to pay increases as personal income grows. The relationship between
growth in personal income and willingness-to-pay to reduce mortality and morbidity
risk is characterized by the income growth factor.
3. Summarized the avoided cases of certain health endpoints for which we either
lacked an appropriate economic value (cardiovascular hospital admissions and
stroke) or in which we no longer had sufficient confidence to retain in our primary
benefits estimate (chronic bronchitis).
4. Compared the valuation of hospitalizations and work loss days using the 2000 AHRQ
database to the 2007 database.
5-15
-------
5.5.5 Distributional Assessment
In the Appendix to this chapter, we characterize the distribution of PM2.5-related
benefits based on the geographic distribution of race and education in areas where the selected
control strategies would reduce PM2.5 concentrations. In this assessment, we aim to answer two
key questions:
1. What is the estimated distribution of PM2.5-related mortality risk based on the race
and education characteristics of the population living within areas projected to
exceed alternative combinations of primary PM2.5 standards?
2. How would air quality improvements within these counties change the distribution
of risk among populations of different races and educational attainment?3
This assessment is generally consistent with the distributional assessments performed in
support of the CSAPR (EPA, 2011c) and the MATS (EPA, 2011c), with one key difference. The
environmental justice analyses accompanying the CSAPR and MATS RIAs applied CMAQ-
modeled PM2.5 predictions that represent the change in air quality after the implementation of
each rule. By contrast, this RIA aims to illustrate the potential benefits and costs of attaining
alternative primary PM2.5 standards; the states will ultimately implement attainment strategies,
which may differ greatly from the least-cost strategy EPA modeled here. Alternative attainment
strategies—particularly those that maximize benefits to human health and provide a more
equitable distribution of risk—are also available to the states, though not modeled here (Fann
etal., 2012b).
5.5.6 Influence Analysis—Quantitative Assessment of Uncertainty
In the past few years, EPA has initiated several projects to improve the characterization
of uncertainty for benefits analysis. In particular, EPA recently completed the first phase of a
quantitative uncertainty analysis of benefits, hereafter referred to as the "Influence Analysis"
(Mansfield et al., 2009). The Influence Analysis diagramed the uncertain components of each
step within the benefits analysis process, identified plausible ranges for a sensitivity analysis,
and assessed the sensitivity to total benefits to changes in each component. While this analysis
does not quite fulfill the goal of a full probabilistic assessment, it accomplished the necessary
first steps and identified the challenges to accomplishing that goal. Below are some of the
preliminary observations from the first phase of the project.
3 In this analysis we assess the change in risk among populations of different race and educational attainment. As
we discuss further in the methodology, we consider this last variable because of the availability of education-
modified PM2.5 mortality risk estimates.
5-16
-------
• The components that contribute the most to uncertainty of the monetized benefits
and mortality incidence (in order of importance) are the value-of-a-statistical-life
(VSL), the concentration-response (C-R) function for mortality, and change in PM2.5
concentration.
• The components that contribute the least to uncertainty of the monetized benefits
and mortality incidence are population, morbidity valuation, and income elasticity.
• The choice of a C-R function for mortality affects the mortality incidence and
monetized benefits more than other sources of uncertainty within each C-R
function.
• Alternative cessation lag structures for mortality have a moderate effect on the
monetized benefits.
• Because the health impact function is essentially linear, the key components show
the same sensitivity across all mortality C-R functions even if the midpoints differ
significantly from one expert to another.
5.5.7 Qualitative Assessment of Uncertainty and Other Analysis Limitations
Although we strive to incorporate as many quantitative assessments of uncertainty as
possible, there are several aspects we are only able to address qualitatively. These aspects are
important factors to consider when evaluating the relative benefits of the attainment strategies
for each of the alternative standards:
The total monetized benefits presented in this chapter are based on our interpretation
of the best available scientific literature and methods and supported by EPA's independent
Science Advisory Board (Health Effects Subcommittee) (SAB-HES) (U.S. EPA- SAB, 2010a) and
the National Academies of Science (NAS) (NRC, 2002). The benefits estimates are subject to a
number of assumptions and uncertainties. For example, the key assumptions underlying the
estimates for premature mortality, which account for over 98% of the total monetized benefits
in this analysis, include the following:
1. We assume that all fine particles, regardless of their chemical composition, are
equally potent in causing premature mortality. This is an important assumption,
because PM2.s varies considerably in composition across sources, but the scientific
evidence is not yet sufficient to allow differentiation of effect estimates by particle
type. The PM ISA, which was twice reviewed by CASAC, concluded that "many
constituents of PM2.s can be linked with multiple health effects, and the evidence is
not yet sufficient to allow differentiation of those constituents or sources that are
more closely related to specific outcomes" (U.S. EPA, 2009b).
5-17
-------
2. We assume that the health impact function for fine particles is log-linear down to
the lowest air quality levels modeled in this analysis. Thus, the estimates include
health benefits from reducing fine particles in areas with varied concentrations of
PM2.5, including both regions that are in attainment with the fine particle standard
and those that do not meet the standard down to the lowest modeled
concentrations.
3. To characterize the uncertainty in the relationship between PM2.5and premature
mortality (which account for over 98% of total monetized benefits in this analysis),
we include a set of twelve estimates based on results of the expert elicitation study
in addition to our core estimates. Even these multiple characterizations omit the
uncertainty in air quality estimates, baseline incidence rates, populations exposed
and transferability of the effect estimate to diverse locations. As a result, the
reported confidence intervals and range of estimates give an incomplete picture
about the overall uncertainty in the PM2.5 estimates. This information should be
interpreted within the context of the larger uncertainty surrounding the entire
analysis.
As previously described, we strive to monetize as many of the benefits anticipated from
the alternative standards as possible given data and resource limitations, but the monetized
benefits estimated in this RIA inevitably only reflect a portion of the benefits. Specifically, only
certain benefits attributable to the health impacts associated with exposure to ambient fine
particles have been monetized in this analysis. Data and methodological limitations prevented
EPA from quantifying or monetizing the benefits from several important health benefit
categories in this RIA, including benefits from reducing ozone exposure, N02 exposure, S02
exposure, and methylmercury exposure (see section 5.6.5 for more information). If we could
fully monetize all of the benefit categories, the total monetized benefits would exceed the costs
by an even greater margin than we currently estimate.
To more fully address all these uncertainties including those we cannot quantify, we
apply a four-tiered approach using the WHO uncertainty framework (WHO, 2008), which
provides a means for systematically linking the characterization of uncertainty to the
sophistication of the underlying risk assessment. EPA has applied similar approaches in analyses
(U.S. EPA, 2010b, 2011a). Using this framework, we summarize the key uncertainties in the
health benefits analysis, including our assessment of the direction of potential bias, magnitude
of impact on the monetized benefits, degree of confidence in our analytical approach, and our
ability to assess the source of uncertainty. More information on this approach and the
uncertainty characterization are available in Appendix 5C. Because this approach reflects a new
5-18
-------
application for regulatory benefits analysis, we request comments on this general approach to
characterizing uncertainty as well as the specific uncertainty assessments.
5.6 Benefits Analysis Data Inputs
In Figure 5-2, we summarized the key data inputs to the health impact and economic
valuation estimate. Below we summarize the data sources for each of these inputs, including
demographic projections, incidence and prevalence rates, effect coefficients, and economic
valuation. We indicate where we have updated key data inputs since the benefits analysis
conducted for the MATS (U.S. EPA, 2011d).
5.6.1 Demographic Data
Quantified and monetized human health impacts depend on the demographic
characteristics of the population, including age, location, and income. We use projections based
on economic forecasting models developed by Woods and Poole, Inc. (Woods and Poole, 2007).
The Woods and Poole (WP) database contains county-level projections of population by age,
sex, and race out to 2030. Projections in each county are determined simultaneously with every
other county in the United States to take into account patterns of economic growth and
migration. The sum of growth in county-level populations is constrained to equal a previously
determined national population growth, based on Bureau of Census estimates (Hollman et al.,
2000). According to WP, linking county-level growth projections together and constraining to a
national-level total growth avoids potential errors introduced by forecasting each county
independently. County projections are developed in a four-stage process:
• First, national-level variables such as income, employment, and populations are
forecasted.
• Second, employment projections are made for 179 economic areas defined by
the Bureau of Economic Analysis (U.S. BEA, 2004), using an "export-base"
approach, which relies on linking industrial-sector production of non-locally
consumed production items, such as outputs from mining, agriculture, and
manufacturing with the national economy. The export-based approach requires
estimation of demand equations or calculation of historical growth rates for
output and employment by sector.
• Third, population is projected for each economic area based on net migration
rates derived from employment opportunities and following a cohort-
component method based on fertility and mortality in each area.
• Fourth, employment and population projections are repeated for counties, using
the economic region totals as bounds. The age, sex, and race distributions for
5-19
-------
each region or county are determined by aging the population by single year of
age by sex and race for each year through 2020 based on historical rates of
mortality, fertility, and migration.
5.6.2 Baseline Incidence and Prevalence Estimates
Epidemiological studies of the association between pollution levels and adverse health
effects generally provide a direct estimate of the relationship of air quality changes to the
relative risk of a health effect, rather than estimating the absolute number of avoided cases. For
example, a typical result might be that a 10 u.g/m3 decrease in daily PM2.5 levels might be
associated with a decrease in hospital admissions of 3%. The baseline incidence of the health
effect is necessary to convert this relative change into a number of cases. A baseline incidence
rate is the estimate of the number of cases of the health effect per year in the assessment
location, as it corresponds to baseline pollutant levels in that location. To derive the total
baseline incidence per year, this rate must be multiplied by the corresponding population
number. For example, if the baseline incidence rate is the number of cases per year per million
people, that number must be multiplied by the millions of people in the total population.
Table 5-3 summarizes the sources of baseline incidence rates and provides average
incidence rates for the endpoints included in the analysis. For both baseline incidence and
prevalence data, we used age-specific rates where available. We applied concentration-
response functions to individual age groups and then summed over the relevant age range to
provide an estimate of total population benefits. In most cases, we used a single national
incidence rate, due to a lack of more spatially disaggregated data. Whenever possible, the
national rates used are national averages, because these data are most applicable to a national
assessment of benefits. For some studies, however, the only available incidence information
comes from the studies themselves; in these cases, incidence in the study population is
assumed to represent typical incidence at the national level. County, state and regional
incidence rates are available for hospital admissions, and county-level data are available for
premature mortality. We have projected mortality rates such that future mortality rates are
consistent with our projections of population growth (Abt Associates, 2011).
The baseline incidence rates for hospital admissions and emergency department visits
reflect the revised rates first applied in the CSAPR RIA (U.S. EPA, 2011c). In addition, we have
also revised the baseline incidence rates for acute myocardial infarction. These revised rates are
more recent (AHRQ, 2007), which provides a better representation of the rates at which
populations of different ages, and in different locations, visit the hospital and emergency
department for air pollution-related illnesses. Also, the new baseline incidence rates are more
5-20
-------
spatially refined. For many locations within the U.S., these data are resolved at the county- or
state-level, providing a better characterization of the geographic distribution of hospital and
emergency department visits than the previous national rates. Lastly, these rates reflect
unscheduled hospital admissions only, which represents a conservative assumption that most
air pollution-related visits are likely to be unscheduled. If air pollution-related hospital
admissions are scheduled, this assumption would underestimate these benefits.
5-21
-------
Table 5-3. Baseline Incidence Rates and Population Prevalence Rates for Use in Impact
Functions, General Population
Rates
Endpoint
Parameter
Value
Source
Mortality
Hospitalizations
ER Visits
Cerebrovascular
events
Chronic Bronchitis
Nonfatal Myocardial
Infarction (heart
attacks)
Asthma Exacerbations
Acute Bronchitis
Lower Respiratory
Symptoms
Upper Respiratory
Symptoms
Work Loss Days
School Loss Days
Daily or annual mortality rate
projected to 2015a
Daily hospitalization rate
Daily ER visit rate for asthma and
cardiovascular events
Incidence of new cerebrovascular
events among populations 50-79
Annual prevalence rate per person
• Aged 18^4
• Aged 45-64
• Aged 65 and older
Annual incidence rate per person
Daily nonfatal myocardial infarction
incidence rate per person, 18+
Incidence among asthmatic African-
American children
• daily wheeze
• daily cough
• daily dyspnea
Annual bronchitis incidence rate,
children
Daily lower respiratory symptom
incidence among children c
Daily upper respiratory symptom
incidence among asthmatic children
Daily WLD incidence rate per person
(18-65)
• Aged 18-24
• Aged 25-44
• Aged 45-64
Rate per person per year, assuming
180 school days per year
Age-, cause-, and county-
specific rate
Age-, region-, state-, county-
and cause-specific rate
Age-, region-, state-, county-
and cause-specific rate
0.0015751
• 0.0315
• 0.0549
• 0.0563
0.00378
Age-, region-, state-, and
county-specific rate
• 0.076
• 0.067
• 0.037
0.043
0.0012
0.3419
• 0.00540
• 0.00678
• 0.00492
9.9
CDC WONDER (2004-2006)
U.S. Census bureau, 2000
2007 HCUP data files b
2007 HCUP data files"
Table 3 of Miller etal. (2007)
American Lung Association
(2010a, Table 4).
Abbey etal. (1993, Table3)
2007 HCUP data files b;
adjusted by 0.93 for
probability of surviving after
28 days (Rosamond et al.,
1999)
Ostro et al. (2001)
American Lung Association
(2002c, Table 11)
Schwartz et al. (1994, Table
2)
Pope etal. (1991, Table 2)
1996 HIS (Adams,
Hendershot, and Marano,
1999, Table 41); U.S. Census
Bureau (2000)
National Center for
Education Statistics (1996)
and 1996 HIS (Adams etal.,
1999, Table 47);
(continued)
5-22
-------
Table 5-3. Baseline Incidence Rates and Population Prevalence Rates for Use in Impact
Functions, General Population (continued)
Rates
Endpoint
Parameter
Value
Source
Minor Restricted- Daily MRAD incidence rate per
Activity Days person
0.02137
Ostro and Rothschild (1989,
p. 243)
Mortality rates are only available at 5-year increments.
b Healthcare Cost and Utilization Program (HCUP) database contains individual level, state and regional-level
hospital and emergency department discharges for a variety of ICD codes (AHRQ, 2007).
c Lower respiratory symptoms are defined as two or more of the following: cough, chest pain, phlegm, and
wheeze.
d Assessed in sensitivity analysis only. The rate numbers may be slightly different from those in Table 4 because
we received more current estimates from ALA.
For the set of endpoints affecting the asthmatic population, in addition to baseline
incidence rates, prevalence rates of asthma in the population are needed to define the
applicable population. Table 5-4 lists the prevalence rates used to determine the applicable
population for asthma symptoms. Note that these reflect current asthma prevalence and
assume no change in prevalence rates in future years. We updated these rates in the CSAPR RIA
(U.S. EPA, 2011c).
Table 5-4. Asthma Prevalence Rates
Population Group
All Ages
<18
5-17
18-44
45-64
65+
African American, 5-17
African American, <18
Value
0.0780
0.0941
0.1070
0.0719
0.0745
0.0716
0.1776
0.1553
Asthma Prevalence Rates
Source
American Lung Association (2010b, Table 7)
American Lung Association (2010b, Table 9)
American Lung Association3
Calculated by ALA for U.S. EPA, based on NHIS data (CDC, 2008).
5-23
-------
5.6.3 Effect Coefficients
The first step in selecting effect coefficients is to identify the health endpoints to be
quantified. We base our selection of health endpoints on consistency with EPA's Integrated
Science Assessments (which replace previous Criteria Documents), with input and advice from
the EPA Science Advisory Board —Health Effects Subcommittee (SAB-HES), a scientific review
panel specifically established to provide advice on the use of the scientific literature in
developing benefits analyses for air pollution regulations. In general, we follow a weight of
evidence approach, based on the biological plausibility of effects, availability of concentration-
response functions from well conducted peer-reviewed epidemiological studies, cohesiveness
of results across studies, and a focus on endpoints reflecting public health impacts (like hospital
admissions) rather than physiological responses (such as changes in clinical measures like
Forced Expiratory Volume (FEV1)).
There are several types of data that can support the determination of types and
magnitude of health effects associated with air pollution exposures. These sources of data
include toxicological studies (including animal and cellular studies), human clinical trials, and
observational epidemiology studies. All of these data sources provide important contributions
to the weight of evidence surrounding a particular health impact. However, only epidemiology
studies provide direct concentration-response relationships that can be used to evaluate
population-level impacts of reductions in ambient pollution levels in a health impact
assessment.
For the data-derived estimates, we relied on the published scientific literature to
ascertain the relationship between PM2.5 and adverse human health effects. We evaluated
epidemiological studies using the selection criteria summarized in Table 5-5. These criteria
include consideration of whether the study was peer-reviewed, the match between the
pollutant studied and the pollutant of interest, the study design and location, and
characteristics of the study population, among other considerations. In general, the use of
concentration-response functions from more than a single study can provide a more
representative distribution of the effect estimate. However, there are often differences
between studies examining the same endpoint, making it difficult to pool the results in a
consistent manner. For example, studies may examine different pollutants or different age
groups. For this reason, we consider very carefully the set of studies available examining each
endpoint and select a consistent subset that provides a good balance of population coverage
and match with the pollutant of interest. In many cases, either because of a lack of multiple
5-24
-------
studies, consistency problems, or clear superiority in the quality or comprehensiveness of one
study over others, a single published study is selected as the basis of the effect estimate.
When several effect estimates for a pollutant and a given health endpoint have been
selected, they are quantitatively combined or pooled to derive a more robust estimate of the
relationship. The BenMAP Manual Technical Appendices provides details of the procedures
used to combine multiple impact functions (Abt Associates, 2011). In general, we used fixed or
random effects models to pool estimates from different single city studies of the same
endpoint. Fixed effects pooling simply weights each study's estimate by the inverse variance,
giving more weight to studies with greater statistical power (lower variance). Random effects
pooling accounts for both within-study variance and between-study variability, due, for
example, to differences in population susceptibility. We used the fixed effects model as our null
hypothesis and then determined whether the data suggest that we should reject this null
hypothesis, in which case we would use the random effects model. Pooled impact functions are
used to estimate hospital admissions and asthma exacerbations. When combining evidence
across multi-city studies (e.g., cardiovascular hospital admission studies), we use equal weights
pooling. The effect estimates drawn from each multi-city study are themselves pooled across a
large number of urban areas. For this reason, we elected to give each study an equal weight
rather than weighting by the inverse of the variance reported in each study. For more details on
methods used to pool incidence estimates, see the BenMAP Manual Appendices (Abt
Associates, 2011).
Effect estimates selected for a given health endpoint were applied consistently across all
locations nationwide. This applies to both impact functions defined by a single effect estimate
and those defined by a pooling of multiple effect estimates. Although the effect estimate may,
in fact, vary from one location to another (e.g., because of differences in population
susceptibilities or differences in the composition of PM), location-specific effect estimates are
generally not available.
The specific studies from which effect estimates for the main analysis are drawn are
included in Table 5-6. We highlight in blue those studies that have been added since the
benefits analysis conducted for the MATS RIA (U.S. EPA, 2011d) and incorporated into the
central benefits estimate. In all cases where effect estimates are drawn directly from
epidemiological studies, standard errors are used as a partial representation of the uncertainty
in the size of the effect estimate. Table 5-7 summarizes those health endpoints and studies we
have included as in sensitivity analyses.
5-25
-------
Table 5-5. Criteria Used When Selecting C-R Functions
Consideration
Comments
Peer-Reviewed
Research
Study Type
Study Period
Population Attributes
Study Size
Study Location
Pollutants Included in
Model
Measure of PM
Economically Valuable
Health Effects
Non-overlapping
Endpoints
Peer-reviewed research is preferred to research that has not undergone the peer-review
process.
Among studies that consider chronic exposure (e.g., over a year or longer), prospective
cohort studies are preferred over ecological studies because they control for important
individual-level confounding variables that cannot be controlled for in ecological studies.
Studies examining a relatively longer period of time (and therefore having more data) are
preferred, because they have greater statistical power to detect effects. Studies that are
more recent are also preferred because of possible changes in pollution mixes, medical
care, and lifestyle over time. However, when there are only a few studies available,
studies from all years will be included.
The most technically appropriate measures of benefits would be based on impact
functions that cover the entire sensitive population but allow for heterogeneity across
age or other relevant demographic factors. In the absence of effect estimates specific to
age, sex, preexisting condition status, or other relevant factors, it may be appropriate to
select effect estimates that cover the broadest population to match with the desired
outcome of the analysis, which is total national-level health impacts. When available,
multi-city studies are preferred to single city studies because they provide a more
generalizable representation of the concentration-response function.
Studies examining a relatively large sample are preferred because they generally have
more power to detect small magnitude effects. A large sample can be obtained in several
ways, including through a large population or through repeated observations on a
smaller population (e.g., through a symptom diary recorded for a panel of asthmatic
children).
U.S. studies are more desirable than non-U.S. studies because of potential differences in
pollution characteristics, exposure patterns, medical care system, population behavior,
and lifestyle.
When modeling the effects of ozone and PM (or other pollutant combinations) jointly, it
is important to use properly specified impact functions that include both pollutants.
Using single-pollutant models in cases where both pollutants are expected to affect a
health outcome can lead to double-counting when pollutants are correlated.
For this analysis, impact functions based on PM2.5 are preferred to PM10 because of the
focus on reducing emissions of PM2.s precursors, and because air quality modeling was
conducted for this size fraction of PM. Where PM2.5 functions are not available, PM10
functions are used as surrogates, recognizing that there will be potential downward
(upward) biases if the fine fraction of PM10 is more (less) toxic than the coarse fraction.
Some health effects, such as forced expiratory volume and other technical
measurements of lung function, are difficult to value in monetary terms. These health
effects are not quantified in this analysis.
Although the benefits associated with each individual health endpoint may be analyzed
separately, care must be exercised in selecting health endpoints to include in the overall
benefits analysis because of the possibility of double-counting of benefits.
(continued)
5-26
-------
Table 5-6. Health Endpoints and Epidemiological Studies Used to Quantify Health Impacts
in the Main Analysis3
Risk Estimate
(95th Percentile Confidence
Endpoint Study Study Population Interval)3
Premature Mortality
Premature mortality-
cohort study, all-cause
Premature mortality,
total exposures
Premature mortality—
all-cause
Krewski et al. (2009)
Laden etal. (2006)
PM2 5 Expert Elicitation (Roman et al., 2008)
Woodruff etal. (1997)
>29 years
>24 years
>24 years
Infant (<1 year)
RR = 1.06 (1.04-1.06) per
10 Mg/m3
RR = 1.16 (1.07-1.26) per
10 Mg/m3
Varies by expert
OR = 1.04 (1.02-1.07) per
10 Mg/m3
Chronic Illness
Nonfatal heart attacks Peters et al. (2001)
Pooled estimate:
Pope et al. (2006)
Sullivan etal. (2005)
Zanobetti et al. (2009)
Zanobetti and Schwartz (2006)
Adults (>18 years) OR = 1.62 (1.13 - 2.34) per
20 Mg/m3
P = 0.00481 (0.00199)
P = 0.00198 (0.00224)
P = 0.00225 (0.000591)
P = 0.0053 (0.00221)
Hospital Admissions
Respiratory
Cardiovascular
Zanobetti et al. (2009)—ICD 460-519 (All
respiratory)
Moolgavkar (2000)-ICD 490-496 (Chronic
lung disease)
Babin et al. (2007)—ICD 493 (asthma)
Pooled estimate:
Zanobetti et al. (2009)—ICD 390-459 (all
cardiovascular)
Peng et al. (2009)—ICD 426-427; 428; 430-
438; 410-414; 429; 440-449 (Cardio-,
cerebro- and peripheral vascular
disease)
Peng et al. (2008)—ICD 426-427; 428; 430-
438; 410-414; 429; 440-449 (Cardio-,
cerebro- and peripheral vascular
disease)
Bell et al. (2008)—ICD 426-427; 428; 430-
438; 410-414; 429; 440-449 (Cardio-,
cerebro- and peripheral vascular
disease)
>64 years
18-64 years
<19
>64 years
P=0.00207 (0.00446)
1.02(1.01—1.03) per
36 Mg/m3
P=0.002 (0.004337)
P=0.00189 (0.000283)
P=0.00068
(0.000214)
P=0.00071
(0.00013)
P=0.0008
(0.000107)
Asthma-related ER
visits
Moolgavkar (2000)— ICD 390-429 (all
cardiovascular)
Pooled estimate:
Mar et al. (2010)
Slaughter etal. (2005)
20-64 years RR=1.04 (t statistic: 4.1) per
10 Mg/m3
AN ages RR = 1.04 (1.01 -1.07) per
7 Mg/m3
RR = 1.03 (0.98 -1.09) per
10 Mg/m3
(continued)
5-27
-------
Endpoint
Study
Study Population
Risk Estimate
(95th Percentile Confidence
Interval)3
Other Health Endpoints
Acute bronchitis
Dockeryetal. (1996)
8-12 years OR = 1.50 (0.91-2.47) per
14.9 Mg/m3
Other Health Endpoints (continued)
Asthma exacerbations Pooled estimate:
6-18 years
Ostro et al. (2001) (cough, wheeze and
shortness of breath)
Work loss days
Acute respiratory
symptoms
Mar et al. (2004) (cough, shortness of
breath)
Ostro (1987)
Ostro and Rothschild (1989) (Minor
restricted activity days)
OR = 1.03 (0.98 -1.07) per
30 Mg/m3
OR = 1.06 (1.01 -1.11) per
30 Mg/m3
OR = 1.08 (1.00 -1.17) per
30 Mg/m3
RR = 1.21 (1-1.47) per
10
18-65 years
18-65 years
RR = 1.13 (0.86 -1.48) per
10 Mg/m3
(3=0.0046 (0.00036)
(3=0.00220 (0.000658)
Studies highlighted in blue represent updates incorporated since the RIA for MATS (U.S. EPA, 2011d).
The original study populations were 8 to 13 for the Ostro et al. (2001) study and 7 to 12 for the Mar et al. (2004)
study. Based on advice from the Science Advisory Board Health Effects Subcommittee (SAB-HES), we extended
the applied population to 6 to 18, reflecting the common biological basis for the effect in children in the broader
age group. See: U.S. EPA-SAB (2004) and NRC (2002).
Table 5-7. Health Endpoints and Epidemiological Studies Used to Quantify Health Impacts
in the Sensitivity Analysisa
Endpoint
Study
Study Population
Chronic Illness
Chronic bronchitis
Stroke
Abbey etal. (1995)
Miller etal. (2007)
>26 years
50-79 years
Hospital Admissions
Cardiovascular ED Visits
Metzger et al. (2004)
Tolbert et al. (2007)
0-99
0-99
a Studies highlighted in blue represent updates incorporated since the RIA for MATS (U.S. EPA, 2011d).
5.6.3.1 PM2,5 Premature Mortality Effect Coefficients
Both long- and short-term exposures to ambient levels of PM2.5 air pollution have been
associated with increased risk of premature mortality. The size of the mortality effect estimates
5-28
-------
from epidemiological studies, the serious nature of the effect itself, and the high monetary
value ascribed to prolonging life make mortality risk reduction the most significant health
endpoint quantified in this analysis.
Although a number of uncertainties remain to be addressed by continued research
(NRC, 2002), a substantial body of published scientific literature documents the correlation
between elevated PM2.5 concentrations and increased mortality rates (U.S. EPA, 2009b). Time-
series methods have been used to relate short-term (often day-to-day) changes in PM2.5
concentrations and changes in daily mortality rates up to several days after a period of elevated
PM2.5 concentrations. Cohort methods have been used to examine the potential relationship
between community-level PM2.5 exposures over multiple years (i.e., long-term exposures) and
community-level annual mortality rates. Researchers have found statistically significant
associations between PM2.5 and premature mortality using both types of studies. In general, the
effect estimates based on the cohort studies are larger than those derived from time-series
studies. When choosing between using short-term studies or cohort studies for estimating
mortality benefits, cohort analyses are thought to capture more of the public health impact of
exposure to air pollution over time because they account for the effects of long-term exposures
as well as some fraction of short-term exposures (Kunzli et al., 2001; NRC, 2002). This section
discusses some of the issues surrounding the estimation of PM2.5-related premature mortality.
To demonstrate the sensitivity of the benefits estimates to various concentration-response
estimates in the epidemiological literature, we present benefits estimates using several relative
risk estimates from the largest long-term epidemiological studies as well as a U.S. EPA-
sponsored expert elicitation (Roman et al. 2008). The epidemiological studies from which these
estimates are drawn are described below. The PM2.5 expert elicitation and the derivation of
effect estimates from the expert elicitation results are described in the 2006 PM2.5 NAAQS RIA
(U.S. EPA, 2006a) and Roman et al. (2008). In the interest of brevity, we do not repeat those
details here. 4
4
In summary, the goal of the study was to elicit from a sample of health experts probabilistic distributions
describing uncertainty in estimates of the reduction in mortality among the adult U.S. population resulting from
reductions in ambient annual average PM2.5 levels. These distributions were obtained through a formal interview
protocol using methods designed to elicit subjective expert judgments. These experts were selected through a
peer-nomination process and included experts in epidemiology, toxicology, and medicine. The elicitation
interview consisted of a protocol of carefully structured questions, both qualitative and quantitative, about the
nature of the PM2.5-mortality relationship questions requiring qualitative responses probed experts' beliefs
concerning key evidence and critical sources of uncertainty and enabled them to establish a conceptual basis
supporting their quantitative judgments. The results of the full-scale study consist of twelve individual
distributions for the coefficient or slope of the C-R function relating changes in annual average PM2.5 exposures
5-29
-------
Over a dozen epidemiological studies demonstrate significant associations between
various measures of long-term exposure to PM2.5 and mortality, beginning with Lave and Seskin
(1977). Most of the published studies found positive (but not always statistically significant)
associations with available PM indices such as total suspended particles (TSP). However,
exploration of alternative model specifications sometimes raised questions about causal
relationships (e.g., Lipfert et al., 1989). These early "ecological cross-sectional" studies (Lave
and Seskin, 1977; Ozkaynak and Thurston, 1987) were criticized for a number of methodological
limitations, particularly for inadequate control at the individual level for variables that are
potentially important in causing mortality, such as wealth, smoking, and diet.
Over the last two decades, several studies using "prospective cohort" designs have been
published that are consistent with the earlier body of literature. These "prospective cohort"
studies reflect a significant improvement over the earlier work because they include individual
level information with respect to health status and residence. Two prospective cohort groups,
often referred to as the Harvard "Six Cities Study" (Dockery et al., 1993; Laden et al., 2006) and
the "American Cancer Society or ACS study" (Pope et al., 1995; Pope et al., 2002; Pope et al.,
2004; Krewski et al., 2009), provide the most extensive analyses of ambient PM2.s
concentrations and mortality. These studies have found consistent relationships between fine
particle indicators and premature mortality across multiple locations in the United States. A
third major data set comes from the California-based 7th Day Adventist Study (e.g., Abbey
et al., 1999), which reported associations between long-term PM2.s exposure and mortality in
men. Results from this cohort, however, have been inconsistent, the air quality results are not
geographically representative of most of the United States, and the lifestyle of the population is
not reflective of much of the U.S. population. Analysis is also available for a cohort of adult
male veterans diagnosed with hypertension (Lipfert et al., 2000, 2003, 2006). The
characteristics of this group also differ from the cohorts in the Six Cities and ACS studies as well
as the 7th Day Adventist study with respect to income, race, health status, and smoking status.
Unlike previous long-term analyses, this study found some associations between mortality and
ozone but found inconsistent results for PM indicators.
Given their consistent results and broad geographic coverage, and importance in
informing the NAAQS development process, the Six Cities and ACS data have been particularly
important in benefits analyses. The credibility of these two studies is further enhanced by the
fact that the initial published studies (Pope et al., 1995; Dockery et al., 1993) were subject to
to annual, adult all-cause mortality. The results have not been combined in order to preserve the breadth and
diversity of opinion on the expert panel. Roman et al (2006).
5-30
-------
extensive reexamination and reanalysis by an independent team of scientific experts
commissioned by the Health Effect Institute (HEI) (Krewski et al., 2000). The final results of the
reanalysis were then independently peer reviewed by a Special Panel of the HEI Health Review
Committee. The results of these reanalyses confirmed and expanded the conclusions of the
original investigators. While the HEI reexamination lends credibility to the original studies, it
also highlights sensitivities concerning the relative impact of various pollutants, such as S02, the
potential role of education in mediating the association between pollution and mortality, and
the influence of spatial correlation modeling. Further confirmation and extension of the
findings of the 1993 Six Cities Study and the 1995 ACS study were recently completed using
more recent air quality and a longer follow-up period for the ACS cohort was published over the
past several years (Pope et al., 2002, 2004; Laden et al., 2006; Krewski et al., 2009). The follow
up to the Harvard Six Cities Study both confirmed the effect size from the first analysis and
provided additional confirmation that reductions in PM2.5 are associated with reductions in the
risk of premature death. This additional evidence stems from the observed reductions in PM2.5
in each city during the extended follow-up period. Laden et al. (2006) found that mortality rates
consistently went down at a rate proportional to the observed reductions in PM2.5.
A number of additional analyses have been conducted on the ACS cohort data (Pope
et al., 2009). These studies have continued to find a strong significant relationship between
PM2.5 and mortality outcomes and life expectancy. Specifically, much of the recent research has
suggested a stronger relationship between cardiovascular mortality and lung cancer mortality
with PM2.5, and a less significant relationship between respiratory-related mortality and PM2.5.
The extended analyses of the ACS cohort data (Krewski et al., 2009) provides additional
refinements to the analysis of PM-related mortality by (a) extending the follow-up period by
2 years to the year 2000, for a total of 18 years; (b) incorporating almost double the number of
urban areas (c) addressing spatial autocorrelation by incorporating ecological, or community-
level, co-variates; (d) performing an extensive spatial analysis using land use regression
modeling in two large urban areas. These enhancements make this analysis well-suited for the
assessment of mortality risk from long-term PM2.5 exposures for EPA benefits analyses.
In developing and improving the methods for estimating and valuing the potential
reductions in mortality risk over the years, EPA consulted with the Health Effects Subcommittee
of the Science Advisory Board (SAB-HES). That panel recommended using long-term prospective
cohort studies in estimating mortality risk reduction (U.S. EPA-SAB, 1999). This
recommendation has been confirmed by a report from the National Research Council, which
stated that "it is essential to use the cohort studies in benefits analysis to capture all important
5-31
-------
effects from air pollution exposure" (NRC, 2002, p. 108). NRC further notes that "the overall
effect estimates may be a combination of effects from long-term exposure plus some fraction
from short-term exposure. The amount of overlap is unknown" (NRC, 2002, p. 108-9). More
specifically, the SAB recommended emphasis on the ACS study because it includes a much
larger sample size and longer exposure interval and covers more locations (e.g., 50 cities
compared to the Six Cities Study) than other studies of its kind. Because of the refinements in
the extended follow-up analysis, the SAB-HES recommended using the Pope et al. (2002) study
as the basis for the main mortality estimate for adults and suggests that alternate estimates of
mortality generated using other cohort and time-series studies could be included as part of the
sensitivity analysis (U.S. EPA-SAB, 2004a). In 2009, the SAB-HES again reviewed the choice of
mortality risk coefficients for benefits analysis, concluding that "[t]he Krewski et al. (2009)
findings, while informative, have not yet undergone the same degree of peer review as have
the aforementioned studies. Thus, the HES recommends that EPA not use the Krewski et al.
(2009) findings for generating the Primary Estimate" (U.S. EPA-SAB, 2010a). Since this time, the
Krewski et al. (2009) has undergone additional peer review, which we believe strengthens the
support for including this study in this RIA. For example, the PM ISA (U.S. EPA, 2009b), which
was twice reviewed by Clean Air Scientific Review Committee (CASAC) (U.S. EPA-SAB, 2009b,
2009c), included this study among the key mortality studies. In addition, the risk assessment
supporting the PM NAAQS (U.S. EPA, 2010b) utilized risk coefficients drawn from the Krewski et
al. (2009) study, the most recent reanalysis of the ACS cohort data. The risk assessment cited a
number of advantages that informed the selection of the Krewski et al. (2009) study as the
source of the core effect estimates, including the extended period of observation, the rigorous
examination of model forms and effect estimates, the coverage for ecological variables, and the
large dataset with over 1.2 million individuals and 156 MSAs (U.S. EPA, 2010b). The CASAC also
provided extensive peer review of the risk assessment and supported the use of effect
estimates from this study (U.S. EPA-SAB, 2009a, 2010b,c).
As both the ACS and Six Cities studies have inherent strengths and weaknesses and the
expert elicitation results encompass within their range the estimates from both the earlier ACS
Study (Pope et al. 2002) and from the Laden et al. (2006) study, we present benefits estimates
using RR estimates from the Krewski et al. (2009) (RR=1.06, 95% confidence intervals 1.04-1.08
per 10u.g/m3 increase in PM2.5) and Laden et al. (2006; RR=1.16, 95% confidence intervals 1.07-
1.26 per 10 u.g/m3 increase in PM2.s) studies. For the ACS Study (Krewski et al., 2009), we use
the all-cause mortality risk estimate based on the random-effects Cox proportional hazard
model that incorporates 44 individual and 7 ecological co-variates, consistent with the
Quantitative Health Risk Assessment for Particulate Matter (U.S. EPA, 2010bJ. Unlike the Pope
5-32
-------
et al. (2002) study, Krewski et al. (2009) do not report a risk estimate based on an average
between the initial monitoring period (1979-1983) and the follow-up period (1999-2000).
When considering each time period from which we could select risk coefficients, we elected to
use the estimate based on the 1999-2000 air quality monitoring period because it reflected
more recent population exposures, a larger number of urban areas (116 vs. 58) and a larger
population cohort (488,000 vs. 343,000). The relative risk estimate (1.06 per 10u.g/m3 increase
in PM2.5) is identical to the risk estimate drawn from the Pope et al. (2002) study, though the
confidence interval around the Krewski et al. (2009) risk estimate is tighter.
Presenting results using both ACS and Six Cities is consistent with other recent RIAs (e.g.,
U.S. EPA, 2006a, 2010c, 2011c, 2011d). EPA's independent SAB also supported using these two
cohorts for benefits, concluding that "the selection of these cohort studies as the underlying
basis for PM mortality benefit estimates to be a good choice. These are widely cited, well
studied and extensively reviewed data sets" (U.S. EPA-SAB, 2010a).
In addition to the ACS and Six Cities cohorts described above, several recent cohort
studies conducted in North America provide evidence for the relationship between long-term
exposure to PM2.5 and the risk of premature death. Many of these additional cohort studies are
described in the PM ISA (U.S. EPA, 2009) (and thus not summarized here). We describe the
newer multi-state studies below.5'6 Table 5-8 provides the effect estimates from each of these
cohort studies (new and included in the PM ISA) for all-cause, cardiovascular, cardiopulmonary,
and ischemic heart disease (IHD) mortality as well as the lowest measured air quality level
(LML) in the study.
Puett et al. (2009) examined the risk of all-cause mortality and fatal congestive heart
disease among a cohort of about 66,000 female nurses in 13 northeastern and midwestern
states (i.e., the Nurses' Health Study cohort) resulting from long-term exposure to PM2.5.
Consistent with findings from previous cohort studies, the researchers found significant
associations between long-term PM2.5 exposure and all-cause mortality and fatal coronary heart
disease. Puett et al. (2011) examined the risk of all-cause and cardiovascular mortality among a
cohort of 17,000 male health professionals with high socioeconomic status in 13 northeastern
and midwestern states. The researchers found no association between long-term PM2.5
5 It is important to note that these newer studies have not been assessed in the context of an Integrated Science
Assessment nor gone through review by the SAB. In addition, only the ACS and H6C cohort studies have been
recommended by the SAB as appropriate for benefits analysis of national rulemakings.
6 In this chapter, we only describe multi-state cohort studies. There are additional cohorts that focus on single
cities, such as Gan et al. (2012) that we have not included. In Appendix 5B, we provide additional information
regarding cohort studies in California, which is the only state for which we identified single state cohorts.
5-33
-------
exposure and mortality, concluding that additional research is needed to determine whether
men with higher socioeconomic status are less susceptible to cardiovascular outcomes
associated with long-term particle exposure.
Table 5-8. Summary of Effect estimates from Associated With Change in Long-Term
Exposure to PM2.5 in Recent Cohort Studies in North America
Hazard Ratios per 10 u.g/ms Change in PM2.5
(95th percentile confidence intervals)
Study
Pope et al.
(2002)
Laden et al.
(2006)
Lipfert et al.
(2006)a
Miller et al.
(2007)b
Eftim et al.
(2008)
Zeger et al.
(2008)°
Krewski et al.
(2009)d
Puettetal.
(2009)b
Grouse et al.
(2011)d'f
Puettetal.
(2011)e
Lepeule et al.
(2012)d
^IVII_
Cohort (age) (u,g/ms)
ACS 7.5
(age >30)
Six Cities 10
(age > 25)
Veterans <14.1
(age 39-63)
WHI 3.4
(age 50-79)
Medicare 6
(age > 65)
Medicare <9.8
(age > 65)
ACS 5.8
(age >30)
NHS 5.8
(age 30-55)
Canadian 1.9
census
Health <14.4
Professionals
(age 40-75)
Six Cities 8
(age > 25)
All Causes
1.06
(1.02-1.11)
1.16
(1.07-1.26)
1.15
(1.05-1.25)
N/A
1.21
(1.15-1.27)
1.068
(1.049-1.087)
1.06
(1.04-1.08)
1.26
(1.02-1.54)
1.06
(1.01-1.10)
0.86
(0.70-1.00)
1.14
(1.07-1.22)
Cardiovascular
1.12
(1.08-1.15)
1.28
(1.13-1.44)
N/A
1.76
(1.25-2.47)
N/A
N/A
N/A
N/A
N/A
1.02
(0.84-1.23)
1.26
(1.14-1.40)
Cardiopulmonary
1.09
(1.03-1.16)
N/A
N/A
N/A
N/A
N/A
1.13
(1.10-1.16)
N/A
N/A
N/A
N/A
IHD
N/A
N/A
N/A
2.21
(1.17-4.16)
N/A
N/A
1.24
(1.19-1.29)
2.02
(1.07-3.78)
N/A
N/A
N/A
Used traffic proximity as a surrogate of exposure.
b Women only.
c Reflects risks in the Eastern U.S. Risks in the Central U.S. were higher, but the authors found no association in
the Western U.S.
d Random effects Cox model with individual and ecologic covariates.
Men with high socioeconomic status only.
Canadian population.
5-34
-------
Grouse et al. (2012) found elevated risks of non-accidental and ischemic heart disease
mortality associated with long-term exposure to PM2.5from the period of 1991 to 2001 among
a cohort of Canadian adults aged 25 and older. This study used a combination of monitored air
quality data and remote-sensing (i.e., satellite) data to assign PM2.5 concentrations to the
population cohort. Notably, the median annual mean PM2.5 levels observed, or modeled, in this
study were 7.4 u.g/m3 and the minimum value was 1.9 u.g/m3. The authors note that these air
quality values are significantly lower than those observed in either the ACS or Six Cities studies,
which provides further evidence that PM effects may occur at very low annual mean PM2.5
levels.
Lepeule et al. (2012) evaluated the sensitivity of previous Six Cities results to model
specifications, lower exposures, and averaging time using eleven additional years of cohort
follow-up that incorporated recent lower exposures. The authors found significant associations
between PM2.5 exposure and increased risk of all-cause, cardiovascular and lung cancer
mortality. The authors also concluded that the concentration-response relationship was linear
down to PM2.5 concentrations of 8 u.g/m3, and that mortality rate ratios for PM2.5 fluctuated
over time, but without clear trends, despite a substantial drop in the sulfate fraction.
As further described in the mortality valuation discussion, we assume that there is a
"cessation" lag between PM exposures and the total realization of changes in health effects.
While the structure of the lag is uncertain, most of the premature deaths occur within the first
couple years after the change in exposure (U.S. EPA, 2004c; Schwartz et al, 2008). Changes in
the cessation lag assumptions do not change the total number of estimated deaths but rather
the timing of those deaths.
In addition to the adult mortality studies described above, several studies provide
evidence for an association between PM exposure and respiratory inflammation and infection
leading to premature mortality in children under 5 years of age. Specifically, the SAB-HES noted
the release of the WHO Global Burden of Disease Study focusing on ambient air, which cites
several recently published time-series studies relating daily PM exposure to mortality in
children (U.S. EPA-SAB, 2004a). The SAB-HES also cites the study by Belanger et al. (2003) as
corroborating findings linking PM exposure to increased respiratory inflammation and
infections in children. Recently, a study by Chay and Greenstone (2003) found that reductions
in TSP caused by the recession of 1981- 1982 were related to reductions in infant mortality at
the county level. With regard to the cohort study conducted by Woodruff et al. (1997), the SAB-
HES notes several strengths of the study, including the use of a larger cohort drawn from a large
number of metropolitan areas and efforts to control for a variety of individual risk factors in
5-35
-------
infants (e.g., maternal educational level, maternal ethnicity, parental marital status, and
maternal smoking status). Based on these findings, the SAB-HES recommends that EPA
incorporate infant mortality into the primary benefits estimate and that infant mortality be
evaluated using an impact function developed from the Woodruff et al. (1997) study (U.S. EPA-
SAB, 2004a). A more recent study by Woodruff et al. (2006) continues to find associations
between PM2.5 and infant mortality. The study also found the most significant relationships with
respiratory-related causes of death. We have not yet sought comment from the SAB on this
more recent study and as such continue to rely on the earlier 1997 analysis.
5.6.3.2 Nonfatal Acute Myocardial Infarctions (AMI) (Heart Attacks)
Nonfatal heart attacks have been linked with short-term exposures to PM2.5 in the
United States (Mustafic et al., 2012; Peters et al., 2001; Sullivan et al., 2005; Pope et al., 2006;
Zanobetti and Schwartz, 2006; Zanobetti et al., 2009) and other countries (Poloniecki et al.,
1997; Barnett et al., 2006; Peters et al., 2005). In previous health impact assessments, we have
relied upon a study by Peters et al. (2001) as the basis for the impact function estimating the
relationship between PM2.5 and nonfatal heart attacks. The Peters et al. (2001) study exhibits a
number of strengths. In particular, it includes a robust characterization of populations
experiencing acute myocardial infarctions (AMIs). The researchers interviewed patients within 4
days of their AMI events and, for inclusion in the study, patients were required to meet a series
of criteria including minimum kinase levels, an identifiable onset of pain or other symptoms and
the ability to indicate the time, place and other characteristics of their AMI pain in an interview.
Since the publication of Peters et al. (2001), a number of other single and multi-city
studies have appeared in the literature. These studies include Sullivan et al. (2005), which
considered the risk of PM2.5-related hospitalization for AMIs in King County, WA; Pope et al.
(2006), based in Wasatch Range, UT; Zanobetti and Schwartz (2006), based in Boston; and,
Zanobetti et al. (2009), a multi-city study of 26 U.S. communities. Each of these single and
multi-city studies, with the exception of Pope et al. (2006), measure AMIs using hospital
discharge rates. Conversely, the Pope et al. (2006) study is based on a large registry with
angiographically characterized patients—arguably a more precise indicator of AMI. Because the
Pope et al. (2006) study reflected both myocardial infarctions and unstable angina, this
produces a more comprehensive estimate of acute ischemic heart disease events than the
other studies. However, unlike the Peters study (Peters et al., 2006), Pope and colleagues did
not measure the time of symptom onset, and PM2.5 data were not measured on an hourly basis.
5-36
-------
As a means of recognizing the strengths of the Peters study while also incorporating the
newer evidence found in the four single and multi-city studies, we present a range of AMI
estimates. The upper end of the range is calculated using the Peters study while the lower end
of the range is the result of an equal-weights pooling of these four newer studies. It is
important to note that when calculating the incidence of nonfatal AMI, the fraction of fatal
heart attacks is subtracted to ensure that there is no double-counting with premature mortality
estimates. Specifically, based on Rosamond et al. (1999), we apply an adjustment factor of 0.93
in the concentration-response function to reflect the probability of survival 28 days after the
heart attack.
5.6.3.3 Hospital Admissions and Emergency Department Visits
Because of the availability of detailed hospital admission and discharge records, there is
an extensive body of literature examining the relationship between hospital admissions and air
pollution. For this reason, we pool together the incidence estimates using several different
studies for many of the hospital admission endpoints. In addition, some studies have examined
the relationship between air pollution and emergency department visits. Since most emergency
department visits do not result in an admission to the hospital (i.e., most people going to the
emergency department are treated and return home), we treat hospital admissions and
emergency department visits separately, taking account of the fraction of emergency
department visits that are admitted to the hospital. Specifically, within the baseline incidence
rates, we parse out the scheduled hospital visits from unscheduled ones as well as the hospital
visits that originated in the emergency department.
The two main groups of hospital admissions estimated in this analysis are respiratory
admissions and cardiovascular admissions. There is not much evidence linking PM2.5 with other
types of hospital admissions. Both asthma- and cardiovascular-related visits have been linked to
PM2.s in the United States, though as we note below, we are able to assign an economic value
to asthma-related events only. To estimate the effects of PM2.5 air pollution reductions on
asthma-related ER visits, we use the effect estimate from a study of children 18 and under by
Mar et al. (2010) and Slaughter et al. (2005). Both studies examine populations 0 to 99 in
Washington State. Mar and colleagues perform their study in Tacoma, while Slaughter and
colleagues base their study in Spokane. We apply random/fixed effects pooling to combine
evidence across these two studies.
To estimate avoided incidences of cardiovascular hospital admissions associated with
PM2.5, we used studies by Moolgavkar (2000), Zanobetti et al. (2009), Peng et al. (2008, 2009)
5-37
-------
and Bell et al., (2008). Only Moolgavkar (2000) provided a separate effect estimate for adults 20
to 64, while the remainder estimate risk among adults over 64.7 Total cardiovascular hospital
admissions are thus the sum of the pooled estimate for adults over 65 and the single study
estimate for adults 20 to 64. Cardiovascular hospital admissions include admissions for
myocardial infarctions. To avoid double-counting benefits from reductions in myocardial
infarctions when applying the impact function for cardiovascular hospital admissions, we first
adjusted the baseline cardiovascular hospital admissions to remove admissions for myocardial
infarctions. We applied equal weights pooling to the multi-city studies assessing risk among
adults over 64 because these studies already incorporated pooling across the city-level
estimates. One potential limitation of our approach is that while the Zanobetti et al. (2009)
study assesses all cardiovascular risk, Bell et al. (2008), and Peng et al., (2008, 2009) studies
estimate a subset of cardiovascular hospitalizations as well as certain cerebro- and peripheral-
vascular diseases. To address the potential for the pooling of these four studies to produce a
biased estimate, we match the pooled risk estimate with a baseline incidence rate that excludes
cerebro- and peripheral-vascular disease. An alternative approach would be to use the
Zanobetti et al. (2009) study alone, though this would prevent us from drawing upon the
strengths of the three multi-city studies.
To estimate avoided incidences of respiratory hospital admissions associated with PM2.5,
we used a number of studies examining total respiratory hospital admissions as well as asthma
and chronic lung disease. We estimated impacts among three age groups: adults over 65, adults
18 to 64 and children 0 to 17. For adults over 65, the multi-city study by Zanobetti et al. (2009)
provides an effect estimate for total respiratory hospital admissions (defined as ICD codes 460-
519). Moolgavkar et al. (2003) examines PM2.5 and chronic lung disease hospital admissions
(less asthma) in Los Angeles, CA among adults 18 to 64. For children 0 to 18, we pool two
studies using random/fixed effects. The first is Babin et al. (2007) which assessed PM2.5 and
asthma hospital admissions in Washington, DC among children 1 to 18; we adjusted the age
range for this study to apply to children 0 to 18. The second is Sheppard et al. (2003) which
assessed PM2.5and asthma hospitalizations in Seattle, Washington, among children 0 to 18.
7 Note that the Moolgavkar (2000) study has not been updated to reflect the more stringent GAM convergence
criteria. However, given that no other estimates are available for this age group, we chose to use the existing
study. Given the very small (<5%) difference in the effect estimates for people 65 and older with cardiovascular
hospital admissions between the original and reanalyzed results, we do not expect this choice to introduce
much bias. For a discussion of the GAM convergence criteria, and how it affected the size of effect coefficients
reported by time series epidemiological studies using NMMAPS data, see:
http://www.healtheffects.org/Pubs/st-timeseries.htm.
5-38
-------
5.6.3.4 Acute Health Events and School/Work Loss Days
In addition to mortality, chronic illness, and hospital admissions, a number of acute
health effects not requiring hospitalization are associated with exposure to PM2.5. The sources
for the effect estimates used to quantify these effects are described below.
Approximately 4% of U.S. children between the ages of 5 and 17 experience episodes of
acute bronchitis annually (ALA, 2002). Acute bronchitis is characterized by coughing, chest
discomfort, slight fever, and extreme tiredness, lasting fora number of days. According to the
MedlinePlus medical encyclopedia,8 with the exception of cough, most acute bronchitis
symptoms abate within 7 to 10 days. Incidence of episodes of acute bronchitis in children
between the ages of 5 and 17 were estimated using an effect estimate developed from Dockery
et al. (1996). Incidences of lower respiratory symptoms (e.g., wheezing, deep cough) in children
aged 7 to 14 were estimated using an effect estimate from Schwartz and Neas (2000).
Because asthmatics have greater sensitivity to stimuli (including air pollution), children
with asthma can be more susceptible to a variety of upper respiratory symptoms (e.g., runny or
stuffy nose; wet cough; and burning, aching, or red eyes). Research on the effects of air
pollution on upper respiratory symptoms has thus focused on effects in asthmatics. Incidences
of upper respiratory symptoms in asthmatic children aged 9 to 11 are estimated using an effect
estimate developed from Pope et al. (1991).
Health effects from air pollution can also result in missed days of work (either from
personal symptoms or from caring for a sick family member). Days of work lost due to PM2.5
were estimated using an effect estimate developed from Ostro (1987). Children may also be
absent from school because of respiratory or other diseases caused by exposure to air
pollution, but we have not quantified these effects for this rule.
Minor restricted activity days (MRAD) result when individuals reduce most usual daily
activities and replace them with less strenuous activities or rest, yet not to the point of missing
work or school. For example, a mechanic who would usually be doing physical work most of the
day will instead spend the day at a desk doing paper and phone work because of difficulty
breathing or chest pain. The effect of PM2.5 on MRAD was estimated using an effect estimate
derived from Ostro and Rothschild (1989).
More recently published literature examining the relationship between short-term PM2.5
exposure and acute respiratory symptoms was available in the PM ISA (U.S. EPA, 2009), but
See http://www.nlm.nih.gov/medlineplus/ency/article/001087.htm, accessed April 2012.
5-39
-------
proved to be unsuitable for use in this benefits analysis. In particular, the best available study
(Patel et al., 2010) specified a population aged 13-20, which overlaps with the population in
which we asses asthma exacerbation. As we describe in detail below, to avoid the chance of
double-counting impacts, we do not estimate changes in acute respiratory symptoms and
asthma exacerbation among populations of the same age.
For this RIA, we have followed the SAB-HES recommendations regarding asthma
exacerbations in developing the main estimate (U.S. EPA-SAB, 2004a). While certain studies of
acute respiratory events characterize these impacts among only asthmatic populations, others
consider the full population, including both asthmatics and non-asthmatics. For this reason,
incidence estimates derived from studies focused only on asthmatics cannot be added to
estimates from studies that consider the full population—to do so would double-count impacts.
To prevent such double-counting, we estimated the exacerbation of asthma among children
and excluded adults from the calculation. Asthma exacerbations occurring in adults are
assumed to be captured in the general population endpoints such as work loss days and
MRADs. Finally, note also the important distinction between the exacerbation of asthma among
asthmatic populations, and the onset of asthma among populations not previously suffering
from asthma; in this RIA, we quantify the exacerbation of asthma among asthmatic populations
and not the onset of new cases of asthma.
To characterize asthma exacerbations in children, we selected two studies (Ostro et al.,
2001; Mar et al., 2004) that followed panels of asthmatic children. Ostro et al. (2001) followed a
group of 138 African-American children in Los Angeles for 13 weeks, recording daily
occurrences of respiratory symptoms associated with asthma exacerbations (e.g., shortness of
breath, wheeze, and cough). This study found a statistically significant association between
PM2.5, measured as a 12-hour average, and the daily prevalence of shortness of breath and
wheeze endpoints. Although the association was not statistically significant for cough, the
results were still positive and close to significance; consequently, we decided to include this
endpoint, along with shortness of breath and wheeze, in generating incidence estimates (see
below).
Mar et al. (2004) studied the effects of various size fractions of particulate matter on
respiratory symptoms of adults and children with asthma, monitored over many months. The
study was conducted in Spokane, Washington, a semiarid city with diverse sources of
particulate matter. Data on respiratory symptoms and medication use were recorded daily by
the study's subjects, while air pollution data was collected by the local air agency and
Washington State University. Subjects in the study consisted of 16 adults—the majority of
5-40
-------
whom participated for over a year—and nine children, all of whom were studied for over eight
months. Among the children, the authors found a strong association between cough symptoms
and several metrics of particulate matter, including PM2.5. However, the authors found no
association between respiratory symptoms and PM of any metric in adults. Mar et al. therefore
concluded that the discrepancy in results between children and adults was due either to the
way in which air quality was monitored, or a greater sensitivity of children than adults to
increased levels of PM air pollution.
We employed the following pooling approach in combining estimates generated using
effect estimates from the two studies to produce a single asthma exacerbation incidence
estimate. First, we used random/fixed effects pooling to combine the Ostro and Mar estimates
for shortness of breath and cough. Next, we pooled the Ostro estimate of wheeze with the
pooled cough and shortness of breath estimates to derive an overall estimate of asthma
exacerbation.
5.6.3.5 Effect Coefficients Selected for the Sensitivity Analyses
Chronic Bronchitis. Chronic bronchitis is characterized by mucus in the lungs and a
persistent wet cough for at least 3 months a year for several years in a row. Chronic bronchitis
affects an estimated 5% of the U.S. population (ALA, 1999). A limited number of studies have
estimated the impact of air pollution on new incidences of chronic bronchitis. Schwartz (1993)
and Abbey et al. (1995) provide evidence that long-term PM2.5 exposure gives rise to the
development of chronic bronchitis in adults in the United States; these remain the two most
recent studies observing a relationship between long-term exposure to PM2.5 and the onset of
chronic bronchitis in the U.S. The absence of newer studies finding a relationship between long-
term PM2.s exposure and chronic bronchitis argues for moving this endpoint from the main
benefits analysis to a sensitivity analysis. In their review of the scientific literature on chronic
obstructive pulmonary disease (COPD), which includes chronic bronchitis and emphysema, the
American Thoracic Society concluded that air pollution is "associated with COPD, but sufficient
criteria for causation were not met" (Eisner et al., 2010).
Stroke. The PM ISA (U.S. EPA, 2009) includes several new studies that have examined
the relationship between PM2.5 exposure and cerebrovascular events (U.S. EPA, 2009). Time-
series studies have generally been inconsistent with several studies showing positive
associations (Dominici et al., 2006; Metzger et al., 2004; Lippman et al., 2000; Lisabeth et al.,
2008). Several other studies have demonstrated null or negative associations (Anderson et al.,
2001; Barnett et al., 2006; Peel et al., 2007). In general, these studies examined cerebrovascular
5-41
-------
disease as a group, though a few studies partition ischemic and hemorrhagic strokes separately
(Lisabeth et al., 2008). A key limitation of these time-series studies is that they use hospital
discharge rates as the diagnosis and relatively short lags (0-2 days)—this is problematic, as
discharge rates are an imperfect diagnosis and strokes may occur several days before admission
to the hospital.
Longer-term prospective cohort studies of PM2.5 and stroke include Miller et al. (2007),
which estimated the change in risk among post-menopausal women enrolled in the Women's
Health Initiative (U.S. EPA, 2009b). After adjusting for age, race, smoking status, educational
level, household income, body-mass index, diabetes, hypertension, and hypercholesterolemia,
hazard ratios were estimated for the first cardiovascular event. Because this study considers
first-time cardiovascular events, a key challenge to incorporating this study into a health impact
assessment would be to match the baseline incidence rate correctly.
Three factors argue for treating this endpoint in the sensitivity analysis: (1) the
epidemiological literature examining PM-related cerebrovascular events is still evolving; (2)
there are special uncertainties associated with quantifying this endpoint; (3) we have not yet
identified an appropriate method for estimating the economic value of this endpoint.
Cardiovascular Emergency Department Visits. A large number of recent U.S.-based
studies provide support for an association between short-term increases in PM2.5 and increased
risk of ED visits for ischemic heart diseases (U.S. EPA, 2009b). Both Metzger et al. (2004) and
Tolbert et al. (2007) published interim results from the Study of Particles and Health in Atlanta
(SOPHIA), finding a relationship between PM2.5 exposure and cardiovascular emergency
department visits. These cardiovascular emergency department visits are distinct from
cardiovascular hospital admissions and non-fatal heart attacks. To ensure no double-counting,
we excluded ICD-9-411 (ischemic heart disease) from the baseline incidence rates for
cardiovascular emergency department visits. The principal challenge to incorporating these
studies is the absence of readily-available economic valuation estimates for cardiovascular
emergency department visits. Until we develop an approach for estimating the economic value
of this endpoint, we will treat these ED visits as a sensitivity analysis.
5.6.4 Unqualified Human Health Benefits
Implementing the illustrative control strategy described in Chapter 4 would reduce
emissions of directly emitted particles, S02, and NOX. Although we have quantified many of the
health benefits associated with reducing exposure to PM2.5, as shown in Table 5-2, we are
5-42
-------
unable to quantify the health benefits associated with reducing ozone exposure, S02 exposure,
N02 exposure or methylmercury exposure due to the absence of air quality modeling data for
these pollutants in this analysis. Although we applied the rollback method to simulate the
impact of attaining alternative combination of standard levels on ambient levels of PM2.5, this
method does not simulate how the illustrative emission reductions would affect ambient levels
of ozone, S02, or N02. Furthermore, the air quality modeling conducted for this analysis did not
assess mercury, so we are unable to estimate mercury deposition associated with the
illustrative controls or subsequent bioaccumulation and exposure. Below we provide a
qualitative description these health benefits. In general, previous analyses have shown that the
monetized value of these additional health benefits is much smaller than PM2.5-related benefits
(U.S. EPA, 2010a, 2010c, 2010d). The extent to which ozone, S02, NOx, and/or methylmercury
would be reduced would depend on the specific control strategy used to reduce PM2.5 in a given
area.
Reducing NOX emissions also reduces ozone concentrations in most areas. Reducing
ambient ozone concentrations is associated with significant human health benefits, including
mortality and respiratory morbidity (U.S. EPA, 2008a, 2010d). Epidemiological researchers have
associated ozone exposure with adverse health effects in numerous toxicological, clinical and
epidemiological studies (U.S. EPA, 2006b). When adequate data and resources are available,
EPA generally quantifies several health effects associated with exposure to ozone (e.g., U.S.
EPA, 2008a, 2010d, 2011a, 2011c). These health effects include respiratory morbidity such as
asthma attacks, hospital and emergency department visits, school loss days, as well as
premature mortality. The scientific literature suggests that exposure to ozone is also associated
with chronic respiratory damage and premature aging of the lungs, but EPA has not quantified
these effects in benefits analyses previously.
Following an extensive evaluation of health evidence from epidemiologic and laboratory
studies, the Integrated Science Assessment for Sulfur Dioxide—Health Criteria (S02 ISA)
concluded that there is a causal relationship between respiratory health effects and short-term
exposure to S02(U.S. EPA, 2008c). The immediate effect of S02on the respiratory system in
humans is bronchoconstriction. Asthmatics are more sensitive to the effects of S02 likely
resulting from preexisting inflammation associated with this disease. A clear concentration-
response relationship has been demonstrated in laboratory studies following exposures to S02
at concentrations between 20 and 100 ppb, both in terms of increasing severity of effect and
percentage of asthmatics adversely affected. Based on our review of this information, we
identified four short-term morbidity endpoints that the S02 ISA identified as a "causal
5-43
-------
relationship": asthma exacerbation, respiratory-related emergency department visits, and
respiratory-related hospitalizations. The differing evidence and associated strength of the
evidence for these different effects is described in detail in the S02 ISA. The S02 ISA also
concluded that the relationship between short-term S02 exposure and premature mortality was
"suggestive of a causal relationship" because it is difficult to attribute the mortality risk effects
to S02 alone. Although the S02 ISA stated that studies are generally consistent in reporting a
relationship between S02 exposure and mortality, there was a lack of robustness of the
observed associations to adjustment for pollutants. We did not quantify these benefits due to
data constraints.
Epidemiological researchers have associated N02 exposure with adverse health effects
in numerous toxicological, clinical and epidemiological studies, as described in the Integrated
Science Assessment for Oxides of Nitrogen—Health Criteria (N02 ISA) (U.S. EPA, 2008b). The
N02 ISA provides a comprehensive review of the current evidence of health and environmental
effects of N02. The N02 ISA concluded that the evidence "is sufficient to infer a likely causal
relationship between short-term N02 exposure and adverse effects on the respiratory system."
These epidemiologic and experimental studies encompass a number of endpoints including
emergency department visits and hospitalizations, respiratory symptoms, airway
hyperresponsiveness, airway inflammation, and lung function. Effect estimates from
epidemiologic studies conducted in the United States and Canada generally indicate a 2-20%
increase in risks for ED visits and hospital admissions and higher risks for respiratory symptoms.
The N02 ISA concluded that the relationship between short-term N02 exposure and premature
mortality was "suggestive but not sufficient to infer a causal relationship" because it is difficult
to attribute the mortality risk effects to N02 alone. Although the N02 ISA stated that studies
consistently reported a relationship between N02 exposure and mortality, the effect was
generally smaller than that for other pollutants such as PM. We did not quantify these benefits
due to data constraints.
5.6.5 Economic Valuation Estimates
Reductions in ambient concentrations of air pollution generally lower the risk of future
adverse health effects for a large population. Therefore, the appropriate economic measure is
WTP for changes in risk of a health effect rather than WTP for a health effect that would occur
with certainty (Freeman, 1993). Epidemiological studies generally provide estimates of the
relative risks of a particular health effect that is avoided because of a reduction in air pollution.
We converted those changes in risk to units of avoided statistical incidence for ease of
5-44
-------
presentation. We calculated the value of avoided statistical incidences by dividing individual
WTP for a risk reduction by the related observed change in risk.
WTP estimates generally are not available for some health effects, such as hospital
admissions. In these cases, we instead used the cost of treating or mitigating the effect to
estimate the economic value. These cost-of-illness (COI) estimates generally understate the
true value of reducing the risk of a health effect, because they reflect the direct expenditures
related to treatment, but not the value of avoided pain and suffering (Harrington and Portney,
1987; Berger, 1987). We provide unit values for health endpoints (along with information on
the distribution of the unit value) in Table 5-9. All values are in constant year 2006 dollars,
adjusted for growth in real income out to 2020 using projections provided by Standard and
Poor's. Economic theory argues that WTP for most goods (such as environmental protection)
will increase if real income increases. Several of the valuation studies used in this analysis were
conducted in the late 1980s and early 1990s, and we are in the process of reviewing the
literature to update these unit values. Because real income has grown since the studies were
conducted, people's willingness to pay for reductions in the risk of premature death and
disease likely has grown as well. We do not have data to adjust the COI estimates for
projections of medical costs in the future, which leads to an inherent though unavoidable
inconsistency between COI- and WTP-based estimates. For these two reasons, these cost-of-
illness estimates may underestimate the economic value of avoided health impacts. The
discussion below provides additional details on valuing PM2.5-related related endpoints.
5-45
-------
Table 5-9. Unit Values for Economic Valuation of Health Endpoints (2006$)a
en
Central Estimate of Value Per Statistical Incidence
Health Endpoint
Premature Mortality (Value of a
Statistical Life)
Nonfatal Myocardial Infarction
(heart attack)
3% discount rate
Age 0-24
Age 25-44
Age 45-54
Age 55-65
Age 66 and over
//o discount rote
Age 0-24
Age 25-44
Age 45-54
Age 55-64
Age 65 and over
1990 Income Level
$6,300,000
$85,000
$96,000
$100,000
$180,000
$85,000
$84,000
$94,000
$98,000
$170,000
$84,000
2020 Income Level
$8,900,000
$85,000
$96,000
$100,000
$180,000
$85,000
$84,000
$94,000
$98,000
$170,000
$84,000
Derivation of Distributions of Estimates
EPA currently recommends a central VSL of $4.8m (1990$, 1990
income) based on a Weibull distribution fitted to 26 published VSL
estimates (5 contingent valuation and 21 labor market studies). The
underlying studies, the distribution parameters, and other useful
information are available in Appendix B of EPA's Guidelines for
Preparing Economic Analyses (U.S. EPA, 2010e).
No distributional information available. Age-specific cost-of-illness
values reflect lost earnings and direct medical costs over a 5-year
period following a nonfatal Ml. Lost earnings estimates are based on
Cropper and Krupnick (1990). Direct medical costs are based on
simple average of estimates from Russell et al. (1998) and Wittels et
al. (1990).
Lost earnings:
Cropper and Krupnick (1990). Present discounted value of 5 years of
lost earnings in 2000$:
age of onset: at 3% at 7%
25-44 $9,000 $8,000
45-54 $13,000 $12,000
55-65 $77,000 $69,000
Direct medical expenses (2000$): An average of:
1. Wittels et al. (1990) ($100,000— no discounting)
2. Russell et al. (1998), 5-year period ($22,000 at 3% discount
rate; $21,000at 7% discount rate)
Hospital Admissions
Chronic Lung Disease (18-64)
$19,000
$19,000 No distributional information available. The COI estimates (lost
earnings plus direct medical costs) are based on ICD-9 code-level
information (e.g., average hospital care costs, average length of
hospital stay, and weighted share of total chronic lung illnesses)
reported in Agency for Healthcare Research and Quality (2007)
(www.ahrq.gov).
(continued)
-------
Table 5-9. Unit Values for Economic Valuation of Health Endpoints (2006$)a (continued)
Central Estimate of Value Per Statistical Incidence
Health Endpoint
2000 Income Level
2020 Income Level
Derivation of Distributions of Estimates
Hospital Admissions (continued)
Asthma Admissions (0-64)
$14,000
$14,000
All Cardiovascular
Age 18-64
Age 65-99
All respiratory (ages 65+)
$37,000
$35,000
$32,000
$37,000
$35,000
$32,000
Emergency Department Visits
for Asthma
$370
$370
No distributional information available. The COI estimates (lost
earnings plus direct medical costs) are based on ICD-9 code-level
information (e.g., average hospital care costs, average length of
hospital stay, and weighted share of total asthma category illnesses)
reported in Agency for Healthcare Research and Quality (2007)
(www.ahrq.gov).
No distributional information available. The COI estimates (lost
earnings plus direct medical costs) are based on ICD-9 code-level
information (e.g., average hospital care costs, average length of
hospital stay, and weighted share of total cardiovascular category
illnesses) reported in Agency for Healthcare Research and Quality
(2007) (www.ahrq.gov).
No distributions available. The COI point estimates (lost earnings plus
direct medical costs) are based on ICD-9 code level information (e.g.,
average hospital care costs, average length of hospital stay, and
weighted share of total respiratory category illnesses) reported in
Agency for Healthcare Research and Quality, 2007 (www.ahrq.gov).
No distributional information available. Simple average of two unit
COI values (2000$):
(1) $310, from Smith et al. (1997) and
(2) $260, from Stanford et al. (1999).
(continued)
-------
Table 5-9. Unit Values for Economic Valuation of Health Endpoints (2006$)a (continued)
Central Estimate of Value Per Statistical Incidence
Health Endpoint
2000 Income Level
2020 Income Level
Derivation of Distributions of Estimates
Respiratory Ailments Not Requiring Hospitalization
Upper Respiratory Symptoms
(URS)
$25
$31
Lower Respiratory Symptoms
(LRS)
$16
$19
CO
Asthma Exacerbations
$43
$53
Combinations of the three symptoms for which WTP estimates are
available that closely match those listed by Pope et al. result in seven
different "symptom clusters," each describing a "type" of URS. A
dollar value was derived for each type of URS, using mid-range
estimates of WTP (lEc, 1994) to avoid each symptom in the cluster
and assuming additivity of WTPs. In the absence of information
surrounding the frequency with which each of the seven types of
URS occurs within the URS symptom complex, we assumed a uniform
distribution between $9.2 and $43 (2000$).
Combinations of the four symptoms for which WTP estimates are
available that closely match those listed by Schwartz et al. result in
11 different "symptom clusters," each describing a "type" of LRS. A
dollar value was derived for each type of LRS, using mid-range
estimates of WTP (lEc, 1994) to avoid each symptom in the cluster
and assuming additivity of WTPs. The dollar value for LRS is the
average of the dollar values for the 11 different types of LRS. In the
absence of information surrounding the frequency with which each
of the 11 types of LRS occurs within the LRS symptom complex, we
assumed a uniform distribution between $6.9 and $25 (2000$).
Asthma exacerbations are valued at $45 per incidence, based on the
mean of average WTP estimates for the four severity definitions of a
"bad asthma day," described in Rowe and Chestnut (1986). This
study surveyed asthmatics to estimate WTP for avoidance of a "bad
asthma day," as defined by the subjects. For purposes of valuation,
an asthma exacerbation is assumed to be equivalent to a day in
which asthma is moderate or worse as reported in the Rowe and
Chestnut (1986) study. The value is assumed have a uniform
distribution between $16 and $71 (2000$).
(continued)
-------
Table 5-9. Unit Values for Economic Valuation of Health Endpoints (2006$)a (continued)
Central Estimate of Value Per Statistical Incidence
Health Endpoint
2000 Income Level
2020 Income Level
Derivation of Distributions of Estimates
Respiratory Ailments Not Requiring Hospitalization (continued)
Acute Bronchitis
$360
$440
ID
Work Loss Days (WLDs)
Minor Restricted Activity Days
(MRADs)
Variable (U.S. median
$140)
$51
Variable (U.S. median
$140)
$63
Assumes a 6-day episode, with the distribution of the daily value
specified as uniform with the low and high values based on those
recommended for related respiratory symptoms in Neumann et al.
(1994). The low daily estimate of $10 is the sum of the mid-range
values recommended by lEc (1994) for two symptoms believed to be
associated with acute bronchitis: coughing and chest tightness. The
high daily estimate was taken to be twice the value of a minor
respiratory restricted-activity day, or $110 (2000$).
No distribution available. Point estimate is based on county-specific
median annual wages divided by 52 and then by 5—to get median
daily wage. U.S. Year 2000 Census, compiled by Geolytics, Inc.
(Geolytics, 2002)
Median WTP estimate to avoid one MRAD from Tolley et al. (1986).
Distribution is assumed to be triangular with a minimum of $22 and a
maximum of $83, with a most likely value of $52 (2000$). Range is
based on assumption that value should exceed WTP for a single mild
symptom (the highest estimate for a single symptom—for eye
irritation—is $16.00) and be less than that for a WLD. The triangular
distribution acknowledges that the actual value is likely to be closer
to the point estimate than either extreme.
All estimates are rounded to two significant digits. Unrounded estimates in 2000$ are available in the Appendix J of the BenMAP user manual (Abt
Associates, 2011).
-------
5.6.6.1 Mortality Valuation
Following the advice of the EEAC of the SAB, EPA currently uses the value of statistical
life (VSL) approach in calculating the main estimate of mortality benefits, because we believe
this calculation provides the most reasonable single estimate of an individual's willingness to
trade off money for reductions in mortality risk (U.S. EPA-SAB, 2000). The VSL approach is a
summary measure for the value of small changes in mortality risk experienced by a large
number of people. Fora period of time (2004-2008), the Office of Air and Radiation (OAR)
valued mortality risk reductions using a VSL estimate derived from a limited analysis of some of
the available studies. OAR arrived at a VSL using a range of $1 million to $10 million (2000$)
consistent with two meta-analyses of the wage-risk literature. The $1 million value represented
the lower end of the interquartile range from the Mrozek and Taylor (2002) meta-analysis of 33
studies. The $10 million value represented the upper end of the interquartile range from the
Viscusi and Aldy (2003) meta-analysis of 43 studies. The mean estimate of $5.5 million (2000$)
was also consistent with the mean VSL of $5.4 million estimated in the Kochi et al. (2006) meta-
analysis. However, the Agency neither changed its official guidance on the use of VSL in rule-
makings nor subjected the interim estimate to a scientific peer-review process through the
Science Advisory Board (SAB) or other peer-review group.
During this time, the Agency continued work to update its guidance on valuing mortality
risk reductions, including commissioning a report from meta-analytic experts to evaluate
methodological questions raised by EPA and the SAB on combining estimates from the various
data sources. In addition, the Agency consulted several times with the Science Advisory Board
Environmental Economics Advisory Committee (SAB-EEAC) on the issue. With input from the
meta-analytic experts, the SAB-EEAC advised the Agency to update its guidance using specific,
appropriate meta-analytic techniques to combine estimates from unique data sources and
different studies, including those using different methodologies (i.e., wage-risk and stated
preference) (U.S. EPA-SAB, 2007).
Until updated guidance is available, the Agency determined that a single, peer-reviewed
estimate applied consistently best reflects the SAB-EEAC advice it has received. Therefore, the
Agency has decided to apply the VSL that was vetted and endorsed by the SAB in the Guidelines
for Preparing Economic Analyses (U.S. EPA, 2000)1 while the Agency continues its efforts to
update its guidance on this issue. This approach calculates a mean value across VSL estimates
1 In the updated Guidelines for Preparing Economic Analyses (U.S. EPA, 2010e), EPA retained the VSL endorsed by
the SAB with the understanding that further updates to the mortality risk valuation guidance would be
forthcoming in the near future.
5-50
-------
derived from 26 labor market and contingent valuation studies published between 1974 and
1991. The mean VSL across these studies is $4.8 million (1990$) or $6.3 million (2000$).2 The
Agency is committed to using scientifically sound, appropriately reviewed evidence in valuing
mortality risk reductions and has made significant progress in responding to the SAB-EEAC's
specific recommendations. In the process, the Agency has identified a number of important
issues to be considered in updating its mortality risk valuation estimates. These are detailed in a
white paper on "Valuing Mortality Risk Reductions in Environmental Policy," which recently
underwent review by EPA's independent Science Advisory Board). A meeting with the SAB on
this paper was held on March 14, 2011 and formal recommendations were transmitted on
July 29, 2011 (U.S. EPA-SAB, 2011). Draft guidance responding to SAB recommendations will be
developed shortly.
The economics literature concerning the appropriate method for valuing reductions in
premature mortality risk is still developing. The adoption of a value for the projected reduction
in the risk of premature mortality is the subject of continuing discussion within the economics
and public policy analysis community. EPA strives to use the best economic science in its
analyses. Given the mixed theoretical finding and empirical evidence regarding adjustments to
VSL for risk and population characteristics, we use a single VSL for all reductions in mortality
risk.
Although there are several differences between the labor market studies EPA uses to
derive a VSL estimate and the PM2.s air pollution context addressed here, those differences in
the affected populations and the nature of the risks imply both upward and downward
adjustments. Table 5-10 lists some of these differences and the expected effect on the VSL
estimate for air pollution-related mortality. In the absence of a comprehensive and balanced
set of adjustment factors, EPA believes it is reasonable to continue to use the $4.8 million
(1990$) value adjusted for inflation and income growth over time while acknowledging the
significant limitations and uncertainties in the available literature.
The SAB-EEAC has reviewed many potential VSL adjustments and the state of the
economics literature. The SAB-EEAC advised EPA to "continue to use a wage-risk-based VSL as
its primary estimate, including appropriate sensitivity analyses to reflect the uncertainty of
these estimates," and that "the only risk characteristic for which adjustments to the VSL can be
made is the timing of the risk" (U.S. EPA-SAB, 2000). In developing our main estimate of the
2 In this analysis, we adjust the VSL to account for a different currency year (2006$) and to account for income
growth to 2020. After applying these adjustments to the $6.3 million value, the VSL is $8.9M.
5-51
-------
benefits of premature mortality reductions, we have followed this advice. For premature
mortality, we assume that there is a "cessation" lag between PM exposures and the total
realization of changes in health effects. We assumed for this analysis that some of the
incidences of premature mortality related to PM2.5 exposures occur in a distributed fashion over
the 20 years following exposure and discounted over the period between exposure and
premature mortality. Although the structure of the lag is uncertain, EPA follows the advice of
the SAB-HES to assume a segmented lag structure characterized by 30% of mortality reductions
in the first year, 50% over years 2 to 5, and 20% over the years 6 to 20 after the reduction in
PM2.5 (U.S. EPA-SAB, 2004c). Additional cessation lag structures are described and assessed in
Appendix 5B of the RIA. To take this into account in the valuation of reductions in premature
mortality, we discount the value of premature mortality occurring in future years using rates of
3% and 7%.3 Changes in the cessation lag assumptions do not change the total number of
estimated deaths but rather the timing of those deaths. As such, the monetized benefits using a
7% discount rate are only approximately 10% less than the monetized benefits using a 3%
discount rate. Further discussion of this topic appears in EPA's Guidelines for Preparing
Economic Analyses (U.S. EPA, 2010e).
Table 5-10. Influence of Applied VSL Attributes on the Size of the Economic Benefits of
Reductions in the Risk of Premature Death (U.S. EPA, 2006a)
Attribute
Expected Direction of Bias
Age
Life Expectancy/Health Status
Attitudes Toward Risk
Income
Voluntary vs. Involuntary
Catastrophic vs. Protracted Death
Uncertain, perhaps overestimate
Uncertain, perhaps overestimate
Underestimate
Uncertain
Uncertain, perhaps underestimate
Uncertain, perhaps underestimate
Uncertainties Specific to Premature Mortality Valuation. The economic benefits
associated with reductions in the risk of premature mortality are the largest category of
monetized benefits of the CSAPR. In addition, in prior analyses, EPA has identified valuation of
mortality-related benefits as the largest contributor to the range of uncertainty in monetized
The choice of a discount rate, and its associated conceptual basis, is a topic of ongoing discussion within the
federal government. To comply with Circular A-4, EPA provides monetized benefits using discount rates of 3%
and 7% (OMB, 2003). A 3% discount reflects reliance on a "social rate of time preference" discounting concept.
A 7% rate is consistent with an "opportunity cost of capital" concept to reflect the time value of resources
directed to meet regulatory requirements.
5-52
-------
benefits (Mansfield et al., 2009).4 Because of the uncertainty in estimates of the value of
reducing premature mortality risk, it is important to adequately characterize and understand
the various types of economic approaches available for valuing reductions in mortality risk.
Such an assessment also requires an understanding of how alternative valuation approaches
reflect that some individuals may be more susceptible to air pollution-induced mortality or
reflect differences in the nature of the risk presented by air pollution relative to the risks
studied in the relevant economics literature.
The health science literature on air pollution indicates that several human
characteristics affect the degree to which mortality risk affects an individual. For example, some
age groups appear to be more susceptible to air pollution than others (e.g., the elderly and
children). Health status prior to exposure also affects susceptibility. An ideal benefits estimate
of mortality risk reduction would reflect these human characteristics, in addition to an
individual's WTP to improve one's own chances of survival plus WTP to improve other
individuals' survival rates. The ideal measure would also take into account the specific nature of
the risk reduction commodity that is provided to individuals, as well as the context in which risk
is reduced. To measure this value, it is important to assess how reductions in air pollution
reduce the risk of dying from the time that reductions take effect onward and how individuals
value these changes. Each individual's survival curve, or the probability of surviving beyond a
given age, should shift as a result of an environmental quality improvement. For example,
changing the current probability of survival for an individual also shifts future probabilities of
that individual's survival. This probability shift will differ across individuals because survival
curves depend on such characteristics as age, health state, and the current age to which the
individual is likely to survive.
Although a survival curve approach provides a theoretically preferred method for
valuing the benefits of reduced risk of premature mortality associated with reducing air
pollution, the approach requires a great deal of data to implement. The economic valuation
literature does not yet include good estimates of the value of this risk reduction commodity. As
a result, in this study we value reductions in premature mortality risk using the VSL approach.
This conclusion was based on an assessment of uncertainty based on statistical error in epidemiological effect
estimates and economic valuation estimates. Additional sources of model error such as those examined in the
PM2.5 mortality expert elicitation (Roman et al., 2008) may result in different conclusions about the relative
contribution of sources of uncertainty.
5-53
-------
Other uncertainties specific to premature mortality valuation include the following:
• Across-study variation: There is considerable uncertainty as to whether the available
literature on VSL provides adequate estimates of the VSLfor risk reductions from air
pollution reduction. Although there is considerable variation in the analytical designs
and data used in the existing literature, the majority of the studies involve the value
of risks to a middle-aged working population. Most of the studies examine
differences in wages of risky occupations, using a hedonic wage approach. Certain
characteristics of both the population affected and the mortality risk facing that
population are believed to affect the average WTP to reduce the risk. The
appropriateness of a distribution of WTP based on the current VSL literature for
valuing the mortality-related benefits of reductions in air pollution concentrations
therefore depends not only on the quality of the studies (i.e., how well they
measure what they are trying to measure), but also on the extent to which the risks
being valued are similar and the extent to which the subjects in the studies are
similar to the population affected by changes in pollution concentrations.
• Level of risk reduction: The transferability of estimates of the VSL from the wage-risk
studies to the context of the PM NAAQS analysis rests on the assumption that,
within a reasonable range, WTP for reductions in mortality risk is linear in risk
reduction. For example, suppose a study provides a result that the average WTP for
a reduction in mortality risk of 1/100,000 is $50, but that the actual mortality risk
reduction resulting from a given pollutant reduction is 1/10,000. If WTP for
reductions in mortality risk is linear in risk reduction, then a WTP of $50 for a
reduction of 1/100,000 implies a WTP of $500 for a risk reduction of 1/10,000
(which is 10 times the risk reduction valued in the study). Under the assumption of
linearity, the estimate of the VSL does not depend on the particular amount of risk
reduction being valued. This assumption has been shown to be reasonable provided
the change in the risk being valued is within the range of risks evaluated in the
underlying studies (Rowlatt et al., 1998).
• Voluntariness of risks evaluated: Although job-related mortality risks may differ in
several ways from air pollution-related mortality risks, the most important
difference may be that job-related risks are incurred voluntarily, or generally
assumed to be, whereas air pollution-related risks are incurred involuntarily. Some
evidence suggests that people will pay more to reduce involuntarily incurred risks
than risks incurred voluntarily. If this is the case, WTP estimates based on wage-risk
studies may understate WTP to reduce involuntarily incurred air pollution-related
mortality risks.
• Sudden versus protracted death: A final important difference related to the nature of
the risk may be that some workplace mortality risks tend to involve sudden,
catastrophic events, whereas air pollution-related risks tend to involve longer
periods of disease and suffering prior to death. Some evidence suggests that WTP to
5-54
-------
avoid a risk of a protracted death involving prolonged suffering and loss of dignity
and personal control is greater than the WTP to avoid a risk (of identical magnitude)
of sudden death. To the extent that the mortality risks addressed in this assessment
are associated with longer periods of illness or greater pain and suffering than are
the risks addressed in the valuation literature, the WTP measurements employed in
the present analysis would reflect a downward bias.
• Self-selection and skill in avoiding risk: Recent research (Shogren and Stamland,
2002) suggests that VSL estimates based on hedonic wage studies may overstate the
average value of a risk reduction. This is based on the fact that the risk-wage trade-
off revealed in hedonic studies reflects the preferences of the marginal worker (i.e.,
that worker who demands the highest compensation for his risk reduction). This
worker must have either a higher workplace risk than the average worker, a lower
risk tolerance than the average worker, or both. However, the risk estimate used in
hedonic studies is generally based on average risk, so the VSL may be upwardly
biased because the wage differential and risk measures do not match.
• Baseline risk and age: Recent research (Smith, Pattanayak, and Van Houtven, 2006)
finds that because individuals reevaluate their baseline risk of death as they age, the
marginal value of risk reductions does not decline with age as predicted by some
lifetime consumption models. This research supports findings in recent stated
preference studies that suggest only small reductions in the value of mortality risk
reductions with increasing age.
5.6.6.2 Nonfatal Myocardial Infarctions Valuation
We were not able to identify a suitable WTP value for reductions in the risk of nonfatal
heart attacks. Instead, we use a COI unit value with two components: the direct medical costs
and the opportunity cost (lost earnings) associated with the illness event. Because the costs
associated with a myocardial infarction extend beyond the initial event itself, we consider costs
incurred over several years. Using age-specific annual lost earnings estimated by Cropper and
Krupnick (1990) and a 3% discount rate, we estimated a rounded present discounted value in
lost earnings (in 2000$) over 5 years due to a myocardial infarction of $8,800 for someone
between the ages of 25 and 44, $13,000 for someone between the ages of 45 and 54, and
$75,000 for someone between the ages of 55 and 65. The rounded corresponding age-specific
estimates of lost earnings (in 2000$) using a 7% discount rate are $7,900, $12,000, and $67,000,
respectively. Cropper and Krupnick (1990) do not provide lost earnings estimates for
populations under 25 or over 65. As such, we do not include lost earnings in the cost estimates
for these age groups.
We found three possible sources in the literature of estimates of the direct medical
costs of myocardial infarction, which provide significantly different values (see Table 5-11):
5-55
-------
Table 5-11. Alternative Direct Medical Cost of Illness Estimates for Nonfatal Heart Attacks3
Study Direct Medical Costs (2006$) Over an x-Year Period, for x =
Wittels et al. (1990) $140,000b 5
Russell et al. (1998) $30,000° 5
Eisenstein et al. (2001) $64,000° 10
Russell et al. (1998) $38,000° 10
All estimates rounded to two significant digits. Unrounded estimates in 2000$ are available in appendix J of the
BenMAP user manual (Abt Associates, 2011).
Wittels et al. (1990) did not appear to discount costs incurred in future years.
° Using a 3% discount rate. Discounted values as reported in the study.
• Wittels et al. (1990) estimated expected total medical costs of myocardial infarction
over 5 years to be $51,000 (rounded in 1986$) for people who were admitted to the
hospital and survived hospitalization. (There does not appear to be any discounting
used.) This estimated cost is based on a medical cost model, which incorporated
therapeutic options, projected outcomes, and prices (using "knowledgeable
cardiologists" as consultants). The model used medical data and medical decision
algorithms to estimate the probabilities of certain events and/or medical procedures
being used. The authors note that the average length of hospitalization for acute
myocardial infarction has decreased over time (from an average of 12.9 days in 1980
to an average of 11 days in 1983). Wittels et al. used 10 days as the average in their
study. It is unclear how much further the length of stay for myocardial infarction
may have decreased from 1983 to the present. The average length of stay for ICD
code 410 (myocardial infarction) in the year-2000 Agency for Healthcare Research
and Quality (AHRQ) HCUP database is 5.5 days (AHRQ, 2000). However, this may
include patients who died in the hospital (not included among our nonfatal
myocardial infarction cases), and whose length of stay was therefore substantially
shorter than it would be if they had not died.
• Eisenstein et al. (2001) estimated 10-year costs of $45,000 in rounded 1997$ (using
a 3% discount rate) for myocardial infarction patients, using statistical prediction
(regression) models to estimate inpatient costs. Only inpatient costs (physician fees
and hospital costs) were included.
• Russell et al. (1998) estimated first-year direct medical costs of treating nonfatal
myocardial infarction of $16,000 (in rounded 1995$) and $1,100 annually thereafter
for a 10-year period.
As noted above, the estimates from these three studies are substantially different, and
we have not adequately resolved the sources of differences in the estimates. Because the
5-56
-------
wage-related opportunity cost estimates from Cropper and Krupnick (1990) cover a 5-year
period, we used estimates for medical costs that similarly cover a 5-year period (i.e., estimates
from Wittels et al. (1990) and Russell et al. (1998). We used a simple average of the two 5-year
estimates, or rounded to $66,000, and added it to the 5-year opportunity cost estimate. The
resulting estimates are given in Table 5-12.
Table 5-12. Estimated Costs Over a 5-Year Period of a Nonfatal Myocardial Infarction (in
2006$)a
Age Group
0-24
25-44
45-54
55-65
>65
Opportunity Cost
$0
$11,000°
$16,000°
$91,000°
$0
Medical Cost b
$85,000
$85,000
$85,000
$85,000
$85,000
Total Cost
$85,000
$96,000
$100,000
$180,000
$85,000
3 All estimates rounded to two significant digits. Unrounded estimates in 2000$ are available in appendix J of the
BenMAP user manual (Abt Associates, 2011).
b An average of the 5-year costs estimated by Wittels et al. (1990) and Russell et al. (1998).
From Cropper and Krupnick (1990), using a 3% discount rate.
5.6.6 Hospital Admissions and Emergency Department Valuation
In the absence of estimates of societal WTP to avoid hospital visits/admissions for
specific illnesses, we derive COI estimates for use in the benefits analysis. The International
Classification of Diseases (WHO, 1977) code-specific COI estimates used in this analysis consist
of estimated hospital charges and the estimated opportunity cost of time spent in the hospital
(based on the average length of a hospital stay for the illness). We based all estimates of
hospital charges and length of stays on statistics provided by the Agency for Healthcare
Research and Quality's Healthcare Utilization Project National Inpatient Sample (NIS) database
(AHRQ, 2007). We estimated the opportunity cost of a day spent in the hospital as the value of
the lost daily wage, regardless of whether the hospitalized individual is in the workforce. To
estimate the lost daily wage, we divided the median weekly wage reported by the 2007
American Community Survey (ACS) by five and deflated the result to year 2006$ using the CPI-U
"all items" (Abt Associates, 2011). The resulting national average lost daily wage is $134. The
total cost-of-illness estimate for an ICD code-specific hospital stay lasting n days, then, was the
mean hospital charge plus $134 multiplied by n. In general, the mean length of stay has
decreased since the 2000 database used in previous version of BenMAP while the mean
5-57
-------
hospital charge has increased. We provide the rounded unit values in 2000$ for the COI
functions used in this analysis in Table 5-13.
Table 5-13. Unit Values for Hospital Admissions
End Point
Age Range Total Cost of Illness
Mean Hospital Mean Length (unit value in
ICD Codes mm. max. Charge (2000$) of Stay (days) 2000$)
HA, All Cardiovascular
HA, All Cardiovascular
HA, All Respiratory
HA, Asthma
HA, Chronic Lung
Disease
390-429
390-429
460-519
493
490-496
18
65
65
0
18
64
99
99
64
64
$27,000
$25,000
$21,000
$9,700
$13,000
4.1
4.9
6.1
3.0
3.9
$27,000
$25,000
$21,000
$10,000
$13,000
* All estimates rounded to two significant digits. Unrounded estimates in 2000$ are available in Appendix J of the
BenMAP user manual (Abt Associates, 2011).
To value asthma emergency department visits, we used a simple average of two
estimates from the health economics literature. The first estimate comes from Smith et al.
(1997), who reported approximately 1.2 million asthma-related emergency department visits in
1987, at a total cost of $186.5 million (1987$). The average cost per visit that year was $155; in
2006$, that cost was $401 (using the CPI-U for medical care to adjust to 2006$). The second
estimate comes from Stanford et al. (1999), who reported the cost of an average asthma-
related emergency department visit at $335, based on 1996-1997 data. A simple average of the
two estimates yields a unit value of $368.
5.6.7 Minor Restricted Activity Days Valuation
No studies are reported to have estimated WTP to avoid a minor restricted activity day.
However, Neumann et al. (1994) derived an estimate of willingness to pay to avoid a minor
respiratory restricted activity day, using estimates from Tolley et al. (1986) of WTP for avoiding
a combination of coughing, throat congestion and sinusitis. This estimate of WTP to avoid a
minor respiratory restricted activity day is $38 (1990$), or about $62 (2006$). Although Ostro
and Rothschild (1989) statistically linked ozone and minor restricted activity days, it is likely that
most MRADs associated with ozone and PM2.s exposure are, in fact, minor respiratory restricted
activity days. For the purpose of valuing this health endpoint, we used the estimate of mean
WTP to avoid a minor respiratory restricted activity day.
5-58
-------
5.6.8 Growth in WTP Reflecting National Income Growth Over Time
Our analysis accounts for expected growth in real income over time. This is a distinct
concept from inflation and currency year. Economic theory argues that WTP for most goods
(such as environmental protection) will increase if real incomes increase. There is substantial
empirical evidence that the income elasticity5 of WTP for health risk reductions is positive,
although there is uncertainty about its exact value. Thus, as real income increases, the WTP for
environmental improvements also increases. Although many analyses assume that the income
elasticity of WTP is unit elastic (i.e., a 10% higher real income level implies a 10% higher WTP to
reduce risk changes), empirical evidence suggests that income elasticity is substantially less
than one and thus relatively inelastic. As real income rises, the WTP value also rises but at a
slower rate than real income.
The effects of real income changes on WTP estimates can influence benefits estimates
in two different ways: through real income growth between the year a WTP study was
conducted and the year for which benefits are estimated, and through differences in income
between study populations and the affected populations at a particular time. Empirical
evidence of the effect of real income on WTP gathered to date is based on studies examining
the former. The Environmental Economics Advisory Committee (EEAC) of the Science Advisory
Board (SAB) advised EPA to adjust WTP for increases in real income over time but not to adjust
WTP to account for cross-sectional income differences "because of the sensitivity of making
such distinctions, and because of insufficient evidence available at present" (U.S. EPA-SAB,
2000). An advisory by another committee associated with the SAB, the Advisory Council on
Clean Air Compliance Analysis, has provided conflicting advice. While agreeing with "the
general principle that the willingness to pay to reduce mortality risks is likely to increase with
growth in real income" and that "[t]he same increase should be assumed for the WTP for
serious nonfatal health effects," they note that "given the limitations and uncertainties in the
available empirical evidence, the Council does not support the use of the proposed adjustments
for aggregate income growth as part of the primary analysis" (U.S. EPA-SAB, 2004b). Until these
conflicting advisories have been reconciled, EPA will continue to adjust valuation estimates to
reflect income growth using the methods described below, while providing sensitivity analyses
for alternative income growth adjustment factors.
Based on a review of the available income elasticity literature, we adjusted the valuation
of human health benefits upward to account for projected growth in real U.S. income. Faced
5 Income elasticity is a common economic measure equal to the percentage change in WTP for a 1% change in
income.
5-59
-------
with a dearth of estimates of income elasticities derived from time-series studies, we applied
estimates derived from cross-sectional studies in our analysis. Details of the procedure can be
found in Kleckner and Neumann (1999). An abbreviated description of the procedure we used
to account for WTP for real income growth between 1990 and 2020 is presented below.
Reported income elasticities suggest that the severity of a health effect is a primary
determinant of the strength of the relationship between changes in real income and WTP. As
such, we use different elasticity estimates to adjust the WTP for minor health effects, severe
and chronic health effects, and premature mortality. Note that because of the variety of
empirical sources used in deriving the income elasticities, there may appear to be
inconsistencies in the magnitudes of the income elasticities relative to the severity of the
effects (a priori one might expect that more severe outcomes would show less income elasticity
of WTP). We have not imposed any additional restrictions on the empirical estimates of income
elasticity. One explanation for the seeming inconsistency is the difference in timing of
conditions. WTP for minor illnesses is often expressed as a short term payment to avoid a single
episode. WTP for major illnesses and mortality risk reductions are based on longer term
measures of payment (such as wages or annual income). Economic theory suggests that
relationships become more elastic as the length of time grows, reflecting the ability to adjust
spending over a longer time period. Based on this theory, it would be expected that WTP for
reducing long term risks would be more elastic than WTP for reducing short term risks. We also
expect that the WTP for improved visibility in Class I areas would increase with growth in real
income. The relative magnitude of the income elasticity of WTP for visibility compared with
those for health effects suggests that visibility is not as much of a necessity as health, thus, WTP
is more elastic with respect to income. The elasticity values used to adjust estimates of benefits
in 2020 are presented in Table 5-14.
Table 5-14. Elasticity Values Used to Account for Projected Real Income Growth3
Benefit Category Central Elasticity Estimate
Minor Health Effect 0.14
Severe and Chronic Health Effects 0.45
Premature Mortality 0.40
3 Derivation of estimates can be found in Kleckner and Neumann (1999). COI estimates are not adjusted for
income growth.
In addition to elasticity estimates, projections of real gross domestic product (GDP) and
populations from 1990 to 2020 are needed to adjust benefits to reflect real per capita income
5-60
-------
growth. For consistency with the emissions and benefits modeling, we used national population
estimates for the years 1990 to 1999 based on U.S. Census Bureau estimates (Hollman, Mulder,
and Kalian, 2000). These population estimates are based on application of a cohort-component
model applied to 1990 U.S. Census data projections (U.S. Bureau of Census, 2000). For the years
between 2000 and 2020, we applied growth rates based on the U.S. Census Bureau projections
to the U.S. Census estimate of national population in 2000. We used projections of real GDP
provided in Kleckner and Neumann (1999) for the years 1990 to 2010.6 We used projections of
real GDP (in chained 1996 dollars) provided by Standard and Poor's (2000) for the years 2010 to
2020.7
Using the method outlined in Kleckner and Neumann (1999) and the population and
income data described above, we calculated WTP adjustment factors for each of the elasticity
estimates listed in Table 5-15. Benefits for each of the categories (minor health effects, severe
and chronic health effects, premature mortality, and visibility) are adjusted by multiplying the
unadjusted benefits by the appropriate adjustment factor. Note that, for premature mortality,
we applied the income adjustment factor to the present discounted value of the stream of
avoided mortalities occurring over the lag period. Because of a lack of data on the dependence
of COI and income and a lack of data on projected growth in average wages, no adjustments are
made to benefits based on the COI approach or to work loss days and worker productivity. This
assumption leads us to underpredict benefits in future years because it is likely that increases in
real U.S. income would also result in increased COI (due, for example, to increases in wages
paid to medical workers) and increased cost of work loss days and lost worker productivity
(reflecting that if worker incomes are higher, the losses resulting from reduced worker
production would also be higher).
6 U.S. Bureau of Economic Analysis, Table 2A—Real Gross Domestic Product (1997) and U.S. Bureau of Economic
Analysis, The Economic and Budget Outlook: An Update, Table 4—Economic Projections for Calendar Years 1997
Through 2007(1997). Note that projections for 2007 to 2010 are based on average GDP growth rates between
1999 and 2007.
7 In previous analyses, we used the Standard and Poor's projections of GDP directly. This led to an apparent
discontinuity in the adjustment factors between 2010 and 2011. We refined the method by applying the
relative growth rates for GDP derived from the Standard and Poor's projections to the 2010 projected GDP
based on the Bureau of Economic Analysis projections.
5-61
-------
Table 5-15. Adjustment Factors Used to Account for Projected Real Income Growth3
Benefit Category 2020
Minor Health Effect 1.07
Severe and Chronic Health Effects 1.22
Premature Mortality 1.20
a Based on elasticity values reported in Table 5-3, U.S. Census population projections, and projections of real GDP
per capita.
5.7 Benefits Results
5.7.1 Benefits of Attaining Alternative Combinations of Primary PM2.s Standards
Applying the impact and valuation functions described previously in this chapter to the
estimated changes in PM2.5 yields estimates of the changes in physical damages (e.g.,
premature mortalities, cases of acute bronchitis and hospital admissions) and the associated
monetary values for those changes. Not all known PM health effects could be quantified or
monetized. The monetized value of these unquantified effects is represented by adding an
unknown "B" to the aggregate total. The estimate of total monetized health benefits is thus
equal to the subset of monetized PM-related health benefits plus B, the sum of the non-
monetized health and welfare benefits; this B represents both uncertainty and a bias in this
analysis, as it reflects those benefits categories that we are unable quantify in this analysis.
Table 5-16 shows the population-weighted air quality change for the alternative
standards averaged across the continental U.S. Tables 5-17 through 5-24 present the benefits
results for the alternative combinations of primary PM2.5 standards. In analyzing the current
15/35 standard (baseline), EPA determined that all counties would meet the 14/35 standard
concurrently with meeting the existing 15/35 standard at no additional cost. Consequently,
there are no incremental costs or benefits for 14/35, and no need to present an analysis of
14/35. Figure 5-3 graphically displays the total monetized benefits of the proposed range of
primary standard combinations (12/35 and 13/35) using alternative concentration-response
functions at discount rates of 3% and 7%. Figure 5-4 graphically displays the cumulative
distribution of total monetized benefits using the 2 epidemiology-derived and the 12 expert-
derived relationships between PM2.5and mortality for 12/35 and 13/35.
5-62
-------
Table 5-16. Population-Weighted Air Quality Change for Adults (30+) for Alternative
Standards Relative to 15/35
Standard Population-Weighted Air Quality Change
13/35 0.0008 ng/m3
12/35 0.0206 ng/m3
11/35 0.0807 ng/m3
11/30 0.1228ng/m3
5.7.2 Uncertainty in Benefits Results
Health benefits account for between 97 and 99% of total benefits depending on the
PM2.5 mortality estimates used, in part because we are unable to quantify most of the non-
health benefits. The next largest benefit is for reductions in chronic illness (nonfatal heart
attacks), although this value is more than an order of magnitude lower than for premature
mortality. Hospital admissions for respiratory and cardiovascular causes, MRADs and work loss
days account for the majority of the remaining benefits. The remaining categories each account
for a small percentage of total benefit; however, they represent a large number of avoided
incidences affecting many individuals. A comparison of the incidence table to the monetary
benefits table reveals that there is not always a close correspondence between the number of
incidences avoided for a given endpoint and the monetary value associated with that endpoint.
For example, we estimate almost 100 times more work loss days would be avoided than
premature mortalities, yet work loss days account for only a very small fraction of total
monetized benefits. This reflects the fact that many of the less severe health effects, while
more common, are valued at a lower level than the more severe health effects. Also, some
effects, such as hospital admissions, are valued using a proxy measure of WTP. As such, the true
value of these effects may be higher than that reported in the tables above.
PM2.5 mortality benefits represent a substantial proportion of total monetized benefits
(over 98% in this analysis), and these estimates have the following key assumptions and
uncertainties.
• Implementation of this new air quality standard is expected to reduce emissions of
directly emitted PM2.5, S02, and NOX. We assume that all fine particles, regardless of
their chemical composition, are equally potent in causing premature mortality. This
is an important assumption, because PM2.5 produced varies considerably in
5-63
-------
composition across sources, but the scientific evidence is not yet sufficient to allow
differential effects estimates by particle type.
• We assume that the health impact function for fine particles is linear within the
range of ambient concentrations under consideration. Thus, the estimates include
health benefits from reducing fine particles in areas with varied concentrations of
PM2.5, including both regions that are in attainment with fine particle standard and
those that do not meet the standard down to the lowest modeled concentrations.
Given that reductions in premature mortality dominate the size of the overall monetized
benefits, more focus on uncertainty in mortality-related benefits gives us greater confidence in
our uncertainty characterization surrounding total benefits.
5-64
-------
Table 5-17. Estimated number of Avoided PM2.5 Health Impacts for Alternative Combinations
of Primary PM2.s Standards (Incremental to Attaining Current Suite of Primary
PM2.5 Standards)3
Alterative Combination of Standards
(95th percentile confidence interval)
Health Effect
Non-fatal heart
attacks
Peters et al. (2001)
(age >18)
Pooled estimate of 4
studies (age >18)
Hospital admissions-
respiratory
(all ages)
Hospital admissions-
cardiovascular
(age > 18)
Emergency
department visits for
asthma
(age < 18)b
Acute bronchitis
(ages 8-12) b
Lower respiratory
symptoms
(ages 7-14)
Upper respiratory
symptoms
(asthmatics ages
9-11)
Asthma exacerbation
(asthmatics ages
6-18)
Lost work days
(ages 18-65)
Minor restricted-
activity days
(ages 18-65)
13 |ig/m3 Annual &
35 u,g/m3 24-hrb
11
(6-19)
1
(1-3)
3
(2-5)
3
(2-6)
6
(2-13)
22
(7-48)
290
(180-450)
410
(220-710)
410
(150-860)
1,800
(1700-2100)
11,000
(9,500-12,000)
12 |ig/m3 Annual &
35 ug/m3 24-hr
320
(80-550)
35
(15-92)
98
(51-150)
95
(43-170)
160
(-29-340)
540
(-120-1,200)
6,900
(2,700-11,000)
9,800
(1,800-18,000)
24,000
(0-180,000)
44,000
(38,000-51,000)
260,000
(210,000-310,000)
11 u,g/m3 Annual &
35 ug/m3 24-hr
1,300
(390-2,100)
140
(64-330)
430
(240-620)
400
(190-0,700)
730
(-56-1,500)
2,000
(-260-4,200)
25,000
(11,000-40,000)
37,000
(9,200-64,000)
89,000
(1,900-570,000)
170,000
(150,000-200,000)
1,000,000
(840,000-1,200,000)
11 |ig/m3 Annual &
30 ug/m3 24-hr
1,900
(590-3,200)
210
(98-510)
620
(350-0,890)
580
(280-1,000)
1,000
(-79-2,100)
3,100
(-400-6,400)
39,000
(17,000-61,000)
56,000
(14,000-98,000)
140,000
(2,900-870,000)
260,000
(220,000-290,000)
1,500,000
(1,300,000-1,800,000)
All incidence estimates are rounded to whole numbers with a maximum of two significant digits.
b The negative estimates at the 5th percentile confidence estimates for these morbidity endpoints reflect the
statistical power of the study used to calculate these health impacts. These results do not suggest that reducing
air pollution results in additional health impacts.
5-65
-------
Table 5-18. Estimated Number of Avoided PM2.5-Related Deaths for Alternative
Combinations of Primary PM2.s Standards (Incremental to Attaining Current Suite
of Primary PM2.s Standards)3
Alterative Combination of Standards
(95th percentile confidence interval)
13 |ig/m3 Annual & 12 |ig/m3 Annual &
35 ug/m3 24-hr 35 ug/m3 24-hr
11 |ig/m3 Annual &
35 ug/m3 24-hr
11 ug/m3 Annual &
30 ug/m3 24-hr
Mortality impact functions derived from the epidemiology literature
Krewski et al.
(2009)
Laden et al.
(2006)
Woodruff et al.
(1997)
(infant mortality)
Mortality impact functions
Expert A
Expert B
Expert C
Expert D
Expert E
Expert F
Expert G
Expert H
Expert 1
Expert J
Expert K
Expert L
11
(9-14)
27
(19-41)
0
(0-0)
derived from the PM2.5
28
(11-55)
22
(4-47)
22
(12-37)
15
(0-26)
36
(23-55)
20
(14-28)
13
(0-24)
16
(0-38)
22
(0-38)
18
(6-39)
3
(0-12)
16
(0-29)
280
(190-370)
730
(330-1100)
1
(o-D
Expert Elicitation
740
(55-1,500)
590
(27-1,300)
580
(140-1000)
410
(0-700)
950
(370-1,500)
530
(310-770)
340
(0-640)
420
(0-1100)
570
(0-1000)
470
(17-1100)
72
(0-330)
400
(0-790)
1,100
(790-1,400)
2,900
(1,400-4,300)
3
(1-5)
(Roman et al., 2006)
2,900
(400-5,700)
2,300
(160-5,000)
2,300
(670-3,900)
1,600
(0-2,700)
3,700
(1,600-5,700)
2,100
(1,300-3,000)
1,300
(0-2,500)
1,700
(0-4,000)
2,300
(0-3,900)
1,800
(170-4,000)
270
(0-1,200)
1,600
(0-3,000)
1,700
(1,200-2,300)
4,500
(2,200-6,700)
4
(2-7)
4,500
(620-8,900)
3,600
(220-7,700)
3,600
(1100-6,100)
2,500
(0-4,200)
5,900
(2,600-9,000)
3,300
(2,000-4,700)
2,100
(0-3,800)
2,600
(0-6,200)
3,500
(0-6,100)
2,900
(270-6,300)
420
(0-1,900)
2,400
(0^1,700)
All incidence estimates are rounded to whole numbers with a maximum of two significant digits.
5-66
-------
Table 5-19. Estimated Monetized PM2.5 Health Impacts for Alternative Combinations of
Primary PM2.s Standards (Incremental to Attaining Current Suite of Primary PM2.s
Standards) (Millions of 2006$, 3% discount rate)3
Alterative Combination of Standards
(95th percentile confidence interval)
13 |ig/ms Annual & 12 u,g/ms Annual & 11 |ig/ms Annual & 11 u,g/ms Annual &
Health Effect
Non-fatal heart attacks
Peters et al. (2001)
(age >18)
Pooled estimate of 4
studies
(age >18)
Hospital admissions-
respiratory
(all ages)
Hospital admissions-
cardiovascular
(age > 18)
Emergency department
visits for asthma (age < 18)
Acute bronchitis
(ages 8-12)b
Lower respiratory
symptoms
(ages 7-14)
Upper respiratory
symptoms
(asthmatics ages 9-11)
Asthma exacerbation
(asthmatics ages 6-18)
Lost work days
(ages 18-65)
Minor restricted-activity
days
(ages 18-65)
35 u,g/m 24-hr
$1.1
($0.19-$2.9)
$0.13
($0.029-$0.43)
$0.081
($0.050-$0.11)
$0.11
($0.058-$0.19)
$0.0023
($0.000065-$0.0050)
$0.010
(-$0.00045-$0.028)
$0.0055
($0.0018-$0.011)
$0.012
($0.0028-$0.030)
$0.022
($0.0019-$0.058)
$0.27
($0.24-$0.31)
$0.67
($0.35-$1.0)
35 ug/m 24-hr
$33
($5.5-$82)
$3.7
($0.85-$12)
$2.4
($1.5-$3.3)
$3.2
($1.7-$5.4)
$0.058
($0.0016-$0.13)
$0.24
(-$0.011-$0.66)
$0.13
($0.044-$0.27)
$0.30
($0.067-$0.74)
$1.3
($0.047-$9.0)
$6.7
($5.8-$7.5)
$16.0
($8.7-$25)
35 ug/m 24-hr
$130
($23-$330)
$15
($3.5-$49)
$10
($6-$15)
$13
($7-$23)
$0.27
($0.008-$0.58)
$0.89
(-$0.040-$2.4)
$0.49
($0.16-$1.0)
$1.1
($0.25-$2.7)
$4.7
($0.18-$34)
$26
($22-$29)
$64
($34-$96)
30 u,g/m 24-hr
$200
($34-$500)
$23
($5.3-$75)
$15
($9-$21)
$20
($10-$33)
$0.38
($0.011-$0.81)
$1.40
(-$0.062-$3.7)
$0.76
($0.25-$1.5)
$1.7
($0.38-$4.2)
$7.2
($0.27-$51)
$39
($34-$44)
$96
($51-$ 140)
All estimates are rounded
in Table 5-2 or Section 5.6
to two significant digits. Estimates do not include unquantified health benefits noted
.5 or welfare benefits noted in Chapter 6.
b The negative estimates at the 5th percentile confidence estimates for this morbidity endpoint reflects the
statistical power of the study used to calculate these health impacts. These results do not suggest that reducing
air pollution results in additional health impacts.
5-67
-------
Table 5-20. Estimated Monetized PM2.5 Health Impacts for Alternative Combinations of
Primary PM2.s Standards (Incremental to Attaining Current Suite of Primary PM2.s
Standards) (Millions of 2006$, 7% discount rate)3
Alterative Combination of Standards
(95th percentile confidence interval)
Health Effect
Non-fatal heart attacks
Peters et al. (2001)
(age >18)
Pooled estimate of
4 studies
(age >18)
Hospital admissions-
respiratory (all ages)
Hospital admissions-
cardiovascular
(age > 18)
Emergency department
visits for asthma
(age < 18)
Acute bronchitis
(ages 8-12)b
Lower respiratory
symptoms
(ages 7-14)
Upper respiratory
symptoms
(asthmatics ages 9-11)
Asthma exacerbation
(asthmatics ages 6-18)
Lost work days
(ages 18-65)
Minor restricted-
activity days (ages 18-
65)
13 ug/m3 Annual &
35 ug/m3 24-hr
$1.1
($0.180-$2.9)
$0.12
($0.027-$0.42)
$0.081
($0.050-$0.11)
$0.11
($0.058-$0.19)
$0.0023
($0.000065-$0.0050)
$0.010
(-$0.00045-$0.028)
$0.0055
($0.0018-$0.011)
$0.012
($0.0028-$0.030)
$0.022
($0.0019-$0.058)
$0.27
($0.24-$0.31)
$0.67
($0.35-$1.0)
12 ug/m3 Annual &
35 ug/m3 24-hr
$32
($5.1-$81)
$3.6
($0.79-$12)
$2.4
($1.5-$3.3)
$3.2
($1.7-$5.4)
$0.058
($0.0016-$0.13)
$0.24
(-$0.011-$0.66)
$0.13
($0.044-$0.27)
$0.30
($0.067-$0.74)
$1.3
($0.047-$9.0)
$6.7
($5.8-$7.5)
$16.0
($8.7-$25)
11 ug/m3 Annual
& 35 ug/m3 24-
hr
$130
($21-$320)
$14
($3.2-$48)
$10
($6-$15)
$13
($7.0-$23)
$0.27
($0.008-$0.58)
$0.89
(-$0.040-$2.4)
$0.49
($0.16-$1.0)
$1.1
($0.25-$2.7)
$4.7
($0.18-$34)
$26
($22-$29)
$64
($34-$96)
11 u,g/m3 Annual &
30 ug/m3 24-hr
$190
($32-$490)
$22
($4.9-$74)
$15
($9-$21)
$20
($10-$33)
$0.38
($0.011-$0.81)
$1.40
(-$0.062-$3.7)
$0.76
($0.25-$1.5)
$1.7
($0.38-$4.2)
$7.2
($0.27-$51)
$39
($34-$44)
$96
($51-$140)
a All estimates are rounded to two significant digits. Estimates do not include unquantified health benefits noted
in Table 5-2 or Section 5.6.5 or welfare benefits noted in Chapter 6.
b The negative estimates at the 5th percentile confidence estimates for this morbidity endpoint reflects the
statistical power of the study used to calculate these health impacts. These results do not suggest that reducing
air pollution results in additional health impacts.
5-68
-------
Table 5-21. Estimated Monetized PM2.5-Related Deaths for Alternative Combinations of
Primary PM2.s Standards (Incremental to Attaining Current Suite of Primary PM2.s
Standards)(Millions of 2006$, 3% discount rate)3
Alterative Combination of Standards
(95th percentile confidence interval)
Health Effect
13 ug/ms Annual &
35 u,g/m3 24-hr
Mortality impact functions derived from the
Krewski et al. (2009)
Laden et al. (2006)
Woodruff et al.
(1997) (infant
mortality)
$86
($8.0-$240)
$220
($19-$640)
$0.28
($0.023-$0.82)
12 u,g/mB Annual &
35 u,g/m3 24-hr
11 ug/ms Annual &
35 u,g/m3 24-hr
11 ug/ms Annual &
30 ug/m3 24-hr
epidemiology literature
$2,300
($210-$6,300)
$5,900
($520-$ 17,000)
$6.9
($0.58-$20)
Mortality impact functions derived from the PM2.5 Expert Elicitation
Expert A
Expert B
Expert C
Expert D
Expert E
Expert F
Expert G
Expert H
Expert 1
Expert J
Expert K
Expert L
$220
($13-$760)
$180
($6.0-$700)
$180
($13-$540)
$120
($3.5-$390)
$290
($25-$860)
$160
($14.0-$460)
$100
($0-$370)
$130
($0-$510)
$170
($6.2-$560)
$140
($6.7-$510)
$23
($0-$140)
$130
($0.63-$450)
$5,900
($330-$20,000)
$4,800
($150-$ 19,000)
$4,700
($350-$ 14,000)
$3,300
($94-$ 10,000)
$7,700
($650-$23,000)
$4,300
($390-$ 12,000)
$2,700
($0-$9,800)
$3,400
($0-$13,000)
$4,600
($170-$ 15,000)
$3,800
($180-$ 13,000)
$580
($0-$3,700)
$3,300
($14-$ 12,000)
$9,000
($840-$25,000)
$23,000
($2,000-$67,000)
$26
($2.2-$78)
(Roman et al., 2008)
$23,000
($1,300-$78,000)
$19,000
($0,560-$73,000)
$18,000
($1,400-$57,000)
$13,000
($370-$41,000)
$30,000
($2,600-$89,000)
$17,000
($1,600-$49,000)
$11,000
($0-$39,000)
$13,000
($0-$53,000)
$18,000
($650-$58,000)
$15,000
($700-$53,000)
$2,200
($0-$ 14,000)
$13,000
($48-$47,000)
$14,000
($1,300-$39,000)
$36,000
($3,200-$100,000)
$39
($3.3-$110)
$37,000
($2,100-$120,000)
$29,000
($0,780-$110,000)
$29,000
($2,200-$88,000)
$20,000
($580-$63,000)
$47,000
($4,000-$140,000)
$26,000
($2,400-$76,000)
$17,000
($0-$6 1,000)
$21,000
($0-$82,000)
$28,000
($1,000-$90,000)
$23,000
($1100-$82,000)
$3,300
($0-$22,000)
$19,000
($53-$72,000)
Rounded to two significant figures. Estimates do not include unquantified health benefits noted in Table 5-2 or
Section 5.6.5 or welfare benefits noted in Chapter 6.The reduction in premature fatalities each year accounts for
over 98% of total monetized benefits in this analysis. Mortality risk valuation assumes discounting over the SAB-
recommended 20-year segmented lag structure.
5-69
-------
Table 5-22. Estimated Monetized PM2.5-Related Deaths for Alternative Combinations of
Primary PM2.s Standards (Incremental to Attaining Current Suite of Primary PM2.s
Standards) (Millions of 2006$, 7% discount rate)3
Alterative Combination of Standards
(95th percentile confidence interval)
Health Effect
Mortality impact functions
Krewski et al. (2009)
Laden et al. (2006)
Woodruff etal. (1997)
(infant mortality)
Mortality impact functions
Expert A
Expert B
Expert C
Expert D
Expert E
Expert F
Expert G
Expert H
Expert 1
Expert J
Expert K
Expert L
13 ug/m3 Annual & 12 ug/m3 Annual &
35 ug/m3 24-hr 35 ug/m3 24-hr
derived from the
$77
($7.2-$210)
$200
($18-$580)
$0.28
($0.023-$0.82)
epidemiology literature
$2,100
($190-$5,600)
$5,300
($460-$ 15,000)
$6.9
($0.58-$20)
11 ug/m3 Annual &
35 ug/m3 24-hr
$8,100
($750-$22,000)
$21,000
($1,800-$60,000)
$26
($2.2-$78)
11 ug/m3 Annual &
30 ug/m3 24-hr
$13,000
($1,200-$35,000)
$32,000
($2,900-$94,000)
$39
($3.3-$110)
derived from the PM2.5 Expert Elicitation (Roman et al., 2008)
$200
($11-$680)
$160
($5.4-$630)
$160
($12-$490)
$110
($3.2-$350)
$260
($22-$770)
$140
($13-$410)
$93
($0-$330)
$120
($0-$460)
$160
($5.6-$500)
$130
($6.0-$460)
$20
($0-$130)
$110
($0.57-$400)
$5,300
($300-$ 18,000)
$4,300
($130-$ 17,000)
$4,200
($320-$ 13,000)
$3,000
($85-$9,300)
$6,900
($590-$20,000)
$3,900
($350-$11,000)
$2,500
($0-$8,900)
$3,100
($0-$ 12,000)
$4,200
($150-$ 13,000)
$3,400
($160-$ 12,000)
$520
($0-$3,400)
$2,900
($12-$11,000)
$21,000
($l,200-$7 1,000)
$17,000
($500-$66,000)
$17,000
($l,300-$5 1,000)
$12,000
($330-$37,000)
$27,000
($2,300-$80,000)
$15,000
($1,400-$44,000)
$9,700
($0-$35,000)
$12,000
($0-$47,000)
$16,000
($590-$52,000)
$13,000
($630-$47,000)
$2,000
($0-$13,000)
$11,000
($43-$42,000)
$33,000
($1,900-$110,000)
$26,000
($0,700-$100,000)
$26,000
($2,000-$80,000)
$18,000
($520-$57,000)
$43,000
($3,600-$130,000)
$24,000
($2,200-$68,000)
$15,000
($0-$55,000)
$19,000
($0-$74,000)
$26,000
($920-$81,000)
$21,000
($980-$74,000)
$3,000
($0-$20,000)
$17,000
($48-$65,000)
Rounded to two significant figures. Estimates do not include unquantified health benefits noted in Table 5-2 or
Section 5.6.5 or welfare benefits noted in Chapter 6.The reduction in premature fatalities each year accounts for
over 98% of total monetized benefits in this analysis. Mortality risk valuation assumes discounting over the SAB-
recommended 20-year segmented lag structure.
5-70
-------
Table 5-23. Total Estimated Monetized Benefits of the for Alternative Combinations of
Primary PM2.s Standards (Incremental to Attaining Current Suite of Primary PM2.s
Standards) (millions of 2006$)a
13 ug/m3 annual & 12 ug/m3 annual & 11 ug/m3 annual & 11 ug/m3 Annual &
Benefits Estimate 35 ug/m3 24-hour 35 ug/m3 24-hour 35 ug/m3 24-hour
30 ug/m 24-hr
Economic value of avoided PM2.5.related morbidities and premature deaths using PM2.5 mortality estimate
from Krewski et al. (2009)
3% discount rate
7% discount rate
$88+B
$79+ B
$2,300 +B
$2,100 +B
$9,200+B
$8,300+B
$14,000 +B
$13,000 +B
Economic value of avoided PM2.5.related morbidities and premature deaths using PM2.5 mortality estimate
from Laden et al. (2006)
3% discount rate
7% discount rate
$220 + B
$200 + B
$5,900 +B
$5,400 +B
$23,000 +B
$21,000 +B
$36,000 +B
$33,000 +B
Rounded to two significant figures. The reduction in premature fatalities each year accounts for over 98% of
total monetized benefits in this analysis. Mortality risk valuation assumes discounting over the SAB-
recommended 20-year segmented lag structure. Not all possible benefits are quantified and monetized in this
analysis. B is the sum of all unquantified health and welfare benefits. Data limitations prevented us from
quantifying these endpoints, and as such, these benefits are inherently more uncertain than those benefits that
we were able to quantify.
Table 5-24. Regional Breakdown of Monetized Benefits Results
Alterative Combination of Standards
Region
East3
California15
Rest of
West
13 ug/m3 annual &
35 ug/m3 24-hour
0%
98%
2%
12 ug/m3 Annual &
35 ug/m3 24-hr
27%
70%
3%
11 ug/m3 Annual &
35 ug/m3 24-hr
53%
44%
3%
11 ug/m3 Annual &
30 ug/m3 24-hr
43%
47%
10%
Includes Texas and those states to the north and east. Several recent rules such as MATS and CSAPR will have
substantially reduced PM2.s levels by 2020 in the East, thus few additional controls would be needed to reach
12/35 or 13/35.
For 12/35 and 13/35, the majority of benefits (occur in California because this highly populated area is where the
most air quality improvement beyond 15/35 is needed to reach these levels.
5-71
-------
$8,000
$7,000
$6,000
$5,000
o
— $4,000
c
o
:=
'^
$100
$50
$0
3% Discount Rate
7% Discount Rate
PM2.5 mortality benefits estimates derived from 2 epidemiology functions and 12 expert functions
3% Discount Rate
7% Discount Rate
PM2.5 mortality benefits estimates derived from 2 epidemiology functions and 12 expert functions
Figure 5-3. Estimated PM2.5-Related Premature Mortalities Avoided According to
Epidemiology or Expert-Derived PM2.5 Mortality Risk Estimate for 12/35 and 13/35
5-72
-------
100%
I
$1 $2 $3 $4
$5 $6 $7 $8 $9 $10 $11 $12 $13 $14 $15 $16 $17 $18 $19 $20
Cumulative Percentage of Monetized Benefits (Billions of 2006$)
100%
"S
••Krewskietal.
• •Ladenetal.
— Expert A
^ExpertB
^ExpertC
^ExpertD
"ExpertE
— Expert F
Expert G
— Expert H
^Expert I
"ExpertJ
— Expert K
Expert L
$0.1
$0.2 $0.3 $0.4 $0.5 $0.6 $0.7 $0.8
Cumulative Percentage of Monetized Benefits (Billions of 2006$)
$0.9
$1.0
Figure 5-4. Total Monetized Benefits Using 2 Epidemiology-Derived and 12-Expert Derived
Relationships Between PM2.5 and Premature Mortality for 12/35 and 13/35
5-73
-------
5.7.3 Estimated Life Years Gained Attributable to Reduced PM2.5 Exposure and Percent of
Total Mortality
In their 2008 review of EPA's approach to estimating ozone-related mortality benefits,
NRC indicated, "EPA should consider placing greater emphasis on reporting decreases in age-
specific death rates in the relevant population and develop models for consistent calculation of
changes in life expectancy and changes in number of deaths at all ages" (NRC, 2008). In
addition, NRC previously noted in an earlier report that "[f]rom a public-health perspective, life-
years lost might be more relevant than annual number of mortality cases" (NRC, 2002). This
advice is consistent with that of the HES, which agreed that "...the interpretation of mortality
risk results is enhanced if estimates of lost life-years can be made" (U.S. EPA-SAB, 2004a). To
address these recommendations, we estimate the number of life years gained and the
reduction in the percentage of all deaths attributable to PM2.5 resulting from attainment of the
alternative combinations of primary PM2.5 standards. EPA included similar estimates of life
years gained in a previous assessment of PM2.5 benefits (U.S. EPA, 2006a, 2011a), the latter of
which was peer reviewed by the HES (U.S. EPA-SAB, 2010a).
Because changes in life years and changes in life expectancy at birth are frequently
conflated, it is important to distinguish these two very different metrics. Life expectancy varies
by age. CDC defines life expectancy as the "average number of years of life remaining for
persons who have attained a given age" (CDC, 2011). In other words, changes in life expectancy
refer to an average change for the entire population, and refer to the future. Over the past 50
years, average life expectancy at birth in the U.S. has increased by 8.4 years (CDC, 2001). Life
years, on the other hand, measure the amount of time that an individual loses if they die before
the age of their life expectancy. Life years refer to individuals, and refer to the past, e.g., when
the individual has already died. For example, life expectancy at birth was estimated in 2007 to
be 77.9 years for an average person born in the U.S., but for people surviving to age 60,
estimated life expectancy is 82.5 years (i.e., 4.6 years more than life expectancy at birth) (CDC,
2011). If a 60-year old individual dies, we estimate about that this individual would lose about
22.5 years of life (i.e., the average population life expectancy for an individual of this age minus
this person's age at death).
In this analysis, we use the same general approach as Hubbell (2006) and Fann et al.
(2012a) for estimating potential life years gained by reducing exposure to PM2.5 in adult
populations. We have not estimated the change in average life expectancy at birth in this RIA.
Hubbell (2006) estimated that reducing exposure to PM2.5from air pollution regulations result
in an average gain of 15 years of life for those adults prematurely dying from PM2.5 exposure. In
5-74
-------
contrast, Pope et al. (2009) estimated changes in average life expectancy at birth over a twenty
year period, finding that reducing exposure to air pollution increased average life expectancy at
birth by approximately 7 months, which was 15% of the overall increase in life expectancy at
birth from 1980 through 2000. These results are not inconsistent because they are reporting
different metrics. Because life expectancy is an average of the entire population (including
those who will not die from PM exposure as well as those who will), average life expectancy
changes associated with PM exposure will always be significantly smaller than the average
number of life years lost by an individual who is projected to die prematurely from PM
exposure.
To calculate the potential distribution of life years gained for populations of different
ages, we use standard life tables available from the CDC (2003) and the following formula:
Total Life Years = 2?=1L£j x Mt (5-2)
where LEj is the remaining life expectancy for age interval i, Mj is the change in number of
deaths in age interval i, and n is the number of age intervals. We binned the life year results by
age range and calculated the average per life lost.
When we estimate the number of avoided premature deaths attributed to changes in
PM2.s exposure in 2020, we apply risk coefficients estimated for all adult populations in
conjunction with age-specific mortality rates. That is, we apply risk coefficients that do not vary
by age, but use baseline mortality rates do. Because mortality rates for younger populations are
much lower than mortality rates for older populations, most but not all, of the avoided deaths
tend to be in older populations. By comparing the projected age distribution of the avoided
premature deaths with the age distribution of life years gained, we observed that about half of
the deaths occur in populations age 75-99, but half of the life years would occur in populations
younger than 65. This is because the younger populations have the potential to lose more life
years per death than older populations based on changes in PM2.s exposure in 2020. On
average, we estimate that the average individual who would otherwise have died prematurely
from PM exposure would gain 16 additional years of life.
When calculating changes in life years, we assume that the life expectancy at birth of
those dying from PM2.s exposure is the same as the general population. In reference to the
most recent Six Cities extended analysis by Laden et al. (2006), Krewski et al. (2009) notes that
"[w]hether PM2.s exposure was modeled as the annual average in the year of death or as the
average over the entire follow-up period, it had similar effects on mortality. The results from
5-75
-------
the study suggest that since PM2.5 exposure may affect sensitive individuals with preexisting
conditions and play a role in the development of chronic disease, as exposure declines so may
the excess mortality related to it." For this reason, we believe that this is a reasonable
assumption.
In addition, this analysis includes an estimate of the percentage of all-cause mortality
attributed to reduced PM2.5 exposure in 2020 as a result of the illustrative control strategies.
The percentage of premature PM2.5-related mortality is calculated by dividing the number of
excess deaths estimated for each alternative standard by the total number of deaths in each
county. We have also binned these results by age range.
Tables 5-25 and 5-26 summarize the estimated number of life years gained and the
reduction in the percentage of all-cause mortality attributable to reduced PM2.5 exposure in
2020 by age range for 12/35. Figure 5-5 bins the potential life years gained and avoided
premature deaths into age ranges for 12/35 for comparison. The number of life years gained
and avoided mortalities would be similar across various combinations of standards on a relative
basis. Because we assume that there is a "cessation" lag between PM exposures and the
reduction in the risk of premature death, there is uncertainty regarding the specific ages that
people die relative to the change in exposure. While the structure of the lag is uncertain, some
studies suggest that most of the premature deaths are avoided within the first three years after
the change in exposure, while others are unable to find a critical window of exposure (U.S. EPA,
2004c; Schwartz, 2008; Krewski et al. 2009). These studies did not examine whether the
cessation lag was modified by either age of exposure or cumulative lifetime exposure.
5-76
-------
Table 5-25. Sum of Life Years Gained by Age Range from Changes in PM2.5 Exposure in 2020
for 12/35a'b
Age Rangeb
25-29
30-34
35^14
45-54
55-64
65-74
75-84
85-99
Total life years gained
Average life years gained per individual
Krewski et al. (2009) Risk
Coefficient c
—
140
410
690
1,100
1,100
740
350
4,500
16.0
Laden et al. (2006)
Risk Coefficient
420
370
1,000
1,800
2,700
2,800
1,900
880
12,000
16.4
Estimates rounded to two significant figures.
b Because we assume that there is a "cessation" lag between PM exposures and the reduction in the risk of
premature death, there is uncertainty regarding the specific ages that people die relative to the change in
exposure.
c The youngest age in the population cohort of this study is 30.
5-77
-------
Avoided Premature Deaths using
Krewski et al. (2009) b
Age
Age
30-34 7; Age
—35'44 45-54
Avoided Premature Deaths
using Laden et al. (2006)
Age Age
25.29 .30-34 Age
Age
85-99
>Age
75-84
Figure 5-5a. Distribution of Estimated Avoided Premature Deaths by the Age at which these
Populations were Exposed in 2020 for 12/35 a
Life Years Gained using Krewski
et al. (2009) b
Age
Life Years Gained using Laden
et al. (2006)
AUP
Age
.30-34
25-29
Figure 5-5b. Distribution of Estimated Life Years Gained by the Age at which these
Populations were Exposed in 2020 for 12/35 a
3 As shown in these charts, slightly more than half of the avoided premature deaths occur in populations age 75-
99, but slightly more than half of the avoided life years occur in populations age <65 due to the fact that the
younger populations would lose more life years per death than older populations. Results would be similar for
other standard levels on a percentage basis. Because we assume that there is a "cessation" lag between PM
exposures and the reduction in the risk of premature death, there is uncertainty regarding the specific ages that
people die relative to the change in exposure.
b The youngest age in the population cohort of this study used to estimate PM2.5 mortality incidence is 30.
5-78
-------
Table 5-26. Estimated Reduction the Percentage of All-Cause Mortality Attributable to PM2.5
for 12/35 from Changes in PM2.5 Exposure in 2020a'b
Age Rangeb
25-29
30-34
35-44
45-54
55-64
65-74
75-84
85-99
Krewski et al. (2009) Risk Coefficient'
—
0.014%
0.013%
0.013%
0.012%
0.011%
0.010%
0.010%
Laden et al. (2006) Risk Coefficient
0.036%
0.035%
0.033%
0.034%
0.031%
0.028%
0.026%
0.025%
a Rounded to two significant figures. Results would be similar for other standard levels on a percentage basis.
b Because we assume that there is a "cessation" lag between PM exposures and the reduction in the risk of
premature death, there is uncertainty regarding the specific ages that people die relative to the change in
exposure.
The youngest age in the population cohort of this study is 30.
5.7.4 Analysis of Mortality Impacts at Various Concentration Benchmarks
In this analysis, we estimate the number of avoided PM2.5-related deaths occurring
down to various PM2.5 concentration benchmarks, including the Lowest Measured Level (LML)
of each long-term PM2.5 mortality study. We include this sensitivity analysis because
assessments quantifying PM2.5 related health impacts generally find that cases of avoided
mortality represent the majority of the monetized benefits. This analysis is one of several
sensitivities that EPA has historically performed that characterize the uncertainty associated
with the PM-mortality relationship and the economic value of reducing the risk of premature
death (Roman et al., 2008; U.S. EPA, 2006a, 2011a; Mansfield, 2009).
Based on our review of the current body of scientific literature, EPA estimated PM-
related mortality without applying an assumed concentration threshold. The PM ISA (U.S. EPA,
2009b), which was reviewed by EPA's Clean Air Scientific Advisory Committee (U.S. EPA-SAB,
2009a; U.S. EPA-SAB, 2009b), concluded that the scientific literature consistently finds that a
no-threshold log-linear model most adequately portrays the PM-mortality concentration-
response relationship while also recognizing potential uncertainty about the exact shape of the
5-79
-------
concentration-response function. Consistent with this finding, we incorporated a LML
assessment, which is a method EPA has employed in several recent RIAs (U.S. EPA, 2010g,
2011c, 2011d). In addition, we have incorporated an assessment using specific concentration
benchmarks identified in EPA's Policy Assessment for Particulate Matter (U.S. EPA, 2011b).
This approach summarizes the distribution of avoided PM2.5-related mortality impacts
according to the baseline (i.e., pre-rule) annual mean PM2.5 levels at which populations are
exposed and the minimum observed air quality level of each long-term cohort study we use to
quantify mortality impacts. In general, our confidence in the estimated number of premature
deaths diminishes as we estimate these impacts in locations where PM2.5 levels are below this
level. This interpretation is consistent with the Policy Assessment (U.S. EPA, 2011b) and advice
from CASAC during their peer review (U.S. EPA-SAB, 2010d). In general, we have greater
confidence in risk estimates based on PM2.5 concentrations where the bulk of the data reside
and somewhat less confidence where data density is lower. However, there are uncertainties
inherent in identifying any particular point at which our confidence in reported associations
becomes appreciably less.
As noted in the preamble to the proposed rule, the Policy Assessment (U.S. EPA, 2011b)
concludes that the range from the 25th to 10th percentiles of the air quality data used in the
epidemiology studies is a reasonable range below which we have appreciably less confidence in
the associations observed in the epidemiological studies.
Although these types of concentration benchmark analyses (e.g., 25th percentile, 10th
percentile, and LML) provide some insight into the level of uncertainty in the estimated PM2.5
mortality benefits, EPA does not view these concentration benchmarks as a concentration
threshold. The central benefits estimates reported in this RIA reflect a full range of modeled air
quality concentrations. In reviewing the Policy Assessment, CASAC confirmed that "[although
there is increasing uncertainty at lower levels, there is no evidence of a threshold (i.e., a level
below which there is no risk for adverse health effects)" (U.S. EPA-SAB, 2010d). In addition, in
reviewing the Costs and Benefits of the Clean Air Act (U.S. EPA, 2011a), the HES noted that
"[t]his [no-threshold] decision is supported by the data, which are quite consistent in showing
effects down to the lowest measured levels. Analyses of cohorts using data from more recent
years, during which time PM concentrations have fallen, continue to report strong associations
with mortality" (U.S. EPA-SAB, 2010a). Therefore, the best estimate of benefits includes
' Fora summary of the scientific review statements regarding the lack of a threshold in the PM2.5-mortality
relationship, see the Technical Support Document (TSD) entitled Summary of Expert Opinions on the Existence
of a Threshold in the Concentration-Response Function for PM2.5-related Mortality (U.S. EPA, 2010f).
5-80
-------
estimates below and above these concentration benchmarks but uncertainty is higher in health
benefits estimated at lower concentrations, with the lowest confidence below the LML.
Estimated health impacts reflecting air quality improvements below and above these
concentration benchmarks are appropriately included in the total benefits estimate. In other
words, our confidence in the estimated benefits above these concentration benchmarks should
not imply an absence of confidence in the benefits estimated below these concentration
benchmarks.
We estimate that most of the avoided PM-related impacts we estimate in this analysis
occur among populations exposed at or above the LML of the Laden et al. (2006) study, while a
majority of the impacts occur at or above the LML of the Krewski et al. (2009) study. We show
the estimated reduction in incidence of premature mortality above and below the LML of these
studies in Tables 5-27 and 5-28, and we graphically display the distribution of PM2.5-related
mortality impacts for 12/35 and 13/35 in Figures 5-6 and 5-7. When interpreting these LML
graphs, it is important to understand that the plots illustrate the avoided PM2.5 deaths
estimated to occur from PM2.5 reductions in the baseline air quality simulation in which we
assume that 15/35 is already met. When simulating attainment with alternative standards, we
do not adjust the PM2.5 concentration in every 12km grid cell to equal the alternative standard.
Instead, we adjust the design value at the monitor to equal the alternative standard and
simulate how that adjustment would be reflected in the surrounding grid cells. As such, there
may be a small number of grid cells with concentrations greater than 15 u.g/m3 in the baseline
even though all monitors meet an annual standard at 15 u.g/m3. Specifically, there is one grid
cell in San Bernardino County with a baseline concentration of 16.4 u.g/m3, which falls in the 16
to 17 u.g/m3 bin. This one grid cell is highly populated and has a relatively high percentage of
the avoided premature mortalities because this area received the most air quality improvement
from the control strategies to reach 12/35 and 13/35. In addition, several recent rules such as
the Mercury and Air Toxics Standard (MATS) and the Cross-State Air Pollution Rule (CSAPR) will
have substantially reduced PM2.5 levels by 2020 in the East, thus few additional controls would
be needed to reach 12/35 or 13/35 in the East.
It is important to note that these estimated benefits reflect specific control measures
and emission reductions that are needed to lower PM2.5 concentrations for monitors projected
to exceed the alternative standard analyzed. The result is that air quality will improve in
counties that exceed the alternative standards as well as surrounding areas that do not exceed
the alternative standards. It is not possible to apply controls that only reduce PM2.5 at the
monitor without affecting surrounding areas. In order to make a direct comparison between
5-81
-------
the benefits and costs of these control strategies, it is appropriate to include all the benefits
occurring as a result of the control strategies applied.
We estimate benefits using modeled air quality data with 12km grid cells, which is
important because the grid cells are smaller than counties and PM2.5 concentrations vary
spatially within a county. Some grid cells in a county can be below the level of the alternative
standard even though the highest monitor value is above the alternative standard. Thus,
emission reductions lead to benefits in grid cells that are below the alternative standards even
within a county with a monitor that exceeds the alternative standard. We have not estimated
the fraction of benefits that occur only in counties that exceed the alternative standards.
Table 5-27. Estimated Reduction in Incidence of Adult Premature Mortality Occurring Above
and Below the Lowest Measured Levels in the Underlying Epidemiology Studies
for 12/35 and 13/35a
Allocation of Reduced Mortality Incidence
Total Reduced
Study and Lowest Measured Level (LML) Mortality Incidence Below LML At or Above LML
12/35
Krewski et al. (2009) 5.8 ng/m3
Laden et al. (2006) 10 ng/m3
280
730
23
360
260
370
13/35
Krewski et al. (2009) 5.8 ng/m3 11 19
Laden et al. (2006) 10 ng/m3 27 10 17
Mortality incidence estimates are rounded to whole numbers and two significant digits, so estimates may not
sum across columns. It is important to emphasize that although we have lower levels of confidence in levels
below the LML for each study, the scientific evidence does not support the existence of a level below which
health effects from exposure to PM2.5 do not occur.
5-82
-------
Table 5-28. Percentage of Avoided Premature Deaths Occurring At or Above the Lowest
Measured Levels in the Underlying Epidemiology Studies for each Alternative
Combination of Primary PM2.s Standards3
Study and Lowest Measured Level (LML)
Krewski et al. (2009) 5.8 ng/m3
Laden et al. (2006) 10 ng/m3
13/35
89%
62%
12/35
92%
51%
11/35
95%
46%
11/30
93%
32%
It is important to emphasize that although we have lower levels of confidence in levels below the LML for each
study, the scientific evidence does not support the existence of a level below which health effects from exposure
to PM2.5 do not occur.
While the LML of each study is important to consider when characterizing and
interpreting the overall level PM2.5-related co-benefits, as discussed earlier in this chapter, EPA
believes that both cohort-based mortality estimates are suitable for use in air pollution health
impact analyses. When estimating PM-related premature deaths avoided using risk coefficients
drawn from the Laden et al. (2006) analysis of the Harvard Six Cities and the Krewski et al.
(2009) analysis of the ACS cohorts there are innumerable other attributes that may affect the
size of the reported effect estimates—including differences in population demographics, the
size of the cohort, activity patterns and particle composition among others. The LML
assessment presented here provides a limited representation of one key difference between
the two studies.
5-83
-------
25%
20%
15%
10%
5%
0%
LMLof Krewski et
al. (2009) study
2-3
5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 13-14 14-15 15-16 16-17 17-18 18-19 19-20
Baseline Annual Mean PM2 5 Level (pg/m3)
Of total PM2.5-Related deaths avoided for 12/35:
92% occur among populations exposed to PM2.5 levels at or above the LML of the Krewski et al. (2009) study.
51% occur among populations exposed to PM2.5 levels at or above the LML of the Laden et al. (2006) study.
Of total PM2.5-Related deaths avoided for 13/35:
89% occur among populations exposed to PM2.5 levels at or above the LML of the Krewski et al. (2009) study.
62% occur among populations exposed to PM2.5 levels at or above the LML of the Laden et al. (2006) study.
Figure 5-6. Number of Premature PM2.5-related Deaths Avoided for 12/35 and 13/35
According to the Baseline Level of PM2.5 and the Lowest Measured Air Quality Levels of Each
Mortality Study
5-84
-------
100%
90%
80%
70%
60%
73
1
50%
if
40%
0.
g
30%
20%
10%
0%
16
17
18
19
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Baseline Annual Mean PM25 Level (ug/m3)
Of total PM2.5-Related deaths avoided for 12/35:
92% occur among populations exposed to PM2 5 levels at or above the LML of the Krewski et al. (2009) study.
51% occur among populations exposed to PM2.5 levels at or above the LML of the Laden et al. (2006) study.
Of total PM2.5-Related deaths avoided for 13/35:
89% occur among populations exposed to PM2.5 levels at or above the LML of the Krewski et al. (2009) study.
62% occur among populations exposed to PM2.5 levels at or above the LML of the Laden et al. (2006) study.
Figure 5-7. Number of Premature PM2.5-related Deaths Avoided for 12/35 According to the
Baseline Level of PM2.5and the Lowest Measured Air Quality Levels of Each Mortality Study
5.7.5 Additional Sensitivity Analyses
The details of these sensitivity analyses are provided in appendix 5B, and summarized
here. The use of an alternate lag structure would change the PM2.5-related mortality benefits
discounted at 3% discounted by between 10% and -27%; when discounted at 7%, these
benefits change by between 22% and -52%. When applying higher and lower income growth
adjustments, the monetary value of PM2.5-related premature mortality changes between 33%
and -14%; the value of acute endpoints changes between 8% and -4%. Using the updated cost-
of-illness functions for hospital admissions, the rounded estimates of total monetized benefits
do not change, but the monetary value of respiratory hospital admissions increases 3.4% and
cardiovascular hospital admissions increase 2.1%. These results on a percentage basis would be
similar for alternative combinations of standards.
5-85
-------
5.8 Discussion
This analysis demonstrates the potential for significant health benefits of the illustrative
emission controls applied to simulate attainment with the alternative combination of primary
PM2.5 standards. We estimate that by 2020 the proposed standards would have reduced the
number of PM2.5-related premature mortalities and produce substantial non-mortality benefits.
This rule promises to yield significant welfare impacts as well, though the quantification of
those endpoints in this RIA is incomplete. Even considering the quantified and unquantified
uncertainties identified in this chapter, we believe that the implementing the proposed
standard would have substantial public health benefits that outweigh the costs.
Inherent in any complex RIA such as this one are multiple sources of uncertainty. Some
of these we characterized through our quantification of statistical error in the concentration
response relationships and our use of the expert elicitation-derived PM2.5 mortality functions.
Others, including the projection of atmospheric conditions and source-level emissions, the
projection of baseline morbidity rates, incomes and technological development are
unquantified. When evaluated within the context of these uncertainties, the health impact and
monetized benefits estimates in this RIA can provide useful information regarding the public
health benefits associated with a revised PM NAAQS.
There are important differences worth noting in the design and analytical objectives of
NAAQS RIAs compared to RIAs for implementation rules, such as the recent MATS rule (U.S.
EPA, 2011d). The NAAQS RIAs illustrate the potential costs and benefits of attaining a revised
air quality standard nationwide based on an array of emission control strategies for different
sources, incremental to implementation of existing regulations and controls needed to attain
current standards. In short, NAAQS RIAs hypothesize, but do not predict, the control strategies
that States may choose to enact when implementing a revised NAAQS. The setting of a NAAQS
does not directly result in costs or benefits, and as such, NAAQS RIAs are merely illustrative and
are not intended to be added to the costs and benefits of other regulations that result in
specific costs of control and emission reductions. By contrast, the emission reductions from
implementation rules are generally for specific, well-characterized sources, such as the recent
MATS rule (U.S. EPA, 2011d). In general, EPA is more confident in the magnitude and location of
the emission reductions for implementation rules. As such, emission reductions achieved under
promulgated implementation rules such as MATS have been reflected in the baseline of this
NAAQS analysis. Subsequent implementation rules will be reflected in the baseline for the next
PM NAAQS review. For this reason, the benefits estimated provided in this RIA and all other
NAAQS RIAs should not be added to the benefits estimated for implementation rules.
5-86
-------
In setting the NAAQS, EPA considers that PM2.5 concentrations vary over space and time.
While the standard is designed to limit concentrations at the highest monitor in an area, it is
understood that emission controls put in place to meet the standard at the highest monitor will
simultaneously result in lower PM2.5 concentrations throughout the entire area. In fact, the
Quantitative Risk and Exposure Assessment for Particulate Matter (U.S. EPA, 2010b) shows how
different standard levels would affect the entire distribution of PM2.5 concentrations, and thus
people's exposures and risk, across urban areas. For this reason, it is inappropriate to use the
NAAQS level as a bright line for health effects.
The NAAQS are not set at levels that eliminate the risk of air pollution completely.
Instead, the Administrator sets the NAAQS at a level requisite to protect public health with an
adequate margin of safety, taking into consideration effects on susceptible populations based
on the scientific literature. The risk analysis prepared in support of this PM NAAQS reported
risks below these levels, while acknowledging that the confidence in those effect estimates is
higher at levels closer to the standard (U.S. EPA, 2010b). While benefits occurring below the
standard may be somewhat more uncertain than those occurring above the standard, EPA
considers these to be legitimate components of the total benefits estimate. Though there are
greater uncertainties at lower PM2.5 concentrations, there is no evidence of a threshold in
PM2.5-related health effects in the epidemiology literature. Given that the epidemiological
literature in most cases has not provided estimates based on threshold models, there would be
additional uncertainties imposed by assuming thresholds or other non-linear concentration-
response functions for the purposes of benefits analysis.
5.9 References
Abbey, David E et al. 1995. "Chronic Respiratory Symptoms Associated with Estimated Long-
Term Ambient Concentrations of Fine Particulates less than 2.5 Microns in aerodynamic
diameter (PM2.5) and Other Air Pollutants." Journal of Exposure Analysis and
Environmental Epidemiology 5(2): 137-158.
Abbey, D.E., N. Nishino, W.F. McDonnell, R.J. Burchette, S.F. Knutsen, W. Lawrence Beeson, and
J.X. Yang. 1999. "Long-term inhalable particles and other air pollutants related to
mortality in nonsmokers." American Journal of Respiratory and Critical Care Medicine
159(2):373-382.
Abt Associates, Inc. 2011. BenMAP User's Manual Appendices. Prepared for U.S. Environmental
Protection Agency Office of Air Quality Planning and Standards. Research Triangle Park,
NC. September. Available on the Internet at <
http://www.epa.gov/air/benmap/models/BenMAPAppendicesSeptember2011.pdf>.
5-87
-------
Abt Associates, Inc. 2010. Environmental Benefits and Mapping Program (Version 4.0).
Bethesda, MD. Prepared for U.S. Environmental Protection Agency Office of Air Quality
Planning and Standards. Research Triangle Park, NC. Available on the Internet at
.
Adams PF, Hendershot GE, Marano MA. 1999. "Current Estimates from the National Health
Interview Survey, 1996." Vital Health Stat 10(200):1-212.
Agency for Healthcare Research and Quality (AHRQ). 2000. HCUPnet, Healthcare Cost and
Utilization Project. Rockville, MD. Available on the Internet at
.
Agency for Healthcare Research and Quality (AHRQ). 2007. HCUPnet, Healthcare Cost and
Utilization Project. Rockville, MD. Available on the Internet at
.
American Lung Association (ALA). 1999. Chronic Bronchitis. Available on the Internet at
.
American Lung Association (ALA). 2002. Trends in Asthma Morbidity and Mortality. American
Lung Association, Best Practices and Program Services, Epidemiology and Statistics Unit.
Available on the Internet at .
American Lung Association (ALA). 2010a. Trends in COPD (Chronic Bronchitis and Emphysema):
Morbidity and Mortality. American Lung Association Epidemiology and Statistics Unit
Research and Program Services Division. February. Table 4. Based on NHIS data (CDC,
2008). Available at http://www.lungusa.org/finding-cures/our-research/trend-
reports/copdtrend-report.pdf.
American Lung Association (ALA). 2010b. Trends in Asthma Morbidity and Mortality. American
Lung Association Epidemiology and Statistics Unit Research and Program Services
Division. February. Table 7. Based on NHIS data (CDC, 2008). Available at
http://www.lungusa.org/finding-cures/our-research/trend-reports/asthma-trend-
report.pdf.
Anderson HR, Bremner SA, Atkinson RW, Harrison RM, Walters S. 2001. "Particulate matter and
daily mortality and hospital admissions in the west midlands conurbation of the UK:
associations with fine and coarse particles, black smoke and sulphate." Occup Environ
Med 58:504-510.
Ayyub, B.M. 2002. Elicitation of Expert Opinions for Uncertainty and Risk. CRC Press, Florida.
Babin, S. M., H. S. Burkom, et al. 2007. "Pediatric patient asthma-related emergency
department visits and admissions in Washington, DC, from 2001-2004, and associations
with air quality, socio-economic status and age group." Environ Health 6: 9.
5-88
-------
Barnett, Adrian G. et al. 2006. "The Effects of Air Pollution on Hospitalizations for
Cardiovascular Disease in Elderly People in Australian and New Zealand Cities."
Environmental Health Perspectives 114(7): 1018-1023.
Belanger, K., W. Beckett, E. Triche, M.B. Bracken, T. Holford, P. Ren, J.E. McSharry, D.R. Gold,
T.A. Platts-Mills, and B.P. Leaderer. 2003. "Symptoms of Wheeze and Persistent Cough
in the First Year of Life: Associations with Indoor Allergens, Air Contaminants, and
Maternal History of Asthma." American Journal of Epidemiology 158:195-202.
Bell, M. L, K. Ebisu, et al. 2008. "Seasonal and Regional Short-term Effects of Fine Particles on
Hospital Admissions in 202 US Counties, 1999-2005." American Journal of Epidemiology
168(11): 1301-1310.
Berger, M.C., G.C. Blomquist, D. Kenkel, and G.S. Tolley. 1987. "Valuing Changes in Health Risks:
A Comparison of Alternative Measures." The Southern Economic Journal 53:977-984.
Centers for Disease Control and Prevention (CDC). 2003. Health, United States. Table 30. Years
of Potential Life Lost Before Age 75 for Selected Causes of Death, According to Sex, Race
and Hispanic Origin: United States, Selected Years 1980-2000. Available on the Internet
at .
Centers for Disease Control and Prevention (CDC). 2008. National Center for Health Statistics.
National Health Interview Survey, 1999-2008.
Centers for Disease Control and Prevention (CDC). 2011. United States Life Tables, 2007
National Vital Statistics Reports. Volume 59, Number 9. September. Available on the
Internet at .
Chay, K.Y., and M. Greenstone. 2003. "The Impact of Air Pollution on Infant Mortality: Evidence
from Geographic Variation in Pollution Shocks Induced by a Recession." Quarterly
Journal of Economics 118(3).
Chen L, Jennison BL, Yang W, Omaye ST. 2000. "Elementary school absenteeism and air
pollution." Inhal Toxicol 12(11):997-1016.
Cropper, M. L. and A. J. Krupnick. 1990. The Social Costs of Chronic Heart and Lung Disease.
Resources for the Future. Washington, DC. Discussion Paper QE 89-16-REV.
Grouse DL, Peters PA, van Donkelaar A, Goldberg MS, Villeneuve PJ, Brion 0, et al. 2012. "Risk
of Non-accidental and Cardiovascular Mortality in Relation to Long-term Exposure to
Low Concentrations of Fine Particulate Matter: A Canadian National-level Cohort Study."
Environ Health Perspect. In press, http://dx.doi.org/10.1289/ehp.1104049.
Dockery, D.W., C.A. Pope, X.P. Xu, J.D. Spengler, J.H. Ware, M.E. Fay, B.C. Ferris, and F.E.
Speizer. 1993. "An Association between Air Pollution and Mortality in Six U.S. Cities."
New England Journal of Medicine 329(24):1753-1759.
5-89
-------
Dockery, D.W., J. Cunningham, A.I. Damokosh, L.M. Neas, J.D. Spengler, P. Koutrakis, J.H. Ware,
M. Raizenne, and F.E. Speizer. 1996. "Health Effects of Acid Aerosols On North American
Children-Respiratory Symptoms." Environmental Health Perspectives 104(5):500-505.
Dominici, F., R.D. Peng, M.L Bell, L. Pham, A. McDermott, S.I. Zeger, J.M. Samet. 2006. "Fine
Particulate Air Pollution and Hospital Admission for Cardiovascular and Respiratory
Diseases." Journal of the American Medical Association 295:1127-1134.
Eftim SE, Samet JM, Janes H, McDermott A, Dominici F. 2008. "Fine particulate matter and mor-
tality: a comparison of the Six Cities and American Cancer Society cohorts with a
Medicare cohort." Epidemiology 19:209-216.
Eisenstein, E.L., L.K. Shaw, K.J. Anstrom, C.L. Nelson, Z. Hakim, V. Hasselblad and D.B. Mark.
2001. "Assessing the Clinical and Economic Burden of Coronary Artery Disease: 1986-
1998. "Medical Care 39(8):824-35.
Eisner M, N Anthonisen, D Coultas et al. 2010. "An official American Thoracic Society public
policy statement: novel risk factors and the global burden of chronic obstructive
pulmonary disease," Am J Respir Crit Care Med 182: 693-718.
Fann N, Lamson A, Wesson K, Risley D, Anenberg SC, Hubbell BJ. 2012a. "Estimating the
National Public Health Burden Associated with Exposure to Ambient PM2.sand Ozone.
Risk Analysis," Risk Analysis. 32(1): 81-95.
Fann N, Roman HR, Fulcher CF, Gentile M, Wesson K, Hubbell BJ, Levy Jl. 2012b. "Maximizing
Health Benefits and Minimizing Inequality: Incorporating Local Scale Data in the Design
and Evaluation of Air Quality Policies." Risk Analysis. 32(6): 908—922.
Freeman (III), AM. 1993. The Measurement of Environmental and Resource Values: Theory and
Methods. Washington, DC: Resources for the Future.
Can, W. Q., H. W. Davies, et al. 2012. "Association of Long-term Exposure to Community Noise
and Traffic-related Air Pollution With Coronary Heart Disease Mortality."/Amer/con
Journal of Epidemiology. In press, doi: 10.1093/aje/kwr424.
GeoLytics Inc. 2002. Geolytics CensusCD® 2000 Short Form Blocks. CD-ROM Release 1.0.
GeoLytics, Inc. East Brunswick, NJ. Available: http://www.geolytics.com/ [accessed 29
September 2004].
Gilliland FD, Berhane K, Rappaport EB, Thomas DC, Avol E, Gauderman WJ, et al. 2001. "The
effects of ambient air pollution on school absenteeism due to respiratory illnesses."
Epidemiology 12(l):43-54.
Hall JV, Brajer V, Lurmann FW. 2003. Economic Valuation of Ozone-related School Absences in
the South Coast Air Basin of California. Contemporary Economic Policy 21(4):407-417.
5-90
-------
Harrington, W., and P.R. Portney. 1987. "Valuing the Benefits of Health and Safety Regulation."
Journal of Urban Economics 22:101-112.
Hollman, F.W., T.J. Mulder, and J.E. Kalian. 2000. Methodology and Assumptions for the
Population Projections of the United States: 1999 to 2100. Population Division Working
Paper No. 38, Population Projections Branch, Population Division, U.S. Census Bureau,
Department of Commerce. January. Available on the Internet at
.
Hubbell, B., A. Hallberg, D.R. McCubbin, and E. Post. 2005. "Health-Related Benefits of Attaining
the 8-Hr Ozone Standard." Environmental Health Perspectives 113:73-82. ICD, 1979
Hubbell BL. 2006. "Implementing QALYs in the analysis of air pollution regulations."
Environmental and Resource Economics 34:365-384.
Industrial Economics, Incorporated (lEc). 2006. Expanded Expert Judgment Assessment of the
Concentration-Response Relationship Between PM2,5 Exposure and Mortality. Prepared
for: Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC. September. Available on the Internet at
.
Kochi, I., B. Hubbell, and R. Kramer. 2006. "An Empirical Bayes Approach to Combining
Estimates of the Value of Statistical Life for Environmental Policy Analysis."
Environmental and Resource Economics. 34: 385-406.
Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi, Y, et al. 2009. "Extended follow-up and
spatial analysis of the American Cancer Society study linking particulate air pollution and
mortality." HEI Research Report, 140, Health Effects Institute, Boston, MA.
Krewski, D., R.T. Burnett, M.S. Goldbert, K. Hoover, J. Siemiatycki, M. Jerrett, M. Abrahamowicz,
and W.H. White. 2000. "Reanalysis of the Harvard Six Cities Study and the American
Cancer Society Study of Particulate Air Pollution and Mortality." Special Report to the
Health Effects Institute. Cambridge MA. July.
Krupnick, A.J., and M.L. Cropper. 1992. "The Effect of Information on Health Risk Valuations."
Journal of Risk and Uncertainty 5(2):29-48.
5-91
-------
Kunzli, N., R. Kaiser, S. Medina, M. Studnicka, 0. Chanel, P. Filliger, M. Herry, F. HorakJr., V.
Puybonnieux-Texier, P. Quenel, J. Schneider, R. Seethaler, J-C Vergnaud, and H. Sommer.
2000. "Public-Health Impact of Outdoor and Traffic-Related Air Pollution: A European
Assessment." The Lancet 356:795-801.
Kunzli, N., S. Medina, R. Kaiser, P. Quenel, F. HorakJr, and M. Studnicka. 2001. "Assessment of
Deaths Attributable to Air Pollution: Should We Use Risk Estimates Based on Time Series
or on Cohort Studies?" American Journal of Epidemiology 153(ll):1050-55.
Laden, F., J. Schwartz, F.E. Speizer, and D.W. Dockery. 2006. "Reduction in Fine Particulate Air
Pollution and Mortality." American Journal of Respiratory and Critical Care Medicine
173:667-672.
Lave, L.B., and E.P. Seskin. 1977. Air Pollution and Human Health. Baltimore: Johns Hopkins
University Press for Resources for the Future.
Lepeule J, Laden F, Dockery D, Schwartz J 2012. "Chronic Exposure to Fine Particles and
Mortality: An Extended Follow-Up of the Harvard Six Cities Study from 1974 to 2009."
Environ Health Perspect In press, http://dx.doi.org/10.1289/ehp.1104660
Levy Jl, Baxter LK, Schwartz J. 2009. "Uncertainty and variability in health-related damages from
coal-fired power plants in the United States." Risk Anal. 29(7) 1000-1014.
Lipfert, F.W., H. Mitchell Perry Jr., J. Philip Miller, Jack D. Baty, Ronald E. Wyzg, and Sharon E.
Carmody. 2000. "The Washington University-EPRI Veterans' Cohort Mortality Study:
Preliminary Results." Inhalation Toxicology 12:41-74.
Lipfert, F.W., S.C. Morris, and R.E. Wyzga. 1989. "Acid Aerosols—the Next Criteria Air
Pollutant." Environmental Science & Technology 23(11):1316-1322.
Lipfert, F.W.; Perry, H.M., Jr.; Miller, J.P.; Baty, J.D.; Wyzga, R.E.; Carmody, S.E. 2003. "Air
Pollution, Blood Pressure, and Their Long-Term Associations with Mortality." Inhalation
Toxicology. 15, 493-512.
Lipfert, F.W.; Wyzga, R.E.; Baty, J.D.; Miller, J.P. 2006. "Traffic Density as a Surrogate Measure of
Environmental Exposures in Studies of Air Pollution Health Effects: Long-Term Mortality
in a Cohort of US Veterans." Atmospheric Environment 40: 154-169.
Lippmann M, Ito K, Nadas A, Burnett RT. 2000. "Association of particulate matter components
with daily mortality and morbidity in urban populations." Res Rep Health Eff Inst.
Aug;(95):5-72, discussion 73-82.
Lisabeth, L.D. et al. 2008. "Ambient air pollution and risk for ischemic stroke and transient
ischemic attack." Annals of Neurology. July (64): 53—59.
5-92
-------
Mansfield, Carol; Paramita Sinha; Max Henrion. 2009. Influence Analysis in Support of
Characterizing Uncertainty in Human Health Benefits Analysis: Final Report. Prepared
for U.S. EPA, Office of Air Quality Planning and Standards. November. Available on the
internet at
-------
Neumann, J.E., M.T. Dickie, and R.E. Unsworth. 1994. "Linkage Between Health Effects
Estimation and Morbidity Valuation in the Section 812 Analysis—Draft Valuation
Document." Industrial Economics Incorporated (lEc) Memorandum to Jim DeMocker,
U.S. Environmental Protection Agency, Office of Air and Radiation, Office of Policy
Analysis and Review. March 31. Available on the Internet at
.
Office of Management and Budget (OMB). 2003. Circular A-4: Regulatory Analysis. Washington,
DC. Available on the Internet at
Ostro, B., M. Lipsett, J. Mann, H. Braxton-Owens, and M. White. 2001. "Air Pollution and
Exacerbation of Asthma in African-American Children in Los Angeles." Epidemiology
12(2):200-208.
Ostro, B.D. 1987. "Air Pollution and Morbidity Revisited: A Specification Test." Journal of
Environmental Economics Management 14:87-98.
Ostro, B.D. and S. Rothschild. 1989. "Air Pollution and Acute Respiratory Morbidity: An
Observational Study of Multiple Pollutants." Environmental Research 50:238-247.
Ozkaynak, H., and G.D. Thurston. 1987. "Associations between 1980 U.S. Mortality Rates and
Alternative Measures of Airborne Particle Concentration." Risk Analysis 7(4):449-461.
Patel, M. M., S. N. Chillrud, J. C. Correa, Y. Hazi, M. Feinberg, D. Kc, S. Prakash, J. M. Ross, D.
Levy and P. L. Kinney. 2010. "Traffic-Related Particulate Matter and Acute Respiratory
Symptoms among New York City Area Adolescents." Environ Health Perspect.
Sep;118(9):1338-43.
Peel JL; Metzger KB; Klein M; Flanders WD; Mulholland JA; Tolbert PE. 2007. "Ambient air
pollution and cardiovascular emergency department visits in potentially sensitive
groups." Am J Epidemiol 165: 625-633.
Peng, R. D., M. L. Bell, et al. 2009. "Emergency admissions for cardiovascular and respiratory
diseases and the chemical composition of fine particle air pollution." Environ Health
Perspect 117(6): 957-63.
Peng, R. D., H. H. Chang, et al. 2008. "Coarse particulate matter air pollution and hospital
admissions for cardiovascular and respiratory diseases among Medicare patients. "JAMA
299(18): 2172-9.
Peters A. 2005. "Particulate matter and heart disease: evidence from epidemiological studies."
ToxicolAppI Pharmacol. Sep 1;207(2 Suppl):477-82.
Peters A. 2006. "When a myocardial infarction comes out of the not-so-blue air." Circulation.
Dec5;114(23):2430-l
5-94
-------
Peters, A., D. W. Dockery, J. E. Muller and M. A. Mittleman. 2001. "Increased particulate air
pollution and the triggering of myocardial infarction." Circulation. Vol. 103 (23): 2810-5.
Poloniecki, J.D., R.W. Atkinson., A.P. de Leon., and H.R. Anderson. 1997. "Daily Time Series for
Cardiovascular Hospital Admissions and Previous Day's Air Pollution in London, UK."
Occupational and Environmental Medicine 54(8):535-540
Pope, C.A., III, R.T. Burnett, G.D. Thurston, M.J. Thun, E.E. Calle, D. Krewski, and J.J. Godleski.
2004. "Cardiovascular Mortality and Long-term Exposure to Particulate Air Pollution."
Circulation 109: 71-77.
Pope, C.A., III, R.T. Burnett, M.J. Thun, E.E. Calle, D. Krewski, K. Ito, and G.D. Thurston. 2002.
"Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate
Air Pollution." Journal of the American Medical Association 287:1132-1141.
Pope, C.A., III, D.W. Dockery, J.D. Spengler, and M.E. Raizenne. 1991. "Respiratory Health and
PMio Pollution: A Daily Time Series Analysis." American Review of Respiratory Diseases
144:668-674.
Pope, CA III, E Majid, D Dockery. 2009. "Fine Particle Air Pollution and Life Expectancy in the
United States." New England Journal of Medicine 360: 376-386.
Pope, C. A., Ill, J. B. Muhlestein, et al. 2006. "Ischemic heart disease events triggered by short-
term exposure to fine particulate air pollution." Circulation 114(23): 2443-8.
Pope, C.A., III, M.J. Thun, M.M. Namboodiri, D.W. Dockery, J.S. Evans, F.E. Speizer, and C.W.
Heath, Jr. 1995. "Particulate Air Pollution as a Predictor of Mortality in a Prospective
Study of U.S. Adults." American Journal of Respiratory Critical Care Medicine 151:669-
674.
Puett RC, Hart JE, Suh H, Mittleman M, Laden F. 2011. "Particulate Matter Exposures, Mortality
and Cardiovascular Disease in the Health Professionals Follow-up Study." Environ Health
Perspect. August; 119(8): 1130-1135.
Puett RC, Hart JE, Yanosky JD, Paciorek C, Schwartz J, Suh H, et al. 2009. "Chronic fine and
coarse particulate exposure, mortality and coronary heart disease in the Nurses' Health
Study." Environ Health Perspect 117:1697-1701.
Roman, Henry A., Katherine D. Walker, Tyra L. Walsh, Lisa Conner, Harvey M. Richmond, Bryan
J. Hubbell, and Patrick L. Kinney. 2008. "Expert Judgment Assessment of the Mortality
Impact of Changes in Ambient Fine Particulate Matter in the U.S." Environ. Sci. Technol.,
42(7):2268-2274.
Rosamond, W., G. Broda, E. Kawalec, S. Rywik, A. Pajak, L. Cooper, and L. Chambless. 1999.
"Comparison of Medical Care and Survival of Hospitalized Patients with Acute
Myocardial Infarction in Poland and the United States." American Journal of Cardiology
83:1180-1185
5-95
-------
Rowe, R.D., and L.G. Chestnut. 1986. "Oxidants and Asthmatics in Los Angeles: A Benefits
Analysis— Executive Summary." Prepared by Energy and Resource Consultants, Inc.
Report to the U.S. Environmental Protection Agency, Office of Policy Analysis. EPA- 230-
09-86-018. Washington, DC.
Rowlatt, P., Spackman, M., Jones, S., Jones-Lee, M., Loomes, G. 1998. Valuation of deaths from
air pollution. A Report for the Department of Environment, Transport and the Regions
and the Department of Trade and Industry. National Economic Research Associates
(NERA), London.
Russell, M.W., D.M. Huse, S. Drowns, E.G. Hamel, and S.C. Hartz. 1998. "Direct Medical Costs of
Coronary Artery Disease in the United States." American Journal of Cardiology
Schwartz, J. 1993. "Particulate Air Pollution and Chronic Respiratory Disease." Environment
Research 62:7-13.
Schwartz, J., and L.M. Neas. 2000. "Fine Particles are More Strongly Associated than Coarse
Particles with Acute Respiratory Health Effects in Schoolchildren." Epidemiology 11:6-
10.
Schwartz J, Dockery DW, Neas LM, Wypij D, Ware JH, Spengler JD, Koutrakis P, Speizer FE, Ferris
BG Jr. 1994. "Acute effects of summer air pollution on respiratory symptom reporting in
children." American Journal of Respiratory and Critical Care Medicine 150(5 Pt 1):1234-
42.
Schwartz J, Coull B, Laden F. 2008. "The Effect of Dose and Timing of Dose on the Association
between Airborne Particles and Survival." Environmental Health Perspectives 116: 64-
69.
Sheppard, L. 2003. "Ambient Air Pollution and Nonelderly Asthma Hospital Admissions in
Seattle, Washington, 1987-1994." In Revised Analyses of Time-Series Studies of Air
Pollution and Health. Special Report. Boston, MA: Health Effects Institute.
Shogren, J., and T. Stamland. 2002. "Skill and the Value of Itfe." Journal of Political Economy
110:1168-1197.
Slaughter, James C et al. 2005. "Association between particulate matter and emergency room
visits, hospital admissions and mortality in Spokane, Washington. " Journal of Exposure
Analysis and Environmental Epidemiology (2005) 15, 153-159.
doi:10.1038/sj.jea.7500382
Smith, D.H., D.C. Malone, K.A. Lawson, L.J. Okamoto, C. Battista, and W.B. Saunders. 1997. "A
National Estimate of the Economic Costs of Asthma." American Journal of Respiratory
and Critical Care Medicine 156(3 Pt l):787-793.
5-96
-------
Smith, K., Pattanayak SK, Van Houtven G. 2006. "Structural benefit transfer: An example using
VSL estimates." Ecological Economics 60: 361-371.
Standard and Poor's. 2000. "The U.S. Economy: The 25 Year Focus." Winter. Available on the
Internet at http://:www.standardandpoors.com
Stanford, R., T. Mclaughlin and L. J. Okamoto. 1999. "The cost of asthma in the emergency
department and hospital." Am J Respir Crit Care Med. 160 (1): 211-5.
Sullivan, J., L. Sheppard, et al. 2005. "Relation between short-term fine-particulate matter
exposure and onset of myocardial infarction." Epidemiology 16(1): 41-8.
Tagaris E, Liao KJ, Delucia AJ, et al. 2009. "Potential impact of climate change on air-pollution
related human health effects." Environ. Sci. Technol. 43: 4979-4988.
Tolbert, P. E., M. Klein, et al. 2007. "Multipollutant modeling issues in a study of ambient air
quality and emergency department visits in Atlanta." J Expo Sci Environ Epidemiol 17
Suppl 2:529-35.
Tolley, G.S. et al. 1986. Valuation of Reductions in Human Health Symptoms and Risks.
University of Chicago. Final Report for the U.S. Environmental Protection Agency, Office
of Policy Analysis. January. Available at
.
U.S. Bureau of Economic Analysis (U.S. BEA). 2004. BEA Economic Areas (EAs). U.S. Department
of Commerce. Available on the Internet at
.
U.S. Bureau of Economic Analysis (U.S. BEA). 1997. Real Gross Domestic Product—Table 2A. U.S.
Department of Commerce.
U.S. Bureau of Economic Analysis (U.S. BEA). 1997. The Economic and Budget Outlook: An
Update, Table 4—Economic Projections for Calendar Years 1997 Through 2007.
September. U.S. Department of Commerce.
U.S. Bureau of the Census. 2001. Statistical Abstract of the United States, 2001, Section 12:
Labor Force, Employment, and Earnings, Table No. 521. Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA). 1999a. The Benefits and Costs of the Clean Air
Act, 1990-2010. Prepared for U.S. Congress by U.S. EPA, Office of Air and
Radiation/Office of Policy Analysis and Review, Washington, DC, November; EPA report
no. EPA-410-R-99-001. Available on the Internet at
.
5-97
-------
U.S. Environmental Protection Agency (U.S. EPA). 2000. Guidelines for Preparing Economic
Analyses. EPA 240-R-00-003. National Center for Environmental Economics, Office of
Policy Economics and Innovation. Washington, DC. September. Available on the Internet
at
.
U.S. Environmental Protection Agency (U.S. EPA). 2004. Air Quality Criteria for Particulate
Matter. National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
EPA/600/P-99/002bF. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2006a. Regulatory Impact Analysis, 2006
National Ambient Air Quality Standards for Particulate Matter, Chapter 5. Office of Air
Quality Planning and Standards, Research Triangle Park, NC. October. Available on the
Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2006b. Air Quality Criteria for Ozone and
Related Photochemical Oxidants (Final). EPA/600/R-05/004aF-cF. Washington, DC: U.S.
EPA. February. Available on the Internet at
http://cfpub.epa.gov/ncea/CFM/recordisplay.cfm?deid=149923.U.S. Environmental
Protection Agency (U.S. EPA). 2008. Regulatory Impact Analysis, 2008 National Ambient
Air Quality Standards for Ground-level Ozone, Chapter 6. Office of Air Quality Planning
and Standards, Research Triangle Park, NC. March. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2008a. Regulatory Impact Analysis, 2008
National Ambient Air Quality Standards for Ground-level Ozone, Chapter 6. Office of Air
Quality Planning and Standards, Research Triangle Park, NC. March. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2008b. Integrated Science Assessment for
Oxides of Nitrogen—Health Criteria (Final Report). National Center for Environmental
Assessment, Research Triangle Park, NC. July. Available on the Internet at
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=194645
U.S. Environmental Protection Agency (U.S. EPA). 2008c. Integrated Science Assessment for
Sulfur Oxides—Health Criteria (Final Report). National Center for Environmental
Assessment, Research Triangle Park, NC. September. Available on the Internet at
.
5-98
-------
U.S. Environmental Protection Agency (U.S. EPA). 2009a. Regulatory Impact Analysis: National
Emission Standards for Hazardous Air Pollutants from the Portland Cement
Manufacturing Industry. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. April. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2009b. Integrated Science Assessment for
Particulate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment—RTP Division. December. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010a. Final Regulatory Impact Analysis (RIA)
for the N02 National Ambient Air Quality Standards (NAAQS). Office of Air Quality
Planning and Standards, Research Triangle Park, NC. January. Available on the Internet
at .
U.S. Environmental Protection Agency (U.S. EPA). 2010b. Quantitative Health Risk Assessment
for Particulate Matter—Final Report. EPA-452/R-10-005. Office of Air Quality Planning
and Standards, Research Triangle Park, NC. September. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010c. Final Regulatory Impact Analysis (RIA)
for theS02National Ambient Air Quality Standards (NAAQS). Office of Air Quality
Planning and Standards, Research Triangle Park, NC. June. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010d. Regulatory Impact Analysis (RIA) for
the Proposed Reconsideration of the Ozone National Ambient Air Quality Standards
(NAAQS). Office of Air Quality Planning and Standards, Research Triangle Park, NC.
January. Available on the Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2010e. Guidelines for Preparing Economic
Analyses. EPA 240-R-10-001. National Center for Environmental Economics, Office of
Policy Economics and Innovation. Washington, DC. December. Available on the Internet
at .
5-99
-------
U.S. Environmental Protection Agency (U.S. EPA). 2010g. Proposed Regulatory Impact Analysis
(RIA)forthe Transport Rule. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. January. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2011a. The Benefits and Costs of the Clean Air
Act 1990 to 2020: EPA Report to Congress. Office of Air and Radiation, Office of Policy,
Washington, DC. March. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2011b. Policy Assessment for the Review of
the Particulate Matter National Ambient Air Quality Standards. EPA-452/D-11-003. April.
Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2011c. Regulatory Impact Analysis for the
Federal Implementation Plans to Reduce Interstate Transport of Fine Particulate Matter
and Ozone in 27 States; Correction of SIP Approvals for 22 States. June. Available on the
Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2011d. Regulatory Impact Analysis for the
Final Mercury and Air Toxics Standards. EPA-452/R-11-011. December. Available on the
Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2011e. Integrated Science Assessment for
Ozone and Related Photochemical Oxidants (Second External Review Draft). EPA/600/R-
10/076B. U.S. Environmental Protection Agency, National Center for Environmental
Assessment—RTP Division. September. Available at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 1999. An SAB
Advisory on the Clean Air Act Section 812 Prospective Study Health and Ecological Initial
Studies. Prepared by the Health and Ecological Effects Subcommittee (HEES) of the
Advisory Council on the Clean Air Compliance Analysis, Science Advisory Board, U.S.
Environmental Protection Agency. Washington DC. EPA-SAB-COUNCIL-ADV-99-005.
Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2000. An SAB
Report on EPA's White Paper Valuing the Benefits of Fatal Cancer Risk Reduction. EPA-
SAB-EEAC-00-013. July. Available on the Internet at
.
5-100
-------
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004a. Advisory
on Plans for Health Effects Analysis in the Analytical Plan for EPA's Second Prospective
Analysis-Benefits and Costs of the Clean Air Act, 1990-2020. EPA-SAB-COUNCIL-ADV-
04-002. March. Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004b. Review
of the Draft Analytical Plan for EPA's Second Prospective Analysis—Benefits and Costs of
the Clean Air Act, 1990-2020: An Advisory by the Advisory Council for Clean Air
Compliance Analysis. EPA-SAB-COUNCIL-ADV-04-004. May. Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004c. Advisory
Council on Clean Air Compliance Analysis Response to Agency Request on Cessation Lag.
EPA-COUNCIL-LTR-05-001. December. Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2007. SAB
Advisory on EPA's Issues in Valuing Mortality Risk Reduction. EPA-SAB-08-001. October.
Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2008.
Characterizing Uncertainty in Particulate Matter Benefits Using Expert Elicitation. EPA-
COUNCIL-08-002. July. Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2009a.
Consultation on EPA's Particulate Matter National Ambient Air Quality Standards: Scope
and Methods Plan for Health Risk and Exposure Assessment. EPA-COUNCIL-09-009. May.
Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2009b. Review
of EPA's Integrated Science Assessment for Particulate Matter (First External Review
Draft, December 2008). EPA-COUNCIL-09-008. May. Available on the Internet at
.
5-101
-------
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2009c. Review
of Integrated Science Assessment for Particulate Matter (Second External Review Draft,
July 2009). EPA-CASAC-10-001. November. Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2010a. Review of
EPA's DRAFT Health Benefits of the Second Section 812 Prospective Study of the Clean Air
Act. EPA-COUNCIL-10-001. June. Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2010b. Review of
Risk Assessment to Support the Review of the Particulate Matter (PM) Primary National
Ambient Air Quality Standards—External Review Draft (September 2009). EPA-CASAC-
10-003. Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2010c. CASAC
Review of Quantitative Health Risk Assessment for Particulate Matter—Second External
Review Draft (February 2010). EPA-CASAC-10-008. Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2010d. CASAC
Review of Policy Assessment for the Review of the PM NAAQS—Second External Review
Draft (June 2010). EPA-CASAC-10-015. Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2011. Review of
Valuing Mortality Risk Reductions for Environmental Policy: A White Paper (December
10, 2010). EPA-SAB-11-011 July. Available on the Internet at
.
Viscusi, V.K., and J.E. Aldy. 2003. "The Value of a Statistical Life: A Critical Review of Market
Estimates throughout the \Nor\d." Journal of Risk and Uncertainty 27(l):5-76.
Viscusi, W.K., W.A. Magat, and J. Huber. 1991. "Pricing Environmental Health Risks: Survey
Assessments of Risk-Risk and Risk-Dollar Trade-Offs for Chronic Bronchitis." Journal of
Environmental Economics and Management 21:32-51.
5-102
-------
Wittels, E.H., J.W. Hay, and A.M. Gotto, Jr. 1990. "Medical Costs of Coronary Artery Disease in
the United States." American Journal of Cardiology 65(7):432-440.
Woodruff, T.J., J. Grille, and K.C. Schoendorf. 1997. "The Relationship Between Selected of
postneonatal infant mortality and particulate air pollution in the United States."
Environmental Health Perspectives. Vol. 105 (6): 608-612.
Woodruff TJ, Parker JD, Schoendorf KC. 2006. "Fine particulate matter (PM2.5) air pollution and
selected causes of postneonatal infant mortality in California." Environmental Health
Perspectives 114(5):786-90.
Woods & Poole Economics Inc. 2007. Population by Single Year of Age. CD-ROM. Woods &
Poole Economics, Inc. Washington, D.C.
World Health Organization (WHO). 1977. International Classification of Diseases, 9th Revision
(ICD-9). Geneva: WHO.
World Health Organization (WHO). 2008. Part 1: Guidance Document on Characterizing and
Communicating Uncertainty in Exposure Assessment, Harmonization Project Document
No. 6. Published under joint sponsorship of the World Health Organization, the
International Labour Organization and the United Nations Environment Programme.
WHO Press: Geneva, Switzerland. Available on the Internet at
.
Zanobetti, A., M. Franklin, et al. 2009. "Fine particulate air pollution and its components in
association with cause-specific emergency admissions." Environmental Health 8: 58-
60.Zeger S; Dominici F; McDermott A; Samet J. 2008. "Mortality in the Medicare
population and chronic exposure to fine particulate air pollution in urban centers (2000-
2005)." Environ Health Perspect 116:1614-1619.
Zanobetti A. and Schwartz, J. 2006. "Air pollution and emergency admissions in Boston, MA." J
Epidemiol Community Health 60(10): 890-5.
5-103
-------
APPENDIX 5.A
DISTRIBUTION OF THE PM2.5-RELATED BENEFITS
5.A.1 Overview
EPA is developing new approaches and metrics to improve its characterization of the
impacts of EPA rules on different populations. This analysis reflects one such approach, which
attempts to answer two questions regarding the distribution of PM2.5-related benefits resulting
from the illustrative attainment of a more stringent National Ambient Air Quality Standards
(NAAQS) for particulate matter (PM):
1. What is the baseline distribution of PM2.5-related mortality risk according to the
race, income and education of the population living within areas projected to exceed
alternative combinations of primary PM2.5 standards?
2. How would air quality improvements within these counties change the distribution
of risk among populations of different races—particularly those populations at
greatest risk in the baseline?1
There are important methodological differences between this distributional analysis and
the Environmental Justice analyses accompanying the Regulatory Impact Analyses (RIAs) for the
Cross-State Air Pollution Rule and the Mercury and Air Toxics Standard that are worth noting
here. These latter two RIAs applied photochemical modeling to characterize the change in
population exposure to PM2.5 after the implementation of well-characterized emission controls
on Electricity Generating Units. By contrast, this RIA aims to illustrate the potential benefits and
costs of attaining alternative primary PM2.5 standards. For this reason, similar to the main
benefits analysis in this RIA, we performed monitor rollbacks to just attain the alternative
combinations of primary standards following the approach described in Chapter 2 of this RIA.
A limitation of this approach to characterizing improvements in PM2.5 air quality is that
populations in each projected nonattainment area share the exposure reductions equally; this
is because simple rollbacks do not reflect the spatial heterogeneity in PM2.5 changes one would
expect from a modeled attainment strategy. However, as EPA demonstrated in the Detroit
multi-pollutant pilot project, states can design attainment strategies to maximize air quality
improvements among those populations at greatest risk of air pollution health impacts—which
both maximizes overall benefits while lowering the level of risk inequality (Fann et al., 2012a).
1 In this analysis we assess the change in risk among populations of different race and educational attainment. As
we discuss further in the methodology, we consider this last variable because of the availability of education-
modified PM2.5 mortality risk estimates.
5.A-1
-------
In this analysis we estimated that in 2020, prior to the attainment of a more stringent
PM standard, the level of PM2.5 mortality risk is not distributed equally among populations of
different levels of educational attainment—though the level of mortality risk appears to be
shared fairly equally among populations of different races. We find that attaining a more
stringent alternative annual PM NAAQS level of 12 u.g/m3 in conjunction with a 24-hour
standard of 35 u.g/m3 (as an illustrative example) would provide air quality improvements, and
lower PM2.5-related mortality risk, by a fairly consistent margin among minority populations.
We note that while the methods used for this analysis have been employed in recent EPA
Regulatory Impact Assessments (U.S. EPA, 2011) and are drawn from techniques described in
the peer reviewed literature (Fann et al., 2012b) EPA will continue to modify these approaches
based on evaluation of the methods.
5.A.2 Methodology
As a first step, we identify the counties exceeding an annual standard of 12 u.g/m3 in
conjunction with a 24-hour standard of 35 u.g/m3 in 2020, using the results of the baseline
CMAQ air quality modeling. This air quality modeling simulation projects PM2.5 levels after the
incorporation of all "on the books" rules (i.e., those promulgated at the time the air quality
modeling was performed), but does not reflect the illustrative attainment strategies. We next
identified the counties whose PM2.5 levels exceed the alternative combinations of PM2.5
standards. We then performed a monitor rollback to adjust the annual PM2.5 levels in each
county such that they attain the alternative combinations of PM2.5 standards. This approach
provides us with baseline and rolled-back PM2.5 levels that attain this combination of annual
and daily PM2.5 standards. Within each county exceeding the this combination of PM2.5
standards, we estimate the level of all-cause PM2.5 mortality risks for adult populations as well
as the level of PM2.5 mortality risk according to the race and educational attainment of the
population.
Our approach to calculating PM2.5 mortality risk is generally consistent with the primary
analysis with two exceptions: the PM mortality risk coefficients used to quantify impacts and
the baseline mortality rates used to calculate education-modified mortality impacts (a detailed
discussion of how both the mortality risk coefficients and baseline incidence rates are used to
estimate the incidence of PM2.5-related deaths may be found in the benefits chapter). Within
both this and other analyses of the ACS cohort (see: Krewski et al., 2000), educational
attainment has been found to be inversely related to the risk of all-cause mortality. That is,
populations with lower levels of education (in particular, < grade 12) are more vulnerable to
PM2.5-related mortality. Krewski and colleagues note that "...the level of education attainment
5.A-2
-------
may likely indicate the effects of complex and multi-factorial socioeconomic processes on
mortality...," factors that we would like to account for in this distributional assessment. When
estimating PM mortality impacts among populations according to level of education, we
applied PM2.5 mortality risk coefficients modified by educational attainment: less than grade 12
(Relative Risk (RR) = 1.082, 95% confidence intervals 1.024—1.144 per 10 u.g/m3 change), grade
12 (RR = 1.072, 95% confidence intervals 1.020—1.127 per 10 u.g/m3 change), and greater than
grade 12 (RR = 1.055, 95% confidence intervals 1.018—1.094 per 10 u.g/m3 change). The Pope
et al. (2002) study, which EPA has frequently relied upon to quantify PM-related mortality, does
not provide education-stratified RR estimates. The principal reason we applied risk estimates
from the Krewski study was to ensure that the risk coefficients used to estimate all-cause
mortality risk and education-modified mortality risk were drawn from a consistent modeling
framework and because the use of the Krewski study is consistent with the primary benefits
analysis.
The other key difference between this distributional analysis and the main benefits
analysis for this rule relates to the baseline mortality rates. As described in Chapter 5 of this
RIA, we calculate PM2.5-related mortality risk relative to baseline mortality rates in each county.
Traditionally, for benefits analysis, we have applied county-level age- and sex-stratified baseline
mortality rates when calculating mortality impacts (Abt Associates, 2010). To calculate PM2.5
impacts by race, we incorporated race-specific (stratified by White/Black/Asian/Native
American) baseline mortality rates. This approach improves our ability to characterize the
relationship between race and PM2.5-related mortality however, we do not have a differential
concentration-response function as we do for education, and as a result, we are not able to
capture the full impacts of race in modifying the benefits associated with reductions in PM2.5.
Table 5.A-1 summarizes the key attributes of the two distributional assessments.
Table 5.A-1. Key Attributes of the Distributional Analyses in this Appendix
Distributional Analysis
Input parameter Education-modified PM Mortality Risk Race-stratified PM mortality risk
Effect coefficient Stratified by education attainment (<12, All-cause, applied to each
=12, >12) population subgroup
Baseline mortality rates Cause, age and sex stratified Cause, age, sex race and ethnicity
stratified
5.A-3
-------
The result of this analysis is a distribution of PM2.5 mortality risk estimates by county,
stratified by each of the two population variables (race and educational attainment). We have
less confidence in county-level estimates of mortality than the national or even state estimates.
However, the modeling down to the county level can be considered reasonable because the
estimates are based on monitored air quality modeling estimates of PM2.5, county level baseline
mortality rates, and a concentration-response function that is derived from county level data.
We next identified the counties projected to exceed the current combination of annual and
daily PM2.5 standards (15/35) ("baseline") in 2020. The second step of the analysis was to repeat
the sequence above by estimating PM2.5 mortality risk in counties projected to exceed an
illustrative combination of PM2.5 standards (12/35) after rolling back monitor values to reach
attainment in 2020.
5.A.3 Results
Figures 5.A-1 and 5.A-2 summarize the change in the median level of PM2.5 mortality risk
among populations stratified by educational attainment and race in non-attaining counties. The
percentage of deaths due to PM2.5 among populations with less than a grade 12 education is
significantly higher than those who have either completed high school or who have attained an
education level greater than high school. This finding is consistent with the relative levels of risk
coefficients for each population, where we apply a much larger risk coefficient for populations
with less than a grade 12 education. The level of risk reduction between the baseline and 12/35
is roughly equal between the three groups.
In Figure 5.A-2, Black and Native American populations are at significantly greater PM2.5
mortality risk in the baseline, as compared to other races. White and Asian populations are at
lower levels of PM2.5 mortality risk. The finding that black populations are at greater PM2.5
mortality risk in the baseline may be due both to the elevated baseline mortality risks or greater
exposure to PM2.5 among this population. After attaining 12/35, populations of all races benefit,
though the reduction in PM mortality risk among whites is within rounding error.
5.A-4
-------
• Arttrrwit with I MS
«V *
i/i
i
5%
JV
< Grade 12 'Grade 12
G duuuhm At ttttnm*n(
Figure 5.A-1. PM2.5 Mortality Risk Modified by Educational Attainment in Counties Projected
to Exceed 12/35
5.A-5
-------
ELiS
» KegL* very &K*i Ira
*
5.1%
&
4.7%
4.6%
Qucfc White Aiiw* rttd^i Afturirtrt
Figure 5.A-2. PM2.s Mortality Risk by Race in Counties Projected to Exceed 12/35
Table 5.A-2. Numerical Values Used for Figures 5.A-1 and 5.A-2 Above3
Year
Impacts by education
< Grade 12
= Grade 12
> Grade 12
Impacts by race
Asian
Black
Native American
White
Scenario, Percent
Baseline
8%
7.1%
5.5%
4.9%
5.2%
5%
4.9%
12/35
6.5%
5.7%
4.4%
4.5%
5%
4.8%
4.9%
Estimates expressed with a greater number of significant digits to facilitate comparisons among values.
5.A-6
-------
5.A.4 Discussion
This analysis is subject to certain limitations, some of which we note above but are
worth repeating here. First, the change in the distribution of PM2.5-related mortality risk we
estimate here depends is influenced strongly by the simulated attainment strategy. While we
performed simple monitor rollbacks to attain a more stringent standard, we describe other
approaches above that may maximize human health benefits while also reducing the level of
risk inequality. The monitor rollback approach employed here simulates improvements in PM2.5
levels in proximity to monitors projected to exceed a tighter PM NAAQS; we would expect an
attainment strategy to achieve air quality improvements over a broader geographic area,
affecting a greater portion the population than we have reflected here.
Notwithstanding these uncertainties, these results suggest that all populations,
irrespective of education attainment or race, living in locations projected to exceed an
illustrative annual standard of 12 u.g/m3 in conjunction with a 24-hour standard of 35 u.g/m3 in
2020 would experience a reduction in PM-related mortality risk. Certain sub-populations,
including those with less than a grade 12 education and Native Americans, area at an elevated
risk in the baseline. Attainment of this illustrative standard in 2020 would reduce the level of
mortality risk among these sub-populations.
5.A.5 References
Abt Associates, Inc. 2010. Environmental Benefits and Mapping Program (Version 4.0).
Bethesda, MD. Prepared for U.S. Environmental Protection Agency Office of Air Quality
Planning and Standards. Research Triangle Park, NC. Available on the Internet at
.
Fann N, Roman HR, Fulcher CF, Gentile M, Wesson K, Hubbell BJ, Levy Jl. 2012a. "Maximizing
Health Benefits and Minimizing Inequality: Incorporating Local Scale Data in the Design
and Evaluation of Air Quality Policies." Risk Analysis 31(6): 908-222.
Fann N, Lamson A, Wesson K, Risley D, Anenberg SC, Hubbell BJ. 2012b. "Estimating the
National Public Health Burden Associated with Exposure to Ambient PM2.sand Ozone."
Risk Analysis 32(1): 81-95.
Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi Y, Turner C, Pope CA, Thurston G, Calle
EE, Thunt MJ. 2009. Extended follow-up and spatial analysis of the American Cancer
Society study linking particulate air pollution and mortality. HEI Research Report, 140,
Health Effects Institute, Boston, MA.
5.A-7
-------
Krewski D, Burnett RT, Goldberg MS, et al. 2000. Reanalysis of the Harvard Six Cities Study and
the American Cancer Society Study of Particulate Air Pollution and Mortality: Special
Report. Cambridge, Mass: Health Effects Institute.
Pope CA, Burnett RT, Thun MJ, Calle EE, Krewski D, Ito K, Thurston GD. 2002. "Lung Cancer,
Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate Air Pollution."
Journal of the American Medical Association, 287: 1132—1141.
5.A-8
-------
APPENDIX 5.B
ADDITIONAL SENSITIVITY ANALYSES RELATED TO THE HEALTH BENEFITS ANALYSIS
The analysis presented in Chapter 5 of this RIA is based on our current interpretation of
the scientific and economic literature. That interpretation requires judgments regarding the
best available data, models, and modeling methodologies and the assumptions that are most
appropriate to adopt in the face of important uncertainties. The majority of the analytical
assumptions used to develop the main estimates of benefits have been reviewed and approved
by EPA's independent Science Advisory Board (SAB). Both EPA and the SAB recognize that data
and modeling limitations as well as simplifying assumptions can introduce significant
uncertainty into the benefit results and that alternative choices exist for some inputs to the
analysis, such as the concentration-response functions for mortality.
This appendix supplements our main analysis of benefits with five additional sensitivity
calculations. The supplemental estimates examine sensitivity to both for physical effects issues
(i.e., the structure of the cessation lag; estimates of the number of avoided cerebrovascular
events, cardiovascular emergency department visits and cases of chronic bronchitis; and
alternate effect estimates for cohorts in California) and valuation issues (i.e., the appropriate
income elasticity, updated cost-of-illness estimates). We conducted these sensitivity analyses
for an annual standard of 12 u.g/m3 in conjunction with a 24-hour standard of 35 u.g/m3 as an
illustrative example. These supplemental estimates are not meant to be comprehensive.
Rather, they reflect some of the key issues identified by EPA or commenters as likely to have a
significant impact on total benefits, or they are health endpoints for which the health data are
still evolving, or for which we lack an appropriate method to estimate the economic value. The
individual income growth and lag adjustments in the tables should not simply be added
together because 1) there may be overlap among the alternative assumptions, and 2) the joint
probability among certain sets of alternative assumptions may be low.
5.B.I Cessation Lag Structure for PM2.5-Related Premature Mortality
Based in part on prior advice from the EPA's independent Science Advisory Board (SAB),
EPA typically assumes that there is a time lag between reductions in particulate matter (PM)
exposures in a population and the full realization of reductions in premature mortality. Within
the context of benefits analyses, this term is often referred to as "cessation lag." The existence
of such a lag is important for the valuation of reductions in premature mortality because
economic theory suggests that dollar-based representations of health effect incidence changes
occurring in the future should be discounted.
5.B-1
-------
Over the last 15 years, there has been a continuing discussion and evolving advice
regarding the timing of changes in health effects following changes in ambient air pollution. It
has been hypothesized that some reductions in premature mortality from exposure to ambient
PM2.5 will occur over short periods of time in individuals with compromised health status, but
other effects are likely to occur among individuals who, at baseline, have reasonably good
health that will deteriorate because of continued exposure. No animal models have yet been
developed to quantify these cumulative effects, nor are there epidemiologic studies bearing on
this question. The SAB-HES has recognized this lack of direct evidence. However, in early advice,
they also note that "although there is substantial evidence that a portion of the mortality effect
of PM is manifest within a short period of time, i.e., less than one year, it can be argued that, if
no lag assumption is made, the entire mortality excess observed in the cohort studies will be
analyzed as immediate effects, and this will result in an overestimate of the health benefits of
improved air quality. Thus some time lag is appropriate for distributing the cumulative mortality
effect of PM in the population" (EPA-SAB-COUNCIL-ADV-00-001, 1999, p. 9). In more recent
advice, the SAB-HES suggests that appropriate lag structures may be developed based on the
distribution of cause-specific deaths within the overall all-cause estimate (EPA-SAB-COUNCIL-
ADV-04-002, 2004). They suggest that diseases with longer progressions should be
characterized by longer-term lag structures, while air pollution impacts occurring in populations
with existing disease may be characterized by shorter-term lags.
A key question is the distribution of causes of death within the relatively broad
categories analyzed in the long-term cohort studies. Although it may be reasonable to assume
the cessation lag for lung cancer deaths mirrors the long latency of the disease, it is not at all
clear what the appropriate lag structure should be for cardiopulmonary deaths, which include
both respiratory and cardiovascular causes. Some respiratory diseases may have a long period
of progression, while others, such as pneumonia, have a very short duration. In the case of
cardiovascular disease, there is an important question of whether air pollution is causing the
disease, which would imply a relatively long cessation lag, or whether air pollution is causing
premature death in individuals with preexisting heart disease, which would imply very short
cessation lags. The SAB-HES provides several recommendations for future research that could
support the development of defensible lag structures, including using disease-specific lag
models and constructing a segmented lag distribution to combine differential lags across causes
of death (EPA-SAB-COUNCIL-ADV-04-002, 2004). The SAB-HES indicated support for using "a
Weibull distribution or a simpler distributional form made up of several segments to cover the
response mechanisms outlined above, given our lack of knowledge on the specific form of the
distributions" (EPA-SAB-COUNCIL-ADV-04-002, 2004, p. 24). However, they noted that "an
5.B-2
-------
important question to be resolved is what the relative magnitudes of these segments should
be, and how many of the acute effects are assumed to be included in the cohort effect
estimate" (EPA-SAB-COUNCIL-ADV-04-002, 2004, p. 24-25). Since the publication of that report
in March 2004, EPA has sought additional clarification from this committee. In its follow-up
advice provided in December 2004, this SAB suggested that until additional research has been
completed, EPA should assume a segmented lag structure characterized by 30% of mortality
reductions occurring in the first year, 50% occurring evenly over years 2 to 5 after the reduction
in PM2.s, and 20% occurring evenly over the years 6 to 20 after the reduction in PM2.s (EPA-
COUNCIL-LTR-05-001, 2004). The distribution of deaths over the latency period is intended to
reflect the contribution of short-term exposures in the first year, cardiopulmonary deaths in the
2- to 5-year period, and long-term lung disease and lung cancer in the 6- to 20-year period.
Furthermore, in their advisory letter, the SAB-HES recommended that EPA include sensitivity
analyses on other possible lag structures. In this appendix, we investigate the sensitivity of
premature mortality-reduction related benefits to alternative cessation lag structures, noting
that ongoing and future research may result in changes to the lag structure used for the main
analysis.
In previous advice from the SAB-HES, they recommended an analysis of 0-, 8-, and 15-
year lags, as well as variations on the proportions of mortality allocated to each segment in the
segmented lag structure (EPA-SAB-COUNCIL-ADV-00-001, 1999, (EPA-COUNCIL-LTR-05-001,
2004). The 0-year lag is representative of EPA's assumption in previous RIAs. The 8- and 15-year
lags are based on the study periods from the Pope et al. (1995) and Dockery et al. (1993)
studies, respectively.1 However, neither the Pope et al. nor Dockery et al. studies assumed any
lag structure when estimating the relative risks from PM exposure. In fact, the Pope et al. and
Dockery et al. analyses do not supporting or refute the existence of a lag. Therefore, any lag
structure applied to the avoided incidences estimated from either of these studies will be an
assumed structure. The 8- and 15-year lags implicitly assume that all premature mortalities
occur at the end of the study periods (i.e., at 8 and 15 years).
In addition to the simple 8- and 15-year lags, we have added several additional
sensitivity analyses examining the impact of assuming different allocations of mortality to the
segmented lag of the type suggested by the SAB-HES. The first alternate lag structure assumes
that more of the mortality impact is associated with chronic lung diseases or lung cancer and
1 Although these studies were conducted for 8 and 15 years, respectively, the choice of the duration of the study
by the authors was not likely due to observations of a lag in effects but is more likely due to the expense of
conducting long-term exposure studies or the amount of satisfactory data that could be collected during this
time period.
5.B-3
-------
less with acute cardiopulmonary causes. This illustrative lag structure ("alternate segmented")
is characterized by 20% of mortality reductions occurring in the first year, 50% occurring evenly
over years 2 to 5 after the reduction in PM2.5, and 30% occurring evenly over the years 6 to 20
after the reduction in PM2.5. The second alternate lag structure ("5-year distributed") assumes
the 5-year distributed lag structure used in previous analyses, which is equivalent to a three-
segment lag structure with 50% in the first 2-year segment, 50% in the second 3-year segment,
and 0% in the 6- to 20-year segment. The third alternate lag structure assumes a smooth
negative exponential relationship between the reduction in exposure and the reduction in
mortality risk, which is described in more detail below.
In 2004, SAB-HES (U.S. EPA-SAB, 2004) urged EPA to consider using smoothed lag
distributions, incorporating information from the smoking cessation literature. In June 2010,
the SAB-HES provided additional advice regarding alternate cessation lags (U.S. EPA-SAB, 2010).
For PM2.5-related benefits, the SAB-HES continued to support the previous 20-year distributed
lag as the main estimate, while recommending that EPA further examine additional exponential
decay functions. Specifically, the SAB-HES suggested varying the rate constant with the risk
coefficient from in the cohort studies. EPA intends to incorporate these new alternate cessation
lag for PM2.5-related benefits in the final PM NAAQS RIA.
In response to these suggestions, EPA identified Roosli et al. (2005) as model that
combines empirical data on the relationship between changes in exposure and changes in
mortality and the timing of the cessation of those effects for the smooth decay function.2
Because an exponential model is often observed in biological systems, Roosli et al. (2005)
developed a dynamic model that assumes that mortality risks decrease exponentially after
exposure termination. This model assumes the form risk=exp"kt, where k is the time constant
and t is the time after t0. The relative risk from air pollution (RR) at a given time (t) can be
calculated from the excess relative risk (ERR) attributable to air pollution from PM cohort
studies (ERR=RR-R0), as follows:
RR(t) = ERR X exp~kt + R0 (5.B.I)
where R0 is the baseline relative risk in the absence of air pollution (Ro=l). After cessation of
exposure, mortality will start to decline and approach the baseline level. The change in
mortality (AM), in units of percent-years, can be derived from Equation (5.B.I) as follows:
2 In the 2006 PM NAAQS RIA (U.S. EPA, 2006), EPA applied equations and the time constant from a conference
presentation by Roosli et al. (2004). We have updated this sensitivity analysis in this assessment to reflect the
published version in Roosli et al. (2005) and generated additional time constants.
5.B-4
-------
AM = ERRxt- fERRxexp-ktdt (5.B.2)
Integrating Equation (5.B.2) gives:
AM = ERR X t - — + — X exp~kt (5.B.3)
K.
In order to calculate values for the time constant, k, we applied the AM values from the
two intervention studies that provide data on the time course of the change in mortality along
with the ERR values from cohort studies on PM2.5-related mortality. We applied the
intervention studies by Clancy et al. (2002), which analyzed the change in mortality following
the ban of coal sales in Dublin, and by Pope et al. (1992), which examined the change in
mortality resulting from the closure of a steel mill in the Utah Valley. We applied effect
estimates from the American Cancer Society (ACS) cohort by Krewski et al. (2002)3 and the Six
Cities cohort by Laden et al. (2006). Applying combinations of these studies to equation 5.B.3
generates four estimates of k that range from 0.05 to 1.24. For additional context, the time
constant calculated using on a smoking cessation study (i.e., Leksell and Rabl (2001)) is in the
middle of this range (k=0.10). For this sensitivity analysis, we applied a time constant of k=0.45
as a reasonable parameter for the exponential decay function, but we acknowledge the range
of estimates that we could have chosen. This k constant is calculated as the average of the
average k constants corresponding to each cohort study.4 Table 5.B. 1 provides the time
constants for each of these combinations and averages, and Figure 5.B.2 illustrates the
exponential decay lag structures.
3 The relative risk estimate from Krewski et al. (2009) (1.06 per 10 u.g/m3 change in average PM2.5 exposure for all-
cause mortality) is the same as the risk estimate from Pope et al. (2002).
The general approach for calculating the time constants based on the combination of the intervention study and
cohort study is consistent with the 812 analysis (U.S. EPA, 2011), which was reviewed by SAB. However, in this
analysis we have applied a single time constant (k=0.45) rather than presenting the monetized benefits results
for every exponential lag function applying the various time constants.
5.B-5
-------
Table 5.B-1. Values of the Time Constant (k) for the Exponential Decay Lag Function
Value of k PM2.5 Cohort Study Intervention Study
0.05 H6C-Laden et al. (2006) Dublin-Clancy et al. (2002)
0.15 ACS-Krewski et al. (2009) Dublin-Clancy et al. (2002)
0.37 H6C-Laden et al. (2006) Utah Valley-Pope et al. (1992)
1.24 ACS-Krewski etal. (2009) Utah Valley-Pope et al. (1992)
0.70 Average k for ACS—Krewski et al. (2009)
0.21 Average k for H6C-Laden et al. (2006)
0.45 Average of average k for each cohort study
The estimated impacts of alternative lag structures on the monetary benefits associated
with reductions in PM-related premature mortality (estimated using the effect estimate from
Krewski et al. (2009)) are presented in Table 5.B-2. These monetized estimates are calculated
using the value of a statistical life (i.e., $6.3 million per incidence adjusted for inflation and
income growth) and are presented for both a 3 and 7% discount rate over the lag period). The
choice of mortality risk study and mortality valuation approach are described in detail in
Chapter 5 of this RIA. Figure 5.B.I illustrates the cumulative distributions of the cessation lags
applied in this appendix.
The results of this sensitivity analyses demonstrate that because of discounting of
delayed benefits, the lag structure may also have a large impact on monetized benefits,
reducing benefits by 27% if an extreme assumption that no effects occur until after 15 years is
applied at a 3% discount rate and 53% at a 7% discount rate. However, for most reasonable
distributed lag structures, differences in the specific shape of the lag function have relatively
small impacts on overall benefits. For example, the overall impact of moving from the previous
5-year distributed lag to the segmented lag recommended by the SAB-HES in 2004 in the main
estimate is relatively modest, reducing benefits by approximately 5% when a 3% discount rate
is used and 9% when a 7% discount rate is used. If no lag is assumed, benefits are increased by
approximately 10% relative to the segmented lag at a 3% discount rate and 22% at a 7%
discount rate.
5.B-6
-------
Table 5.B-2. Sensitivity of Monetized PM2.5-Related Premature Mortality Benefits to
Alternative Cessation Lag Structures, Using Effect Estimate from Krewski et al.
(2009)
12/35
Alternative Lag Structures for PM-Related Premature Mortality
Value Percent
(billion Difference from
2006$)a b Base Estimate
SAB Segmented 30% of incidences occur in 1st year, 50% in years 2 to 5,
(Main estimate) and 20% in years 6 to 20
3% discount rate
7% discount rate
$2.3
$2.1
N/A
N/A
No lag Incidences all occur in the first year
3% discount rate
7% discount rate
$2.5
$2.5
10.4%
22.5%
8-year Incidences all occur in the 8th year
3% discount rate
7% discount rate
$2.1
$1.6
-10.3%
-23.7%
15-year Incidences all occur in the 15th year
3% discount rate
7% discount rate
$1.7
$1.0
-27.0%
-52.5%
Alternative 20% of incidences occur in 1st year, 50% in years 2 to 5,
Segmented and 30% in years 6 to 20
3% discount rate
7% discount rate
$2.2
$1.9
-3.2%
-6.6%
5-Year Distributed 50% of incidences occur in years 1 and 2 and 50% in years
2 to 5
3% discount rate
7% discount rate
Exponential Decay Incidences occur at an exponentially declining rate
(k=0.45)
3% discount rate
7% discount rate
a Dollar values rounded to two significant digits. The percent difference using effect
$2.4
$2.3
$2.4
$2.3
estimates from
4.9%
9.4%
5.0%
9.9%
Laden et al.
would be identical, but the value would be approximately 2.5 times higher.
5.B-7
-------
100%
o,
ro
75%
8
>,
1
o
SAB segmented lag (mam estimate)
Alternate Segmented Lag
Exponential decay model
5-yeardistributed lag
50%
25%
1st 2nd 3rd 4th 5th ah 7th 8th 9th 10th 11th 12th 13th 14th 15th 16th 17th 18th 19th 20th
year year year year year year year year year year year year year year year year year year year year
Year Following Reduction in PM2.5
Figure 5.B-1. Alternate Lag Structures for PM2.s Premature Mortality (Cumulative)
3s
D
IS
O
100%
75%
50%
25%
k=0.05 (H6C/Dublin)
k=0.37 (H6C/Utah)
k=0.15(ACS/Dublin)
k=1.24(ACS/Utah)
k=0.70 (Average ACS)
k=0.21 (Average H6C)
k=0.45 (Overall Average)
SAB Lag (Main Estimate)
Year Following Reduction in
Figure 5.B-2. Exponential Lag Structures for PM2.s Premature Mortality (Cumulative)
5.B-8
-------
5.B.2 Income Elasticity of Willingness to Pay
As discussed in Chapter 5, our estimates of monetized benefits account for growth in
real GDP per capita by adjusting the WTP for individual endpoints based on the central estimate
of the adjustment factor for each of the categories (minor health effects, severe and chronic
health effects, premature mortality, and visibility). We examined how sensitive the estimate of
total benefits is to alternative estimates of the income elasticities. Table 5.B-3 lists the ranges of
elasticity values used to calculate the income adjustment factors, while Table 5.B-4 lists the
ranges of corresponding adjustment factors. The results of this sensitivity analysis, giving the
monetized benefit subtotals for the four benefit categories, are presented in Table 5.B-5.
Table 5.B-3. Ranges of Elasticity Values Used to Account for Projected Real Income Growth3
Benefit Category
Minor Health Effect
Premature Mortality
Lower Sensitivity Bound
0.04
0.08
Upper Sensitivity Bound
0.30
1.00
Derivation of these ranges can be found in Kleckner and Neumann (1999). COI estimates are assigned an
adjustment factor of 1.0.
Table 5.B-4. Ranges of Adjustment Factors Used to Account for Projected Real Income
Growth3
Benefit Category
Minor Health Effect
Premature Mortality
Lower Sensitivity Bound
1.018
1.037
Upper Sensitivity Bound
1.147
1.591
Based on elasticity values reported in Table C-4, U.S. Census population projections, and projections of real GDP
per capita.
Table 5.B-5. Sensitivity of Monetized Benefits to Alternative Income Elasticities3
Benefit Category
Minor Health Effect
Premature Mortality15
Benefits Incremental to 15/35 Attainment Strategy (Millions of 2006$)
12/35
Lower Sensitivity Bound
$18
$2,200
Upper Sensitivity Bound
$20
$3,300
a All estimates rounded to two significant digits.
Using mortality effect estimate from Krewski et al. (2009) and 3% discount rate. Results using Laden et al. (2006)
or a 7% discount rate would show the same proportional range.
5.B-9
-------
Consistent with the impact of mortality on total benefits, the adjustment factor for
mortality has the largest impact on total benefits. The value of mortality in 2020 ranges from
96% to 108% of the main estimate based on the lower and upper sensitivity bounds on the
income adjustment factor. The effect on the value of minor health effects is much less
pronounced, ranging from 86% to 133% of the main estimate for minor effects.
5.B.3 Analysis of Cardiovascular Emergency Department Visits, Cerebrovascular Events and
Chronic Bronchitis
Below we summarize the results of a sensitivity analysis of three health endpoints:
cardiovascular emergency department visits, cerebrovascular events (stroke) and chronic
bronchitis (Table 5.B-6). While in the benefits chapter we provide a full description of the
rationale for treating these endpoints as a sensitivity, it is worth summarizing these reasons
here. In the case of cardiovascular emergency department visits, we lack the necessary
economic valuation functions to quantify the monetary value of these avoided cases. We treat
cerebrovascular events as a sensitivity for three reasons: (1) the epidemiological literature
examining PM-related cerebrovascular events is still evolving; (2) there are special uncertainties
associated with quantifying this endpoint; (3) we have not yet identified an appropriate means
for estimating the economic value of this endpoint. Finally, we now quantify chronic bronchitis
as a sensitivity because of the absence of newer studies finding a relationship between long-
term PM2.5 exposure and this endpoint.
To quantify cardiovascular hospital admissions, we apply risk coefficient drawn from
Metzger et al. (2004) (RR= 1.033, 95% confidence intervals 1.01-1.056 per 10 u.g/m3 PM2.5) and
Tolbert et al. (2007) (RR= 1.005, 95% confidence intervals 0.993-1.017 per 10 ug/m3 PM2.5). To
estimate cerebrovascular events, we apply a risk coefficient drawn from Miller et al. (2007)
(RR= 1.28, 95% confidence intervals 1.02-1.61 per 10 u.g/m3 PM2.5). To estimate chronic
bronchitis, we use a risk coefficient drawn from Abbey et al. (1995) (RR= 1.81, 95% confidence
intervals 0.98-3.25 per 45 u.g/m3 PM2.5). Additional information, including the rationale for
incorporating these new endpoints into the analysis, the baseline incidence rates for these
endpoints, and the prevalence rate for chronic bronchitis are described in Chapter 5 of this RIA.
5.B-10
-------
Table 5.B-6. Avoided Cases of Cardiovascular Emergency Department Visits, Stroke and
Chronic Bronchitis in 2020 (95th percentile confidence intervals)*
Endpoint 12/35
Cardiovascular hospital admissions
Metzger et al. (2004) 300
(ages 0-99) (180-480)
Tolbert et al. (2007) 42
(ages 0-99) (-16-130)
Stroke
Miller etal. (2007) 77
(ages 50-79) (36-140)
Chronic Bronchitis
Abbey et al. (1995) 220
(ages 27-99) (99-420)
* All estimates rounded to two significant digits.
5.B.4 New Hospitalization Cost-of-lllness Functions and Median Wage Data
As described in Chapter 5 of this RIA, we updated the cost-of-illness functions for
hospitalizations. Specifically, we updated the estimates of hospital charges and lengths of
hospital stays were based on discharge statistics provided by the Agency for Healthcare
Research and Quality's Healthcare Utilization Project National Inpatient Sample (NIS) database
for 2000 (AHRQ, 2000) to 2007 (AHRQ, 2007). In addition, we updated the county-level median
wage data reported by the 2007 American Community Survey (ACS) (Abt Associates, 2011).
Using cost-of-illness functions for hospital admissions, which include updated charges, length of
stay, and median wages, the rounded estimates of total monetized benefits do not change, but
the monetary value of respiratory hospital admissions increases 3% and cardiovascular hospital
admissions increase 2%. Because the median wages were updated, the valuation also changed
the valuation for work loss days. It is important to note that while the national average median
daily wage slightly decreased (i.e., approximately 2% in 2000$), the county-level median income
increased slightly in the locations where PM2.5 levels improved for 12/35. Tables 5.B.7 and 5.B.8
show the previous and current unit values, respectively. Table 5.B.9 shows the sensitivity of the
monetized hospitalization benefits to this update.
5.B-11
-------
Table 5.B-7. Unit Values for Hospital Admissions in BenMAP 4.0.51 (Abt Associates, 2011)a
Age Range Total Cost of Illness
Mean Hospital Mean Length (Unit value in
End Point ICD Codes mm. max. Charge (2000$) of Stay (days) 2000$)
HA, All Cardiovascular
HA, All Cardiovascular
HA, All Respiratory
HA, Asthma
HA, Chronic Lung Disease
390-429
390-429
460-519
493
490-496
18
65
65
0
18
64
99
99
64
64
$26,654
$24,893
$20,667
$9,723
$12,836
4.12
4.88
6.07
3.00
3.90
$27,119
$25,444
$21,351
$10,051
$13,276
a National average median daily wage is $112.86 (2000$).
Table 5.B-8. Unit Values for Hospital Admissions in BenMAP 4.0.43 (Abt Associates, 2010)a
End Point
Age Range Total Cost of
Mean Hospital Mean Length Illness (Unit value
ICD Codes min. max. Charge (2000$) of Stay (days) in 2000$)
HA, All Cardiovascular
HA, All Cardiovascular
HA, All Respiratory
HA, Asthma
HA, Chronic Lung Disease
390-429
390-429
460-519
493
490-496
20
65
65
0
20
64
99
99
64
64
$22,300
$20,607
$17,600
$7,448
$10,194
4.15
5.07
6.88
2.95
$5.92
$22,778
$21,191
$18,393
$7,788
$15,375
a National average median daily wage is $115.20 (2000$).
Table 5.B-9. Change in Monetized Hospitalization Benefits for 12/35
Endpoint
2000 AHRQ
(millions of 2006$)
2007 AHRQ
(millions of 2006$)
Percent Change
Respiratory hospital admissions
Cardiovascular hospital admissions
Work loss days
$2.3
$3.1
$6.7
$2.4
$3.2
$6.7
3.4%
2.1%
0.02%
* All estimates rounded to two significant digits.
5.B.5 Long-term PM2.s Mortality Estimates using Cohort Studies in California
In Chapter 5, we described the multi-state cohort studies we used to estimate the PM2.5-
related mortality (i.e., Krewski et al., 2009; Laden et al., 2006), as well as summarized the effect
estimates for additional cohort studies. In this appendix, we provide additional information
5.B-12
-------
about cohort studies in California.1 As shown in Table 5.x in the health benefits chapter, a large
percentage of the monetized human health benefits associated with the illustrative control
strategy to attain the alternative combination of standards are projected to occur in California.
Specifically, for an annual PM2.5 standard of 12 u.g/m3 in conjunction with retaining the 24-hour
standard of 35 u.g/m3, 70% of the total monetized benefits were estimated to occur in California
and 98% for an annual PM2.5 standard of 13 u.g/m3. For this reason, we determined that it was
appropriate to consider the sensitivity of the benefits results using effect estimates for cohorts
in California specifically. Although we have not calculated the benefits results using these
cohort studies, it is possible to use the effect estimates themselves to determine how much the
monetized benefits in California would have changed if we used effect estimates from the
California cohorts. Each of the California cohort studies are summarized in the PM ISA (and thus
not summarized here) with the exception of the Ostro et al. (2010, 2011) studies, which we
describe below. Table 5.B.10 provides the effect estimates from each of these cohort studies
for all-cause, cardiovascular, cardiopulmonary, and ischemic heart disease (IHD) mortality for
each of the California cohort studies.
Ostro et al. (2010) characterize the risk of premature death associated with long-term
exposure to PM2.5 in California among a cohort of about 134,000 current and former female
public school professionals (i.e., the California Teacher's Study (CTS)). In this prospective cohort
study, Ostro and colleagues estimated long-term PM exposure to several PM constituents,
including elemental carbon, organic carbon, sulfates, nitrates, iron, potassium, silicon and zinc.
In an erratum, Ostro et al. (2011) modified their approach to assigning PM2.5 levels to the
cohort populations, noting that they "reanalyzed the CTS data using time-dependent pollution
metrics—in which the exposure estimates for everyone remaining alive in the risk set were
recalculated at the time of each death—in order to compare their average exposures up to that
time with that of the individual who had died. In this way, decedents and survivors comprising
the risk set had similar periods of pollution exposure, without subsequent pollution trends
influencing the surviving women's exposure estimates." This change in assumption attenuated
the hazard ratios significantly, though hazard ratios remained significant for cardiovascular
mortality and total PM2.5 mass and certain constituents, nitrate and sulfate; no association was
observed between all-cause mortality and total PM2.5 mass or its constituents. The authors note
that these revised results are generally consistent with other long-term PM cohort studies,
including the ACS and H6C studies.
1 In addition to cohorts studies conducted in California, we have also identified a cross-sectional studies (Hankey
et al., 2012). However, we have not summarized that study here.
5.B-13
-------
Table 5.B-10. Summary of Effect Estimates From Associated With Change in Long-Term
Exposure to PM2.s in Recent Cohort Studies in California
Hazard Ratios per 10 |ig/ms Change in PM2.5
Authors
McDonnell et al. (2000)a
Jerrett et al. (2005)b
Chen etal. (2005)°
Enstrom et al. (2005)d
Krewski et al. (2009)e
Ostro et al. (2010)°
Ostro et al. (2011)c'f
Cohort
Adventist Health Study (AHS)
cohort (age > 27)
Subset of the ACS cohort living
in the Los Angeles
metropolitan area (age > 30)
Adventist Health Study (AHS)
cohort living in San Francisco,
South Coast (i.e., Los Angeles
and eastward), and San Diego
air basins (age >25)
California Prevention Study
(age >65)
Subset of the ACS cohort living
in the 5-county Los Angeles
Metropolitan Statistical Area
(age > 30)
California Teacher's study.
Current and former female
public school professionals
(age > 22)
(95th
All Causes
1.09
(.98-1.24)
1.15
(1.03-1.29)
N/A
1.04
(1.01-1.07)
1.42
(1.26-1.27)
1.84
(1.66-2.05)
1.06
(0.96-1.16)
percentile confidence
Cardiopulmonary
N/A
1.10
(0.94-1.28)
N/A
N/A
1.11
(0.95-1.23)
2.05
(1.80-2.36)
1.19
(1.05-1.36)
interval)
Ischemic
Heart Disease
N/A
1.32
(1.03-1.29)
1.42
(1.06-1.90)
N/A
1.32
(1.06-1.64)
2.89
(2.27-3.67)
1.55
(1.24-1.93)
a Table 3, adjusted for 10 ng/m3 change in PM2.5.
Table 1. 44 individual-level co-variates + all social (i.e., ecologic) factors specified (principal component analysis).
Women only.
d Represents deaths occurring from 1973-1982, but no significant associations were reported with deaths in later
time periods. The PM ISA (U.S. EPA, 2009) concludes that the use of average values for California counties as
exposure surrogates likely leads to significant exposure error, as many California counties are large and quite
topographically variable.
e Table 23. 44 individual-level co-variates + all social (i.e., ecologic) factors specified.
Erratum Table 2.
5.B-14
-------
As shown in Table 5.B.10, most of the cohort studies conducted in California report
central effect estimates similar to the (nation-wide) all-cause mortality risk estimate we applied
from Krewski et al. (2009) and Laden et al. (2006) albeit with wider confidence intervals. A
couple cohort studies conducted in California indicate higher risks than the risk estimates we
applied.
5.B.6 References
Abbey, David E et al. 1995. "Chronic Respiratory Symptoms Associated with Estimated Long-
Term Ambient Concentrations of Fine Particulates less than 2.5 Microns in aerodynamic
diameter (PM2.5) and Other Air Po\\utar\ts." Journal of Exposure Analysis and
Environmental Epidemiology 5(2): 137-158.
Abt Associates, Inc. 2010. BenMAP User's Manual Appendices. Prepared for U.S. Environmental
Protection Agency Office of Air Quality Planning and Standards. Research Triangle Park,
NC. Available on the Internet at
.
Agency for Healthcare Research and Quality (AHRQ). 2000. HCUPnet, Healthcare Cost and
Utilization Project. Rockville, MD. Available on the Internet at
.
Agency for Healthcare Research and Quality (AHRQ). 2007. HCUPnet, Healthcare Cost and
Utilization Project. Rockville, MD. Available on the Internet at
.
Chen LH, Knutsen SF, Shavlik D, Beeson WL, Petersen F, Ghamsary M, et al. 2005. "The associa-
tion between fatal coronary heart disease and ambient particulate air pollution: are
females at greater risk?" Environ Health Perspect 113:1723-1729.
Clancy, L, P. Goodman, H. Sinclair, and D.W. Dockery. 2002. "Effect of Air-pollution Control on
Death Rates in Dublin, Ireland: An Intervention Study." Lancet Oct 19;360(9341):1210-4.
Desvousges, W.H., F.R. Johnson, and H.S. Banzhaf. 1998. Environmental Policy Analysis With
Limited Information: Principles and Applications of the Transfer Method (New Horizons
in Environmental Economics.) Edward ElgarPub: London.
5.B-15
-------
Dockery, D.W., C.A. Pope, X.P. Xu, J.D. Spengler, J.H. Ware, M.E. Fay, B.C. Ferris, and F.E.
Speizer. 1993. "An Association between Air Pollution and Mortality in Six U.S. Cities."
New England Journal of Medicine 329(24):1753-1759.
Enstrom JE 2005. "Fine particulate air pollution and total mortality among elderly Californians,
1973-2002." Inhal Toxicol 17: 803-816.
Hankey, S., J. D. Marshall, et al. (2011). "Health Impacts of the Built Environment: Within-Urban
Variability in Physical Inactivity, Air Pollution, and Ischemic Heart Disease Mortality."
Environ Health Perspect 120(2).
Jerrett M, Burnett RT, Ma R, Pope CA III, Krewski D, Newbold KB, et al. 2005. "Spatial analysis of
air pollution and mortality in Los Angeles." Epidemiology 16(6):7'27'-7'36.
Kleckner, N., and J. Neumann. 1999. "Recommended Approach to Adjusting WTP Estimates to
Reflect Changes in Real Income." Memorandum to Jim Democker, US EPA/OPAR. June 3,
Available on the Internet at
.
Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi, Y, et al. 2009. "Extended follow-up and
spatial analysis of the American Cancer Society study linking particulate air pollution and
mortality." HEI Research Report, 140, Health Effects Institute, Boston, MA.
Laden F, Schwartz J, Speizer FE, Dockery DW. 2006. "Reduction in fine particulate air pollution
and mortality: extended follow-up of the Harvard Six Cities study." Am J Respir Crit Care
Med 173:667-672.
McDonnell WF; Nishino-lshikawa N; Petersen FF; Chen LH; Abbey DE. 2000. "Relationships of
mortality with the fine and coarse fractions of long-term ambient PM10 concentrations
in nonsmokers." J Expo Sci Environ Epidemiol, 10: 427-436.
Metzger, K. B., P. E. Tolbert, et al. (2004). "Ambient air pollution and cardiovascular emergency
department visits." Epidemiology 15(1): 46-56.
Miller, Kristin A., David S. Siscovick, Lianne Sheppard, Kristen Shepherd, Jeffrey H. Sullivan,
Garnet L. Anderson, and Joel D. Kaufman. 2007. "Long-Term Exposure to Air Pollution
and Incidence of Cardiovascular Events in Women." New England Journal of Medicine.
356 (5) :447-458.
Ostro B, Lipsett M, Reynolds P, Goldberg D, Hertz A, Garcia C, et al. 2010. "Long-Term Exposure
to Constituents of Fine Particulate Air Pollution and Mortality: Results from the
California Teachers Study." Environ Health Perspect 118:363-369.
Ostro B, Reynolds P, Goldberg D, Hertz A, et al. 2011. "Assessing Long-Term Exposure in the
California Teachers Study." Environ Health Perspect June; 119(6): A242-A243.
5.B-16
-------
Peters, A., D. W. Dockery, J. E. Muller and M. A. Mittleman. 2001. "Increased particulate air
pollution and the triggering of myocardial infarction." Circulation. Vol. 103 (23): 2810-5.
Pope, C.A., III, M.J. Thun, M.M. Namboodiri, D.W. Dockery, J.S. Evans, F.E. Speizer, and C.W.
Heath, Jr. 1995. "Particulate Air Pollution as a Predictor of Mortality in a Prospective
Study of U.S. Adults." American Journal of Respiratory Critical Care Medicine 151:669-
674.
Roosli M, Kunzli N, Braun-Fahrlander C. 2004. "Use of Air Pollution 'Intervention-Type' Studies
in Health Risk Assessment." 16th Conference of the International Society for
Environmental Epidemiology, New York, August 1-4, 2004.
Roosli M, Kunzli N, Braun-Fahrlander C, Egger M. 2005. "Years of life lost attributable to air
pollution in Switzerland: dynamic exposure-response model." International Journal of
Epidemiology 34(5):1029-35.
Tolbert, P. E., M. Klein, et al. 2007. "Multipollutant modeling issues in a study of ambient air
quality and emergency department visits in Atlanta." J Expo Sci Environ Epidemiol 17
Suppl 2:529-35.
U.S. Environmental Protection Agency—Science Advisory Board (EPA-SAB). 1999a. The Clean Air
Act Amendments (CAAA) Section 812 Prospective Study of Costs and Benefits (1999):
Advisory by the Health and Ecological Effects Subcommittee on Initial Assessments of
Health and Ecological Effects. Parti. EPA-SAB-COUNCIL-ADV-99-012. July. Available on
the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (EPA-SAB). 1999b. The Clean Air
Act Amendments (CAAA) Section 812 Prospective Study of Costs and Benefits (1999):
Advisory by the Health and Ecological Effects Subcommittee on Initial Assessments of
Health and Ecological Effects. Part 2. EPA-SAB-COUNCIL-ADV-00-001. October. Available
on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (EPA-SAB). 2001. Review of the
Draft Analytical Plan for EPA's Second Prospective Analysis—Benefits and Costs of the
Clean Air Act 1990-2020: An Advisory by a Special Panel of the Advisory Council on Clean
Air Compliance Analysis. EPA-SAB-COUNCIL-ADV-01-004. September. Available on the
Internet at
.
5.B-17
-------
U.S. Environmental Protection Agency—Science Advisory Board (EPA-SAB). 2004. Advisory on
Plans for Health Effects Analysis in the Analytical Plan for EPA's Second Prospective
Analysis—Benefits and Costs of the Clean Air Act, 1990-2020: Advisory by the Health
Effects Subcommittee of the Advisory Council on Clean Air Compliance Analysis. EPA-
SAB-COUNCIL-ADV-04-002. March. Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2010. Review of
EPA's DRAFT Health Benefits of the Second Section 812 Prospective Study of the Clean Air
Act. EPA-COUNCIL-10-001. June. Available on the Internet at
.
5.B-18
-------
APPENDIX 5.C
QUALITATIVE ASSESSMENT OF UNCERTAINTY
Although we strive to incorporate as many quantitative assessments of uncertainty as
possible, there are several aspects we are only able to address qualitatively. These aspects are
important factors to consider when evaluating the relative benefits of the attainment strategies
for each of the alternative standards:
To more fully address all these uncertainties including those we cannot quantify, we
apply a four-tiered approach using the WHO uncertainty framework (WHO, 2008), which
provides a means for systematically linking the characterization of uncertainty to the
sophistication of the underlying risk assessment. EPA has applied similar approaches in peer-
reviewed analyses of PM2.5-related impacts (U.S. EPA, 2010b, 2011a). In addition to the WHO
uncertainty framework, we also include an assessment of how each aspect of uncertainty could
affect the benefits results, including the direction of potential bias, the magnitude of impact on
results, and the degree of confidence in our approach. In Table 5.C-1, we summarize the key
uncertainties in the health benefits analysis, including our assessment of the direction of
potential bias, magnitude of impact on the monetized benefits, degree of confidence in our
analytical approach, and our ability to assess the source of uncertainty. Because this approach
reflects a new application for regulatory benefits analysis, we request comments on this
general approach as well as the specific uncertainty assessments.
5.C.I Description of Classifications Applied in the Uncertainty Characterization
Uncertainty Characterization Tiers
The WHO framework (2008) defines 4 tiers of uncertainty characterization, which vary
depending on the degree of quantification. In Table 5.C-1, we apply these tiers considering the
degree of quantification of uncertainty we have conducted in this analysis or that we plan to
conduct for the final RIA. Ultimately, the tier decision is professional judgment based on the
availability of information.
• Tier 0—Screening level, generic qualitative characterization
• Tier 1—Scenario-specific qualitative characterization
• Tier 2—Scenario-specific sensitivity analysis
5.C-1
-------
• Tier 3—Scenario-specific probabilistic assessment of individual and combined
uncertainty
Magnitude of Impact
The magnitude of impact is an assessment of how much a plausible alternative
assumption or approach could influence the overall monetary benefits. Similar classifications
have been included in a previous analyses (U.S. EPA, 2010b, 2011a), but we have revised the
category names and the cut-offs here.1 We note that PM2.5-related mortality benefits comprise
over 98% of the monetized benefits in this analysis, thus alternative assumptions affecting
mortality have the potential to have higher impacts on the total monetized benefits. Including
currently omitted categories of benefits would lead to a reduction in the fraction of monetized
benefits attributable to lower mortality risk. Ultimately, the magnitude decision is professional
judgment based on the experience with various sensitivity analyses.
• High—If this uncertainty could influence the total monetized benefits by more than 25%
• Medium—If this uncertainty could influence the total monetized benefits by 5% to 25%
• Low—If this uncertainty could influence the total monetized benefits by less than 5%
Degree of Confidence in Our Analytic Approach
The degree of confidence is an assessment based on our assessment of the available
body of evidence. That is, based on the given available evidence, how certain we are that the
selected assumption is the most plausible of the alternatives. Similar classifications have been
included in a previous analyses (U.S. EPA, 2010b, 2011a).2 Ultimately, the degree of confidence
is professional judgment based on the volume and consistence of supporting evidence, much of
1 In The Benefits and Costs of the Clean Air Act from 1990 to 2020 (U.S. EPA, 2011a), EPA applied a classification of
"potentially major" if a plausible alternative assumption or approach could influence the overall monetary
benefit estimate by 5% percent or more and "probably minor." if an alternative assumption or approach is likely
to change the total benefit estimate by less than five percent. In the Quantitative Health Risk Assessment for
Paniculate Matter (U.S. EPA, 2010b), EPA applied classifications of "low" if the impact would not be expected to
impact the interpretation of risk estimates in the context of the PM NAAQS review, "medium" if the impact had
the potential to change the interpretation; "high" if it was are likely to influence the interpretation of risk in the
context of the PM NAAQS review.
2 We have applied the same classification as The Benefits and Costs of the Clean Air Act from 1990 to 2020 (U.S.
EPA, 2011a) in this analysis. In the Quantitative Health Risk Assessment for Particulate Matter (U.S. EPA, 2010b),
EPA assessed the degree of uncertainty (low, medium, or high) associated with the knowledge-base (i.e.,
assessed how well we understand each source of uncertainty), but did not provide specific criteria for the
classification.
5.C-2
-------
which has been evaluated in the PM ISA.
• High—The current evidence is plentiful and strongly supports the selected approach
• Medium—Some evidence exists to support the selected approach, but data gaps are
present
• Low—Limited data exists to support the selected approach
5.C-3
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with PM2i5 Concentration Changes
Projections of future levels of
emissions and emissions
reductions necessary to attain
alternative standards
Responsiveness of air quality
model to changes in precursor
emissions from control scenarios
Air quality model chemistry,
particularly for formation of
ambient nitrate concentrations
Post-processing of air quality
modeled concentrations to
estimate future-year PM2.5
design value and spatial fields of
PM2.5 concentrations.
Both Medium
Future expected emissions are
difficult to predict because they
depend on many independent
factors. Emission inventories are
aggregated from many spatially
and technically diverse sources
of emissions, so simplifying
assumptions are necessary to
make estimating the future
tractable.
Both Medium-high
Both Medium
Both High
Medium
Tierl
See Chapters
Medium
High
High
Tierl
See Chapters
Tierl
See Chapter 3
Tierl
See Chapter 3
Uncertainties Associated with Concentration-Response Functions
Causal relationship between
PM2.sexposure and premature
mortality
Overestimate, if no causal
relationship
High
PM-mortality effects are the
largest contributor to the
monetized benefits. If the
PM2.5/mortality relationship
were not causal, benefits would
be significantly overestimated.
High
The PM ISA (U.S. EPA, 2009b),
which was twice peer reviewed
by CASAC, evaluated the entire
body of scientific literature and
concluded that the relationship
between both short-term and
long-term exposure to PM2 5and
mortality is causal.
Tier3
Experts included likelihood of
causal relationship, so causality
addressed in results derived from
PM2 5expert elicitation (Roman et
al., 2008).
(continued)
5.C-4
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Concentration-Response Functions (continued)
Modification of Mortality C-R
function by socio-economic
status (SES)
Exposure misclassification in
epidemiology studies
Spatial matching of air quality
estimates from epidemiology
studies to air quality estimates
from air quality modeling
Potential underestimate for ACS
cohort (Krewski et al., 2009)
because of the demographics of
that study population. Unknown
for H6C cohort (Laden et al.,
2006)
Underestimate (generally)
Reducing exposure error can
result in stronger associations
between pollutants and effect
estimates than generally
observed in studies having less
exposure detail.
Unknown
Epidemiology studies often
create a composite air quality
monitor that is assumed to be
representative of an entire urban
area to estimate health risks,
while benefits are often
calculated using air quality
modeling conducted at 12 km
spatial resolution. This spatial
mismatch could introduce
uncertainty.
Potentially medium-high for ACS
cohort
Unknown for H6C cohort
Medium-high
Recent analyses reported in
Krewski et al. (2009)
demonstrate the potentially
significant effect that this source
of uncertainty can have on effect
estimates. These analyses also
illustrate the complexity and site-
specific nature of this source of
uncertainty.
Medium
Medium
We only have mortality risk
coefficients modified by
educational attainment (Krewski,
2000), not other risk modifiers
such as income or race.
High
The results from Krewski et al.
(2009) and Jerrett et al. (2005)
suggest that exposure error
underestimates effect estimates
(U.S. EPA, 2009b).
Medium-Low
Tier 2
Effect modification for educational
attainment evaluated in
distributional analysis in Appendix
5A.
Tierl
Tierl
(continued)
5.C-5
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Concentration-Response Functions (continued)
Variation in effect estimates
reflecting differences in PM2.s
composition (mixtures)
Differential toxicity of particle
components
Both
Unknown
We assume that all fine particles,
regardless of their chemical
composition, are equally potent
in causing premature mortality.
Depending on the toxicity of
each PM species reduced, this
could over or underestimate
benefits.
Medium- High
Epidemiology studies examining
regional differences in PM2.s-
related health effects have found
differences in the magnitude of
those effects. While these may
be the result of factors other
than composition (e.g., different
degrees of exposure
misclassification), composition
remains one potential
explanatory factor.
Medium
If the benefits are due to a
variety of PM species reduced,
the magnitude of this
uncertainty is likely to be small. If
only one PM species is reduced,
this uncertainty may have larger
magnitude.
Medium
Tierl
Medium-Low
The PM ISA (U.S. EPA, 2009b),
which was twice peer reviewed
by CASAC, evaluated the entire
body of scientific literature and
concluded that because there is
no clear scientific evidence that
would support the development
of differential effects estimates
by particle type (U.S. EPA,
2009b).
Tier 2
To be assessed in final RIA
Application of C-R relationships
only to those subpopulations
matching the original study
population
Underestimate
The C-R functions for several
health endpoints were applied
only to subgroups of the U.S.
population (e.g., adults 30+ for
mortality, children 8-12 for acute
bronchitis), and thus this may
underestimate the whole
population benefits of reductions
in pollutant exposures.
Low
The baseline mortality rate for
PM-related health effects is
significantly lower in those under
the age of 30. Mortality
valuation generally dominates
monetized benefits.
High
Our approach follows
recommendations from the NAS
(NRC, 2002)
Tier 2
To be assessed in final RIA
(continued)
5.C-6
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Concentration-Response Functions (continued)
Shape of the C-R functions,
particularly at low
concentrations
Impact of historical exposure on
long-term effect estimates
Both
If there is a threshold (i.e., a level
of exposure below which health
effects do not occur), then the
relative risk (i.e., steeper slope)
estimates would be higher within
the range of observed effects.
Both
Long-term studies of mortality
suggest that different time
periods of PM exposure can
produce significantly different
effects estimates, raising the
issue of uncertainty in relation to
determining which exposure
window is most strongly
associated with mortality.
Medium
For PM2.5-related long-term
mortality, the PM ISA concludes
that a log-linear non-threshold
model is best supported in the
scientific literature (U.S. EPA,
2009b). Although consideration
for alternative model forms
(Krewski et al., 2009) does
suggest that different models
can impact effect estimates to a
certain extent, generally this
appears to be a moderate source
of overall uncertainty.
Medium
The Reanalysis II study (HEI,
2009) which looked at exposure
windows (1979-1983 and 1999-
2000) for long-term exposure in
relation to mortality, did not
draw any conclusions as to which
window was more strongly
associated with mortality.
However, the study did suggest
that moderately different effects
estimates are associated with
the different exposure periods
(with the more recent period
having larger estimates). Overall,
the evidence for determining the
window over which the mortality
effects of long-term pollution
exposures occur suggests a
latency period of up to five years,
with the strongest results
observed in the first few years
after intervention (PM ISA,
section 7.6.4. p. 7-95).
High
Our approach follows
recommendations from the SAB
(U.S. EPA-SAB, 2010a)
TierS
Assessed in LML assessment and
the results derived from the expert
elicitation
Medium
See PM risk assessment (U.S.
EPA, 2010b)
Tier 2
To be assessed in final RIA
(continued)
5.C-7
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Concentration-Response Functions (continued)
Confounding by co-pollutants
Both
Confounding by ecologic factors,
such as SES or smoking
Exclusion of C-R functions from
short-term exposure studies in
PM mortality calculations
Both
Underestimate
Medium
For long-term health endpoints,
the final ISA states, "Given
similar sources for multiple
pollutants (e.g., traffic),
disentangling the health
responses of co-pollutants is a
challenge in the study of ambient
air pollution." The PM ISA also
notes that in some instances,
consideration of co-pollutants
can have a significant impact on
effect estimates.
For morbidity, the PM ISA
concludes that observed
associations are fairly robust to
the inclusion of co-pollutants in
the predictive models (see PM
ISA).
Medium
Medium
PM/mortality is the top
contributor to the benefits
estimate. If short-term functions
contribute substantially to the
overall PM-related mortality
estimate, then the benefits could
be underestimated.
Medium
Tierl
Medium-High
To minimize confounding, we
selected the risk coefficient that
controlled for ecologic factors
from Krewski et al. (2009).
Medium
Long-term PM exposure studies
likely capture a large part of the
impact of short-term peak
exposure on mortality; however,
the extent of overlap between
the two study types is unclear.
Tierl
Tierl
(continued)
5.C-8
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Valuation
Value-of-a-Statistical-Life(VSL)
Cessation lag structure for long
term PM mortality
Both
Some studies suggest that EPA's
VSL is too high, while other
studies suggest that it is too low.
The VSL used by EPA is based on
26 labor market and stated
preference studies published
between 1974 and 1991.
Underestimate
Recent studies (Schwartz, 2008)
have shown that the majority of
the risk occurs within 2 years of
reduced exposure. EPA's current
lag structure assumption was
provided by the SAB, and it
estimates that 30% of mortality
reductions in the first year, 50%
over years 2 to 5, and 20% over
the years 6 to 20 after the
reduction in PM2.5(U.S. EPA-SAB,
2004c).
High
Mortality valuation generally
dominates monetized benefits.
Medium
Although the cessation lag does
not affect the number of
premature deaths attributable to
PM2.s exposure, it affects the
timing of those deaths and thus
the discounted monetized
benefits.
Medium
EPA is in the process of reviewing
this estimate and will issue
revised guidance based on the
most up-to-date literature and
recommendations from the SAB-
EEAC in the near future.
Medium
The main cessation lag applied
was recently confirmed by the
SAB (U.S. EPA-SAB, 2010a).
Tier 2
Assessed uncertainty in mortality
valuation using a Weibull
distribution.
Tier 2
Assessed in sensitivity analysis
(continued)
5.C-9
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Valuation (continued)
Morbidity valuation
Income growth adjustments
Underestimate
Morbidity benefits such as
hospital admissions and heart
attacks are calculated using cost-
of-illness (COI) estimates, which
studies have shown (Alberini and
Krupnick, 2000) are generally half
as much as willingness-to-pay to
avoid the illness, in addition, the
quantified morbidity impacts do
not reflect physiological
responses or sequelae events,
such as increased susceptibility
for future morbidity.
Both
Income growth increases
willingness-to-pay valuation
estimates, including mortality,
over time. From 1997 to 2010,
personal income and GDP growth
have begun to diverge If this
trend continues, the assumption
that per capita GDP growth is a
reasonable proxy for income
growth may lead to an
overstatement of benefits. (lEc,
2012).
Low
Mortality valuation generally
dominates monetized benefits.
Medium
Income growth from 1990 to
2020 increases mortality
valuation by 20%. Alternate
estimates for this adjustment
vary by20%(IEc, 2012).
Low
Although the COI estimates for
hospitalizations reflect recent
data, other COI estimates such as
for AMI have not yet been
updated. Nevertheless, even
current COI valuation estimates
do not capture the full valuation
of these morbidity impacts.
Tierl
Medium
Adjusting for income growth is
consistent with SAB
recommendations (U.S. EPA,-
SAB, 2000). Difficult to forecast
future income growth. However,
in the absence of readily
available income data
projections, per capita GDP is the
best available option.
Tier 2
To be assessed in final RIA
(continued)
5.C-10
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Baseline Incidence and Population
Uncertainty in projecting
baseline incidence rates for
mortality
Uncertainty in projecting
baseline incidence rates and
prevalence rates or morbidity
Both
Baseline mortality rates are at
the county level and projected
for 5-year increments for
multiple age groups. Due to data
suppression for small numbers of
specific
age/gender/race/ethnicity
combinations, many counties
have missing baseline mortality
rates.
Both
Morbidity baseline incidence is
available for year 2000 only (i.e.,
no projections available).
Medium
Mortality valuation generally
dominates monetized benefits.
The county-level baseline
mortality rates reflect recent
databases (i.e., 2004-2006). Also,
the mortality rates projections for
future years are internally
consistent with population
projections in that they reflect
changes in mortality patterns as
well as population growth.
Low
Mortality valuation generally
dominates monetized benefits.
The magnitude of uncertainty
associated with projections of
morbidity baseline incidence
varies with the health endpoint.
Some endpoints such as
hospitalizations and ER visits have
more recent data (i.e., 2007)
stratified by age and geographic
location. Other endpoints, such as
respiratory symptoms reflect one
national average.
Medium-High
The mortality rate databases
(CDC, 2008) are generally
considered to have relatively low
uncertainty. These projections
account for both spatial and
temporal changes.
Tierl
Low
There is no current method for
projecting baseline morbidity
rates beyond 2000.
Asthma prevalence rates reflect
recent increases in baseline
asthma rates (i.e., 2008).
Tierl
(continued)
5.C-11
-------
Table 5.C-1. Summary of Qualitative Uncertainty for Key Modeling Elements in PM2.5 Benefits (continued)
Potential Source of Uncertainty
Direction of Potential Bias
Magnitude of Impact on
Monetized Benefits
Degree of Confidence in Our
Analytical Approach
Ability to Assess Uncertainty
Uncertainties Associated with Baseline Incidence and Population (continued)
Population estimates and
projections
Both
The monetized benefits would
change in the same direction as
the over or underestimate in
population projections in areas
where exposure changes.
Low-Medium
Monetized benefits are
substantially affected by
population density. Comparisons
using historical census data show
that population projections are
+/- 5% nationally, but projections
accuracy can vary by locality.
Medium
These projections cannot account
for future population migration
due to possible catastrophic
events.
Tierl
Uncertainties Associated with Omitted Categories
Unquantified PM health benefit Underestimate
categories, such as pulmonary
function, cerebrovascular events
or low birth weight
Unquantified health benefit Underestimate
categories for components of
PM, such as air toxics (organics
and metals)
Medium
Mortality valuation generally
dominates monetized benefits,
but it is possible that some of
these omitted categories could
be significant, especially for
morbidity.
Medium
Studies have found air toxics
cancer risks to be orders of
magnitude lower than overall
risks from criteria pollutants.
However, air toxics can also be
associated with cardiovascular,
reproductive, respiratory,
developmental, and neurological
risks with potentially synergistic
effects.
Low Tier 1
Current data and methods are
insufficient to develop (and
value) national quantitative
estimates of the health effects of
these pollutants.
Low Tier 1
Current data and methods are
insufficient to develop (and
value) national quantitative
estimates of the health effects of
these pollutants.
5.C-12
-------
5.C.2 References
Alberini, Anna and Alan Krupnick. 2000. "Cost-of-lllness and Willingness-to-Pay Estimates of the
Benefits of Improved Air Quality: Evidence from Taiwan. "Land Economics
Vol. 76, No. 1 (Feb), pp. 37-53.
Centers for Disease Control and Prevention (CDC). 2008. National Center for Health Statistics.
National Health Interview Survey, 1999-2008.
Industrial Economics, Incorporated (lEc). 2012. Updating BenMAP Income Elasticity Estimates-
Literature Review. Memo to Neal Fann. March. Available on the Internet at
Jerrett M, Burnett RT, Ma R, et al. 2005. "Spatial analysis of air pollution and mortality in Los
Angeles." Epidemiology 16:727-736.
Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi, Y, et al. 2009. "Extended follow-up and
spatial analysis of the American Cancer Society study linking particulate air pollution and
mortality." HEI Research Report, 140, Health Effects Institute, Boston, MA.
Laden, F., J. Schwartz, F.E. Speizer, and D.W. Dockery. 2006. "Reduction in Fine Particulate Air
Pollution and Mortality." American Journal of Respiratory and Critical Care Medicine
173:667-672.
National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed Air
Pollution Regulations. Washington, DC: The National Academies Press. Washington, DC.
Roman, Henry A., Katherine D. Walker, Tyra L. Walsh, Lisa Conner, Harvey M. Richmond, Bryan
J. Hubbell, and Patrick L. Kinney. 2008. "Expert Judgment Assessment of the Mortality
Impact of Changes in Ambient Fine Particulate Matter in the U.S." Environ. Sci. Technol.,
42(7):2268-2274.
Schwartz J, Coull B, Laden F. 2008. "The Effect of Dose and Timing of Dose on the Association
between Airborne Particles and Survival." Environmental Health Perspectives 116: 64-
69.
U.S. Environmental Protection Agency (U.S. EPA). 2009b. Integrated Science Assessment for
Particulate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment—RTP Division. December. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010b. Quantitative Health Risk Assessment
for Particulate Matter—Final Report. EPA-452/R-10-005. Office of Air Quality Planning
and Standards, Research Triangle Park, NC. September. Available on the Internet at
.
5.C-13
-------
U.S. Environmental Protection Agency (U.S. EPA). 2011a. The Benefits and Costs of the Clean Air
Act 1990 to 2020: EPA Report to Congress. Office of Air and Radiation, Office of Policy,
Washington, DC. March. Available on the Internet at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2010a. Review of
EPA's DRAFT Health Benefits of the Second Section 812 Prospective Study of the Clean Air
Act. EPA-COUNCIL-10-001. June. Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004c. Advisory
Council on Clean Air Compliance Analysis Response to Agency Request on Cessation Lag.
EPA-COUNCIL-LTR-05-001. December. Available on the Internet at
.
World Health Organization (WHO). 2008. Part 1: Guidance Document on Characterizing and
Communicating Uncertainty in Exposure Assessment, Harmonization Project Document
No. 6. Published under joint sponsorship of the World Health Organization, the
International Labour Organization and the United Nations Environment Programme.
WHO Press: Geneva, Switzerland. Available on the Internet at
.
5.C-14
-------
CHAPTER 6
WELFARE BENEFITS ANALYSIS APPROACH
6.1 Important Caveats Regarding this Chapter
Due to data limitations for this proposed rule, we were unable to calculate changes in
light extinction associated with emission reductions from the illustrative control strategies,
which are necessary for calculating the visibility benefits. Instead, this chapter and associated
appendix describe in detail the methodology for calculating visibility benefits to encourage
comment on the revised approach.
6.2 Synopsis
Emission reductions associated with the illustrative control strategies to attain
alternative combinations of the PM NAAQS have numerous documented effects on
environmental quality that affect human welfare, including changes in visibility, materials
damage, ecological effects from PM deposition, ecological effects from nitrogen and sulfur
emissions, vegetation effects from ozone exposure, ecological effects from mercury deposition,
and climate effects. Even though the primary standards are designed to protect against adverse
effects to human health, the emission reductions have welfare co-benefits in addition to the
direct human health benefits. Due to data limitations for this proposed rule, we are unable to
estimate the recreational visibility and residential visibility benefits associated for alternative
standard combinations in 2020 even though we have a complete methodology for estimating
benefits for scenarios with light extinction estimates for both the baseline and control
scenarios. We intend to apply the approach described in this chapter in the RIA accompanying
the final rulemaking, and as such, we solicit comment here. Despite our goal to quantify and
monetize as many of the benefits as possible, welfare benefits remain unquantified and
nonmonetized in this analysis due to data, methodology, and resource limitations. The
monetized value of these unquantified effects is represented by adding an unknown "B" to the
aggregate total benefits. These unquantified welfare benefits may be substantial, although the
magnitude of these benefits is highly uncertain.
6.3 Introduction to Welfare Benefits Analysis
Emission reductions associated with the illustrative control strategies to attain
alternative combinations of the PM NAAQS have numerous documented effects on
environmental quality that affect human welfare. We define welfare effects to include any non-
health effects, including direct damages to property, either through impacts on material
structures or by soiling of surfaces, direct economic damages in the form of lost productivity of
6-1
-------
crops and trees, indirect damages through alteration of ecosystem functions, and indirect
economic damages through the loss in value of recreational experiences or the existence value
of important resources. EPA's Integrated Science Assessments for PM (hereafter, "PM ISA")
(U.S. EPA, 2009b) and NOX/SOX—Ecological Criteria (U.S. EPA, 2008), as well as the Criteria
Document for ozone (U.S. EPA, 2006) identify numerous physical and ecological effects known
to be causally linked to these pollutants. This chapter describes these individual effects and
how we would quantify and monetize them if there is enough data to do so. These welfare
effects include changes in visibility, materials damage, ecological effects from PM deposition,
ecological effects from nitrogen and sulfur emissions, vegetation effects from ozone exposure,
ecological effects from mercury deposition, and climate effects.
These welfare benefits are associated with reductions in emissions of specific pollutants
resulting from emissions controls applied to attain the suite of PM standards, not the form or
intent of any specific standard. Even though the primary standards are designed to protect
against adverse effects to human health, the emission reductions have welfare co-benefits in
addition to the direct human health benefits.
The impacts of emission reductions associated with the illustrative control strategies can
be grouped into four categories: directly emitted PM (e.g., metals, organic compounds, dust),
reductions of PM2.s precursors (e.g., NOx, SOx, VOCs), other ancillary reductions from
illustrative control strategies (e.g., mercury and C02), and secondary co-pollutant formation
from PM precursors (e.g., ozone from NOx and VOCs). Regardless of the category, these
emission changes are anticipated to affect ambient concentrations and deposition, and
consequently affect public welfare. It is therefore appropriate and reasonable to include all the
benefits associated with these emission reductions to provide a comprehensive understanding
of the likely public impacts of attaining alternative standard level combinations. Table 6-1
shows the welfare effects associated with the various pollutants (either directly or as a
precursor to secondary formation of PM or ozone) that would be reduced by the illustrative
control strategies to attain the alternative standard level combinations.
Based on previous EPA analyses, we believe the welfare benefits associated with these
non-health benefit categories could be significant (U.S. EPA, 2011b). Despite our goal to
quantify and monetize as many of the benefits as possible, welfare benefits remain
unquantified and nonmonetized in this analysis due to data, methodology, and resource
limitations. For the final rulemaking, we anticipate that visibility would be the only welfare
benefit category with sufficient data to quantify monetized benefits. Although it is possible to
estimate some of the acidification and ozone vegetation benefits, we are limited by the time
6-2
-------
Table 6-1. Welfare Effects by Pollutants Potentially Affected by Attainment of the PM
NAAQS
Pollutant
Direct
PM2.5
NOx
S02
VOCs
PM10
Hg
CO2
Atmospheric Effects
Vegetation Vegetation
Visibility Injury Injury
Impairment (SO2) (Ozone)
'
S S
s s
s s
s
Atmospheric and
Deposition Effects
Materials
Damage Climate
S ,
S S
s s
s
s s
'
Deposition Effects
Ecosystem
Effects—
(Organics Acidification Nitrogen Mercury
& Metals) (freshwater) Enrichment Methylation
'
S S
s s
'
s s
•S = Welfare category affected by this pollutant.
and resources available, and we do not anticipate being able to quantify these benefits in the
final rulemaking. The other welfare effects have additional data and methodology limitations
that preclude us from monetizing those benefits. Therefore, the total benefits would be larger
than we have estimated in this analysis. The monetized value of these unquantified effects is
represented by adding an unknown "B," which includes both unmonetized health and welfare
effects, to the aggregate total for the cost-benefit comparison. These unquantified benefits
may be substantial, although the magnitude of these benefits is highly uncertain. For these
categories of welfare benefits that we are unable to quantify in this analysis, we include a
qualitative analysis of the anticipated effects in this chapter to characterize the type and
potential extent of those benefits. In Table 6-2, we identify the quantified and unquantified
welfare benefits.
The remainder of this chapter is organized as follows: Section 6.3 provides the
methodology for the visibility benefits analysis. Sections 6.4 through 6.6 provide qualitative
benefits for the unquantified benefits categories of materials damage, climate, and ecosystem
benefits. References are provided in Section 6.7. Additional information regarding technical
details of the visibility benefits analysis is provided in Appendix 6a.
6-3
-------
Table 6-2. Quantified and Unquantified Welfare Benefits
Benefits Category
Specific Effect
Effect Has
Been
Quantified
Effect Has
Been
Monetized
More
Information
Improved Environment
Reduced visibility impairment
Visibility in Class I areas in SE, SW,
and CA regions
Visibility in Class I areas in other
regions
Visibility in 8 cities
Visibility in other residential areas
Section 6.3
Section 6.3
Section 6.3
Section 6.3
Reduced climate effects
Global climate impacts from CO2
Climate impacts from ozone and
PM
Other climate impacts (e.g., other
GHGs, other impacts)
SCCTSD1
Ozone CD, Draft
Ozone ISA, PM
ISA2
IPCC
Reduced effects on materials
Household soiling
Materials damage (e.g., corrosion,
increased wear)
PM ISA
PM ISA
Reduced effects from PM
deposition (metals and
organics)
Effects on Individual organisms
and ecosystems
PM ISA
Reduced vegetation and
ecosystem effects from
exposure to ozone
Visible foliar injury on vegetation
Reduced vegetation growth and
reproduction
Yield and quality of commercial
forest products and crops
Damage to urban ornamental
plants
Carbon sequestration in terrestrial
ecosystems
Recreational demand associated
with forest aesthetics
Other non-use effects
Ecosystem functions (e.g., water
cycling, biogeochemical cycles, net
primary productivity, leaf-gas
exchange, community
composition)
Ozone CD, Draft
Ozone ISA2
Ozone CD, Draft
Ozone ISA1
Ozone CD, Draft
Ozone ISA1'3
Ozone CD, Draft
Ozone ISA2
Ozone CD, Draft
Ozone ISA2
Ozone CD, Draft
Ozone ISA2
Ozone CD, Draft
Ozone ISA2
Ozone CD, Draft
Ozone ISA2
(continued)
6-4
-------
Table 6-2. Quantified and Unquantified Welfare Benefits (continued)
Benefits Category Specific Effect
Effect Has
Been
Quantified
Effect Has
Been
Monetized
More
Information
Improved Environment (continued)
Reduced effects from acid
deposition
Recreational fishing
Tree mortality and decline
Commercial fishing and forestry
effects
Recreational demand in terrestrial
and aquatic ecosystems
Other non-use effects
Ecosystem functions (e.g.,
biogeochemical cycles)
NOx SOx ISA
NOx SOx ISA2
NOx SOx ISA
NOx SOx ISA
NOx SOx ISA
NOx SOx ISA
Reduced effects from
nutrient enrichment
Species composition and
biodiversity in terrestrial and
estuarine ecosystems
Coastal eutrophication
Recreational demand in terrestrial
and estuarine ecosystems
Other non-use effects
Ecosystem functions (e.g.,
biogeochemical cycles, fire
regulation)
NOx SOx ISA
NOx SOx ISA
NOx SOx ISA
NOx SOx ISA
NOx SOx ISA
Reduced vegetation effects
from ambient exposure to
SO2 and NOx
Injury to vegetation from SO2
exposure
Injury to vegetation from NOX
exposure
NOx SOx ISA
NOx SOx ISA
Reduced ecosystem effects
from exposure to
methylmercury (through the
role of sulfate in
methylation)
Effects on fish, birds, and
mammals (e.g., reproductive
effects)
Commercial, subsistence and
recreational fishing
Mercury Study
RTC
2,3
Mercury Study
RTC2
1 We assess these benefits qualitatively due to time and resource limitations for this analysis.
2 We assess these benefits qualitatively because we do not have sufficient confidence in available data or methods.
3 We assess these benefits qualitatively because current evidence is only suggestive of causality or there are other
significant concerns over the strength of the association.
4 We quantify these benefits in a sensitivity analysis, but not the main analysis.
6-5
-------
6.4 Visibility Benefits
6.4.1 Visibility and Light Extinction
The illustrative strategies designed to attain alternative standard level combinations
would reduce emissions of directly emitted PM2.5 as well as precursor emissions such as NOX
and S02. These emission reductions would improve the level of visibility throughout the United
States because these suspended particles and gases impair visibility by scattering and absorbing
light (U.S. EPA, 2009b).1 Visibility is also referred to as visual air quality (VAQ),2 and it directly
affects people's enjoyment of a variety of daily activities (U.S. EPA, 2009b). Good visibility
increases quality of life where individuals live and work, and where they travel for recreational
activities, including sites of unique public value, such as the Great Smoky Mountains National
Park (U. S. EPA, 2009b). This section discusses the economic benefits associated with improved
visibility as a result of emission reductions associated with the alternative PM2.5 standard level
combinations.
Air pollution affects light extinction, a measure of how much the components of the
atmosphere scatter and absorb light. More light extinction means that the clarity of visual
images and visual range is reduced, all else held constant. Light extinction is the optical
characteristic of the atmosphere that occurs when light is either scattered or absorbed, which
converts the light to heat. Particulate matter and gases can both scatter and absorb light. Fine
particles with significant light-extinction efficiencies include sulfates, nitrates, organic carbon,
elemental carbon, and soil (Sisler, 1996). The extent to which any amount of light extinction
affects a person's ability to view a scene depends on both scene and light characteristics. For
example, the appearance of a nearby object (e.g., a building) is generally less sensitive to a
change in light extinction than the appearance of a similar object at a greater distance. See
Figure 6-1 for an illustration of the important factors affecting visibility.
According to the PM ISA, there is strong and consistent evidence that PM is the
overwhelming source of visibility impairment in both urban and remote areas (U.S. EPA,
2009b). After reviewing all of the evidence, the PM ISA concluded that the evidence was
sufficient to conclude that a causal relationship exists between PM and visibility impairment.
1 The visibility benefits results shown in this section only reflect the emission reductions associated with attaining
the alternative PM2.5 primary standards. Visibility benefits results associated with attaining alternative
secondary PM NAAQS levels are provided in Chapter 13 of this RIA.
2 We use the term VAQ to refer to the visibility effects caused solely by air quality conditions, excluding fog.
6-6
-------
Figure 6-1. Important Factors Involved in Seeing a Scenic Vista (Malm, 1999)
Visibility is commonly measured as either light extinction (3ext), which is defined as the
loss of light per unit of distance in terms of inverse megameters (Mm"1), or using the deciview
(dv) metric, which is a logarithmic function of extinction (Pitchford and Malm, 1994). Deciviews,
a unitless measure of visibility, are standardized for a reference distance in such a way that one
deciview corresponds to a change of about 10% in available light.3 Pitchford and Malm (1994)
characterize a change of one deciview as "a small but perceptible scenic change under many
circumstances."4 Extinction and deciviews are both physical measures of the amount of
visibility impairment (e.g., the amount of "haze"), with both extinction and deciview increasing
as the amount of haze increases. Using the relationships derived by Pitchford and Malm (1994),
Note that deciviews are inversely related to visual range, such that a decrease in deciviews implies an increase in
visual range (i.e., improved visibility). Conversely, an increase in deciviews implies a decrease in visual range
(i.e., decreased visibility). Deciview, in effect, is a measure of the lack of visibility.
An instantaneous change of less than 1 deciview (i.e., less than 10% in the light extinction budget) represents a
measurable improvement in visibility but may not be perceptible to the eye. The visibility benefits analysis
described in this chapter reflects annual average changes in visibility, which are likely made up of periods with
changes less than one deciview and periods with changes exceeding one deciview. Annual averages appear to
more closely correspond to the economic literature relied upon for valuation of visibility changes in this
analysis. The secondary PM NAAQS uses a different averaging time than the benefits analysis (see Chapter 13).
6-7
-------
/391\ /B \
Deciviews = 10 * In — =10 * In -^
\ VR ) V 10 /
where VR denotes visual range (in kilometers) and 3ext denotes light extinction (in Mm"1).5
Annual average visibility conditions (reflecting light extinction due to both
anthropogenic and non-anthropogenic sources) vary regionally across the U.S. and by season
(U.S. EPA, 2009b). Particulate sulfate is the dominant source of regional haze in the eastern U.S.
(>50% of the particulate light extinction) and an important contributor to haze elsewhere in the
country (>20% of particulate light extinction) (U.S. EPA, 2009b). Particulate nitrate is an
important contributor to light extinction in California and the upper Midwestern U.S.,
particularly during winter (U.S. EPA, 2009b). Smoke plumes from large wildfires dominate many
of the worst haze periods in the western U.S., while Asian dust only caused a few of the worst
haze episodes, primarily in the more northerly regions of the west (U.S. EPA, 2009b). Higher
visibility impairment levels in the East are due to generally higher concentrations of fine
particles, particularly sulfates, and higher average relative humidity levels (U.S. EPA, 2009b).
Humidity increases visibility impairment because some particles such as ammonium sulfate and
ammonium nitrate absorb water and form droplets that become larger when relative humidity
increases, thus resulting in increased light scattering (U.S. EPA, 2009b).
Reductions in air pollution from implementation of various programs associated with
the Clean Air Act Amendments of 1990 (CAAA) provisions have resulted in substantial
improvements in visibility, and will continue to do so in the future. Because trends in haze are
closely associated with trends in particulate sulfate and nitrate due to the simple relationship
between their concentration and light extinction, visibility trends have improved as emissions
of S02 and NOx have decreased over time due to air pollution regulations such as the Acid Rain
Program (U.S. EPA, 2009b). For example, Figure 6-2 shows that visual range increased nearly
50% in the eastern U.S. since 1992.6 Recent EPA regulations such as the Cross-State Air
Pollution Rule (U.S. EPA, 2011c) and the Mercury and Air Toxics Standard (U.S. EPA, 2011d) are
anticipated to reduce S02 emissions down to 2 million tons nationally, which would lead to
substantial further improvement in visibility levels in the Eastern U.S. Calculated from light
5 It has been noted that, for a given deciview value, there can be many different visual ranges, depending on the
other factors that affect visual range—such as light angle and altitude. See Appendix 6a for more detail.
i Figure 6-2, the "best days" are defined as the best 20% of days, the "mid-range days" are defined as the
20%, and the "worst days" are defined as the worst 20% of days (IMPROVE, 2010).
6-8
-------
A. WtatnU.S,
B tanernu.s.
•?v.
£ 1V>
| 150
m •'.
»
•5
Bcst^lb.^*^ -
^ , --• " **^
— ^^
Mi:l-i9no» Ability d«ys
"fcVrrd ki^ihihhj rU»A
Z '94 « « BD TO TW TO T]
C ^
/;;).
BfRtutHfaliyttays^, -^-^
MtfHwMWtt^ye
Worst viability days
BV«'BB'WT(Z1MK»I0
'A- - h; -rij '.;<•••
: 30 monitoring SteS h ilifi -*KEIRIII US. and 11 mQnitaiiny Sties in IT* cistern U S wilh
flstg
EV sua' ranges are calculated TOTI the maasired tevdE of d^rent CDrnpcnantB wlhn alrfaome parclas and ^^ *> .
Pus?. cnm(if\fttfA&'
Oafs iflurt*: HtffflltrrtF.
Figure 6-2. Visibility in Selected National Parks and Wilderness Areas in the U.S., 1992-
2008a'b
(Source: U.S. EPA (2008) updated, IMPROVE (2010))
extinction efficiencies from Trijonis et al. (1987, 1988), annual average visual range under
natural conditions in the East is estimated to be 150 km ± 45 km (i.e., 65 to 120 miles) and 230
km ± 35 km (i.e., 120 to 165 miles) in the West (Irving, 1991). Figure 6-2 reflects the average
trends in visual ranges at select monitors in the eastern and western areas of the U.S. since
1992 using data from the IMPROVE monitoring network (U.S. EPA (2008) updated; IMPROVE
(2010)). As an illustration of the improvements in visibility attributable to the CAAA, Figure 6-3
depicts the modeled improvements in visibility associated with all the CAAA provisions in 2020
compared to a counterfactual scenario without the CAAA (U.S. EPA, 2011b). While visibility
trends have improved in most National Parks, the recent data show that these areas continue
to suffer from visibility impairment beyond natural background levels (U.S. EPA, 2009b).
For the final rulemaking, we would generate light extinction estimates using the CMAQ
model in conjunction with the IMPROVE (Interagency Monitoring of Protected Visual
Environments) algorithm that estimates light extinction as a function of PM concentrations and
relative humidity levels (U.S. EPA, 2009b).7 The procedure for calculating light extinction
7 According to the PM ISA, the algorithm performs reasonably well despite its simplicity (U.S. EPA, 2009b).
6-9
-------
/
\
Without CAAA
I*
With CAAA
.
V
Visibility in Declviews, 2020
036 9 15 15 18 21 24 27 SO
Best Wore!
Figure 6-3. Estimated Improvement in Annual Average Visibility Levels Associated with the
CAAA Provisions in 2020
Source: U.S. EPA, 2011b8
' It is important to note that visibility levels shown in these maps were modeled differently than the modeling
conducted for this analysis, including coarser grid resolution (i.e., 36 km instead of 12 km). In addition, these
maps present annual average visibility levels, which are different than the short-term averages being
considered for the secondary standard.
6-10
-------
associated with alternative standard level combinations is described in detail in Chapter 3 of
this RIA. In addition, Appendix 6a describes how the spatial resolution of the light extinction
estimates was then adjusted for the benefits analysis.
It is important to note that the light extinction estimates used in this benefits analysis
represent annual averages, which is different from the averaging times currently being
considered for the secondary PM NAAQS. While the annual averages are influenced by days
with extremely impaired visibility, the light extinction data is not sufficient to provide higher
temporal resolution than quarterly averages. While we suspect that the most impaired days
would have disproportionately improved visibility from the emission reductions associated with
attaining the alternative standard level combinations, we are not able to quantify those
impacts. These data gaps result in an underestimate of visibility benefits associated with
extreme days. We recognize that recent advice from EPA's Science Advisory Board recommends
estimating visibility benefits considering daytime visibility on days with severe impairment (U.S.
EPA-SAB, 2010), but the available data and valuation studies do not allow such fine temporal
resolution.
While we have made substantial improvements in estimating light extinction nationally
in this analysis, we are still developing a method to estimate coarse particle concentrations for
the entire continental U.S. for estimating light extinction. As an interim solution, we provide
sensitivity analyses to show the potential impact of omitting coarse particles from the
recreational and residential visibility benefits analysis. For this sensitivity analysis, we selected
the levels of coarse particles to represent the full range of possible annual concentrations from
a recent report on the IMPROVE monitoring network (Debell et al., 2006). We estimate the
sensitivity of impacts on recreational and residential visibility benefits using four levels of
coarse particles: no coarse particles, 5 u.g/m3 nationwide, 15 u.g/m3 in the Southwest with
5 u.g/m3 in the rest of the country, and 15 u.g/m3 in the Southwest with 8 u.g/m3 in the rest of
the country.9
In Table 6-3, we also provide a qualitative assessment of how key assumptions in the
estimation of light extinction affect the visibility benefits.
9 We define "Southwest" for this sensitivity analysis to be the states of California, Nevada, Utah, Arizona, New
Mexico, Colorado, and Texas.
6-11
-------
Table 6-3. Key Assumptions in the Light Extinction Estimates Affecting the Visibility Benefits
Analysis3
Key Assumption
The light extinction estimates are annual averages to
correspond with the valuation studies. People may value large
changes to the haziest days differently than small changes to
many days. We assume that annual average light extinction is
the most appropriate temporal scale for estimating visibility
benefits.
Coarse particles are a component of light extinction, but we
were unable to include coarse particles in the light extinction
estimates. We provide sensitivity analyses with up to 15 ng/m
in the Southwest and 8 ng/m3 in the rest of the country.
Direction of Bias
Potential
Underestimate
Potential
Overestimate
Magnitude of Effect
Medium
Very Low
a A description of the classifications for magnitude of effects can be found in Appendix 5C of this RIA.
6.4.2 Visibility Valuation Overview
In the Clean Air Act Amendments of 1977, the U.S. Government recognized visibility's
value to society by establishing a national goal to protect national parks and wilderness areas
from visibility impairment caused by manmade pollution.10 Air pollution impairs visibility in
both residential and recreational settings, and an individual's willingness to pay (WTP) to
improve visibility differs in these two settings. Benefits of residential visibility relate to the
impact of visibility changes on an individual's daily life (e.g., at home, at work, and while
engaged in routine recreational activities). Benefits of recreational visibility relate to the impact
of visibility changes manifested at parks and wilderness areas that are expected to be
experienced by recreational visitors.
Both recreational and residential visibility benefits consist of use values and nonuse
values. Use values include the aesthetic benefits of better visibility, improved road and air
safety, and enhanced recreation in activities like hunting and birdwatching. Nonuse values are
based on a belief that the environment ought to exist free of human-induced haze. This
includes the value of better visibility for use by others now and in the future (bequest value).
Nonuse values may be more important for recreational areas, particularly national parks and
monuments.
The relationship between a household's WTP and changes in visibility can be derived
from a number of contingent valuation (CV) studies published in the peer-reviewed economics
literature. The studies used to estimate the residential and recreational visibility benefits
10 See Section 169(a) of the Clean Air Act.
6-12
-------
associated with alternative standard level combinations are described in the following sections.
In addition to CV studies, hedonic valuation studies (Beron et al., 2001, 2004) also demonstrate
that visibility has value, but we are unable to apply these valuation estimates in the context of
estimating the visibility benefits associated with national regulations that reduce air pollution
(Leggett and Neumann, 2004).
In this approach, we assume that individuals value visibility for aesthetic reasons rather
than viewing visibility as a proxy for other impacts associated with air pollution, such as health
or ecological improvements. Some studies in the literature indicate that individuals may have
difficulty distinguishing visibility from other aspects of air pollution (e.g., McClelland et al.,
1993; Chestnut and Rowe, 1990c; Carson, Mitchell, and Rudd, 1990). Because visual air quality
is inherently multi-attribute, it is a challenge for all visibility valuation studies to isolate the
value of visibility from the collection of intertwined benefits. Each study used in this analysis
attempts to isolate visibility from other effect categories, but the different studies take
different approaches (U.S. EPA, 2009b).n Because we believe that residual potential for double-
counting visibility and health effects is relatively minimal, we do not further adjust the benefits
to account for potentially embedded health effects beyond what the studies have already done.
Similarly, it is important to try to distinguish residential visibility from recreational
visibility benefits, specifically whether these can these be treated as distinct and additive
benefit categories based on the available literature. In this analysis, we assume that residential
and recreational visibility benefits are distinct and separable. It is conceivable that respondents
to the recreational visibility survey may have partially included values for their own residential
visibility when evaluating changes at national parks and wilderness areas located in their region
of the country. However, we believe that the potential for double-counting recreational and
residential visibility is minimal for several reasons. First, we only include a subset of areas in the
primary estimates of recreational and residential visibility benefits, which overlap in only a few
places.12 Second, a number of the overlapping counties are wilderness areas, which contribute
little to the overall monetized benefits due to low visitation rates, rather than highly visited
national parks. For example, Los Angeles County is home to the San Gabriel Wilderness Area,
which has 10 thousand annual visitors (NFS, 2008). If we exclude the residential visibility
benefits that accrue to 10 million residents in Los Angeles County and only include the very
small recreational visibility benefits for the wilderness area, we would be substantially biasing
11 See Leggett and Neumann (2004) for a more detailed discussion of this issue.
12 As described in detail in Sections 6.3.3 and 6.3.4, we only include a subset of visibility benefits in the primary
benefits estimates, while providing the rest of the visibility benefits in sensitivity analyses.
6-13
-------
the overall estimates downward. For these reasons, we believe that the potential for double-
counting is minimal.
In the next sections, we describe the methodology and limitations of the recreational
and residential visibility analysis. Consistent with the health benefits analysis, the monetized
visibility benefits would be adjusted for inflation and income growth. These benefits would be
specific to the analysis year, and as population and income increase over time, these benefits
can be expected to increase each year for the same incremental change in light extinction.
6.4.3 Recreational Visibility
6.4.3.1 Methodology
The methodology for estimating recreational visibility benefits in this RIA follows a well-
established approach that has been used in numerous EPA analyses (U.S. EPA, 2006; U.S. EPA,
2005; U.S. EPA, 2010; U.S. EPA, 1999; U.S. EPA, 2011b). For the purposes of this analysis,
recreational visibility benefits apply to Class I areas, such as National Parks and Wilderness
Areas.13 Although other recreational settings such as National Forests, state parks, or even
hiking trails or roadside areas have important scenic vistas, a lack of suitable economic
valuation literature to identify these other areas and/or a lack of visitation data prevents us
from generating estimates for those recreational vista areas.
Under the 1999 Regional Haze Rule (64 FR 35714), states are required to set goals
develop long-term strategies to improve visibility in Class I areas, with the goal of achieving
natural background visibility levels by 2064. In conjunction with the U.S. National Park Service
(NPS), the U.S. Forest Service (USFS), other Federal land managers, and State organizations in
the U.S., the U.S. EPA has supported visibility monitoring in national parks and wilderness areas
since 1988. The monitoring network known as IMPROVE includes 156 sites that represent the
Class I areas across the country (U.S. EPA, 2009b).14 The IMPROVE monitoring network
measures fine particles, coarse particles, and key PM2.s constituents that affect visibility, such
as sulfate, nitrate, organic and elemental carbon, soil dust, and several other elements.
Figure 6-4 identifies where each of these parks are located in the U.S.
13 Hereafter referred to as Class I areas, which are defined as areas of the country such as national parks, national
wilderness areas, and national monuments that have been set aside under Section 169(a) of the Clean Air Act
to receive the most stringent degree of air quality protection. Mandatory Class I federal lands fall under the
jurisdiction of three federal agencies, the National Park Service, the Fish and Wildlife Service, and the Forest
Service. EPA has designated 156 areas as mandatory Class I federal areas for visibility protection, including
national parks that exceed 6,000 acres and wilderness areas that exceed 5,000 acres (40 CFR §81.400).
14 The formula used to estimate light extinction from concentrations of PM constituents and relative humidity is
referred to as the IMPROVE algorithm.
6-14
-------
Figure 6-4. Mandatory Class I Areas in the U.S.
For recreational visibility, EPA has determined that only one existing study provides
adequate monetary estimates of the value of changes in recreational visibility: a contingent
valuation (CV) survey conducted by Chestnut and Rowe in 1988 (1990a; 1990b). Although there
are several other studies in the literature on recreational visibility valuation, they are older and
use less robust methods. In EPA's judgment, the Chestnut and Rowe study contains many of the
elements of a valid CV study and is sufficiently reliable to serve as the basis for monetary
estimates of the benefits of visibility changes in recreational areas.15 This study serves as an
essential input to our estimates of the benefits from improving recreational visibility.
In this analysis, we assume that the household WTP is higher if the Class I recreational
area is located close to the person's home (i.e., in the same region of the country). People
appear to be willing to pay more for visibility improvements at parks and wilderness areas that
In 1999, EPA's SAB stated, "many members of the Council believe that the Chestnut and Rowe study is the best
available" study for recreational visibility valuation" (U.S. EPA-SAB, 1999). In July 2010, the SAB stated that the
studies were dated, but EPA "used what the Council understands to be the only relevant studies." (U.S. EPA-
SAB, 2010)
6-15
-------
are in the same region as their household than at those that are not in the same region as their
household (Chestnut and Rowe, 1990a, 1990b). This is plausible, because people are more likely
to visit, be familiar with, and care about parks and wilderness areas in their own part of the
country. However, studies have also found many people who had never visited and never
planned to visit the parks still had positive values for visibility improvements in those locations
(Chestnut and Rowe, 1990b).
The Chestnut and Rowe survey measured the demand for visibility in Class I areas
managed by the NFS in three broad regions of the country: California, the Colorado Plateau
(Southwest), and the Southeast.16 Respondents in five states were asked about their WTP to
protect national parks or NPS-managed wilderness areas within a particular region. The survey
used photographs reflecting different visibility levels in the specified recreational areas. The
authors used the survey data to estimate household WTP values for improved visibility in each
region.
The separate regions were developed to capture differences in household WTP values
based on proximity to recreational areas. Chestnut (1997) also concluded that, fora given
region, a substantial proportion of the WTP is attributable to one specific park within the
region. This so called "indicator park" is the most well-known and frequently visited park within
a particular region. The indicator parks for the three studied park regions are Yosemite National
Park for the California region, the Grand Canyon National Park for the Southwest region, and
Shenandoah National Park for the Southeast region. In accordance with the methodology in
Chestnut (1997), this analysis calculates the benefits from households for a particular region for
a given change in visibility at a particular Class I area. In theory, summing benefits from
households in all regions would yield the total monetary benefits associated with a given
visibility improvement at a particular park, which could then be summed with other parks and
regions to estimate national benefits. Because recreational visibility benefits may reflect the
value an individual places on visibility improvements regardless of whether the person plans to
visit the park, all households in the U.S. are assumed to derive some benefit from
improvements to Class I areas.
To value recreational visibility improvements associated with its rulemakings, EPA
developed a valuation WTP equation function based on the baseline of visibility, the magnitude
16 The Colorado Plateau (Southwest) region is defined as the states of Colorado, New Mexico, Arizona, and Utah.
The Southeast region is defined as the states of West Virginia, Virginia, North Carolina, South Carolina, Georgia,
Florida, Alabama, Mississippi, Louisiana, Tennessee, and Kentucky. The California region includes the state of
California and one wilderness area in Nevada.
6-16
-------
of the visibility improvement, and household income. This function requires light extinction
estimates measured as visual range. The behavioral parameters of this equation were taken
from analysis of the Chestnut and Rowe survey (1990a, 1990b). These parameters were used to
calibrate WTP for the visibility changes resulting from this rule.17 As an example, household
WTP for a visibility improvement at a park in their region takes the following form:
= m - mP + yik * <.fc - Q
where:
i indexes region,
k indexes park,
m = household income,
p = shape parameter (0.1),
V = parameter corresponding to the visibility at in-region parks,
do = starting visibility, and
Qi = visibility after change.
As discussed in more detail in Appendix 6a of this RIA, this approach to valuing
recreational visibility changes is an application of the Constant Elasticity of Substitution (CES)
utility function approach and is based on the preference calibration method developed by
Smith, Van Houtven, and Pattanayak (2002). 18 Available evidence indicates that households are
willing to pay more for a given visibility improvement as their income increases (Chestnut,
1997). Using the income elasticity calculated by Chestnut (1997), the visibility benefits assume a
1% increase in income is associated with a 0.9% increase in WTP for a given change in visibility.
WTP responses reported in Chestnut and Rowe (1990a, 1990b) were also region-specific, rather
than park-specific. As visibility improvements are not constant across all parks in a region, we
must infer park-specific visibility parameters in order to calculate WTP for projected visibility
changes. As the quantity and quality of parks differs between regions, we apportion the
regional WTP parameters based on relative visitation rates at the different parks, because this
statistic likely captures both park quality (more people visit parks with more desirable
attributes, so collective WTP is likely higher) and quantity (more people visit parks in a region if
17 The parameters for each region are available in Appendix 6a of this RIA.
18 The Constant Elasticity of Substitution utility function has been chosen for use in this analysis due to its flexibility
when illustrating the degree of substitutability present in various economic relationships (in this case, the
tradeoff between income and improvements in visibility).
6-17
-------
the parks are more numerous, so collective WTP is likely higher).19 We also adjust the benefits
for inflation and growth in real income.
Recreational visibility benefits are calculated as the sum of the household WTPs for
changes in light extinction. We assume that each household is valuing the first or only visibility
change that occurs in a particular area. The benefits at particular areas can be calculated by
assuming that the subset of visibility changes of interest is the first or the only set of changes
being valued by households. Estimating benefit components in this way will yield slightly
upwardly biased estimates of benefits, because disposable income is not reduced by the WTPs
for any prior visibility improvements. The upward bias should be extremely small, however,
because all of the WTPs for visibility changes are very small relative to income.
The primary estimate for recreational visibility only includes benefits for 86 Class I areas
in the original study regions (i.e., California, the Southwest, and the Southeast).20 These
benefits reflect the value to households living in the same region as the Class I area as well as
values for all households in the United States living outside the state containing the Class I area.
The Chestnut and Rowe study did not measure values for visibility improvement in Class
I areas in the Northwest, Northern Rockies, and Rest of U.S. regions.21 Their study covered 86 of
the 156 Class I areas in the United States. We can infer the value of visibility changes in the 70
additional Class I areas by transferring values of visibility changes at Class I areas in the study
regions.22 In order to obtain estimates of WTP for visibility changes for parks in these additional
regions, we have to transfer the WTP values from the studied regions. This benefits transfer
approach introduces additional uncertainty into the estimates. However, we have taken steps
to adjust the WTP values to account for the possibility that a visibility improvement in parks
within one region may not necessarily represent the same visibility improvement at parks
within a different region in terms of environmental improvement. This may be due to
19 We use 2008 park visitation data from the National Park Service Statistical Abstracts (IMPS, 2008), as this is the
most current data available. Where the data for a particular park was not representative of normal visitation
rates at that park (for example due to fire damage that occurred during that year), we substitute data from the
prior year. We use 1997 visitation data for those wilderness areas not included in the National Park Service
Statistical Abstracts, as more current data is not readily available. As visitation rates for Wilderness Areas are
small compared to visitation rates in National Parks, the inaccuracies generated by using 1997 data are likely to
also be small.
20 The 86 Class I areas in the three studied park regions represented 68% of the total visitor days to Class I areas in
2008 (NPS, 2008).
21 The Northwest region is defined as the states of Washington and Oregon. The Northern Rockies region include
the states of Idaho, Montana, Wyoming, North Dakota, and South Dakota. The Rest of the U.S. region includes
all other states not included in the other 5 regions.
22 The 70 additional Class I areas represented 32% of the total visitor days to Class I areas in 2008 (NPS, 2008).
6-18
-------
differences in the scenic vistas at different parks, uniqueness of the parks, or other factors, such
as public familiarity with the park resource. To account for this potential difference, we
adjusted the transferred WTP being transferred by the ratio of visitor days in the two regions.23
A complete description of the benefits transfer method used to infer values for visibility
changes in Class I areas outside the study regions is provided in Appendix 6a of this RIA.
Table 6-4 indicates which studied park regions we used to estimate the value in the non-
studied park regions. Figure 6-5 shows how the visitation rates vary across Class I areas and
regions and indicates whether each Class 1 area is located within one of the studied regions.
6.4.3.2 Recreational Visibility Limitations, Caveats, and Uncertainties
This analysis relies upon several data sources as inputs, including emission inventories,
air quality data from models (with their associated parameters and inputs), relative humidity
measurements, park information, economic data and assumptions for monetizing benefits.
Each of these inputs may contain uncertainty that would affect the recreational visibility
benefits estimates. Though we are unable to quantify the cumulative effect of all of these
uncertainties in this analysis, we do provide information on uncertainty based on the available
data, including model evaluation24 and sensitivity analyses to characterize major omissions (i.e.,
benefits from parks in non-studied park regions and inclusion of coarse particles). Although we
strive to incorporate as many quantitative assessments of uncertainty as possible, we are
severely limited by the available data, and there are several aspects that we are only able to
address qualitatively. A summary of the key assumptions including direction and magnitude of
bias is provided in Table 6-5.
One major source of uncertainty for the recreational visibility benefits estimate is the
benefits transfer process. Choices regarding the functional form and key parameters of the
estimating equation for WTP for the affected population could have significant effects on the
magnitude of the estimates. Assumptions about how individuals respond to changes in visibility
that are either very small or outside the range covered in the Chestnut and Rowe study could
also affect the estimates.
23 For example, if total park visitation in a transfer region was less than visitation in a study region, transferred WTP
would be adjusted downward by the ratio of the two.
24 See Chapter 4 for more information on model evaluation.
6-19
-------
Table 6-4. WTP for Visibility Improvements in Class I Areas in Non-Studied Park Regions
Park Region
Source of WTP Estimate
1. Northwest
benefits transfer from California
2. Northern Rockies
benefits transfer from Colorado Plateau
3. Rest of U.S.
benefits transfer from Southeast
o
O
* 0-30.KO
q HJ.DQI -WO/MO
i.m».H»
1 CW OTI • 3.CTO.OTO
3C030CI •
Figure 6-5. Visitation Rates and Park Regions for Class I Areas*
* The colors in this map correspond to the park regions used in the valuation study and the extrapolation to parks
in other regions. Red = California, light red = Northwest (extrapolated from California), blue = Colorado Plateau,
light blue = Northern Rockies (extrapolated from Colorado Plateau), green = Southeast, light green = Rest of U.S.
(extrapolated from Southeast).
6-20
-------
Table 6-5. Summary of Key Assumptions in the Recreational Visibility Benefits3
Key Assumption
Potential Magnitude
Direction of Bias of Effect
Chestnut and Rowe study covers parks in three regions: California,
Southwest, and Southeast. Benefits to other regions in the U.S.
are not included in the primary benefits estimate.
Underestimate
Medium
Benefits to other recreational settings, such as National Forests
and state parks, are not included in this analysis.
Chestnut and Rowe study conducted on populations in five states.
These results are applied to the entire U.S. population.
Individuals have a greater WTP for visibility changes in parks
within their region.
WTP values reflect only visibility improvements and not overall air
quality improvements.
We assume that there are 2.68 people per household. Because
this estimate has been decreasing over time, this may
underestimate the number of households.
Underestimate
Unclear
Unclear
Potential
Overestimate
Potential
Underestimate
Medium-Low
Unclear
Unclear
Unclear
Medium-Low
a A description of the classifications for magnitude of effects can be found in Appendix 5C of this RIA.
Since the valuation of recreational visibility benefits relies upon one study (Chestnut and
Rowe, 1990a; 1990b), all of the uncertainties within that study also pertain to this analysis. In
general, the survey design and implementation reflect the period in which the Chestnut and
Rowe study was conducted. Since that time, many improvements to the design of stated
preference surveys have been developed (e.g., Arrow, 1993), but we are currently unaware of
newer studies that we could incorporate into our visibility benefits methodology. Although
Chestnut and Rowe still offers the best available WTP estimates, the study has a number of
limitations, including:
• The vintage of the survey (late 1980s) invites questions whether the values would
still be valid for current populations, or more importantly for this analysis, future
populations in 2020.
• The survey focused on visibility improvements in and around specific national parks
and wilderness areas. Given that national parks and wilderness areas exhibit unique
characteristics, it is not clear whether the WTP estimate obtained from this survey
can be transferred to other national parks and wilderness areas, even other parks
within the studied park regions, without introducing additional uncertainty.
6-21
-------
• The survey focused only on populations in five states, so the application of the
estimated values to populations outside those states requires that preferences of
populations in the five surveyed states be similar to those of non-surveyed states.
• There is an inherent difficulty in separating values expressed for visibility
improvements from an overall value for improved air quality. The survey attempted
to control for this by informing respondents that "other households are being asked
about visibility, human health, and vegetation protections in urban areas and at
national parks in other regions." However, most of the respondents did not feel that
they were able to segregate recreational visibility at national parks entirely from
residential visibility and health effects.
• It is not clear exactly what visibility improvements the respondents to the survey
were valuing. The WTP question asked about changes in average visibility, but the
survey respondents were shown photographs of only daytime, summer conditions,
when visibility is generally at its worst. It is possible that the respondents believed
those visibility conditions held year-round, in which case they would have been
valuing much larger overall improvements in visibility than what otherwise would be
the case. For the purpose of the benefits analysis for this rule, EPA assumed that
respondents provided values for changes in annual average visibility. Because most
policies would result in a shift in the distribution of visibility (usually affecting the
worst days more than the best days), the annual average may not be the most
relevant metric for policy analysis.
• The survey did not include reminders of possible substitutes (e.g., visibility at other
parks) or budget constraints. These reminders are considered to be best practice for
stated preference surveys.
6.4.4 Residential Visibility
6.4.4.1 Methodology
Residential visibility benefits are those that occur from visibility changes in urban,
suburban, and rural areas where people live. These benefits are important because some
people living in certain urban areas may place a high value on unique scenic resources in or
near these areas that are outside of Class I areas. For example, the State of Colorado
established a local visibility standard for the Denver metropolitan area in 1990 (Ely et al., 1991).
For the purposes of this analysis, residential visibility improvements are defined as those that
occur specifically in Metropolitan Statistical Areas (MSAs).
In the Urban-focused Visibility Assessment (U.S. EPA, 2010b) and the Policy Assessment
for the Review of the PM NAAQS (U.S. EPA, 2011a), several preference studies provide the
6-22
-------
foundation for the secondary PM NAAQS.25 The three completed survey studies (all in the west)
included Denver, Colorado (Ely et al., 1991), one in the lower Fraser River valley near
Vancouver, British Columbia (BC), Canada (Pryor, 1996), and one in Phoenix, Arizona (BBC
Research & Consulting, 2003). A pilot focus group study was conducted in Washington, DC on
behalf of EPA to inform the 2006 PM NAAQS review (Abt Associates Inc., 2001). While these
studies indicate that visual air quality associated with ambient levels of air pollution in urban
areas have been deemed unacceptable, they do not provide sufficient information on which to
develop monetized benefits estimates. Specifically, the public perception studies do not
provide preferences expressed in dollar values, even though they do provide additional
evidence that the benefits associated with improving residential visibility are not zero.
A wide range of published, peer-reviewed literature supports a non-zero value for
residential visibility (Brookshire et al., 1982; Rae, 1983; Tolley et al., 1984; Chestnut and Rowe,
1990c; McClelland et al., 1993; Loehman et al., 1994). Furthermore, Chestnut and Rowe (1990c)
conclude that residential visibility benefits are likely to be at least as high as recreational
visibility benefits because of the quantity of time most people spend in and near their homes
and the substantial number of people affected. In previous assessments, EPA used a study on
residential visibility valuation conducted in 1990 (McClelland et al., 1993). Consistent with
advice from EPA's Science Advisory Board (SAB), EPA designated the McClelland et al. study as
significantly less reliable for regulatory benefit-cost analysis, although it does provide useful
estimates on the order of magnitude of residential visibility benefits (U.S. EPA-SAB, 1999).26 In
order to estimate residential visibility benefits in this analysis, we have replaced the previous
methodology with a new benefits transfer approach and incorporated additional valuation
studies. This new approach was developed for The Benefits and Costs of the Clean Air Act 1990
to 2020: EPA Report to Congress (U.S. EPA, 2011)27 and reviewed by the SAB (U. S. EPA-SAB,
2004, 2010).
25 For more detail about these preference studies, including information about study designs and sampling
protocols, please see Section 2 of the Particulate Matter Urban-Focused Visibility Assessment (U.S. EPA, 2010b).
26 EPA's Advisory Council on Clean Air Compliance Analysis noted that the McClelland et al. (1993) study may not
incorporate two potentially important adjustments. First, their study does not account for the "warm glow"
effect, in which respondents may provide higher willingness to pay estimates simply because they favor "good
causes" such as environmental improvement. Second, while the study accounts for non-response bias, it may
not employ the best available methods. As a result of these concerns, the Council recommended that
residential visibility be omitted from the overall primary benefits estimate. (U.S. EPA-SAB, 1999)
27 This report is also known as the Second Prospective 812 analysis.
6-23
-------
To value residential visibility improvements, the new approach draws upon information
from the Brookshire et al. (1979), Loehman et al. (1985) and Tolley et al. (1984) studies.28 Each
of the studies provides estimates of household WTP to improve visibility conditions. While
uncertainty exists regarding the precision of these older, stated-preference residential
valuation studies, we believe their results support the argument that individuals have a non-
zero value for residential visibility improvements. These studies provide primary visibility values
for Atlanta, Boston, Chicago, Denver, Los Angeles, Mobile, San Francisco, and Washington
D.C.29
In accordance with Chestnut and Rowe (1990c), we utilize the WTP estimates and the
associated change in visual range from each study to estimate the 3 parameter for the eight
study areas. The 3 parameter represents the WTP for a specific improvement in visibility in a
specific location. Where studies provide multiple estimates for visual range improvements, we
estimate 3 by regressing the natural log of the ratio of visual range following and prior to
improvement against WTP. To express these value estimates in comparable terms across study
locations, we express household WTP for a change in visual range in a specific MSA using the
following function:
WTP(AVR)=3*ln(^)
where:
VR0 = mean annual visual range in miles before the improvement,
VRi = mean annual visual range in miles after the improvement, and
3 = parameter.
Total residential visibility benefits within a particular MSA are driven by visibility
improvements, population density, and the WTP value applied. Only those people living within
28 Loehman et al. (1985) and Brookshire et al. (1979) were subsequently published in peer-reviewed journals (see
Loehman et al. (1994) and Brookshire et al. (1982). The specific details need to compute visibility benefits using
Tolley et al. (1984) were not subsequently published, but the overall work including study and survey design
was subject to peer review during study development, (see Leggett et al, 2004 and Patterson et al., 2005) In
addition, Tolley et al. subsequently published a book based on this research, which notes in the preface that
the methods were critiqued throughout by various external economists (Tolley et al., 1988).
29 Recognizing potential fundamental issues associated with data collected in Cincinnati and Miami (e.g., see
Chestnut et al. (1986) and Chestnut and Rowe (1990c), we do not include values for these cities in our analysis.
The 8 MSAs where the valuation studies were conducted represent 15% of the total US population in 2020 (U.S.
Census).
6-24
-------
in an MSA are assumed to receive benefits from improved residential visibility. In other words,
unlike recreational visibility, we do not assume a non-use value by people who live outside the
MSA for residential visibility. Table 6-6 provides a summary of these valuation estimates for
each study location, as well as an illustrative implied WTP value for a 10% improvement in
visual range. As shown, the implied annual per-household WTP estimates for a hypothetical
10% improvement ranges from $21 to $220, depending on the study area. It is not surprising
that such a range of values exists, as these study areas all feature different landscapes and
vistas, populations and prevailing visibility conditions.
Table 6-6. Summary of Residential Visibility Valuation Estimates
City
Atlanta
Boston
Chicago
Denver
Los Angeles
Mobile
San Francisco
Washington, DC
Study
Tolleyetal. (1984)
Tolleyetal. (1984)
Tolleyetal. (1984)
Tolleyetal. (1984)
Brookshireetal. (1979)
Tolleyetal. (1984)
Loehman etal. (1985)
Tolleyetal. (1984)
Implied WTP for 10%
Improvement in Visual
Range (1990$, 1990
P Estimate income)
321
398
310
696
94
313
989
614
$31
$38
$30
$66
$9
$30
$94
$59
Implied WTP for 10%
Improvement in Visual
Range (2006$, 2020
income)
$72
$89
$69
$155
$21
$70
$220
$137
a The table assumes full attainment of the alternative standard level combinations. Because these benefits occur
within the analysis year, the monetized benefits are the same for all discount rates. These benefits reflect the
WTP for households who live in MSAs.
Similar to recreational visibility benefits, we then incorporate preference calibration
using the method developed by Smith, Van Houtven, and Pattanayak (2002), which is discussed
in more detail in Appendix 6a of this RIA. To express these "preference-calibrated" value
estimates across study locations, we express household WTP for a change in visual range in a
specific MSA using the following function:
6-25
-------
WTP(AVK) = m - [mP + 9 * (VRPQ - VRP^]\P
where:
m = household income,
p = shape parameter (0.1),
0 = WTP parameter corresponding to the visibility at MSA,
VR0 = starting visibility, and
VRi = visibility after change.
While the primary estimate for residential visibility includes benefits in only the eight
MSAs included in the valuation studies, people living in other urban areas also have non-zero
values for residential visibility. For this reason, the sensitivity analysis for residential visibility
includes the benefits extrapolated to the 351 additional MSAs.30 Because there is considerable
uncertainty about the validity of this benefit transfer approach, these extrapolated benefits are
included in a sensitivity analysis only.
There are many factors that could influence WTP for residential visibility, and these
factors vary across urban areas. For the purpose of this analysis, we utilize the benefit transfer
approach developed for the Second Prospective 812 analysis, but we recognize that there are
alternative methods that we could have used. We assigned a valuation study area to each MSA
based on two factors: geographic proximity to one of the eight study cities and elevation. Any
MSA with a county elevation above 1,500 meters was assigned the Denver valuation instead of
the nearest study area.31 Because residents of Denver have a dramatic view of the Rocky
Mountains that is rarely obstructed by trees, it is plausible that they might have a greater
interest in protecting visibility than a city without nearby mountains. The geographic proximity
factor is constrained in two areas. The San Francisco valuation study is only assigned to the six
counties in the San Francisco Bay area MSAs because the study is unique among the three
regarding the temporal description of visibility conditions, landscape/vistas, and prevailing
weather conditions. In addition, the Los Angeles valuation was assigned to the Riverside MSA
30 The 351 additional MSAs plus the 8 study area MSAs represent 84% of the total US population in 2020 (U.S.
Census).
31 Elevation data represent the county-level maximum, which were calculated using the ArcGIS Spatial Analyst tool
"Zonal Statistics" using the geographic database HYDRO1K for North America (U.S. Geological Survey, 1997).
This dataset and associated documentation are available on the Internet at
DEMhttp://eros.usgs.gov/#/Find_Data/Products_and_Data_Available/gtopo30/hydro/namerica.
6-26
-------
despite exceeding the elevation threshold.32 Figure 6-6 indicates the study cities as well as the
assignment of the other MSAs to the study cities.
6.4.4.2 Residential Visibility Limitations, Caveats, and Uncertainties
Similar to recreational visibility benefits, there are many data inputs into the residential
visibility benefits that contribute to overall uncertainty. We provide sensitivity analyses to
characterize major omissions (i.e., benefits in other MSAs and coarse particles). A summary of
the key assumptions including direction and magnitude of bias is provided in Table 6-7.
Figure 6-6. Residential Visibility Study City Assignment
Riverside MSA is assigned to the Los Angeles study area because a significant portion of Riverside County itself is
located in the South Coast Air Quality Management District, which can be considered by to be part of the same
regulated airshed as Los Angeles. The geographic assignment is preserved despite exceeding the elevation
threshold because Riverside is adjacent to one of the study cities and this region has a particular set of location-
specific characteristics that set it apart from Denver.
6-27
-------
Table 6-7. Summary of Key Assumptions in the Residential Visibility Benefits3
Key Assumption
Direction of Bias
Magnitude of
Effect
Residential and recreational visibility benefits are distinct and
separable.
Potential
Overestimate
Medium-Low
Estimates residential visibility benefits are limited to populations
within the boundaries of MSAs. Areas outside of an MSA are not
included in this analysis.
Underestimate
Low
WTP values reflect only visibility improvements and not overall air Potential
quality improvements. Overestimate
Medium-Low
WTP values from studies in Atlanta, Boston, Chicago, Denver, Los
Angeles, Mobile, San Francisco, and Washington D.C. can be
accurately transferred to MSAs across the U.S. based on proximity
and elevation
Unclear
Unclear
We assume that there are 2.68 people per household. Because
this estimate has been decreasing over time, this may
underestimate the number of households.
Potential
Underestimate
Medium-Low
a A description of the classifications for magnitude of effects can be found in Appendix 5C of this RIA.
The valuation studies relied upon for the residential visibility benefits, although
representing the best available estimates, have a number of limitations. These include the
following:
• The survey design and implementation reflects the period in which the surveys were
conducted. Since that time, many improvements to the stated preference methods
have been developed.
• The vintage of the surveys (1970s and 1980s) invites questions whether the values
are still valid for current populations, or more importantly for this analysis, future
populations in 2020.
• The survey focused only on populations in eight cities, so the transfer of the WTP
estimates values to populations outside those cities requires that their preferences
be similar to those in non-surveyed cities, as well as the visibility attributes be
similar across study and transfer MSAs.
• There is an inherent difficulty in separating values expressed for visibility
improvements from an overall value for improved air quality. The studies attempted
to control for this, but most of the respondents did not feel that they were able to
segregate residential visibility entirely from recreational visibility and health effects.
6-28
-------
6.4.5 Discussion of Visibility Benefits
As described in the previous sections of this chapter, the estimation of visibility benefits
is complex and suffers from unavoidable limitations. While we are confident that the underlying
scientific literature supports a non-zero estimate for visibility benefits attributable to emission
reductions, we are less confident in the magnitude of those benefits outside of previously
studied locations. While acknowledging these limitations, it is important to emphasize that
these valuation studies have withstood intense scrutiny (U.S. EPA-SAB, 2010). To minimize
uncertainties related to extrapolation and double counting, we only include a subset of
monetized visibility benefits in the primary benefits estimate to correspond with our higher
level of confidence in recreational benefits within the study regions and residential benefits
within the study cities. Although we are confident that visibility benefits extend beyond these
studied areas, we are less confident about the magnitude of those benefits.
In the approach described here, we have revised several aspects of the visibility benefits
analysis since previous RIAs, including light extinction estimation methods, visitation data for
Class I areas (used in extrapolating benefits), valuation studies for residential visibility benefits,
and the benefit transfer technique for residential benefits. Including residential visibility
benefits in the primary benefits estimates reflects an evolution in our understanding of the
nature and importance of the effect on public welfare from visibility impairment to a more
multifaceted approach that includes non-Class I areas, such as urban areas. This evolution has
occurred in conjunction with the expansion of available PM data and information from
associated studies of public perception, valuation and personal comfort and well-being. While
visibility preference studies (Abt Associates Inc., 2001, Ely et al., 1991, Pryor, 1996, BBC
Research & Consulting, 2003) also provide support for a non-zero benefits estimate, these
surveys did not include questions that would enable monetization of those preferences.
Despite these improvements, we are limited by the available peer-reviewed studies on
visibility benefits, which have not undergone a similar expansion as the health literature. In
fact, to our knowledge, no peer-reviewed studies have been published in the past 10 years on
visibility valuation that we are able to incorporate into this analysis.33 When EPA's Scientific
Advisory Council reviewed the visibility benefits analysis for the Second Prospective 812
analysis, they also lamented on the need for additional research to improve methods and
estimates (U.S. EPA-SAB, 2010). Because of time and resource constraints, performing original
33 While several studies using hedonic valuation techniques to value air quality have been published in the last 10
years (Kim et al., 2010; Bayer et al., 2009; Anselin and Lozano-Gracia, 2006, 2008; Chay and Greenstone, 2005),
these studies do not provide any mechanism to distinguish visibility from health or ecosystem effects.
6-29
-------
research for regulatory analyses of specific policy actions is infeasible. Therefore, we actively
encourage and solicit new research to address many of the limitations in our analysis. Most
importantly, we are interested in recently published national-scale visibility valuation studies
that incorporate current CV best practices, as the existing studies are limited to specific subset
of geographic areas. Other important research questions that remain unresolved include
identifying factors that affect valuation preferences in order to facilitate benefits transfer from
the original studies to transfer sites across localities, disentangling health and ecosystem
valuation from visibility valuation, usefulness of preference calibration, and potential role of
hedonic valuation approaches. Many of these same research needs were identified by Cropper
(2000), but they have yet to be addressed by the research community.
For these reasons, EPA requests public comment on the approach taken here to
quantify the monetary value of changes in recreational and residential visibility. Specifically, we
request comment on additional valuation studies, methods for benefit transfer, and methods
for characterizing uncertainty.
6.5 Materials Damage Benefits
Building materials including metals, stones, cements, and paints undergo natural
weathering processes from exposure to environmental elements (e.g., wind, moisture,
temperature fluctuations, sunlight, etc.). Pollution can worsen and accelerate these effects.
Deposition of PM is associated with both physical damage (materials damage effects) and
impaired aesthetic qualities (soiling effects). Wet and dry deposition of PM can physically affect
materials, adding to the effects of natural weathering processes, by potentially promoting or
accelerating the corrosion of metals, by degrading paints and by deteriorating building
materials such as stone, concrete and marble (U.S. EPA, 2009b). The effects of PM are
exacerbated by the presence of acidic gases and can be additive or synergistic due to the
complex mixture of pollutants in the air and surface characteristics of the material. Acidic
deposition has been shown to have an effect on materials including zinc/galvanized steel and
other metal, carbonate stone (as monuments and building facings), and surface coatings
(paints) (Irving, 1991). The effects on historic buildings and outdoor works of art are of
particular concern because of the uniqueness and irreplaceability of many of these objects.
The PM ISA concludes that evidence is sufficient to support a causal relationship
between PM and effects on materials (U.S. EPA, 2009b). Considerable research has been
conducted on the effects of air pollutants on metal surfaces due to the economic importance of
these materials, especially steel, zinc, aluminum, and copper. Moisture is the single greatest
6-30
-------
factor promoting metal corrosion; however, deposited PM can have additive, antagonistic or
synergistic effects. In general, S02 is more corrosive than NOx although mixtures of NOx, S02
and other particulate matter corrode some metals at a faster rate than either pollutant alone
(U.S. EPA, 2008). Metal structures are usually coated by alkaline corrosion product layers and
thus are subject to increased corrosion by acidic deposition. In addition, research has
demonstrated that iron, copper, and aluminum-based products are subject to increased
corrosion due to pollution (Irving, 1991). Information from both the PM ISA (U.S. EPA, 2009b)
and NOx/SOx ISA (U.S. EPA, 2008) suggest that the extent of damage to metals due to ambient
PM is variable and dependent upon the type of metal, prevailing environmental conditions, rate
of natural weathering and presence or absence of other pollutants
In addition, the deposition of PM can cause soiling, which is the accumulation of dirt,
dust, and ash on exposed surfaces such as metal, glass, stone and paint. Particles consisting
primarily of carbonaceous compounds can cause soiling of commonly used building materials
and culturally important items such as statues and works of art. Soiling occurs when PM
accumulates on an object and alters the optical characteristics (appearance). The reflectivity of
a surface may be changed or presence of particulates may alter light transmission. These
effects can reduce the aesthetic value of a structure or result in reversible or irreversible
damage to statues, artwork and architecturally or culturally significant buildings. Due to soiling
of building surfaces by PM, the frequency and duration of cleaning or repainting may be
increased. In addition to natural factors, exposure to PM may give painted surfaces a dirty
appearance. Pigments in works of art can be degraded or discolored by atmospheric pollutants,
especially sulfates (U.S. EPA, 2008). Previous assessments estimated household soiling benefits
based on the Manuel et al. (1982) study of consumer expenditures on cleaning and household
maintenance. However, the data used to estimate household soiling damages in the Manuel et
al. study is from a 1972 consumer expenditure survey and as such may not accurately represent
consumer preferences in the future. In light of this significant limitation, we believe that this
study cannot provide reliable estimates of the likely magnitude of the benefits of reduced PM
household soiling.
In order to estimate the monetized benefits associated with reducing materials damage
and household soiling, quantitative relationships are needed between particle size,
concentration, chemical concentrations and frequency of maintenance and repair. Such an
analysis would require three steps:
1. Develop a national inventory of sensitive materials;
6-31
-------
2. Derive concentration-response functions that relate material damage to change in
pollution concentration or deposition; and,
3. Estimate the value of lost materials and/or repair of damage.
Due to data limitations and uncertainties inherent in each of these steps, we have
chosen not to include a monetized estimate of materials damage and household soiling in this
analysis. The PM ISA concluded that there is considerable uncertainty with regard to interaction
of co-pollutants in regards to materials damage and soiling processes (U.S. EPA, 2009b).
Previous EPA benefits analyses have provided quantitative estimates of materials damage (U.S.
EPA, 2011b) and household soiling damage (U.S. EPA, 1999). Consistent with SAB advice (U.S.
EPA, 1998), we determined that the existing data are not sufficient to calculate a reliable
estimate of future year household soiling damages (U.S. EPA, 1998). These previous analyses
have shown that materials damage benefits are significantly smaller than the health benefits
associated with reduced exposure to PM2.s and ozone, or even visibility benefits. However,
studies of materials damage to historic buildings and outdoor artwork in Sweden (Grosclaude
and Soguel, 1994) indicate that these benefits could be an order of magnitude larger than
household soiling benefits.
In the absence of quantified benefits, we provide a qualitative description of the
avoided damage associated with reducing PM and PM precursor pollutants. Table 6-8 shows
the effect of various PM2.s precursor pollutants and other co-pollutants on various materials.
6-32
-------
Table 6-8. Materials Damaged by Pollutants Affected by this Rule (U.S. EPA, 2011b)
Pollutant
Hydrogen ion and
nitrogen oxides
Carbon dioxide
Unquantified Effects/ Damage to:
Sulfur oxides Infrastructural materials—galvanized and painted carbon steel
Commercial buildings—carbonate stone, metal, and painted wood surfaces
Residential buildings—carbonate stone, metal, and painted wood surfaces
Monuments—carbonate stone and metal
Structural aesthetics
Automotive finishes—painted metal
Infrastructural materials—galvanized and painted carbon steel
Zinc-based metal products, such as galvanized steel
Commercial and residential buildings—carbonate stone, metal, and wood surfaces
Monuments—carbonate stone and metal
Structural aesthetics
Automotive finishes—painted metal
Zinc-based metal products, such as galvanized steel
Formaldehyde
Zinc-based metal products, such as galvanized steel
Particulate matter
Household cleanliness (i.e., household soiling)
Ozone
Rubber products (e.g., tires)
6.6 Climate Benefits
Actions taken by state and local governments to implement the proposed PM2.5
standards are likely to have implications for climate change because emission controls
ultimately implemented to meet the standard may have impacts on emissions of long-lived
greenhouse gas (GHG) such as carbon dioxide (C02), short-lived climate forcers such as black
carbon (BC), and cooling aerosols like organic carbon (OC). Our ability to quantify the climate
effects of these proposed standards is quite limited due to lack of available information on the
co-controlled GHG emission reductions, the energy and associated climate gas implications of
control technologies assumed in the illustrative regulatory alternatives, and remaining
uncertainties regarding the impact of long-lived and short-lived climate forcer impacts on
climate change. For this RIA, we discuss qualitatively the implications of potential emission
reductions in warming and cooling aerosols and changes in long-lived GHG emissions such as
C02 for the regulatory alternatives. Implementation strategies undertaken by state and local
governments to comply with the standards may differ from the illustrative control strategies in
this RIA. It is important to note that the net climate forcing depends on the specific
6-33
-------
combinations of emission reductions chosen to meet the proposed standards because of the
differences in warming and cooling potential of the difference pollutants.
6.6.1 Climate Effects of Short Lived Climate Forcers
Pollutants that affect the energy balance of the earth are referred to as climate forcers.
A pollutant that increases the amount of energy in the Earth's climate system is said to exert
"positive radiative forcing," which leads to warming and climate change. In contrast, a pollutant
that exerts negative radiative forcing reduces the amount of energy in the Earth's system and
leads to cooling.
Long-lived gases such as C02 differ from short-lived pollutants such as BC in the length
of time they remain in the atmosphere affecting the earth's energy balance. Long-lived gases
remain in the atmosphere for hundreds to thousands of years. Short-lived climate forcers
(SLCFs), in contrast, remain in the atmosphere for short periods of time ranging from days to
weeks. The potential to affect near-term climate change and the rate of climate change with
policies to address these emissions is gaining attention nationally and internationally (e.g., Black
Carbon Report to Congress (currently undergoing peer review), Arctic Council Task Force,
Global Methane Initiative, and Convention on Long-Range Trans-boundary Air Pollution of the
United Nations Economic Commission for Europe). A recent United Nations Environmental
Programme (UNEP) study provides the most comprehensive analysis to date of the benefits of
measures to reduce SLCFs including methane, ozone, and black carbon assessing the health,
climate, and agricultural benefits of a suite of mitigation technologies. The report concludes
that the climate is changing now, and these changes have the potential to "trigger abrupt
transitions such as the release of carbon from thawing permafrost and biodiversity loss." While
reducing long-lived GHGs such as C02 is necessary to protect against long-term climate change,
reducing SLCF gases including BC and ozone is beneficial and will slow the rate of climate
change within the first half of this century (UNEP, 2011).
6.6.1.1 Climate Effects of Black Carbon
Black carbon is the most strongly light-absorbing component of PM2.5, and is formed by
incomplete combustion of fossil fuels, biofuels, and biomass. The short atmospheric lifetime of
BC lasting from days to weeks and the mechanisms by which BC affects climate distinguish it
from long-lived GHGs like C02. This means that actions taken to reduce the BC constituents in
direct PM2.5 will have almost immediate effects on climate change. Emissions sources and
ambient concentrations of BC vary geographically and temporally resulting in climate effects
that are more regionally and seasonally dependent than the effects of long-lived, well-mixed
6-34
-------
GHGs. Likewise, mitigation actions for BC will produce different climate impacts depending on
the region, season, and emission source category affected.
BC influences climate in multiple ways: directly, indirectly, and through snow and ice
albedo. Specifically, BC affects climate directly by absorbing both incoming and outgoing
radiation of all wavelengths. In contrast, GHGs mainly trap outgoing infrared radiation from the
earth's surface. Per unit of mass in the atmosphere, BC can absorb a million times more energy
than C02(Bond and Sun 2005). This strong absorptive capacity is the property most relevant to
its potential to affect the Earth's climate. BC also affects climate indirectly by altering the
properties of clouds, affecting cloud reflectivity, precipitation, and surface dimming. These
indirect impacts of BC are associated with all ambient particles and may lead to cooling, but are
not associated with long-lived well mixed GHGs. Finally, when BC is deposited on snow and ice,
it darkens the surface and decreases reflectivity, thereby increasing absorption and accelerating
melting.
The illustrative control strategies evaluated for this proposal include reductions in BC
emissions that will tend to have a beneficial cooling effect on the atmosphere. BC and
elemental carbon (EC) (or particulate elemental carbon (PEC)) are used interchangeably in this
report because EPA traditionally estimates EC emissions rather than BC and for the purpose of
this analysis these measures are essentially equivalent. Emissions reductions discussed below
are from the modeled scenarios for each regulatory alterative, and not from the full attainment
scenarios. This is because speciated PM2.s data were not available for emissions reductions
beyond known controls.
The snow/ice albedo effects from BC deposition have been linked to accelerated snow
and ice melting (Wiscombe and Warren, 1980). While many glaciers around the world and
Arctic sea ice have receded in recent decades, determining whether this phenomenon is
attributable to BC is challenging due to other contributing factors. Emissions north of the 40th
parallel latitude are thought to be particularly important for BC's climate related effects in the
Arctic (Shindell, 2007; Ramanathan and Carmichael, 2008).
Snow and ice cover in the Western U.S. has also been affected by BC. Specifically,
deposition of BC on mountain glaciers and snow packs produces a positive snow and ice albedo
effect, contributing to the melting of snowpack earlier in the spring and reducing the amount of
snowmelt that normally would occur later in the spring and summer (Hadley et al. 2010). This
has implications for freshwater resources in regions of the U.S. dependent on snow-fed or
glacier-fed water systems. In the Sierra Nevada mountain range, Hadley et al. (2010) found BC
6-35
-------
at different depths in the snowpack, deposited over the winter months by snowfall. In the
spring, the continuous uncovering of the BC contributed to the early melt. A model capturing
the effects of soot on snow in the western U.S. shows significant decreases in snowpack
between December and May (Qian et al., 2009). Snow water equivalent (the amount of water
that would be produced by melting all the snow) is reduced 2-50 millimeters (mm) in
mountainous areas, particularly over the Central Rockies, Sierra Nevadas, and western Canada.
A study found that biomass burning emissions in Alaska and the Rocky Mountain region during
the summer can enhance snowmelt. Dust deposition on snow, at high concentrations, can have
similar effects to BC (Koch et al., 2007). Similarly, a study done by Painter et al. (2007) in the
San Juan Mountains in Colorado indicated a decrease in snow cover duration of 18-35 days as a
result of dust transported from non-local desert sources. National elemental carbon and
organic carbon deposition maps are included in Appendix 6B to this report.
6.6.1.2 Climate Effects of Nitrates, Sulfate, and Organic Carbon (excluding BC)
The composition of the total emissions mixture is also relevant as to whether emissions
are warming or cooling to the atmosphere. Pollutants such as S02, NOX, and most OC particles
tend to produce a cooling influence on climate. Exceptions include OC deposition on snow and
ice, which leads to increased melting.
In addition, it is important to account for the indirect effects of all PM constituents on
climate: all aerosols (including BC) affect climate indirectly by changing the reflectivity and
lifetime of clouds. The net indirect effect of all aerosols is very uncertain but is thought to be a
net cooling influence.
6.6.1.3 Climate Effects of Ozone
Ozone changes due to this proposed regulation are not estimated for this analysis but
may occur due to the NOx reductions estimated. Ozone is a well-known SLCF (U.S. EPA, 2006).
Stratospheric ozone (the upper ozone layer) is beneficial because it protects life on Earth from
the sun's harmful ultraviolet (UV) radiation. In contrast, tropospheric ozone (ozone in the lower
atmosphere) is a harmful air pollutant that adversely affects human health and the
environment and contributes significantly to regional and global climate change. Due to its
short atmospheric lifetime, tropospheric ozone concentrations exhibit large spatial and
temporal variability (U.S. EPA, 2009). The discernable influence of ground level ozone on
climate leads to increases in global surface temperature and changes in hydrological cycles.
While reducing long-lived GHGs such as C02 is necessary to protect against long-term climate
6-36
-------
change, reducing SLCF gases including ozone is beneficial and will slow the rate of climate
change within the first half of this century (UNEP, 2011).
6.6.1.4 SLCFs Summary and Conclusions
Assessing the net climate impact of SLCFs for the illustrative emission control strategies
is outside the scope of this regulatory analysis and requires climate atmospheric modeling not
undertaken due to time and resource constraints. Information about the amount of BC relative
to non-BC constituents emitted from a source is important. In general, these non-BC
constituents are emitted in greater volume than BC, counteracting the warming influence of BC.
Qualitatively, it seems likely that BC emission reductions associated with direct emitted PM2.5
controls will be beneficial for the climate in terms of reduced radiative forcing and deposition
on snow and ice. Reductions in OC, sulfates and nitrates are likely to produce warming in the
atmosphere. The indirect impacts of aerosols on clouds and precipitation remain the subject of
great uncertainty making it more difficult to estimate the quantitative impact of aerosol
reductions on climate.
6.6.2 Climate Effects of Long-Lived Greenhouse Gases
The importance of mitigating long-lived climate gases such as C02 has been stressed by
the Intergovernmental Panel on Climate Change (IPCC 2007). While addressing short-lived
climate forcers may result in more immediate climate benefits in specific areas, long-term
policies must deal with long lived GHGs to address long-term climate change. We are unable to
quantify the impact of the illustrative control strategies for this rulemaking on long-lived
climate gases due lack of available data. However, State and Local governments may want to
consider human health, welfare, and climate implications of regulatory strategies undertaken
to implement the promulgated PM standards.
6.7 Ecosystem Benefits and Services
The effects of air pollution on the health and stability of ecosystems are potentially very
important. At present, it is difficult to measure the impact of reducing air pollution in a national
scale analysis across different types of ecosystems and different pollutant effects. Previous EPA
science assessments (U.S. EPA, 2006a; U.S. EPA, 2008c; U.S. EPA, 2009b) have determined that
air pollution can be directly linked to aquatic and terrestrial acidification, nutrient enrichment,
vegetation injury, and metal bioaccumulation in animals. Ecosystem services are a useful
conceptual framework for analyzing the impact of ecosystem changes on public welfare.
6-37
-------
Ecosystem services can be generally defined as the benefits that individuals and
organizations obtain from ecosystems. EPA has defined ecological goods and services as the
"outputs of ecological functions or processes that directly or indirectly contribute to social
welfare or have the potential to do so in the future. Some outputs may be bought and sold, but
most are not marketed" (U.S. EPA, 2006c). Figure 6-7 provides the Millennium Ecosystem
Assessment's schematic demonstrating the connections between the categories of ecosystem
services and human well-being. The interrelatedness of these categories means that any one
ecosystem may provide multiple services. Changes in these services can affect human well-
being by affecting security, health, social relationships, and access to basic material goods
(MEA, 2005).
CONSTITUENTS OF WELL-BEING
ECOSrSlEM SERVICES
rtOVttton Itxj
rvc9t •»..•
Supporting
MUTUFKT CHEUNG
>
Cultural
.
EMLftTONM.
UFE ON EAKTH - WOOVCBSITV
Security
PBWONM-5*ITTT
lor Qood lite
»Bf j HIT Ln
sumeewr wurrmoue roea
•W.TB*
- -
. .
iMumooiMa
W10!
-------
4. Supporting: Services necessary for the production of all other ecosystem services,
such as nutrient cycles and crop pollination
The monetization of ecosystem services generally involves estimating the value of
ecological goods and services based on what people are willing to pay (WTP) to increase
ecological services or by what people are willing to accept (WTA) in compensation for
reductions in them (U.S. EPA, 2006c). There are three primary approaches for estimating the
monetary value of ecosystem services: market-based approaches, revealed preference
methods, and stated preference methods (U.S. EPA, 2006c). Because economic valuation of
ecosystem services can be difficult, nonmonetary valuation using biophysical measurements
and concepts also can be used. An example of a nonmonetary valuation method is the use of
relative-value indicators (e.g., a flow chart indicating uses of a water body, such as beatable,
fishable, swimmable, etc.). It is necessary to recognize that in the analysis of the environmental
responses associated with any particular policy or environmental management action, only a
subset of the ecosystem services likely to be affected are readily identified. Of those ecosystem
services that are identified, only a subset of the changes can be quantified. Within those
services whose changes can be quantified, only a few will likely be monetized, and many will
remain nonmonetized. The stepwise concept leading up to the valuation of ecosystems services
is graphically depicted in Figure 6-8.
Flanrtnti and u tbleni fomnuleflon
Bonds tfd
urn*!,;(.,;!
Figure 6-8. Schematic of the Benefits Assessment Process (U.S. EPA, 2006c)
6-39
-------
6.7.1 Ecosystem Benefits for Metallic and Organic Constituents of PM
Several significant ecological effects are associated with deposition of chemical
constituents of ambient PM such as metals and organics (U.S. EPA, 2009b). The trace metal
constituents of PM include cadmium, copper, chromium, mercury, nickel, zinc, and lead. The
organics include persistent organic pollutants (POPs), polyaromatic hydrocarbons (PAHs) and
polybromiated diphenyl ethers (PBDEs). Exposure to PM for direct effects occur via deposition
(e.g., wet, dry or occult) to vegetation surfaces, while indirect effects occur via deposition to
ecosystem soils or surface waters where the deposited constituents of PM then interacts with
biological organisms. While both fine and coarse-mode particles may affect plants and other
organisms, more often the chemical constituents drive the ecosystem response to PM (Grantz
et al., 2003). Ecological effects of PM include direct effects to metabolic processes of plant
foliage; contribution to total metal loading resulting in alteration of soil biogeochemistry and
microbiology, plant and animal growth and reproduction; and contribution to total organics
loading resulting in bioaccumulation and biomagnification across trophic levels.
The PM ISA concludes that a causal relationship is likely to exist between deposition of
PM and a variety of effects on individual organisms and ecosystems (U.S. EPA 2009b). Most
direct ecosystem effects associated with particulate pollution occur in severely polluted areas
near industrial point sources (quarries, cement kilns, metal smelting) (U.S. EPA, 2009b).
However the PM ISA also finds, in many cases, it is difficult to characterize the nature and
magnitude of effects and to quantify relationships between ambient concentrations of PM and
ecosystem response due to significant data gaps and uncertainties as well as considerable
variability that exists in the components of PM and their various ecological effects (U.S. EPA,
2009b).
Particulate matter can adversely impact plants and ecosystem services provided by
plants by deposition to vegetative surfaces (U.S. EPA, 2009b). Particulates deposited on the
surfaces of leaves and needles can block light, altering the radiation received by the plant. PM
deposition near sources of heavy deposition can obstruct stomata limiting gas exchange,
damage leaf cuticles and increase plant temperatures (U.S. EPA, 2009b). Plants growing on
roadsides exhibit impact damage from near-road PM deposition, having higher levels of
organics and heavy metals, and accumulate salt from road de-icing during winter months (U.S.
EPA, 2009b). In addition, atmospheric PM can convert direct solar radiation to diffuse radiation,
which is more uniformly distributed in a tree canopy, allowing radiation to reach lower leaves
(U.S. EPA, 2009b). Decreases in crop yields (a provisioning service) due to reductions in solar
6-40
-------
radiation have been attributed to regional scale air pollution in other counties with especially
severe regional haze (Chameides et al., 1999).
In addition to damage to plant surfaces, deposited PM can be taken up by plants from
soil or foliage. Copper, zinc, and nickel have been shown to be directly toxic to vegetation under
field conditions (U.S. EPA, 2009b). The ability of vegetation to take up heavy metals is
dependent upon the amount, solubility and chemical composition of the deposited PM. Uptake
of PM by plants from soils and vegetative surfaces can disrupt photosynthesis, alter pigments
and mineral content, reduce plant vigor, decrease frost hardiness and impair root development.
Particulate mattercan also contain organic air toxic pollutants, including PAHs, which
are a class of polycyclic organic matter (POM). PAHs can accumulate in sediments and
bioaccumulate in freshwater, flora and fauna. The uptake of organics depends on the plant
species, site of deposition, physical and chemical properties of the organic compound and
prevailing environmental conditions (U.S. EPA, 2009b). Different species can have different
uptake rates of PAHs. For example, zucchini (Cucurbita pepo) accumulated significantly more
PAHs than related plant species (Parrish et al., 2006). PAHs can accumulate to high enough
concentrations in some coastal environments to pose an environmental health threat that
includes cancer in fish populations, toxicity to organisms living in the sediment and risks to
those (e.g., migratory birds) that consume these organisms (Simcik et al., 1996; Simcik et al.,
1999). Atmospheric deposition of particles is thought to be the major source of PAHs to the
sediments of Lake Michigan, Chesapeake Bay, Tampa Bay and other coastal areas of the U.S.
(Arzavus, Dickhut, and Canuel, 2001).
Contamination of plant leaves by heavy metals can lead to elevated concentrations in
the soil. Trace metals absorbed into the plant, frequently bind to the leaf tissue, and then are
lost when the leaf drops. As the fallen leaves decompose, the heavy metals are transferred into
the soil (Cotrufo et al., 1995; Niklinska et al., 1998). Many of the major indirect plant responses
to PM deposition are chiefly soil-mediated and depend on the chemical composition of
individual components of deposited PM. Upon entering the soil environment, PM pollutants
can alter ecological processes of energy flow and nutrient cycling, inhibit nutrient uptake to
plants, change microbial community structure and, affect biodiversity. Accumulation of heavy
metals in soils depends on factors such as local soil characteristics, geologic origin of parent
soils, and metal bioavailability. Heavy metals, such as zinc, copper, and cadmium, and some
pesticides can interfere with microorganisms that are responsible for decomposition of soil
litter, an important regulating ecosystem service that serves as a source of soil nutrients (U.S.
EPA, 2009b). Surface litter decomposition is reduced in soils having high metal concentrations.
6-41
-------
Soil communities have associated bacteria, fungi, and invertebrates that are essential to soil
nutrient cycling processes. Changes to the relative species abundance and community
composition are associated with deposited PM to soil biota (U.S. EPA, 2009b).
Atmospheric deposition can be the primary source of some organics and metals to
watersheds. Deposition of PM to surfaces in urban settings increases the metal and organic
component of storm water runoff (U.S. EPA, 2009b). This atmospherically-associated pollutant
burden can then be toxic to aquatic biota. The contribution of atmospherically deposited PAHs
to aquatic food webs was demonstrated in high elevation mountain lakes with no other
anthropogenic contaminant sources (U.S. EPA, 2009b). Metals associated with PM deposition
limit phytoplankton growth, affecting aquatic trophic structure. Long-range atmospheric
transport of 47 pesticides and degradation products to the snowpack in seven national parks in
the Western U.S. was recently quantified indicating PM-associated contaminant inputs to
receiving waters during spring snowmelt (Hageman et al., 2006).
The recently completed Western Airborne Contaminants Assessment Project (WACAP) is
the most comprehensive database on contaminant transport and PM depositional effects on
sensitive ecosystems in the Western U.S. (Landers et al., 2008). In this project, the transport,
fate, and ecological impacts of anthropogenic contaminants from atmospheric sources were
assessed from 2002 to 2007 in seven ecosystem components (air, snow, water, sediment,
lichen, conifer needles and fish) in eight core national parks. The study concluded that
bioaccumulation of semi-volatile organic compounds occurred throughout park ecosystems, an
elevational gradient in PM deposition exists with greater accumulation in higher altitude areas,
and contaminants accumulate in proximity to individual agriculture and industry sources, which
is counter to the original working hypothesis that most of the contaminants would originate
from Eastern Europe and Asia.
Although there is considerable data on impacts of PM on ecological receptors, few
studies link ambient PM levels to observed effect. This is due, in part, to the nature, deposition,
transport and fate of PM in ecosystems. Some of the difficulties in quantifying the ecosystem
benefits associated with reduced PM deposition include the following:
• PM is not a single pollutant, but a heterogeneous mixture of particles differing in
size, origin and chemical composition. Since vegetation and other ecosystem
components are affected more by particulate chemistry than size fraction, exposure
to a given mass concentration of airborne PM may lead to widely differing plant or
ecosystem responses, depending on the particular mix of deposited particles.
6-42
-------
• Composition of ambient PM varies in time and space and the particulate mixture
may have synergistic, antagonistic or additive effects on ecological receptors
depending upon the chemical species present.
• Presence of co-pollutants makes it difficult to attribute observed effects to
ecological receptors to PM alone or one component of deposited PM.
• Ecosystem effects linked to PM are difficult to determine because the changes may
not be observed until pollutant deposition has occurred for many decades.
Furthermore, many PM components bioaccumulate over time in organisms or
plants, making correlations to ambient levels of PM difficult.
• Multiple ecological stressors can confound attempts to link specific ecosystem
responses to PM deposition. These stressors can be anthropogenic (e.g., habitat
destruction, eutrophication, other pollutants) or natural (e.g., drought, fire, disease).
Deposited PM interacts with other stressors to affect ecosystem patterns and
processes.
• Each ecosystem has a unique topography, underlying bedrock, soils, climate,
meteorology, hydrologic regime, natural and land use history, and species
composition. Sensitivity of ecosystem response can be highly variable in space and
time. Because of this variety and lack of data for most ecosystems, extrapolating
these effects from one ecosystem to another is highly uncertain.
6.7.2 Ecosystem Benefits from Reductions in Nitrogen and Sulfur Emissions
Emissions of the PM precursors, such as nitrogen and sulfur oxides occur over large
regions of North America. Once these pollutants are lofted to the middle and upper
troposphere, they typically have a much longer lifetime and, with the generally stronger winds
at these altitudes, can be transported long distances from their source regions. The length scale
of this transport is highly variable owing to differing chemical and meteorological conditions
encountered along the transport path (U.S. EPA, 2008c). Secondary particles are formed from
NOX and S02 gaseous emissions and associated chemical reactions in the atmosphere.
Deposition can occur in either a wet (i.e., rain, snow, sleet, hail, clouds, or fog) or dry form (i.e.,
gases or particles). Together these emissions are deposited onto terrestrial and aquatic
ecosystems across the U.S., contributing to the problems of acidification, nutrient enrichment,
and methylmercury production as represented in Figures 6-9 and 6-10. Although there is some
evidence that nitrogen deposition may have positive effects on agricultural and forest output
through passive fertilization, it is likely that the overall value is very small relative to other
health and welfare effects. In addition to deposition effects, S02 can affect vegetation at
ambient levels near pollution sources.
6-43
-------
Atmospheric
Fate and
Transport
Deposition
Process
Acidification
Nutrient
Enrichment
_L
Aquatic
Ecosystems
Terrestrial
Ecosystems
Aquatic
Ecosystems
Terrestrial
Ecosystems
Atmospheric
Fate and
Transport
Deposition
Process
Acidification
-
Methylmercury
Production
Aquatic
Ecosystems
Terrestrial
Ecosystems
Figure 6-9. Schematics of Ecological Effects of Nitrogen and Sulfur Deposition
6-44
-------
Figure 6-10. Nitrogen and Sulfur Cycling, and Interactions in the Environment
Source: U.S. EPA, 2008c
The atmospheric lifetimes of particles vary with particle size. Accumulation-mode
particles such as sulfates are kept in suspension by normal air motions and have a lower
deposition velocity than coarse-mode particles; they can be transported thousands of
kilometers and remain in the atmosphere for a number of days. They are removed from the
atmosphere primarily by cloud processes. Particulates affect acid deposition by serving as cloud
condensation nuclei and contribute directly to the acidification of rain. In addition, the gas-
phase species that lead to the dry deposition of acidity are also precursors of particles.
Therefore, reductions in NOX and S02 emissions will decrease both acid deposition and PM
concentrations, but not necessarily in a linear fashion (U.S. EPA, 2008c). Sulfuric acid is also
deposited on surfaces by dry deposition and can contribute to environmental effects (U.S. EPA,
2008c).
6.7.2.1 Ecological Effects of Acidification
Deposition of nitrogen and sulfur can cause acidification, which alters biogeochemistry
and affects animal and plant life in terrestrial and aquatic ecosystems across the U.S. Soil
acidification is a natural process, but is often accelerated by acidifying deposition, which can
decrease concentrations of exchangeable base cations in soils (U.S. EPA, 2008c). Major
6-45
-------
terrestrial effects include a decline in sensitive tree species, such as red spruce (Picea rubens)
and sugar maple (Acer saccharum) (U.S. EPA, 2008c). Biological effects of acidification in
terrestrial ecosystems are generally linked to aluminum toxicity and decreased ability of plant
roots to take up base cations (U.S. EPA, 2008c). Decreases in the acid neutralizing capacity and
increases in inorganic aluminum concentration contribute to declines in zooplankton, macro
invertebrates, and fish species richness in aquatic ecosystems (U.S. EPA, 2008c).
Geology (particularly surficial geology) is the principal factor governing the sensitivity of
terrestrial and aquatic ecosystems to acidification from nitrogen and sulfur deposition (U.S.
EPA, 2008c). Geologic formations having low base cation supply generally underlie the
watersheds of acid-sensitive lakes and streams. Other factors contribute to the sensitivity of
soils and surface waters to acidifying deposition, including topography, soil chemistry, land use,
and hydrologic flow path (U.S. EPA, 2008c).
Aquatic Acidification. Aquatic effects of acidification have been well studied in the U.S.
and elsewhere at various trophic levels. These studies indicate that aquatic biota have been
affected by acidification at virtually all levels of the food web in acid sensitive aquatic
ecosystems. The ISA for NOx/SOx—Ecological Criteria concluded that the evidence is sufficient
to infer a causal relationship between acidifying deposition and effects on biogeochemistry
related to aquatic ecosystems and biota in aquatic ecosystems (U.S. EPA, 2008c). Effects have
been most clearly documented for fish, aquatic insects, other invertebrates, and algae.
Biological effects are primarily attributable to a combination of low pH and high inorganic
aluminum concentrations. Such conditions occur more frequently during rainfall and snowmelt
that cause high flows of water and less commonly during low-flow conditions, except where
chronic acidity conditions are severe. Biological effects of episodes include reduced fish
condition factor34, changes in species composition and declines in aquatic species richness
across multiple taxa, ecosystems and regions. These conditions may also result in direct fish
mortality (Van Sickle et al., 1996). Biological effects in aquatic ecosystems can be divided into
two major categories: effects on health, vigor, and reproductive success; and effects on
biodiversity. Surface water with ANC values greater than 50 u.eq/L generally provides moderate
protection for most fish (i.e., brook trout, others) and other aquatic organisms (U.S. EPA,
2009c). Table 6-9 provides a summary of the biological effects experienced at various ANC
levels.
34Condition factor is an index that describes the relationship between fish weight and length, and is one measure
of sublethal acidification stress that has been used to quantify effects of acidification on an individual fish (U.S.
EPA, 2008f).
6-46
-------
Table 6-9. Aquatic Status Categories
Category Label ANC Levels Expected Ecological Effects
Acute
Concern
Severe
Concern
Elevated
Concern
Moderate
Concern
Low Concern
<0 micro
equivalent per
Liter (u.eq/L)
0-20 u.eq/L
20-50 u.eq/L
50-100 u.eq/L
>100 u.eq/L
Near complete loss of fish populations is expected. Planktonic communities have
extremely low diversity and are dominated by acidophilic forms. The number of
individuals in plankton species that are present is greatly reduced.
Highly sensitive to episodic acidification. During episodes of high acidifying
deposition, brook trout populations may experience lethal effects. Diversity and
distribution of zooplankton communities decline sharply.
Fish species richness is greatly reduced (i.e., more than half of expected species
can be missing). On average, brook trout populations experience sublethal
effects, including loss of health, reproduction capacity, and fitness. Diversity and
distribution of zooplankton communities decline.
Fish species richness begins to decline (i.e., sensitive species are lost from lakes).
Brook trout populations are sensitive and variable, with possible sublethal
effects. Diversity and distribution of zooplankton communities also begin to
decline as species that are sensitive to acidifying deposition are affected.
Fish species richness may be unaffected. Reproducing brook trout populations
are expected where habitat is suitable. Zooplankton communities are unaffected
and exhibit expected diversity and distribution.
A number of national and regional assessments have been conducted to estimate the
distribution and extent of surface water acidity in the U.S. (U.S. EPA, 2008c). As a result, several
regions of the U.S. have been identified as containing a large number of lakes and streams that
are seriously impacted by acidification. Figure 6-11 illustrates those areas of the U.S. where
aquatic ecosystems are at risk from acidification.
Because acidification primarily affects the diversity and abundance of aquatic biota, it
also affects the ecosystem services that are derived from the fish and other aquatic life found in
these surface waters.
While acidification is unlikely to have serious negative effects on, for example, water
supplies, it can limit the productivity of surface waters as a source of food (i.e., fish). In the
northeastern United States, the surface waters affected by acidification are not a major source
of commercially raised or caught fish; however, they are a source of food for some recreational
and subsistence fishermen and for other consumers. For example, there is evidence that certain
population subgroups in the northeastern United States, such as the Hmong and Chippewa
ethnic groups, have particularly high rates of self-caught fish consumption (Hutchison and Kraft,
1994; Peterson etal., 1994). However, it is not known if and how their consumption patterns
are affected by the reductions in available fish populations caused by surface water
acidification.
6-47
-------
Figure 6-11. Areas Potentially Sensitive to Aquatic Acidification
Source: U.S. EPA, 2008c
Inland surface waters support several cultural services, including aesthetic and
educational services and recreational fishing. Recreational fishing in lakes and streams is among
the most popular outdoor recreational activities in the northeastern United States. Based on
studies conducted in the northeastern United States, Kaval and Loomis (2003) estimated
average consumer surplus values per day of $36 for recreational fishing (in 2007 dollars);
therefore, the implied total annual value of freshwater fishing in the northeastern United States
was $5.1 billion in 2006.35 For recreation days, consumer surplus value is most commonly
measured using recreation demand, travel cost models.
Another estimate of the overarching ecological benefits associated with reducing lake
acidification levels in Adirondacks National Park can be derived from the contingent valuation
(CV) survey (Banzhaf et al., 2006), which elicited values for specific improvements in
acidification-related water quality and ecological conditions in Adirondack lakes. The survey
described a base version with minor improvements said to result from the program, and a
These estimates reflect the total value of the service, not the marginal change in the value of the service as a
result of the emission reductions achieved by this rule.
6-48
-------
scope version with large improvements due to the program and a gradually worsening status
quo. After adapting and transferring the results of this study and converting the 10-year annual
payments to permanent annual payments using discount rates of 3% and 5%, the WTP
estimates ranged from $48 to $107 per year per household (in 2004 dollars) for the base
version and $54 to $154 for the scope version. Using these estimates, the aggregate annual
benefits of eliminating all anthropogenic sources of NOX and SOX emissions were estimated to
range from $291 million to $829 million (U.S. EPA, 2009c).36
In addition, inland surface waters provide a number of regulating services associated
with hydrological and climate regulation by providing environments that sustain aquatic food
webs. These services are disrupted by the toxic effects of acidification on fish and other aquatic
life. Although it is difficult to quantify these services and how they are affected by acidification,
some of these services may be captured through measures of provisioning and cultural services.
Terrestrial Acidification. Acidifying deposition has altered major biogeochemical
processes in the U.S. by increasing the nitrogen and sulfur content of soils, accelerating nitrate
and sulfate leaching from soil to drainage waters, depleting base cations (especially calcium and
magnesium) from soils, and increasing the mobility of aluminum. Inorganic aluminum is toxic to
some tree roots. Plants affected by high levels of aluminum from the soil often have reduced
root growth, which restricts the ability of the plant to take up water and nutrients, especially
calcium (U. S. EPA, 2008c). These direct effects can, in turn, influence the response of these
plants to climatic stresses such as droughts and cold temperatures. They can also influence the
sensitivity of plants to other stresses, including insect pests and disease (Joslin et al., 1992)
leading to increased mortality of canopy trees. In the U.S., terrestrial effects of acidification are
best described for forested ecosystems (especially red spruce and sugar maple ecosystems)
with additional information on other plant communities, including shrubs and lichen (U.S. EPA,
2008c). The ISA for NOx/SOx—Ecological Criteria concluded that the evidence is sufficient to
infer a causal relationship between acidifying deposition and effects on biogeochemistry
related to terrestrial ecosystems and biota in terrestrial ecosystems (U.S. EPA, 2008c).
Certain ecosystems in the continental U.S. are potentially sensitive to terrestrial
acidification, which is the greatest concern regarding nitrogen and sulfur deposition U.S. EPA
(2008c). Figure 6-12 depicts the areas across the U.S. that are potentially sensitive to terrestrial
acidification.
36These estimates reflect the total value of the service, not the marginal change in the value of the service as a
result of the emission reductions achieved by this rule.
6-49
-------
Area of Higas! Potential Sensitivity
Top OuaitilB N
Top Quartile S
250 500 750 1,000
^m ^^^» I km
Figure 6-12. Areas Potentially Sensitive to Terrestrial Acidification
Source: U.S. EPA, 2008c
Both coniferous and deciduous forests throughout the eastern U.S. are experiencing
gradual losses of base cation nutrients from the soil due to accelerated leaching from acidifying
deposition. This change in nutrient availability may reduce the quality of forest nutrition over
the long term. Evidence suggests that red spruce and sugar maple in some areas in the eastern
U.S. have experienced declining health because of this deposition. For red spruce, (Picea
rubens) dieback or decline has been observed across high elevation landscapes of the
northeastern U.S., and to a lesser extent, the southeastern U.S., and acidifying deposition has
been implicated as a causal factor (DeHayes et al., 1999). Figure 6-13 shows the distribution of
red spruce (brown) and sugar maple (green) in the eastern U.S.
Terrestrial acidification affects several important ecological endpoints, including
declines in habitat for threatened and endangered species (cultural), declines in forest
aesthetics (cultural), declines in forest productivity (provisioning), and increases in forest soil
erosion and reductions in water retention (cultural and regulating).
6-50
-------
Figure 6-13. Distribution of Red Spruce (pink) and Sugar Maple (green) in the Eastern U.S.
Source: U.S. EPA, 2008c
Forests in the northeastern United States provide several important and valuable
provisioning services in the form of tree products. Sugar maples are a particularly important
commercial hardwood tree species, providing timber and maple syrup. In the United States,
sugar maple saw timber was nearly 900 million board feet in 2006 (USFS, 2006), and annual
production of maple syrup was nearly 1.4 million gallons, accounting for approximately 19% of
worldwide production. The total annual value of U.S. production in these years was
approximately $160 million (NASS, 2008).37 Red spruce is also used in a variety of products
including lumber, pulpwood, poles, plywood, and musical instruments. The total removal of red
spruce saw timber from timberland in the United States was over 300 million board feet in
2006 (USFS, 2006).
Forests in the northeastern United States are also an important source of cultural
ecosystem services—nonuse (i.e., existence value for threatened and endangered species),
These estimates reflect the total value of the service, not the marginal change in the value of the service as a
result of the emission reductions achieved by this rule.
6-51
-------
recreational, and aesthetic services. Red spruce forests are home to two federally listed species
and one delisted species:
1. Spruce-fir moss spider (Microhexura montivaga)—endangered
2. Rock gnome lichen (Gymnoderma lineare)—endangered
3. Virginia northern flying squirrel (Glaucomys sabrinusfuscus)—de\'\sled, but
important
Forestlands support a wide variety of outdoor recreational activities, including fishing,
hiking, camping, off-road driving, hunting, and wildlife viewing. Regional statistics on
recreational activities that are specifically forest based are not available; however, more
general data on outdoor recreation provide some insights into the overall level of recreational
services provided by forests. More than 30% of the U.S. adult population visited a wilderness or
primitive area during the previous year and engaged in day hiking (Cordell et al., 2008). From
1999 to 2004, 16% of adults in the northeastern United States participated in off-road vehicle
recreation, for an average of 27 days per year (Cordell et al., 2005). The average consumer
surplus value per day of off-road driving in the United States was $25 (in 2007 dollars), and the
implied total annual value of off-road driving recreation in the northeastern United States was
more than $9 billion (Kaval and Loomis, 2003). More than 5% of adults in the northeastern
United States participated in nearly 84 million hunting days (U.S. FWS and U.S. Census Bureau,
2007). Ten percent of adults in northeastern states participated in wildlife viewing away from
home on 122 million days in 2006. For these recreational activities in the northeastern United
States, Kaval and Loomis (2003) estimated average consumer surplus values per day of $52 for
hunting and $34 for wildlife viewing (in 2007 dollars). The implied total annual value of hunting
and wildlife viewing in the northeastern United States was, therefore, $4.4 billion and $4.2
billion, respectively, in 2006 (U.S. EPA, 2009c).38
As previously mentioned, it is difficult to estimate the portion of these recreational
services that are specifically attributable to forests and to the health of specific tree species.
However, one recreational activity that is directly dependent on forest conditions is fall color
viewing. Sugar maple trees, in particular, are known for their bright colors and are, therefore,
an essential aesthetic component of most fall color landscapes. A survey of residents in the
Great Lakes area found that roughly 30% of residents reported at least one trip in the previous
year involving fall color viewing (Spencer and Holecek, 2007). In a separate study conducted in
38These estimates reflect the total value of the service, not the marginal change in the value of the service as a
result of the emission reductions achieved by this rule.
6-52
-------
Vermont, Brown (2002) reported that more than 22% of households visiting Vermont in 2001
made the trip primarily for viewing fall colors.
Two studies estimated values for protecting high-elevation spruce forests in the
southern Appalachian Mountains. Kramer et al. (2003) conducted a contingent valuation study
estimating households' WTP for programs to protect remaining high-elevation spruce forests
from damages associated with air pollution and insect infestation. Median household WTP was
estimated to be roughly $29 (in 2007 dollars) for a smaller program, and $44 for the more
extensive program. Jenkins et al. (2002) conducted a very similar study in seven Southern
Appalachian states on a potential program to maintain forest conditions at status quo levels.
The overall mean annual WTP for the forest protection programs was $208 (in 2007 dollars).
Multiplying the average WTP estimate from these studies by the total number of households in
the seven-state Appalachian region results in an aggregate annual range of $470 million to $3.4
billion for avoiding a significant decline in the health of high-elevation spruce forests in the
Southern Appalachian region (U.S. EPA, 2009c).39
Forests in the northeastern United States also support and provide a wide variety of
valuable regulating services, including soil stabilization and erosion control, water regulation,
and climate regulation. The total value of these ecosystem services is very difficult to quantify
in a meaningful way, as is the reduction in the value of these services associated with total
nitrogen and sulfur deposition. As terrestrial acidification contributes to root damages, reduced
biomass growth, and tree mortality, all of these services are likely to be affected; however, the
magnitude of these impacts is currently very uncertain.
6.7.2.2 Ecological Effects from Nitrogen Enrichment
Aquatic Enrichment. The ISA for NOx/SOx—Ecological Criteria concluded that the
evidence is sufficient to infer a causal relationship between nitrogen deposition and the
alteration of species richness, species composition, and biodiversity in wetland, freshwater
aquatic and coastal marine ecosystems (U.S. EPA, 2008c).
One of the main adverse ecological effects resulting from nitrogen deposition,
particularly in the Mid-Atlantic region of the United States, is the effect associated with nutrient
enrichment in estuarine waters. A recent assessment of 141 estuaries nationwide by the
National Oceanic and Atmospheric Administration (NOAA) concluded that 19 estuaries (13%)
suffered from moderately high or high levels of eutrophication due to excessive inputs of both
39 These estimates reflect the marginal value of the service for the hypothetical program described in the survey,
not the marginal change in the value of the service as a result of the emission reductions achieved by this rule.
6-53
-------
N and phosphorus, and a majority of these estuaries are located in the coastal area from North
Carolina to Massachusetts (NOAA, 2007). For estuaries in the Mid-Atlantic region, the
contribution of atmospheric distribution to total N loads is estimated to range between 10%
and 58% (Valigura et al., 2001).
Eutrophication in estuaries is associated with a range of adverse ecological effects. The
conceptual framework developed by NOAA emphasizes four main types of eutrophication
effects—low dissolved oxygen (DO), harmful algal blooms (HABs), loss of submerged aquatic
vegetation (SAV), and low water clarity. Low DO disrupts aquatic habitats, causing stress to fish
and shellfish, which, in the short-term, can lead to episodic fish kills and, in the long-term, can
damage overall growth in fish and shellfish populations. Low DO also degrades the aesthetic
qualities of surface water. In addition to often being toxic to fish and shellfish, and leading to
fish kills and aesthetic impairments of estuaries, HABs can, in some instances, also be harmful
to human health. SAV provides critical habitat for many aquatic species in estuaries and, in
some instances, can also protect shorelines by reducing wave strength; therefore, declines in
SAV due to nutrient enrichment are an important source of concern. Low water clarity is in part
the result of accumulations of both algae and sediments in estuarine waters. In addition to
contributing to declines in SAV, high levels of turbidity also degrade the aesthetic qualities of
the estuarine environment.
Estuaries in the eastern United States are an important source of food production, in
particular fish and shellfish production. The estuaries are capable of supporting large stocks of
resident commercial species, and they serve as the breeding grounds and interim habitat for
several migratory species. To provide an indication of the magnitude of provisioning services
associated with coastal fisheries, from 2005 to 2007, the average value of total catch was $1.5
billion per year. It is not known, however, what percentage of this value is directly attributable
to or dependent upon the estuaries in these states.
In addition to affecting provisioning services through commercial fish harvests,
eutrophication in estuaries may also affect the demand for seafood. For example, a well-
publicized toxic pfiesteria bloom in the Maryland Eastern Shore in 1997, which involved
thousands of dead and lesioned fish, led to an estimated $56 million (in 2007 dollars) in lost
seafood sales for 360 seafood firms in Maryland in the months following the outbreak (Lipton,
1999).
Estuaries in the United States also provide an important and substantial variety of
cultural ecosystem services, including water-based recreational and aesthetic services. The
6-54
-------
water quality in the estuary directly affects the quality of these experiences. For example, there
were 26 million days of saltwater fishing coastal states from North Carolina to Massachusetts in
2006 (FWA and Census, 2007). Assuming an average consumer surplus value for a fishing day at
$36 (in 2007 dollars) in the Northeast and $87 in the Southeast (Kaval and Loomis, 2003), the
aggregate value was approximately $1.3 billion (in 2007 dollars) (U.S. EPA, 2009c).40 In addition,
almost 6 million adults participated in motorboating in coastal states from North Carolina to
Massachusetts, for a total of nearly 63 million days annually during 1999-2000 (Leeworthy and
Wiley, 2001). Using a national daily value estimate of $32 (in 2007 dollars) for motorboating
(Kaval and Loomis (2003), the aggregate value of these coastal motorboating outings was $2
billion per year (U.S. EPA, 2009c).41 Almost 7 million participated in birdwatching for 175 million
days per year, and more than 3 million participated in visits to non-beach coastal waterside
areas.
Estuaries and marshes have the potential to support a wide range of regulating services,
including climate, biological, and water regulation; pollution detoxification; erosion prevention;
and protection against natural hazards from declines in SAV (MEA, 2005). SAV can help reduce
wave energy levels and thus protect shorelines against excessive erosion, which increases the
risks of episodic flooding and associated damages to near-shore properties or public
infrastructure or even contribute to shoreline retreat.
We are unable to provide an estimate of the aquatic enrichment benefits associated
with the alternative standard level combinations due to data, time, and resource limitations.
Terrestrial Enrichment. Terrestrial enrichment occurs when terrestrial ecosystems
receive N loadings in excess of natural background levels, through either atmospheric
deposition or direct application. Evidence presented in the Integrated Science Assessment (U.S.
EPA, 2008c) supports a causal relationship between atmospheric N deposition and
biogeochemical cycling and fluxes of N and carbon in terrestrial systems. Furthermore,
evidence summarized in the report supports a causal link between atmospheric N deposition
and changes in the types and number of species and biodiversity in terrestrial systems.
Nitrogen enrichment occurs over a long time period; as a result, it may take as much as 50
years or more to see changes in ecosystem conditions and indicators. This long time scale also
affects the timing of the ecosystem service changes. The ISA for NOx/SOx—Ecological Criteria
40These estimates reflect the total value of the service, not the marginal change in the value of the service as a
result of the emission reductions achieved by this rule.
Fhese estimates reflect the total value of the service, not
result of the emission reductions achieved by this rule.
41 These estimates reflect the total value of the service, not the marginal change in the value of the service as a
6-55
-------
concluded that the evidence is sufficient to infer a causal relationship between nitrogen
deposition and the alteration of species richness, species composition, and biodiversity in
terrestrial ecosystems (U.S. EPA, 2008c).
One of the main provisioning services potentially affected by N deposition is grazing
opportunities offered by grasslands for livestock production in the Central U.S. Although N
deposition on these grasslands can offer supplementary nutritive value and promote overall
grass production, there are concerns that fertilization may favor invasive grasses and shift the
species composition away from native grasses. This process may ultimately reduce the
productivity of grasslands for livestock production. Losses due to invasive grasses can be
significant; for example, based on a bioeconomic model of cattle grazing in the upper Great
Plains, Leitch, Leistritz, and Bangsund (1996) and Leistritz, Bangsund, and Hodur (2004)
estimated $130 million in losses due to a leafy spurge infestation in the Dakotas, Montana, and
Wyoming.42 However, the contribution of N deposition to these losses is still uncertain.
Terrestrial nutrient enrichment also affects cultural and regulating services. For
example, in California, Coastal Sage Scrub (CSS) habitat concerns focus on a decline in CSS and
an increase in nonnative grasses and other species, impacts on the viability of threatened and
endangered species associated with CSS, and an increase in fire frequency. Changes in Mixed
Conifer Forest (MCF) include changes in habitat suitability and increased tree mortality,
increased fire intensity, and a change in the forest's nutrient cycling that may affect surface
water quality through nitrate leaching (U.S. EPA, 2008c). CSS and MCF are an integral part of
the California landscape, and together the ranges of these habitats include the densely
populated and valuable coastline and the mountain areas. Numerous threatened and
endangered species at both the state and federal levels reside in CSS and MCF. The value that
California residents and the U.S. population as a whole place on CSS and MCF habitats is
reflected in the various federal, state, and local government measures that have been put in
place to protect these habitats, including the Endangered Species Act, conservation planning
programs, and private and local land trusts. CSS and MCF habitat are showcased in many
popular recreation areas in California, including several national parks and monuments. In
addition, millions of individuals are involved in fishing, hunting, and wildlife viewing in California
every year (DOI, 2007). The quality of these trips depends in part on the health of the
ecosystems and their ability to support the diversity of plants and animals found in important
habitats found in CSS or MCF ecosystems and the parks associated with those ecosystems.
42These estimates reflect the total value of the service, not the marginal change in the value of the service as a
result of the emission reductions achieved by this rule.
6-56
-------
Based on analyses in the NOx SOx REA average values of the total benefits in 2006 from fishing,
hunting, and wildlife viewing away from home in California were approximately $950 million,
$170 million, and $3.6 billion, respectively (U.S. EPA, 2009c).43 In addition, data from California
State Parks (2003) indicate that in 2002, 69% of adult residents participated in trail hiking for an
average of 24 days per year. The aggregate annual benefit for California residents from trail
hiking in 2007 was $11 billion (U.S. EPA, 2009c).44 It is not currently possible to quantify the loss
in value of services due to nitrogen deposition as those losses are already reflected in the
estimates of the contemporaneous total value of these recreational activities. Restoration of
services through decreases in nitrogen deposition would likely increase the total value of
recreational services.
Fire regulation is also an important regulating service that could be affected by nutrient
enrichment of the CSS and MCF ecosystems by encouraging growth of more flammable grasses,
increasing fuel loads, and altering the fire cycle. Over the 5-year period from 2004 to 2008,
Southern California experienced, on average, over 4,000 fires per year burning, on average,
over 400,000 acres per year (National Association of State Foresters [NASF], 2009). It is not
possible at this time to quantify the contribution of nitrogen deposition, among many other
factors, to increased fire risk.
We are unable to provide an estimate of the terrestrial nutrient enrichment benefits
associated with the alternative standard level combinations due to data, time, and resource
limitations. Methods are not yet available to allow estimation of changes in ecosystem services
due to nitrogen deposition.
6.7.2.3 Vegetation Effects Associated with Gaseous Sulfur Dioxide
Uptake of gaseous sulfur dioxide in a plant canopy is a complex process involving
adsorption to surfaces (leaves, stems, and soil) and absorption into leaves. S02 penetrates into
leaves through to the stomata, although there is evidence for limited pathways via the cuticle
(U.S. EPA, 2008c). Pollutants must be transported from the bulk air to the leaf boundary layer in
order to get to the stomata. When the stomata are closed, as occurs under dark or drought
conditions, resistance to gas uptake is very high and the plant has a very low degree of
susceptibility to injury. In contrast, mosses and lichens do not have a protective cuticle barrier
to gaseous pollutants or stomates and are generally more sensitive to gaseous sulfur and
43These estimates reflect the total value of the service, not the marginal change in the value of the service as a
44
result of the emission reductions achieved by this rule.
These estimates reflect the total value of the service, not the marginal change in the value of the service as a
result of the emission reductions achieved by this rule.
6-57
-------
nitrogen than vascular plants (U.S. EPA, 2008c). Acute foliar injury usually happens within hours
of exposure, involves a rapid absorption of a toxic dose, and involves collapse or necrosis of
plant tissues. Another type of visible injury is termed chronic injury and is usually a result of
variable S02 exposures over the growing season. Besides foliar injury, chronic exposure to low
S02 concentrations can result in reduced photosynthesis, growth, and yield of plants (U.S. EPA,
2008c). These effects are cumulative over the season and are often not associated with visible
foliar injury. As with foliar injury, these effects vary among species and growing environment.
S02 is also considered the primary factor causing the death of lichens in many urban and
industrial areas (Hutchinson et al., 1996). The ISA for NOx/SOx—Ecological Criteria concluded
that the evidence is sufficient to infer a causal relationship between S02 injury to vegetation
(U.S. EPA, 2008c).
6.7.2A Mercury-Related Benefits Associated with the Role ofSulfate in Mercury Methylation
Mercury is a persistent, bioaccumulative toxic metal that is emitted from in three forms:
gaseous elemental Hg (Hg°), oxidized Hg compounds (Hg+2), and particle-bound Hg (HgP).
Methylmercury (MeHg) is formed by microbial action in the top layers of sediment and soils,
after Hg has precipitated from the air and deposited into waterbodies or land. Once formed,
MeHg is taken up by aquatic organisms and bioaccumulates up the aquatic food web. Larger
predatory fish may have MeHg concentrations many times, typically on the order of one million
times, that of the concentrations in the freshwater body in which they live.
The NOx SOx ISA—Ecological Criteria concluded that evidence is sufficient to infer a
causal relationship between sulfur deposition and increased mercury methylation in wetlands
and aquatic environments (U.S. EPA, 2008c). Specifically, there appears to be a relationship
between S042" deposition and mercury methylation; however, the rate of mercury methylation
varies according to several spatial and biogeochemical factors whose influence has not been
fully quantified (see Figure 6-14). Therefore, the correlation between S042" deposition and
MeHg could not be quantified for the purpose of interpolating the association across
waterbodies or regions. Nevertheless, because changes in MeHg in ecosystems represent
changes in significant human and ecological health risks, the association between sulfur and
mercury cannot be neglected (U.S. EPA, 2008c).
6-58
-------
Soaaa Factare
LtWrwm WflflWKfc
Upsfc^am Foraitofl Land
Upstnwm Erosion
UtHtorcrn Urt«r> Land
SRQ
•t-
^ il--.ru
Organir. Mwt«r
Temperatura
Methyirnercury
Figure 6-14. Spatial and Biogeochemical Factors Influencing MeHg Production
As research evolves and the computational capacity of models expands to meet the
complexity of mercury methylation processes in ecosystems, the role of interacting factors may
be better parsed out to identify ecosystems or regions that are more likely to generate higher
concentrations of MeHg. Figure 6-15 illustrates the type of current and forward-looking
research being developed by the U.S. Geological Survey (USGS) to synthesize the contributing
factors of mercury and to develop a map of sensitive watersheds. The mercury score
referenced in Figure 6-15 is based on S042" concentrations, acid neutralizing capacity (ANC),
levels of dissolved organic carbon and pH, mercury species concentrations, and soil types to
gauge the methylation sensitivity (Myers et al., 2007).
6-59
-------
Mercury sensitivity
score lunitlcsi)
Figure 6-15. Preliminary USGS Map of Mercury Methylation-Sensitive Watersheds
Source: Myers et al., 2007
Interdependent biogeochemical factors preclude the existence of simple sulfate-related
mercury methylation models. It is clear that decreasing sulfate deposition is likely to result in
decreased MeHg concentrations. Future research may allow for the characterization of a usable
sulfate-MeHg response curve; however, no regional or classification calculation scale can be
created at this time because of the number of confounding factors.
Decreases in S042" deposition have already shown promising reductions in MeHg.
Observed decreases in MeHg fish tissue concentrations have been linked to decreased
acidification and declining S042" and mercury deposition in Little Rock Lake, Wl (Hrabik and
Watras, 2002), and to decreased S042" deposition in Isle Royale in Lake Superior, Ml (Drevnick et
al., 2007). Although the possibility exists that reductions in S042" emissions could generate a
pulse in MeHg production because of decreased sulfide inhibition in sulfate-saturated waters,
this effect would likely involve a limited number of U.S. waters (Harmon et al., 2007). Also,
because of the diffusion and outward flow of both mercury-sulfide complexes and S042",
increased mercury methylation downstream may still occur in sulfate-enriched ecosystems with
increased organic matter and/or downstream transport capabilities.
6-60
-------
Remediation of sediments heavily contaminated with mercury has yielded significant
reductions of MeHg in biotic tissues. Establishing quantitative relations in biotic responses to
MeHg levels as a result of changes in atmospheric mercury deposition, however, presents
difficulties because direct associations can be confounded by all of the factors discussed in this
section. Current research does suggest that the levels of MeHg and total mercury in ecosystems
are positively correlated, so that reductions in mercury deposited into ecosystems would also
eventually lead to reductions in MeHg in biotic tissues. Ultimately, an integrated approach that
involves the reduction of both sulfur and mercury emissions may be most efficient because of
the variability in ecosystem responses. Reducing SOX emissions could have a beneficial effect on
levels of MeHg in many waters of the United States.
MeHg is the only form of mercury that biomagnifies in the food web. Concentrations of
MeHg in fish are generally on the order of a million times the MeHg concentration in water. In
addition to mercury deposition, key factors affecting MeHg production and accumulation in fish
include the amount and forms of sulfur and carbon species present in a given waterbody. Thus,
two adjoining water bodies receiving the same deposition can have significantly different fish
mercury concentrations.
Methylmercury builds up more in some types offish and shellfish than in others. The
levels of methylmercury in high and shellfish vary widely depending on what they eat, how long
they live, and how high they are in the food chain. Most fish, including ocean species and local
freshwater fish, contain some methylmercury. In general, higher mercury concentrations are
expected in top predators, which are often large fish relative to other species in a waterbody.
The ecosystem service most directly affected by sulfate-mediated mercury methylation
is the provision of fish for consumption as a food source. This service is of particular importance
to groups engaged in subsistence fishing, pregnant women and young children.
6.7.3 Ecosystem Benefits from Reductions in Mercury Emissions
Deposition of mercury to waterbodies can also have an impact on ecosystems and
wildlife. Mercury contamination is present in all environmental media with aquatic systems
experiencing the greatest exposures due to bioaccumulation. Bioaccumulation refers to the net
uptake of a contaminant from all possible pathways and includes the accumulation that may
occur by direct exposure to contaminated media as well as uptake from food.
Atmospheric mercury enters freshwater ecosystems by direct deposition and through
runoff from terrestrial watersheds. Once mercury deposits, it may be converted to organic
6-61
-------
methylmercury mediated primarily by sulfate-reducing bacteria. Methylation is enhanced in
anaerobic and acidic environments, greatly increasing mercury toxicity and potential to
bioaccumulate in aquatic foodwebs. A number of key biogeochemical controls influence the
production of methylmercury in aquatic ecosystems. These include sulfur, pH, organic matter,
iron, mercury "aging," and bacteria type and activity (Munthe et al., 2007).
Wet and dry deposition of oxidized mercury is a dominant pathway for bringing mercury
to terrestrial surfaces. In forest ecosystems, elemental mercury may also be absorbed by plants
stomatally, incorporated by foliar tissues and released in litterfall (Ericksen etal., 2003).
Mercury in throughfall, direct deposition in precipitation, and uptake of dissolved mercury by
roots (Rea et al., 2002) are also important in mercury accumulation in terrestrial ecosystems.
Soils have significant capacity to store large quantities of atmospherically deposited
mercury where it can leach into groundwater and surface waters. The risk of mercury exposure
extends to insectivorous terrestrial species such as songbirds, bats, spiders, and amphibians
that receive mercury deposition or from aquatic systems near the forest areas they inhabit
(Bergeron et al., 2010a, b; Cristol et al., 2008; Rimmer et al., 2005; Wada et al., 2009 & 2010).
Numerous studies have generated field data on the levels of mercury in a variety of wild
species. Many of the data from these environmental studies are anecdotal in nature rather than
representative or statistically designed studies. The body of work examining the effects of these
exposures is growing but still incomplete given the complexities of the natural world. A large
portion of the adverse effect research conducted to date has been carried out in the laboratory
setting rather than in the wild; thus, conclusions about overarching ecosystem health and
population effects are difficult to make at this time. In the sections that follow numerous
effects have been identified at differing exposure levels.
6.7.3.1 Mercury Effects on Fish
A review of the literature on effects of mercury on fish (Crump and Trudeau, 2009)
reports results for numerous species including trout, bass (large and smallmouth), northern
pike, carp, walleye, salmon and others from laboratory and field studies. The effects studied are
reproductive and include deficits in sperm and egg formation, histopathological changes in
testes and ovaries, and disruption of reproductive hormone synthesis. These studies were
conducted in areas from New York to Washington and while many were conducted by adding
MeHg to water or diet many were conducted at current environmental levels. While we cannot
determine at this time whether these reproductive deficits are affecting fish populations across
the United States it should be noted that it is possible that overtime reproductive deficits could
6-62
-------
have an effect on populations. Lower fish populations would conceivably impact the ecosystem
services like recreational fishing derived from having healthy aquatic ecosystems quite apart
from the effects of consumption advisories due to the human health effects of mercury.
6.7.3.2 Mercury Effects on Birds
In addition to effects on fish, mercury also affects avian species. In previous reports (U.S.
EPA (1997); U.S. EPA (2005)), much of the focus has been on large piscivorous species in
particular the common loon. The loon is most visible to the public during the summer breeding
season on northern lakes and they have become an important symbol of wilderness in these
areas (Mclntyre and Barr, 1997). A multitude of loon watch, preservation, and protection
groups have formed over the past few decades and have been instrumental in promoting
conservation, education, monitoring, and research of breeding loons (Mclntyre and Evers,
2000; Evers, 2006). Significant adverse effects on breeding loons from mercury have been
found to occur including behavioral (reduced nest-sitting), physiological (flight feather
asymmetry) and reproductive (chicks fledged/territorial pair) effects (Evers, 2008). Additionally
Evers, et al (2008) report that they believe that the weight of evidence indicates that
population-level effects occur in parts of Maine and New Hampshire, and potentially in broad
areas of the loon's range.
Recently attention has turned to other piscivorous species such as the white ibis, and
great snowy egret. While considered to be fish-eating generally these wading birds have a very
wide diet including crayfish, crabs, snails, insects and frogs. These species are experiencing a
range of adverse effects due to exposure to mercury. The white ibis has been observed to have
decreased foraging efficiency (Adams and Frederick, 2008). Additionally ibises have been shown
to exhibit decreased reproductive success and altered pair behavior (Frederick and Jayasena,
2010). These effects include significantly more unproductive nests, male/male pairing, reduced
courtship behavior (head bobbing and pair bowing) and lower nestling production by exposed
males. In this study, a worst-case scenario suggested by the results could involve up to a 50%
reduction in fledglings due to MeHg in diet. These estimates may be conservative if male/male
pairing in the wild it could result in a shortage of partners for females and the effect of
homosexual breeding would be magnified. In egrets, mercury has been implicated in the
decline of the species in south Florida (Sepulveda, et al., 1999) and Hoffman (2010) has shown
that egrets show liver and possibly kidney effects. While ibises and egrets are most abundant in
coastal areas and these studies were conducted in south Florida and Nevada the ranges of
ibises and egrets extend to a large portion of the United States. Ibis territory can range inland
6-63
-------
to Oklahoma, Arkansas and Tennessee. Egret range covers virtually the entire United States
except the mountain west.
Insectivorous birds have also been shown to suffer adverse effects due to mercury
exposure. These songbirds such as Bicknell's thrush, tree swallows and the great tit have shown
reduced reproduction, survival, and changes in singing behavior. Exposed tree swallows
produced fewer fledglings (Brasso, 2008), lower survival (Hallinger, 2010) and had
compromised immune competence (Hawley, 2009). The great tit has exhibited reduced singing
behavior and smaller song repertoire in an area of high contamination in the vicinity of a
metallurgic smelter in Flanders (Gorissen, 2005).
6.7.3.3 Mercury Effects on Mammals
In mammals, adverse effects have been observed in mink and river otter, both fish
eating species. For otter from Maine and Vermont maximum concentrations on Hg in fur nearly
equal or exceed a concentration associated with mortality and concentration in liver for mink in
Massachusetts/Connecticut and the levels in fur from mink in Maine exceed concentrations
associated with acute mortality (Yates, 2005). Adverse sublethal effects may be associated with
lower Hg concentrations and consequently be more widespread than potential acute effects.
These effects may include increased activity, poorer maze performance, abnormal startle reflex,
and impaired escape and avoidance behavior (Scheuhammer et al., 2007).
6.7.3.4 Mercury Ecological Conclusions
The studies cited here provide a glimpse of the scope of mercury effects on wildlife
particularly reproductive and survival effects. These effects range across species from fish to
mammals and spatially across a wide area of the United States. The literature is far from
complete however. Much more research is required to establish a link between the ecological
effects on wildlife and the effect on ecosystem services (services that the environment provides
to people) for example recreational fishing, bird watching and wildlife viewing. EPA is not,
however, currently able to quantify or monetize the benefits of reducing mercury exposures
affecting provision of ecosystem services.
6.7.4 Vegetation Benefits from Reductions in Ambient Ozone
Control strategies that include emission reductions of NOx would affect ambient ozone
concentrations. Ozone causes discernible injury to a wide array of vegetation (U.S. EPA, 2006a;
Fox and Mickler, 1996). Air pollution can affect the environment and affect ecological systems,
leading to changes in the ecological community and influencing the diversity, health, and vigor
6-64
-------
of individual species (U.S. EPA, 2006a). In terms of forest productivity and ecosystem diversity,
ozone may be the pollutant with the greatest potential for regional-scale forest impacts (U.S.
EPA, 2006a). Studies have demonstrated repeatedly that ozone concentrations commonly
observed in polluted areas can have substantial impacts on plant function (De Steiguer et al.,
1990; Pye, 1988).
When ozone is present in the air, it can enter the leaves of plants, where it can cause
significant cellular damage. Like carbon dioxide (C02) and other gaseous substances, ozone
enters plant tissues primarily through the stomata in leaves in a process called "uptake"
(Winner and Atkinson, 1986). Once sufficient levels of ozone (a highly reactive substance), or its
reaction products, reaches the interior of plant cells, it can inhibit or damage essential cellular
components and functions, including enzyme activities, lipids, and cellular membranes,
disrupting the plant's osmotic (i.e., water) balance and energy utilization patterns (U.S. EPA,
2006a; Tingey and Taylor, 1982). With fewer resources available, the plant reallocates existing
resources away from root growth and storage, above ground growth or yield, and reproductive
processes, toward leaf repair and maintenance, leading to reduced growth and/or
reproduction. Studies have shown that plants stressed in these ways may exhibit a general loss
of vigor, which can lead to secondary impacts that modify plants' responses to other
environmental factors. Specifically, plants may become more sensitive to other air pollutants,
or more susceptible to disease, pest infestation, harsh weather (e.g., drought, frost) and other
environmental stresses, which can all produce a loss in plant vigor in ozone-sensitive species
that overtime may lead to premature plant death. Furthermore, there is evidence that ozone
can interfere with the formation of mycorrhiza, essential symbiotic fungi associated with the
roots of most terrestrial plants, by reducing the amount of carbon available for transfer from
the host to the symbiont (U.S. EPA, 2006a).
This ozone damage may or may not be accompanied by visible injury on leaves, and
likewise, visible foliar injury may or may not be a symptom of the other types of plant damage
described above. Foliar injury is usually the first visible sign of injury to plants from ozone
exposure and indicates impaired physiological processes in the leaves (Grulke, 2003). When
visible injury is present, it is commonly manifested as chlorotic or necrotic spots, and/or
increased leaf senescence (accelerated leaf aging). Visible foliar injury reduces the aesthetic
value of ornamental vegetation and trees in urban landscapes and negatively affects scenic
vistas in protected natural areas.
Ozone can produce both acute and chronic injury in sensitive species depending on the
concentration level and the duration of the exposure. Ozone effects also tend to accumulate
6-65
-------
over the growing season of the plant, so that even lower concentrations experienced for a
longer duration have the potential to create chronic stress on sensitive vegetation. Not all
plants, however, are equally sensitive to ozone. Much of the variation in sensitivity between
individual plants or whole species is related to the plant's ability to regulate the extent of gas
exchange via leaf stomata (e.g., avoidance of ozone uptake through closure of stomata) and the
relative ability of species to detoxify ozone-generated reactive oxygen free radicals (U.S. EPA,
2006a; Winner, 1994). After injuries have occurred, plants may be capable of repairing the
damage to a limited extent (U.S. EPA, 2006a). Because of the differing sensitivities among
plants to ozone, ozone pollution can also exert a selective pressure that leads to changes in
plant community composition. Given the range of plant sensitivities and the fact that numerous
other environmental factors modify plant uptake and response to ozone, it is not possible to
identify threshold values above which ozone is consistently toxic for all plants.
Because plants are at the base of the food web in many ecosystems, changes to the
plant community can affect associated organisms and ecosystems (including the suitability of
habitats that support threatened or endangered species and below ground organisms living in
the root zone). Ozone impacts at the community and ecosystem level vary widely depending
upon numerous factors, including concentration and temporal variation of tropospheric ozone,
species composition, soil properties and climatic factors (U.S. EPA, 2006a). In most instances,
responses to chronic or recurrent exposure in forested ecosystems are subtle and not
observable for many years. These injuries can cause stand-level forest decline in sensitive
ecosystems (U.S. EPA, 2006a, McBride et al., 1985; Miller et al., 1982). It is not yet possible to
predict ecosystem responses to ozone with certainty; however, considerable knowledge of
potential ecosystem responses is available through long-term observations in highly damaged
forests in the U.S. (U.S. EPA, 2006a).
6.7.4.1 Ozone Effects on Forests
Air pollution can affect the environment and affect ecological systems, leading to
changes in the ecological community and influencing the diversity, health, and vigor of
individual species (U.S. EPA, 2006a). Ozone has been shown in numerous studies to have a
strong effect on the health of many plants, including a variety of commercial and ecologically
important forest tree species throughout the United States (U.S. EPA, 2007b).
In the U.S., this data comes from the U.S. Department of Agriculture (USDA) Forest
Service Forest Inventory and Analysis (FIA) program. As part of its Phase 3 program (formerly
known as Forest Health Monitoring), FIA looks for visible foliar injury of ozone-sensitive forest
plant species at each ground monitoring site across the country (excluding woodlots and urban
6-66
-------
trees) that meets certain minimum criteria. Because ozone injury is cumulative over the course
of the growing season, examinations are conducted in July and August, when ozone
concentrations and associated injury are typically highest.
Monitoring of ozone injury to plants by the U.S. Forest Service has expanded over the
last 15 years from monitoring sites in 10 states in 1994 to nearly 1,000 monitoring sites in 41
states in 2002. Since 2002, the monitoring program has further expanded to 1,130 monitoring
sites in 45 states. Figure 6-16 shows the results of this monitoring program for the year 2002
broken down by U.S. EPA Regions.45 Figure 6-17 identifies the counties that were included in
Figure 6-16, and provides the county-level data regarding the presence or absence of ozone-
related injury. As shown in Figure 6-16, large geographic areas of EPA Regions 6, 8, and 10 were
not included in the assessment. Ozone damage to forest plants is classified using a subjective
five-category biosite index based on expert opinion, but designed to be equivalent from site to
site. Ranges of biosite values translate to no injury, low or moderate foliar injury (visible foliar
injury to highly sensitive or moderately sensitive plants, respectively), and high or severe foliar
injury, which would be expected to result in tree-level or ecosystem-level responses,
respectively (U.S. EPA, 2006a; Coulston, 2004). The highest percentages of observed high and
severe foliar injury, which are most likely to be associated with tree or ecosystem-level
responses, are primarily found in the Mid-Atlantic and Southeast regions. While the assessment
showed considerable regional variation in ozone injury, this assessment targeted different
ozone-sensitive species in different parts of the country with varying ozone sensitivity, which
contributes to the apparent regional differences. It is important to note that ozone can have
other, more significant impacts on forest plants (e.g., reduced biomass growth in trees) prior to
showing signs of visible foliar injury (U.S. EPA, 2006a).
Assessing the impact of ground-level ozone on forests in the U.S. involves understanding
the risks to sensitive tree species from ambient ozone concentrations and accounting for the
prevalence of those species within the forest. As a way to quantify the risks to particular plants
from ground-level ozone, scientists have developed ozone-exposure/tree-response functions by
exposing tree seedlings to different ozone levels and measuring reductions in growth as
"biomass loss." Typically, seedlings are used because they are easy to manipulate and measure
their growth loss from ozone pollution. The mechanisms of susceptibility to ozone within the
leaves of seedlings and mature trees are identical, and the decreases predicted using the
seedlings should be related to the decrease in overall plant fitness for mature trees, but the
45 The data are based on averages of all observations collected in 2002, which is the last year for which data are
publicly available. For more information, please consult EPA's 2008 Report on the Environment (U.S. EPA,
2008b).
6-67
-------
0«tirea oF injury-
r-f ::!:• |
Ragtonz
(42 srtes)
Ragicri 3
fill sues )
Region 4
(227 sn*si
Ragion 5
('80 sites)
Rag ion 6
(99 srtes)
Hagton7
163 srtes)
Raglon 8
(72a»t«S)
Bagton9
ISO srtes)
Ragion 10
(57 srtes)
Nor* Lam Moderate I High I Severe
Nfo*M of monitorini sites in «aeb uitegory:
63-S | H3.7 11 1 -3.7
61.9 21.4 IMJTJ
309 H.O U4 | |r;> - •
755 W.17.Dfl^g
7S.6 11.3 S1
94 9 =J5.1
85 7 9.5 "JI^ £
1000 |
76.3 12,5 83 |] I
1000
945 monitoring sites.
located in 41 stales
LToials ntay not add to 100%
rouixt)n
-------
Absent
Pi»s«nt
Point IH)Lliy
Figure 6-17. Presence and Absence of Visible Foliar Injury, as Measured by U.S. Forest
Service, 2002
Source: U.S. EPA, 2007b
magnitude of the effect may be higher or lower depending on the tree species (Chappelka and
Samuelson, 1998). In areas where certain ozone-sensitive species dominate the forest
community, the biomass loss from ozone can be significant. Experts have identified 2% annual
biomass loss as a level of concern, which would cause long term ecological harm as the short-
term negative effects on seedlings compound to affect long-term forest health (Heck and
Cowling, 1997).
Ozone damage to the plants including the trees and understory in a forest can affect the
ability of the forest to sustain suitable habitat for associated species particularly threatened and
endangered species that have existence value—a nonuse ecosystem service—for the public.
Similarly, damage to trees and the loss of biomass can affect the forest's provisioning services
in the form of timber for various commercial uses. In addition, ozone can cause discoloration of
leaves and more rapid senescence (early shedding of leaves), which could negatively affect fall-
color tourism because the fall foliage would be less available or less attractive. Beyond the
aesthetic damage to fall color vistas, forests provide the public with many other recreational
and educational services that may be affected by reduced forest health including hiking, wildlife
6-69
-------
viewing (including bird watching), camping, picnicking, and hunting. Another potential effect of
biomass loss in forests is the subsequent loss of climate regulation service in the form of
reduced ability to sequester carbon and alteration of hydrologic cycles.
Some of the common tree species in the United States that are sensitive to ozone are
black cherry (Prunus serotina), tulip-poplar (Liriodendron tulipifera), and eastern white pine
(Pinus strobus). Ozone-exposure/tree-response functions have been developed for each of
these tree species, as well as for aspen (Populus tremuliodes), and ponderosa pine (Pinus
ponderosa) (U.S. EPA, 2007b).
6.7.4.2 Ozone Effects on Crops
Laboratory and field experiments have shown reductions in yields for agronomic crops
exposed to ozone, including vegetables (e.g., lettuce) and field crops (e.g., cotton and wheat).
Damage to crops from ozone exposures includes yield losses (i.e., in terms of weight, number,
or size of the plant part that is harvested), as well as changes in crop quality (i.e., physical
appearance, chemical composition, or the ability to withstand storage) (U.S. EPA, 2007b). The
most extensive field experiments, conducted under the National Crop Loss Assessment
Network (NCLAN) examined 15 species and numerous cultivars. The NCLAN results show that
"several economically important crop species are sensitive to ozone levels typical of those
found in the United States" (U.S. EPA, 2006a). In addition, economic studies have shown
reduced economic benefits as a result of predicted reductions in crop yields, directly affecting
the amount and quality of the provisioning service provided by these crops, associated with
observed ozone levels (Kopp et al., 1985; Adams et al., 1986; Adams et al., 1989). In addition,
visible foliar injury by itself can reduce the market value of certain leafy crops (such as spinach,
lettuce). According to the Ozone Staff Paper, there has been no evidence that crops are
becoming more tolerant of ozone (U.S. EPA, 2007b). Using the Agriculture Simulation Model
(AGSIM) (Taylor, 1994) to calculate the agricultural benefits of reductions in ozone exposure,
U.S. EPA estimated that attaining a W126 standard of 13 ppm-hr would produce monetized
benefits of approximately $400 million to $620 million in 2006 (inflated to 2006 dollars) (U.S.
EPA, 2007b).46
46 These estimates illustrate the value of vegetation effects from a substantial reduction of ozone concentrations,
not the marginal change in ozone concentrations anticipated a result of the emission reductions achieved by
this rule.
6-70
-------
6.7.4.3 Ozone Effects on Ornamental Plants
Urban ornamental plants are an additional vegetation category likely to experience
some degree of negative effects associated with exposure to ambient ozone levels. Several
ornamental species have been listed as sensitive to ozone (Abt Associates, 1995). Because
ozone causes visible foliar injury, the aesthetic value of ornamental plants (such as petunia,
geranium, and poinsettia) in urban landscapes would be reduced (U.S. EPA, 2007b). Sensitive
ornamental species would require more frequent replacement and/or increased maintenance
(fertilizer or pesticide application) to maintain the desired appearance because of exposure to
ambient ozone (U.S. EPA, 2007b). In addition, many businesses rely on healthy-looking
vegetation for their livelihoods (e.g., horticulturalists, landscapers, Christmas tree growers,
farmers of leafy crops, etc.). The ornamental landscaping industry is a multi-billion dollar
industry that affects both private property owners/tenants and governmental units responsible
for public areas (Abt Associates, 1995). Preliminary data from the 2007 Economic Census
indicate that the landscaping services industry, which is primarily engaged in providing
landscape care and maintenance services and installing trees, shrubs, plants, lawns, or gardens,
was valued at $53 billion (U.S. Census Bureau, 2010). Therefore, urban ornamentals represent a
potentially large unquantified benefit category. This aesthetic damage may affect the
enjoyment of urban parks by the public and homeowners' enjoyment of their landscaping and
gardening activities. In addition, homeowners may experience a reduction in home value or a
home may linger on the market longer due to decreased aesthetic appeal. In the absence of
adequate exposure-response functions and economic damage functions for the potential range
of effects relevant to ornamental plants, we cannot conduct a quantitative analysis to estimate
these effects.
We are unable to provide an estimate of the ozone crop benefits associated with the
alternative standard level combinations due to data, time, and resource limitations.
6.8 References
Abt Associates, Inc. 1995. Urban ornamental plants: sensitivity to ozone and potential economic
losses. U.S. EPA, Office of Air Quality Planning and Standards, Research Triangle Park.
Under contract to RADIAN Corporation, contract no. 68-D3-0033, WA no. 6. pp. 9-10.
Available on the Internet at
.
6-71
-------
Abt Associates, Inc. 2001. Assessing Public Opinions on Visibility Impairment due to Air
Pollution: Summary Report. Prepared for EPA Office of Air Quality Planning and
Standards; funded under EPA Contract No. 68-D-98-001. Abt Associates Inc., Bethesda,
MD. January. Available on the Internet at
.
Adams, Evan M., and Frederick, Peter C. 2008. "Effects of methylmercury and spatial complexity
on foraging behavior and foraging efficiency in juvenile white ibises (Eudocimus albus)."
Environmental Toxicology and Chemistry. Vol 27, No. 8.
Adams, R. M., Glyer, J. D., Johnson, S. L, McCarl, B. A. 1989. "A reassessment of the economic
effects of ozone on U.S. agriculture." Journal of the Air Pollution Control Association, 39,
960-968'
Adams, R. M., Hamilton, S. A., McCarl, B. A. 1986. "The benefits of pollution control: the case of
ozone and U.S. agriculture." American Journal of Agricultural Economics, 34, 3-19.
Anselin, L., Le Gallo, J. 2006. "Interpolation of air quality measures in hedonic house price
models: Spatial aspects." Spatial Economic Analysis 1: 31-52
Anselin, L., Lozano-Gracia, N. 2009. "Errors in variables and spatial effects in hedonic house
price models of ambient air quality." Empirical Economics 34: 5-34.
Arrow, K.J., et al. 1993. Report of the NOAA Panel on Contingent Valuation. Federal Register.
Volume 58, Number 10. Pages 4601 to 4614. Available on the Internet at
.
Arzavus K.M., Dickhut R.M., and Canuel E.A. 2001. "Fate of Atmospherically Deposited
Polycyclic Aromatic Hydrocarbons (PAHs) in Chesapeake Bay." Environmental Science &
Technology, 35, 2178-2183.
Bond, T.C., and Sun H.L. (2005) Can reducing BC emissions counteract global warming?
Environmental Science & Technology 39 (16).
Banzhaf, S., D. Burtraw, D. Evans, and A. Krupnick. 2006. "Valuation of Natural Resource
Improvements in the Adirondacks." Land Economics 82:445-464.
Bayer, P., Keohane, N.O. and Timmins, C. 2009. "Migration and hedonic valuation: the case of
air quality," Journal of Environmental Economics and Management, 58 (2009) 1-14.
BBC Research & Consulting. 2003. Phoenix Area Visibility Survey. Draft Report. Available on the
Internet at .
Bergeron, CM., Bodinof, CM., Unrine, JM., Hopkins, WA. 2010a. "Mercury accumulation along a
contamination gradient and nondestructive indices of bioaccumulation in amphibians."
Environmental Toxicology and Chemistry 29(4), 980-988.
6-72
-------
Bergeron, CM., Bodinof, CM., Unrine, JM., Hopkins, WA. 2010b. "Bioaccumulation and maternal
transfer of mercury and selenium in amphibians." Environmental Toxicology and
Chemistry 29(4), 989-997.
Beron KJ, Murdoch J, Thayer M. 2001. "The benefits of visibility improvement: New evidence
from Los Angeles metropolitan area." J Real Estate Finance Econ 22(2-3):319-337.
Beron KJ, Hanson Y, Murdoch JC, Thayer MA. 2004. "Hedonic price functions and spatial
dependence: implications for the demand for urban air quality." In: Anselin L, Florax RJ,
Rey SJ (eds) Advances in spatial econometrics: methodology, tools and applications.
Springer-Verlag, Berlin, pp 267-281.
Brasso, Rebecka L., and Cristol, Daniel A. 2008. "Effects of mercury exposure in the reproductive
success of tree swallows (Tachycineta bicolor)." Ecotoxicology. 17:133-141.
Brookshire, D.S., R.C. d'Arge, W.D. Schulze and M.A. Thayer. 1979. Methods Development for
Assessing Tradeoffs in Environmental Management, Vol. II: Experiments in Valuing Non-
Market Goods: A Case Study of Alternative Benefit Measures of Air Pollution Control in
the South Coast Air Basin of Southern California. Prepared for the U.S. Environmental
Protection Agency, Office of Research and Development. Available on the Internet at
.
Carson, R.T., R.C. Mitchell, and P.A. Rudd. 1990. "Valuing Air Quality Improvements: Simulating
a Hedonic Pricing Equation in the Context of a Contingent Valuation Scenario." In
Visibility and Fine Particles, Transactions of an AWMA/EPA International Specialty
Conference, C.V. Mathai, ed. Air and Waste Management Association, Pittsburgh.
Available on the Internet at
.
Chameides. W.L., Yu, H., Liu, S.C., Bergin, M., Zhou, X., Mearns, L., Wang, G., Kiang, C.S., Saylor,
R.D., Luo, C. Huang, Y., Steiner, A., and Giorgi, F. 1999. "Case study of the effects of
atmospheric aerosols and regional haze on agriculture: an opportunity to enhance crop
yields in China through emission controls?" PNAS, 96: 13626-13633.
6-73
-------
Chappelka, A.M., Samuelson, L.J. 1998. "Ambient ozone effects on forest trees of the eastern
United States: a review." New Phytologist, 139, 91-108.
Chay KY, Greenstone M. 2005. "Does air quality matter? evidence form the housing market." J
PolitEcon 113(2):376-424.
Chestnut, L.G. 1997. Methodology for Estimating Values for Changes in Visibility in National
Parks. Prepared for Office of Policy, Planning and Evaluation, May. Available on the
Internet at
.
Chestnut, L.G. and R. Dennis. 1997. "Economic Benefits of Improvements in Visibility: Acid Rain
Provisions of the 1990 Clean Air Act Amend ments." Journal of Air and Waste
Management Association 47: 395-402.
Chestnut, L.G., and R.D. Rowe. 1990a. "New National Park Visibility Value Estimates." In
Visibility and Fine Particles, Transactions of an AWMA/EPA International Specialty
Conference, C.V. Mathai, ed. Air and Waste Management Association, Pittsburgh.
Chestnut, L.G. and R.D. Rowe. 1990b. Preservation Values for Visibility Protection at the
National Parks. Prepared for U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, and National Park Service, Air Quality Management Division.
Available on the Internet at
.
Chestnut, L.G. and R.D. Rowe. 1990c. "Economic Valuation of Changes in Visibility: A State of
the Science Assessment for NAPAP." Section B5 in Irving, Patricia M., e.d., 1991. Acid
Deposition: State of Science and Technology, Volume IV, Control Technologies, Future
Emissions, and Effects Valuation, The U.S. National Acid Precipitation Assessment
Program, Report 27, page 27-153.
Chestnut, L.G., R.D. Rowe and J. Murdoch. 1986. Review of 'Establishing and Valuing the Effects
of Improved Visibility in Eastern United States.' Prepared for the U.S. Environmental
Protection Agency. October. Available on the Internet at
.
6-74
-------
Cordell, H.K., C.J. Betz, G. Green, and M. Owens. 2005. Off-Highway Vehicle Recreation in the
United States, Regions and States: A National Report from the National Survey on
Recreation and the Environment (NSRE). Prepared for the U.S. Department of
Agriculture Forest Service, Southern Research Station, National OHV Policy and
Implementation Teams, Athens, GA. Available on the Internet at
Cordell, K., B. Leeworthy, G.T. Green, C. Betz, B. Stephens. 2008. The National Survey on
Recreation & the Environment. Research Work Unit 4953. Pioneering Research on
Changing Forest Values in the South and Nation, U.S. Department of Agriculture Forest
Service, Southern Research Station, Athens, GA. Available on the Internet at
.
Cotrufo, M.F., De Santo A.V., Alfani A., Bartoli G., De Cristofaro A. 1995. "Effects of urban heavy
metal pollution on organic matter decomposition in Quercus ilex L. Woods."
Environmental Pollution, 89(1), 81-87.
Coulston, J.W., Riitters, K.H., Smith, G.C. 2004. "A preliminary assessment of the Montreal
process indicators of air pollution for the United States." Environmental Monitoring and
Assessment, 95, 57-74.
Cristol D. A., Brasso R. L., Condon A. M., Fovargue R. E., Friedman S. L., Hallinger K. K., Monroe
A. P., White A. E. 2008. "The movement of aquatic mercury through terrestrial food
webs." Science 320, 335-335.
Cropper, Maureen. 2000. "Has Economic Research Answered the Needs of Environmental
Policy?" Journal of Environmental Economics and Management 39, 328-350.
Crump, Kate L., and Trudeau, Vance L. 2009. "Mercury-induced reproductive impairment in
fish." Environmental Toxicology and Chemistry. Vol. 28, No. 5.
De Steiguer, J., Pye, J., Love, C. 1990. "Air Pollution Damage to U.S. Forests." Journal of Forestry,
88(8), 17-22.
DeBell, L. J.; Gebhart, K. A.; Hand, J. L.; Malm, W. C; Pitchford, M. L.; Schichtel, B. A.; White, W.
H. 2006. Spatial and Seasonal Patterns and Temporal Variability of Haze and its
Constituents in the United States, Report IV; Cooperative Institute for Research in the
Atmosphere, Nov; ISSN 0737-5352-74. Available on the Internet at
.
DeHayes, D.H., P.G. Schaberg, G.J. Hawley, and G.R. Strimbeck. 1999. "Acid rain impacts on
calcium nutrition and forest health." e/osc/ence49(10):789-800.
6-75
-------
Ely, D.W., Leary, J.T., Stewart, T.R., Ross, D.M. 1991. "The Establishment of the Denver Visibility
Standard." Presented at Air and Waste Management Association 84th Annual Meeting,
June 16-21, Paper 91-48.4, 17pp. Available on the Internet at
.
Ericksen, J. A., Gustin, M. S., Schorran, D. E., Johnson, D. W., Lindberg, S. E., & Coleman, J. S.
(2003). "Accumulation of atmospheric mercury in forest foliage." Atmospheric
Environment 37(12), 1613-1622.
Evers, D.C., 2006. Status assessment and conservation plan for the common loon (Gavia immer)
in North America. U.S. Fish and Wildlife Service, Hadley, MA, USA.
Evers, David C., Savoy, Lucas J., DeSorbo, Christopher R., Yates, David E., Hanson, William,
Taylor, Kate M., Siegel, Lori S., Cooley, John H. Jr., Bank, Michael S., Major, Andrew,
Munney, Kenneth, Mower, Barry F., Vogel, Harry S., Schoch, Nina, Pokras, Mark,
Goodale, Morgan W., Fair, Jeff. 2008. "Adverse effects from environmental mercury
loads on breeding common loons." Ecotoxicology. 17:69-81.
Fox, S., Mickler, R. A. (Eds.). 1996. Impact of Air Pollutants on Southern Pine Forests, Ecological
Studies. (Vol. 118, 513 pp.) New York: Springer-Verlag.
Frederick, Peter, and Jayasena, Nilmini. 2010. "Altered pairing behavior and reproductive
success in white ibises exposed to environmentally relevant concentrations of
methylmercury." Proceedings of the Royal Society B. doi: 10-1098.
Gorissen, Leen, Snoeijs, Tinne, Van Duyse, Els, and Eens, Marcel. 2005. "Heavy metal pollution
affects dawn singing behavior in a small passerine bird." Oecologia. 145: 540-509.
Grantz DA; Garner JHB; Johnson DW. 2003. "Ecological effects of particulate matter." Environ
Int, 29: 213-239.
Grosclaude P. and N. Soguel. 1994. "Valuing damage to historic buildings using a contingent
market: a case study of road traffic externalities," Journal of Environmental Planning and
Management 37 (1994) (3), pp. 279-287.
Grulke, N.E. 2003. "The physiological basis of ozone injury assessment attributes in Sierran
conifers." In A. Bytnerowicz, M.J. Arbaugh, & R. Alonso (Eds.), Ozone air pollution in the
Sierra Nevada: Distribution and effects on forests, (pp. 55-81). New York, NY: Elsevier
Science, Ltd.
Hadley O.L., Corrigan C.E., Kirchstetter T.W., Cliff S.S., and Ramanathan V. (2010) Measured
black carbon deposition on the Sierra Nevada snow pack and implication for snow pack
retreat. Atmos. Chem. Phys. Discuss. 10, 10463-10485
(doi:10.5194/acpd-10-10463-2010).
6-76
-------
Hageman KJ; Simonich SL; Campbell DH; Wilson GR; Landers DH. 2006. "Atmospheric deposition
of current-use and historic-use pesticides in snow at national parks in the western
United States." Environ Sci Techno!, 40: 3174-3180
Hallinger, Kelly K., Cornell, Kerri L, Brasso, Rebecka L, and Cristol, Daniel A. 2010. "Mercury
exposure and survival in free-living tree swallows (Tachycineta bicolor)." Ecotoxicology.
Doi: 10.1007/S10646-010-0554-4.
Hawley, Dana M., Hallinger, Kelly K., Cristol, Daniel A. 2009. "Compromised immune
competence in free-living tree swallows exposed to mercury." Ecotoxicology. 18:499-
503.
Heck, W.W. &Cowling E.B. 1997. "The need for a long term cumulative secondary ozone
standard—an ecological perspective." Environmental Management, January, 23-33.
Hoffman, David J., Henny, Charles J., Hill, Elwood F., Grover, Robert A., Kaiser, James L.,
Stebbins, Katherine R. 2009. "Mercury and drought along the lower Carson River,
Nevada: III. Effects on blood and organ biochemistry and histopathology of snowy egrets
and black-crowned night-herons on Lahontan Reservoir, 2002-2006." Journal of
Toxicology and Environmental Health, Part A. 72: 20, 1223-1241.
Hutchison, R., and C.E. Kraft. 1994. "Hmong Fishing Activity and Fish Consumption." Journal of
Great Lakes Research 20(2):471-487.
Intergovernmental Panel on Climate Change 4th Assessment Report: Climate Change 2007
Synthesis Report Summary for Policy Makers Pachauri, R.K. and Reisinger, A. (Eds.).
Available on the Internet at
.
IMPROVE (Interagency Monitoring of Protected Visual Environments). 2010.1992-2008 data
from the IMPROVE network based on the "New IMPROVE algorithm" (updated
November, 2008). Available on the Internet at
Irving, P.M., e.d. 1991. Acid Deposition: State of Science and Technology, Volume III, Terrestrial,
Materials, Health, and Visibility Effects, The U.S. National Acid Precipitation Assessment
Program, Chapter 24, page 24-76.
Joslin, J.D., Kelly, J.M., and van Miegroet, H. 1992. "Soil chemistry and nutrition of North
American spruce-fir stands: evidence for recent change. " Journal of Environmental
Quality, 21, 12-30.
6-77
-------
Kaval, P., and J. Loomis. 2003. Updated Outdoor Recreation Use Values With Emphasis On
National Park Recreation. Final Report October 2003, under Cooperative Agreement CA
1200-99-009, Project number IMDE-02-0070. Available on the Internet at
.
Kim, CW, Phipps, T, and Anselin L. 2003. "Measuring the benefits of air quality improvement: a
spatial hedonic approach." J Environ Econ Manage 45:24-39.
Koch, D., Bond, T.C., Streets D., and Unger N. (2007) Linking future aerosol radiative forcing to
shifts in source activities. Geophys. Res. Lett. 34 (L05821), AGU
(doi:10.1029/2006GL028360, 2007). Available on the Internet at
.
Koch D., Bond T.C., Streets D., Unger N., and van der Werf G.R. (2007) Global impacts of
aerosols from particular source regions and sectors. Journal of Geophysical
Research-Atmospheres 112 (D02205).
Kopp, R. J., Vaughn, W. J., Hazilla, M., Carson, R. 1985. "Implications of environmental policy for
U.S. agriculture: the case of ambient ozone standards." Journal of Environmental
Management, 20, 321-331.
Kramer, A., T. Holmes, and M. Haefel. 2003. Contingent valuation of forest ecosystem
protection. Pp. 303-320 in Forests in a Market Economy. Edited by E.G. Sills and K.L. Abt.
Dordrecht, The Netherlands: Kluwer Academic Publishers.
Landers DH; Simonich SL; Jaffe DA; Geiser LH; Campbell DH; Schwindt AR; Schreck CB; Kent ML;
Hafner WD; Taylor HE; Hageman KJ; Usenko S; Ackerman LK; Schrlau JE; Rose NL; Blett
TF; Erway MM. (2008). The Fate, Transport and Ecological Impacts of Airborne
Contaminants in Western National Parks (USA). EPA/600/R-07/138. U.S. Environmental
Protection Agency, Office of Research and Development, NHEERL, Western Ecology
Division. Corvallis, Oregon.
Leeworthy, V.R., and P.C. Wiley. 2001. Current Participation Patterns in Marine Recreation.
Silver Spring, MD: U.S. Department of Commerce, National Oceanic and Atmospheric
Administration, National Ocean Service, Special Projects. November. Available on the
Internet at < http://www.naturetourismplanning.com/pdfs/NSRE_2.pdf>.
Leggett, C.G. and J.E. Neumann. 2004. Responding to SAB Council Comments on the May 2003
Draft Analytical Plan for the Section 812 Second Prospective—Visibility Benefits.
Prepared for U.S. Environmental Protection Agency OPAR. Available on the Internet at
.
6-78
-------
Leistritz, F. L, D. A. Bangsund, and N. M. Hodur. 2004. "Assessing the Economic Impact of
Invasive Weeds: The Case of Leafy Spurge (Euphorbia esula)." Weed Technology
18:1392-1395.
Leitch, J. A, F. L Leistritz, and D. A Bangsund. 1996. "Economic effect of leafy spurge in the
upper Great Plains: methods, models, and results." Impact Assessment 14: 419-434.
Lipton, D. W. 1999. "Pfiesteria's economic impact on seafood industry sales and recreational
fishing, p. 35-38." In B. L. Gardner and L. Koch (ed.), Proceedings of the Conference,
Economics of Policy Options for Nutrient Management and Pfiesteria. Center for
Agricultural and Natural Resource Policy, University of Maryland, College Park.
Loehman, E.T., D. Boldt, K. Chaikin. 1985. Measuring the Benefits of Air Quality Improvements
in the San Francisco Bay Area. From Methods Development for Environmental Control
Benefits Assessment, Volume IV. Prepared for the U.S. Environmental Protection
Agency, Office of Policy, Planning and Evaluation, September. Grant #R805059-01-0
Available on the Internet at
.
Loehman, E.T., S. Park, and D. Boldt. 1994. "Willingness to Pay for Gains and Losses in Visibility
and Health." Land Economics 70(4): 478-498.
Malm, WC. 1999. Introduction to visibility. Colorado State University, Fort Collins, CO, USA,
1983, Revised edition, 1999. Available on the Internet at
.
McBride, J.R., Miller, P.R., Laven, R.D. 1985. "Effects of oxidant air pollutants on forest
succession in the mixed conifer forest type of southern California." In: Air Pollutants
Effects on Forest Ecosystems, Symposium Proceedings, St. P, 1985, p. 157-167.
McClelland, G., W. Schulze, D. Waldman, J. Irwin, D. Schenk, T. Stewart, L. Deck and M. Thayer.
1993. Valuing Eastern Visibility: A Field Test of the Contingent Valuation Method.
Prepared for U.S. Environmental Protection Agency, Office of Policy, Planning and
Evaluation. September. Available on the Internet at
.
Mclntyre, J.W., Barr, J.F. 1997. "Common Loon (Gavia immer)" in: Pool A, Gill F (eds.) The Birds
of North America. Academy of Natural Sciences, Philadelphia, PA, 313
Mclntrye, J.W., and Evers, D.C.,(eds). 2000. "Loons: old history and new finding." Proceedings of
a Symposium from the 1997 meeting, American Ornithologists' Union. North American
Loon Fund, 15 August 1997, Holderness, NH, USA.
6-79
-------
Millennium Ecosystem Assessment Board (MEA). 2005. Ecosystems and Human Well-being:
Synthesis. Washington, DC: World Resources Institute. Available on the Internet at
.
Miller, P.R., O.C. Taylor, R.G. Wilhour. 1982. Oxidant air pollution effects on a western
coniferous forest ecosystem. Corvallis, OR: U.S. Environmental Protection Agency,
Environmental Research Laboratory (EPA600-D-82-276).
Mitro, Matthew G., Evers, David C., Meyer, Michael W., and Piper, Walter H. 2008. "Common
loon survival rates and mercury in New England and Wisconsin." Journal of Wildlife
Management 72(3): 665-673.
Munthe, J., Bodaly, R. A., Branfireun, B. A., Driscoll, C. T., Gilmour, C. C, Harris, R., et al. (2007).
"Recovery of Mercury-Contaminated Fisheries." Environmental Science & Technology
36(1), 33-44.
National Agricultural Statistics Service (NASS). 2008. Maple Syrup—June 12, 2008: Maple Syrup
Production Up 30 Percent Nationwide. U.S. Department of Agriculture, National
Agricultural Statistics Service, New England Agricultural Statistics, Concord, NH.
National Association of State Foresters (NASF). 2009. Quadrennial Fire Review
2009.Washington, DC: NASF. Quadrennial Fire and Fuel Review Final Report 2009.
National Wildfire Coordinating Group Executive Board. January.
National Oceanic and Atmospheric Administration (NOAA). 2007. Annual Commercial Landing
Statistics. August. Available on the Internet at .
National Park Service (NPS). 2008. Statistical Abstract: 2008. Department of Interior, National
Park Service Social Science Program, Public Use Statistics Office, Denver, Colorado.
Available on the Internet at
.
Neumann, J.E. and R.W. Patterson. 2009. Alternative Approach to Estimating Monetized
Benefits of Residential Visibility for the Section 812 Second Prospective. Prepared for U.S.
Environmental Protection Agency OAR/OPAR. Available on the Internet at
-------
Niklinska M., Laskowski R., Maryanski M. (1998). "Effect of heavy metals and storage time on
two types of forest litter: basal respiration rate and exchangeable metals"
Ecotoxicological Environmental Safety, 41, 8-18.
Office of Management and Budget (OMB). 2003. Circular A-4: Regulatory Analysis. Washington,
DC. Available on the internet at .
Painter, et al. 2007. "Impact of disturbed desert soils on duration of mountain snow cover."
Geophys. Res. Lett. 34, L12502.
Parrish ZD; White JC; Isleyen M; Gent MPN; lannucci-Berger W; Eitzer BD; Kelsey JW; Mattina
Ml. 2006. "Accumulation of weathered polycyclic aromatic hydrocarbons (PAHs) by
plant and earthworm species." Chemosphere, 64: 609-618.
Patterson, R.W., J.E. Neumann, and C.G. Leggett. 2005. Recommended Residential Visibility
Values for the Section 812 Second Prospective Analysis. Prepared for U.S. Environmental
Protection Agency OPAR. Available on the Internet at
.
Pryor, S.C. 1996. "Assessing Public Perception of Visibility for Standard Setting Exercises."
Atmospheric Environment 30, no. 15, pp. 2705-2716.
Pye, J.M. 1988. "Impact of ozone on the growth and yield of trees: A review." Journal of
Environmental Quality, 17, 347-360.
Qian Y., Gustafson W.I., Leung L.R., and Ghan S.J. 2009. "Effects of soot-induced snow albedo
change on snowpack and hydrological cycle in western United States based on Weather
Research Forecasting Chemistry and Regional Climate Simulations." Journal of
Geophysical Research 114 (D02108).
Rae, D.A. 1983. Benefits of Visual Air Quality in Cincinnati Results of a Contingent Ranking
Survey. Prepared for Electric Power Research Institute. May.
6-81
-------
Ramanathan V. and Carmichael G. (2008) Global and regional climate changes due to black
carbon Nature Geoscience 1, 221-227.
Rea, A. W., Lindberg, S. E., Scherbatskoy, T., & Keeler, G. J. 2002. "Mercury Accumulation in
Foliage over Time in Two Northern Mixed-Hardwood Forests." Water, Air, & Soil
Pollution 133(1), 49-67.
Rimmer, C. C, McFarland, K. P., Evers, D. C, Miller, E. K., Aubry, Y., Busby, D., et al. 2005.
"Mercury Concentrations in Bicknell's Thrush and Other Insectivorous Passerines in
Montane Forests of Northeastern North America." Ecotoxicology 14(1), 223-240.
Scheuhammer, Anton M., Meyer Michael W., Sandheinrich, Mark B., and Murray, Michael W.
2007. "Effects of environmental methylmercury on the health of wild birds, mammals,
and fish," Ambio Vol.36, No.l.
Sepulveda, Maria S., Frederick, Peter C., Spalding, Marilyn G., and Williams, Gary E. Jr. 1999.
"Mercury contamination in free-ranging great egret nestlings (Ardea albus) from
southern Florida, USA," Environmental Toxicology and Chemistry Vol. 18, No.5.
Shindell, Drew (2007) Local and remote contributions to Arctic warming. Geophysical Research
Letters, Vol 34, L14704, doi:10.1029/2007GL030221.
Simcik M.F., Eisenreich S.J., Lioy P.J. 1999. "Source apportionment and source/sink relationship
of PAHs in the coastal atmosphere of Chicago and Lake Michigan."/4tmosp/7er/c
Environment, 33, 5071-5079.
Simcik M.F., Eisenreich, S.J., Golden K.A., et al. 1996. "Atmospheric Loading of Polycyclic
Aromatic Hydrocarbons to Lake Michigan as Recorded in the Sediments." Environmental
Science and Technology, 30, 3039-3046.
Sisler, J.F. 1996. Spatial and seasonal patterns and long-term variability of the composition of
the haze in the United States: an analysis of data from the IMPROVE network. CIRA
Report, ISSN 0737-5352-32, Colorado State University.
Smith, G., Coulston, J., Jepsen, E., Prichard, T. 2003. "A national ozone biomonitoring program-
results from field surveys of ozone sensitive plants in Northeastern forests (1994-
2000)." Environmental Monitoring and Assessment, 87, 271-291.
Smith, V.K., G. Van Houtven, and S.K. Pattanayak. 2002. "Benefit Transfer via Preference
Calibration." Land Economics 78:132-152.
Spencer, D.M., and D.F. Holecek. 2007. "Basic characteristics of the fall tourism market."
Tourism Management 28:491-504.
Taylor R. 1994. "Deterministic versus stochastic evaluation of the aggregate economic effects of
price support programs." Agricultural Systems 44: 461-473.
6-82
-------
Tingey, D.T., and Taylor, G.E. 1982. "Variation in plant response to ozone: a conceptual model
of physiological events." In M.H. Unsworth & D.P. Omrod (Eds.), Effects of Gaseous Air
Pollution in Agriculture and Horticulture, (pp.113-138). London, UK: Butterworth
Scientific.
Tolley, G., A. Randall, G. Blomquist, M. Brien, R. Fabian, G. Fishelson, A. Frankel, M. Grenchik, J.
Hoehn, A. Kelly, R. Krumm, E. Mensah, and T. Smith. 1984. Establishing and Valuing the
Effects of Improved Visibility in Eastern United States. Prepared for U.S. Environmental
Protection Agency, Office of Policy, Planning and Evaluation. March. U.S. Environmental
Protection Agency Grant #807768-01-0. Available on the Internet at
.
United Nations Environmental Programme, World Meteorological Organization and Stockholm
Environment Institute (UNEP). 2011. Integrated Assessment of Black Carbon and
Tropospheric Ozone Summary for Decision Makers. Available on the Internet at
.
U.S. Census Bureau. 2010. "Sector 56: EC075612: Administrative Support and Waste
Management and Remediation Services: Industry Series: Preliminary Comparative
Statistics for the United States (2002 NAICS Basis): 2007 and 2002." American FactFinder
Available on the Internet at http://factfinder.census.gov (accessed August 25, 2010)
U.S. Department of the Interior (DOI). 2007. 2006 National Survey of Fishing, Hunting, and
Wildlife-Associated Recreation. U.S. Department of the Interior, Fish and Wildlife
Service, and U.S. Department of Commerce, U.S. Census Bureau. Washington, DC.
Available on the Internet at .
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 1999. The Clean
Air Act Amendments (CAAA) Section 812 Prospective Study of Costs and Benefits (1999):
Advisory by the Advisory Council on Clean Air Compliance Analysis: Costs and Benefits of
the CAAA. EPA-SAB-COUNCIL-ADV-00-002. October. Available on the Internet at
.
6-83
-------
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004. Review of
the Draft Analytical Plan for EPA's Second Prospective Analysis—Benefits and Costs of
the Clean Air Act, 1990-2020: An Advisory by the Advisory Council for Clean Air
Compliance Analysis. EPA-SAB-COUNCIL-ADV-04-004. May. Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2007. Clean Air
Scientific Advisory Committee's (CASAC) Review of the Agency's Final Ozone Staff Paper.
EPA-CASAC-07-002. Available on the Internet at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2010. Advisory
on a Preliminary Draft of the Second Section 812 Prospective Study of the Benefits and
Costs of the Clean Air Act (April 2010). EPA-COUNCIL-10-004. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 1997. Mercury Study Report to Congress.
Volume V: Health Effects of Mercury and Mercury Compounds. EPA-452/R-97-007. U.S.
EPA Office of Air Quality Planning and Standards, and Office of Research and
Development. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 1998. Advisory Council on Clean Air
Compliance Analysis Advisory on the Clean Air Act Amendments (CAAA) of 1990 Section
812 Prospective Study: Overview of Air Quality and Emissions Estimates: Modeling,
Health and Ecological Valuation Issues Initial Studies. EPA-SAB-COUNCIL-ADV-98-003.
Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 1999a. Regional Haze Regulations, July 1,
1999. (64 FR 35714, July 1, 1999).
U.S. Environmental Protection Agency (U.S. EPA). 1999b. The Benefits and Costs of the Clean Air
Act, 1990-2010. Prepared for U.S. Congress by U.S. EPA, Office of Air and Radiation,
Office of Policy Analysis and Review, Washington, DC, November; EPA report no.
EPA410-R-99-001. Available on the Internet at .
6-84
-------
U.S. Environmental Protection Agency (U.S. EPA). 2004. Air Quality Criteria for Pa rticu late
Matter Volume II of II. National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC EPA/600/P-99/002bF. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2005. Regulatory Impact Analysis of the Final
Clean Air Mercury Rule. Office of Air Quality Planning and Standards, Research Triangle
Park, NC, March; EPA report no. EPA-452/R-05-003. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2006a. Air Quality Criteria for Ozone and
Related Photochemical Oxidants (Final). EPA/600/R-05/004aF-cF. Washington, DC: U.S.
EPA. February. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2006b. Regulatory Impact Analysis, 2006
National Ambient Air Quality Standards for Particulate Matter. Office of Air Quality
Planning and Standards, Research Triangle Park, NC. October. Available on the Internet
at < http://www.epa.gov/ttn/ecas/regdata/RIAs/Appendix%20l-
Visibility%20Benefits%20Methodology.pdf>.
U.S. Environmental Protection Agency (U.S. EPA). 2006c. Ecological Benefits Assessment
Strategic Plan. EPA-240-R-06-001. Office of the Administrator. Washington, DC. October.
Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2007a. Guidance on the Use of Models and
Other Analyses for Demonstrating Attainment of Air Quality Goals for Ozone, PM2.s, and
Regional Haze. Office of Air Quality Planning and Standards. EPA-454/B-07-002. April.
.
U.S. Environmental Protection Agency (U.S. EPA). 2007b. Review of the National Ambient Air
Quality Standards for Ozone: Policy assessment of scientific and technical information.
Staff Paper. Office of Air Quality Planning and Standards. EPA-452/R-07-007a. July.
Available on the Internet at
-------
U.S. Environmental Protection Agency (U.S. EPA). 2008a. Regulatory Impact Analysis, 2008
National Ambient Air Quality Standards for Ground-level Ozone, Chapter 6. Office of Air
Quality Planning and Standards, Research Triangle Park, NC. March. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2008b. U.S. EPA's 2008 Report on the
Environment (Final Report). Washington, D.C., EPA/600/R-07/045F (NTIS PB2008-
112484). May. Available on the Internet at
. Updated in
December 2010, available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2008c. Integrated Science Assessment for
Oxides of Nitrogen and Sulfur—Ecological Criteria National (Final Report). National
Center for Environmental Assessment, Research Triangle Park, NC. EPA/600/R-08/139.
December. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2009a. The NOx Budget Trading Program:
2008 Environmental Results. Clean Air Markets Division. September. Available on the
Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2009b. Integrated Science Assessment for
Particulate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment—RTP Division. December. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2009c. Risk and Exposure Assessment for
Review of the Secondary National Ambient Air Quality Standards for Oxides of Nitrogen
and Oxides of Sulfur (Final Report). EPA-452/R-09-008a. Office of Air Quality Planning
and Standards, Research Triangle Park, NC. September. Available on the Internet at
.
6-86
-------
U.S. Environmental Protection Agency (U.S. EPA). 2010b. Particulate Matter Urban-Focused
Visibility Assessment—Final Document (Corrected). Office of Air Quality Planning and
Standards, Research Triangle Park, NC. July. Available on the Internet at .
U.S. Environmental Protection Agency (U.S. EPA). 2011b. The Benefits and Costs of the Clean Air
Act 1990 to 2020: EPA Report to Congress. Office of Air and Radiation, Washington, DC.
March. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2011c. Regulatory Impact Analysis for the
Federal Implementation Plans to Reduce Interstate Transport of Fine Particulate Matter
and Ozone in 27 States; Correction of SIP Approvals for 22 States. June. Available on the
Internet at:
U.S. Environmental Protection Agency (U.S. EPA). 2011d. Regulatory Impact Analysis for the
Final Mercury and Air Toxics Standards. EPA-452/R-11-011. December. Available on the
Internet at:
U.S. Fish and Wildlife Service and U.S. Census Bureau (FWS and Census). 2007. 2006 National
Survey of Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR). FHW/06-NAT.
U.S. Department of the Interior, U.S. Fish and Wildlife Service, Washington, DC, and U.S.
Department of Commerce, U.S. Census Bureau, Washington, DC.
U.S. Forest Service (USFS). 2006. Forest Inventory and Analysis National Program: Forest
Inventory Data Online. Online database. U.S. Department of Agriculture Forest Service,
Forest Inventory and Analysis, Arlington, VA. Available on the Internet at
.
Valigura, R.A., R.B. Alexander, M.S. Castro, T.P. Meyers, H.W. Paerl, P.E. Stacy, and R.E. Turner.
2001. Nitrogen Loading in Coastal Water Bodies: An Atmospheric Perspective.
Washington, DC: American Geophysical Union.
Van Sickle, J., Baker, J.P., Simonin, H.A., Baldigo, B.P., Kretser, W.A., Sharpe, W.E. 1996.
"Episodic acidification of small streams in the northeastern United States: Fish mortality
in field bioassays." Ecological Applications, 6, 408-421.
Wada, H. and Cristol, D.A. and McNabb, F.M.A. and Hopkins, W.A. 2009. "Suppressed
adrenocortical responses and thyroid hormone levels in birds near a mercury-
contaminated river." Environmental Science & Technology 43(15), 6031-6038.
6-87
-------
Wada, H., Yates, DE., Evens, DC., Taylor, RJ., Hopkins, WA. 2010. "Tissue mercury concentrations
and adrenocortical responses of female big brown bats (Eptesicus fuscus) near a
contaminated river." Ecotoxicology 19(7), 1277-1284.
Winner, W.E. 1994. "Mechanistic analysis of plant responses to air pollution." Ecological
Applications, 4(4), 651-661.
Winner, W.E., and C.J. Atkinson. 1986. "Absorption of air pollution by plants, and consequences
for growth." Trends in Ecology and Evolution 1:15-18.
Wiscombe W. and Warren S. 1980. "A model for the spectral albedo of snow. Snow containing
atmospheric aerosols." J.Atmos. Sci. 37, 2734-2274.
Yates, David E., Mayack, David T., Munney, Kenneth, Evers David C., Major, Andrew, Kaur,
Taranjit, and Taylor, Robert J. 2005. "Mercury levels in mink (Mustela vison) and river
otter (Lonra canadensis) from northeastern North America," Ecotoxicology 14, 263-274.
6-88
-------
APPENDIX 6.A
ADDITIONAL DETAILS REGARDING THE VISIBILITY BENEFITS METHODOLOGY
6.A.1 Introduction
Economic benefits may result from two broad categories of changes in light extinction:
(1) changes in "residential" visibility—i.e., the visibility in and around the locations where
people live; and (2) changes in "recreational" visibility at Class I areas—i.e., visibility at Class I
national parks and wilderness areas.1 In this analysis, only those recreational and residential
benefits in areas that have been directly studied in the valuation literature are included in the
primary presentation of benefits; recreational benefits in other U.S. Class I regions and
residential benefits in other metropolitan areas are presented as sensitivity analyses of visibility
benefits.
In Chapter 6 of this RIA, we provide an overview of the visibility benefits methodology
and results. This appendix provides additional detail regarding specific aspects of the visibility
benefits methodology and is organized as follows. Section 6.A.2 describes the process we used
to convert the modeled light extinction data to match the spatial scale of the visibility benefits
assessment. We present the basic utility model in Section 6.A.3. In Section 6.A.4 we discuss the
measurement of visibility, and the mapping from environmental "bads" to environmental
"goods." In Sections 6.A.5 and 6.A.6 we summarize the methodology for estimating the
parameters of the model corresponding to visibility at in-region and out-of-region Class I areas,
and visibility in residential areas, respectively, and we describe the methods used to estimate
these parameters. Section 6.A.7 describes the process for aggregating the recreational and
residential visibility benefits. Section 6.A.8 describes the adjustment to reflect income growth
over time. Section 6.A.9 provides all the parameters used to calculate visibility benefits.
6.A.2 Converting Modeled Light Extinction Estimates
To calculate visibility benefits, we use light extinction estimates generated by the CMAQ
model.2 Modeled light extinction estimates are measured in units of inverse megameters
(Mm"1). Because the valuation studies measure visibility in terms of visual range, we convert the
light extinction units from Mm"1 to visual range (in km) for both recreational and residential
1 Hereafter referred to as Class I areas, which are defined as areas of the country such as national parks, national
wilderness areas, and national monuments that have been set aside under Section 169(a) of the Clean Air Act
to receive the most stringent degree of air quality protection. Class I federal lands fall under the jurisdiction of
three federal agencies, the National Park Service, the Fish and Wildlife Service, and the Forest Service.
2 For more information regarding the CMAQ modeling conducted for the PM NAAQS RIA, please see Chapter 3 of
this RIA.
6.A-1
-------
visibility benefits. Using the relationships derived by Pitchford and Malm (1994), the formulas
for this conversion are
Deciviews = 10 * In ( —) = 10 * In ( —
where VR denotes visual range (in kilometers) and 3ext denotes light extinction (in Mm"1).
Because we leverage the tools and data prepared for previous analyses (U.S. EPA, 2011), we use
a two-step process to convert from Mm"1 to VR using deciviews as an intermediate conversion
instead of converting directly. Therefore, the full formula incorporating the two-step
conversion is
VR = 391 * e-.*
The spatial scale of the modeled light extinction estimates must also be adjusted to
correspond with the design of the valuation studies and the underlying population and
economic data. For the residential visibility benefits analysis, we convert the spatial resolution
of the light extinction estimates from 12-km grid to county-level. We use county-level light
extinction to match the MSA boundaries, population data, and household income data. We
used the geographic centroids of each 12-km grid cell with the Veronoi Neighborhood
Averaging (VNA) interpolation method in the BenMAP model for this conversion (Abt
Associates, 2010).
For the recreational visibility benefits analysis, we use the light extinction estimates
from 12-km grid cell located at the geographic center of the Class I area. Although we
considered using the IMPROVE monitor location instead, we selected the park centroid for
three reasons:
1. Consistency with previous method for estimating recreational visibility benefits
2. Not all Class I areas have monitors, and shared monitors may be outside park
3. Siting criteria for IMPROVE monitors do not include iconic scenic vista location
6.A.3 Basic Utility Model
Within the category of recreational visibility, further distinctions have been made. There
is evidence (Chestnut and Rowe, 1990) that an individual's WTP for improvements in visibility at
a Class I area is influenced by whether it is in the region in which the individual lives, or whether
it is somewhere else. In general, people appear to be willing to pay more for visibility
6.A-2
-------
improvements at parks and wilderness areas that are "in-region" than at those that are "out-of-
region." This is plausible, because people are more likely to visit, be familiar with, and care
about parks and wilderness areas in their own part of the country.
To value estimated changes in visibility, we use an approach that is consistent with
economic theory. Below we discuss an application of the Constant Elasticity of Substitution
(CES) utility function approach3 to value both residential visibility improvements and visibility
improvements at Class I areas in the United States. This approach is based on the preference
calibration method developed by Smith, Van Houtven, and Pattanayak (2002).
We begin with a CES utility function in which a household derives utility from
1. "all consumption goods," X,
2. visibility in the residential area in which the household is located ("residential
visibility"),4
3. visibility at Class I areas in the same region as the household ("in-region recreational
visibility"), and
4. visibility at Class I areas outside the household's region ("out-of-region recreational
visibility").
We have specified a total of six recreational visibility regions,5 so there are five regions
for which any household is out of region. The utility function of a household in the nth
residential area and the ;th region of the country is:
u,,, -- ( x> + az; + r*Ql +
k=\ j*i k=\
3 The constant elasticity of substitution utility function has been chosen for use in this analysis because of its
flexibility when illustrating the degree of substitutability present in various economic relationships (in this case,
the trade-off between income and improvements in visibility).
4 We remind the reader that, although residential and recreational visibility benefits estimation is discussed
simultaneously in this section, benefits are calculated and presented separately for each visibility category.
5 See Section 6.3.4 of this RIA for a description of the different recreational visibility considered in this analysis.
6.A-3
-------
where
Zn = the level of visibility in the nth residential area;
Qik = the level of visibility at the kth in-region park (i.e., the kth park in the ith region);
Qjk = the level of visibility at the kth park in the jth region (for which the household is
out of region), jVi;
Nj = the number of Class I areas in the ith region;
Nj = the number of Class I areas in the jth region (for which the household is out of
region), jVi; and
0, the y's and 6's are parameters of the utility function corresponding to the visibility
levels at residential areas, and at in-region and out-of-region Class I areas,
respectively.
In particular, the yik's are the parameters corresponding to visibility at in-region Class I
areas; the Si's are the parameters corresponding to visibility at Class I areas in region 1
(California), if i^l; the 62's are the parameters corresponding to visibility at Class I areas in
region 2 (Colorado Plateau), if i^2, and so forth. Because the model assumes that the
relationship between residential visibility and utility is the same everywhere, there is only one
0. The parameter p in this CES utility function is an important determinant of the slope of the
marginal WTP curve associated with any of the environmental quality variables. When p=l, the
marginal WTP curve is horizontal. When p
-------
Given estimates of p, 0, the v's and the 6's, the household's utility function and the
corresponding WTP functions are fully specified. The household's WTP for any set of changes in
the levels of visibility at in-region Class I areas, out-of-region Class I areas, and the household's
residential area can be shown to be:
rlk(Q0Plk ~
The household's WTP for a single visibility improvement will depend on its order in the
series of visibility improvements the household is valuing. If it is the first visibility improvement
to be valued, the household's WTP for it follows directly from the previous equation. For
example, the household's WTP for an improvement in visibility at the first in-region park, from
Qn = Qoii to Qn = Qm, is
if this is the first (or only) visibility change the household values.
6.A.4 Measure of Visibility: Environmental "Goods" Versus "Bads"
In the above model, Qand Z are environmental "goods." As the level of visibility
increases, utility increases. The utility function and the corresponding WTP function both have
reasonable properties. The first derivative of the indirect utility function with respect to Q (or Z)
is positive; the second derivative is negative. WTP for a change from Q0 to a higher (improved)
level of visibility, Qi, is therefore a concave function of Qi, with decreasing marginal WTP.
The measure of visibility that is currently preferred by air quality scientists is the
deciview, which increases as visibility decreases. Deciview, in effect, is a measure of the lack of
visibility. As deciviews increase, visibility, and therefore utility, decreases. The deciview, then, is
a measure of an environmental "bad." There are many examples of environmental "bads"— all
types of pollution are environmental "bads." Utility decreases, for example, as the
concentration of particulate matter in the atmosphere increases.
One way to value decreases in environmental bads is to consider the "goods" with
which they are associated, and to incorporate those goods into the utility function. In
particular, if B denotes an environmental "bad," such that:
6.A-5
-------
and the environmental "good," Q, is a function of B,
=
then the environmental "bad" can be related to utility via the corresponding environmental
"good":6
V= V(m,Q)= V(m,F(By) .
The relationship between Q and B, F(B), is an empirical relationship that must be estimated.
There is a potential problem with this approach, however. If the function relating B and
Q is not the same everywhere (i.e., if for a given value of B, the value of Q depends on other
factors as well), then there can be more than one value of the environmental good
corresponding to any given value of the environmental bad, and it is not clear which value to
use. This has been identified as a problem with translating deciviews (an environmental "bad")
into visual range (an environmental "good"). It has been noted that, for a given deciview value,
there can be many different visual ranges, depending on the other factors that affect visual
range— such as light angle and altitude. We note here, however, that this problem is not unique
to visibility, but is a general problem when trying to translate environmental "bads" into
"goods."7
In order to translate deciviews (a "bad") into visual range (a "good"), we use a
relationship derived by Pitchford and Malm (1994) in which
397
DV= 10 * ln( - ,) ,
' VR'
where DV denotes deciview and VR denotes visual range (in kilometers). Solving for VR as a
function of DV yields
6 There may be more than one "good" related to a given environmental "bad." To simplify the discussion, however,
we assume only a single "good."
7 Another example of an environmental "bad" is particulate matter air pollution (PM). The relationship between
survival probability (Q) and the ambient PM level is generally taken to be of the form
2=1- OK . where V denotes the mortality rate (or level) when there is no ambient PM (i.e., when
PM=0). However, a is implicitly a function of all the factors other than PM that affect mortality. As these factors
change (e.g., from one location to another), awill change (just as visual range changes as light angle changes). It
is therefore possible to have many values of Q corresponding to a given value of PM, as the values of V vary.
6.A-6
-------
VR= 391*e-°JDV .
This conversion is based on specific assumptions characterizing the "average" conditions
of those factors, such as light angle, that affect visual range. To the extent that specific
locations depart from the average conditions, the relationship will be an imperfect
approximation.8
6.A.5 Estimating the Parameters for Visibility at Class I Areas: the y's and 6's
As noted in Section 6.A.3, if we consider a particular visibility change as the first or the
only visibility change valued by the household, the household's WTP for that change in visibility
can be calculated, given income (m), the "shape" parameter, p, and the corresponding
recreational visibility parameter. For example, a Southeast household's WTP for a change in
visibility at in-region parks (collectively) from Qi = Qoi to Qi = Qn is:
WTP(DQJ = m- [mr + gl(Q'01 - Qrn)]1/r
if this is the first (or only) visibility change the household values.
Alternatively, if we have estimates of m as well as WTPi"1 and WTPi°ut of in-region and
out-of-region households, respectively, for a given change in visibility from Q0i to Qn in
Southeast parks, we can solve for YI and 61 as a function of our estimates of m, WTPi"1 and
WTPi°ut, for any given value of p. Generalizing, we can derive the values of y and 6 for the jth
region as follows:
(m- WTPY - mp
and
(m -WTP°ut)p-mp
Chestnut and Rowe (1990) and Chestnut (1997) estimated WTP (per household) for
specific visibility changes at national parks in three regions of the United States— both for
households that are in-region (in the same region as the park) and for households that are out-
' Ideally, we would want the location-, time-, and meteorological condition-specific relationships between
deciviews and visual range, which could be applied as appropriate. This is probably not feasible, however.
6.A-7
-------
of-region. The Chestnut and Rowe study asked study subjects what they would be willing to pay
for each of three visibility improvements in the national parks in a given region. Study subjects
were shown a map of the region, with dots indicating the locations of the parks in question. The
WTP questions referred to the three visibility improvements in all the parks collectively; the
survey did not ask subjects' WTP for these improvements in specific parks individually.
Responses were categorized according to whether the respondents lived in the same region as
the parks in question ("in-region" respondents) or in a different region ("out-of-region"
respondents). The areas for which in-region and out-of-region WTP estimates are available
from Chestnut and Rowe (1990), and the sources of benefits transfer-based estimates that we
employ in the absence of estimates, are summarized in Table 6.A-1. In all cases, WTP refers to
WTP per household.
Table 6.A-1. Available Information on WTP for Visibility Improvements in National Parks
Region of Household
Region of Park In Region3 Out of Regionb
1. California WTP estimate from study WTP estimate from study
2. Colorado Plateau WTP estimate from study WTP estimate from study
3. Southeast United States WTP estimate from study WTP estimate from study
4. Northwest United States (based on benefits transfer from California)
5. Northern Rockies (based on benefits transfer from Colorado Plateau)
6. Rest of United States (based on benefits transfer from Southeast U.S.)
a In-region" WTP is WTP for a visibility improvement in a park in the same region as that in which the household is
located. For example, in-region WTP in the "Southeast" row is the estimate of the average Southeast
household's WTP for a visibility improvement in a Southeast park.
b Out-of-region" WTP is WTP for a visibility improvement in a park that is not in the same region in which the
household is located. For example, out-of-region WTP in the "Southeast" row is the estimate of WTP for a
visibility improvement in a park in the Southeast by a household outside of the Southeast.
In the primary calculation of visibility benefits for this analysis, only visibility changes at
parks within visibility regions for which a WTP estimate was available from Chestnut and Rowe
(1990) are considered (for both in- and out-of-region benefits). Primary estimates will not
include visibility benefits calculated by transferring WTP values to visibility changes at parks not
included in the Chestnut and Rowe study. Transferred benefits at parks located outside of the
Chestnut and Rowe visibility regions will, however, be included as an alternative calculation.
6.A-8
-------
The values of the parameters in a household's utility function will depend on where the
household is located. The region-specific parameters associated with visibility at Class I areas
(that is, all parameters except the residential visibility parameter) are arrayed in Table 6.A-2.
The parameters in columns 1 through 3 can be directly estimated using WTP estimates from
Chestnut and Rowe (1990) (the columns labeled "Region 1," "Region 2," and "Region 3").
Table 6.A-2. Summary of Region-Specific Recreational Visibility Parameters to be Estimated in
Household Utility Functions
Region of Household
Region 1
Region 2
Region 3
Region 4
Region 5
Region 6
Region 1
via
6i
61
61
61
6!
Region
Region 2 Region 3
62 63
V2 63
62 Vs
62 63
62 63
62 63
of Park
Region 4
64
64
64
V4
64
64
Region 5
65
65
65
65
Vs
65
Region 6
66
66
66
66
66
Ve
a The parameters arrayed in this table are region-specific rather than park-specific or wilderness area-specific. For
example, 6! is the parameter associated with visibility at "Class I areas in region 1" for a household in any region
other than region 1. The benefits analysis must derive Class I area-specific parameters (e.g., 6lk, for the kth Class I
area in the first region).
For the three regions covered in Chestnut and Rowe (1990a) (California, the Colorado
Plateau, and the Southeast United States), we can directly use the in-region WTP estimates
from the study to estimate the parameters in the utility functions corresponding to visibility at
in-region parks (YI); similarly, we can directly use the out-of-region WTP estimates from the
study to estimate the parameters for out-of-region parks (61). For the other three regions not
covered in the study, however, we must rely on benefits transfer to estimate the necessary
parameters.
While Chestnut and Rowe (1990) provide useful information on households' WTP for
visibility improvements in national parks, there are several significant gaps remaining between
the information provided in that study and the information necessary for the benefits analysis.
First, as noted above, the WTP responses were not park specific, but only region specific.
Because visibility improvements vary from one park in a region to another, the benefits analysis
must value park-specific visibility changes. Second, not all Class I areas in each of the three
regions considered in the study were included on the maps shown to study subjects. Because
6.A-9
-------
the focus of the study was primarily national parks, most Class I wilderness areas were not
included. Third, only three regions of the United States were included, leaving the three
remaining regions without direct WTP estimates.
In addition, Chestnut and Rowe (1990) elicited WTP responses for three different
visibility changes, rather than a single change. In theory, if the CES utility function accurately
describes household preferences, and if all households in a region have the same preference
structure, then households' three WTP responses corresponding to the three different visibility
changes should all produce the same value of the associated recreational visibility parameter,
given a value of p and an income, m. In practice, of course, this is not the case.
In addressing these issues, we take a three-phase approach:
1. We estimate region-specific parameters for the region in the modeled domain
covered by Chestnut and Rowe (1990a) (California, the Colorado Plateau, and the
Southeast)—YI, y2, and y3 and 61, 62, and 63.
2. We infer region-specific parameters for those regions not covered by the Chestnut
and Rowe study (the Northwest United States, the Northern Rockies, and the rest of
the U.S.)—v4, Y5, and Ye and 64, 65, and 66.
3. We derive park- and wilderness area-specific parameters within each region (YIR and
6ik, for k=l, ..., NI; Y2k and 62i<, for k=l, ..., N2; and so forth).
The question that must be addressed in the first phase is how to estimate a single
region-specific in-region parameter and a single region-specific out-of-region parameter for
each of the three regions covered in Chestnut and Rowe (1990) from study respondents' WTPs
for three different visibility changes in each region. All parks in a region are treated collectively
as if they were a single "regional park" in this first phase. In the second phase, we infer region-
specific recreational visibility parameters for regions not covered in the Chestnut and Rowe
study (the Northwest United States, the Northern Rockies, and the rest of the United States). As
in the first phase, we ignore the necessity to derive park-specific parameters at this phase.
Finally, in the third phase, we derive park- and wilderness area-specific parameters for each
region.
6.A.5.1 Estimating Region-Specific Recreational Visibility Parameters for the Region Covered
in the Chestnut and Rowe Study (Regions 1, 2, and 3)
Given a value of p and estimates of m and in-region and out-of-region WTPs for a
change from Q0 to Qi in a given region, the in-region parameter, Y, and the out-of-region
parameter, 6, for that region can be solved for. Chestnut and Rowe (1990), however,
6.A-10
-------
considered not just one, but three visibility changes in each region, each of which results in a
different calibrated y and a different calibrated 6, even though in theory all the y's should be
the same and similarly, all the 6's should be the same. For each region, however, we must have
only a single y and a single 6.
Denoting Y j as our estimate of y for the jth region, based on all three visibility changes,
we chose y j to best predict the three WTPs observed in the study for the three visibility
improvements in the jth region. First, we calculated YJI, i=l, 2, 3, corresponding to each of the
three visibility improvements considered in the study. Then, using a grid search method
beginning at the average of the three's f ji, we chose to minimize the sum of the squared
differences between the WTPs we predict using YJ and the three region-specific WTPs
observed in the study. That is, we selected to minimize:
;=1
where WTPy and WTPjjQ are the observed and the predicted WTPs for a change in visibility in
the jth region from Q0 = Qoi to Qi= On, i=l, ..., 3. An analogous procedure was used to select an
optimal 6, for each of the three regions in the Chestnut and Rowe study.
6.A.5.2 Inferring Region-Specific Recreational Visibility Parameters for Regions Not Covered
in the Chestnut and Rowe Study (Regions 4, 5, and 6)
One possible approach to estimating region-specific parameters for regions not covered
by Chestnut and Rowe (1990a) (y4, y5, and y6 and 64, 65, and 66) is to simply assume that
households' utility functions are the same everywhere, and that the environmental goods being
valued are the same—e.g., that a change in visibility at national parks in California is the same
environmental good to a Californian as a change in visibility at national parks in Minnesota is to
a Minnesotan.
For example, to estimate 64 in the utility function of a California household,
corresponding to visibility at national parks in the Northwest United States, we might assume
that out-of-region WTP for a given visibility change at national parks in the Northwest United
States is the same as out-of-region WTP for the same visibility change at national parks in
California (income held constant). Suppose, for example, that we have an estimated mean WTP
of out-of-region households for a visibility change from Q0i to On at national parks in California
(region 1), denoted WTPi°ut. Suppose the mean income of the out-of-region subjects in the
6.A-11
-------
study was m. We might assume that, for the same change in visibility at national parks in the
Northwest United States, WTP4out = WTPi°ut among out-of-region individuals with income m.
We could then derive the value of 64, given a value of p as follows:
(m- WTP°utY -mp
64 ~ (?£ - Qu
where Q04 = Qoi and Qi4 = Qn, (i.e., where it is the same visibility change in parks in region 4
that was valued at parks in the region 1).
This benefits transfer method assumes that (1) all households have the same preference
structures and (2) what is being valued in the Northwest United States (by a California
household) is the same as what is being valued in the California (by all out-of-region
households). While we cannot know the extent to which the first assumption approximates
reality, the second assumption is clearly problematic. National parks in one region are likely to
differ from national parks in another region in both quality and quantity (i.e., number of parks).
One statistic that is likely to reflect both the quality and quantity of national parks in a
region is the average annual visitation rate to the parks in that region. A reasonable way to
gauge the extent to which out-of-region people would be willing to pay for visibility changes in
parks in the Northwest United States versus in California might be to compare visitation rates in
the two regions.9 Suppose, for example, that twice as many visitor-days are spent in California
parks per year as in parks in the Northwest United States per year. This could be an indication
that the parks in California are in some way more desirable than those in the Northwest United
States and/or that there are more of them—i.e., that the environmental goods being valued in
the two regions ("visibility at national parks") are not the same.
A preferable way to estimate 64, then, might be to assume the following relationship:
WTPout n
VY -L -L ^ /J^
out
WFP
(income held constant), where ni = the average annual number of visitor-days to California
parks and n4 = the average annual number of visitor-days to parks in the Northwest United
States. This implies that
9 We acknowledge that reliance on visitation rates does not get at nonuse value.
6.A-12
-------
for the same change in visibility in region 4 parks among out-of-region individuals with income
m. If, for example, ni = 2n4, WTP4out would be half of WTPi°ut. The interpretation would be the
following: California national parks have twice as many visitor-days per year as national parks in
the Northwest United States; therefore they must be twice as desirable/plentiful; therefore,
out-of-region people would be willing to pay twice as much for visibility changes in California
parks as in parks in the Northwest United States; therefore a Californian would be willing to pay
only half as much for a visibility change in national parks in the Northwest United States as an
out-of-region individual would be willing to pay for the same visibility change in national parks
in California. This adjustment, then, is based on the premise that the environmental goods
being valued (by people out of region) are not the same in all regions.
The parameter 64 is estimated as shown above, using this adjusted WTP4out. The same
procedure is used to estimate 65 and 66. We estimate y4; y, and y6 in an analogous way, using
the in-region WTP estimates from the transfer regions, e.g.,
WTP'" = ^-^WTP'" .
ni
6.A.5.3 Estimating Park- and Wilderness Area-Specific Parameters
As noted above, Chestnut and Rowe (1990) estimated WTP for a region's national parks
collectively, rather than providing park-specific WTP estimates. The E's and E's are the
the parameters that would be in household utility functions if there were only a single park in
each region, or if the many parks in a region were effectively indistinguishable from one
another. Also noted above is the fact that the Chestnut and Rowe study did not include all Class
I areas in the regions it covered, focusing primarily on national parks rather than wilderness
areas. Most Class I wilderness areas were not represented on the maps shown to study
subjects. In California, for example, there are 31 Class I areas, including 6 national parks and 25
wilderness areas. The Chestnut and Rowe study map of California included only 10 of these
Class I areas, including all 6 of the national parks. It is unclear whether subjects had in mind "all
parks and wilderness areas" when they offered their WTPs for visibility improvements, or
whether they had in mind the specific number of (mostly) parks that were shown on the maps.
The derivation of park- and wilderness area-specific parameters depends on this.
6.A-13
-------
6.A.5.4 Derivation of Region-Specific WTP for National Parks and Wilderness Areas
If study subjects were lumping all Class I areas together in their minds when giving their
WTP responses, then it would be reasonable to allocate that WTP among the specific parks and
wilderness areas in the region to derive park- and wilderness area-specific v's and 6's for the
region. If, on the other hand, study subjects were thinking only of the (mostly) parks shown on
the map when they gave their WTP response, then there are two possible approaches that
could be taken. One approach assumes that households would be willing to pay some
additional amount for the same visibility improvement in additional Class I areas that were not
shown, and that this additional amount can be estimated using the same benefits transfer
approach used to estimate region-specific WTPs in regions not covered by Chestnut and Rowe
(1990a).
However, even if we believe that households would be willing to pay some additional
amount for the same visibility improvement in additional Class I areas that were not shown, it is
open to question whether this additional amount can be estimated using benefits transfer
methods. A third possibility, then, is to simply omit wilderness areas from the benefits analysis.
For this analysis we calculate visibility benefits assuming that study subjects lumped all Class I
areas together when stating their WTP, even if these Class I areas were not present on the map.
6.A.5.5 Derivation of Park- and Wilderness Area-Specific WTPs, Given Region-Specific WTPs
for National Parks and Wilderness Areas
The first step in deriving park- and wilderness area-specific parameters is the estimation
of park- and wilderness area-specific WTPs. To derive park and wilderness area-specific WTPs,
we apportion the region-specific WTP to the specific Class I areas in the region according to
each area's share of the region's visitor-days. For example, if WTPi"1 and WTPi°ut denote the
mean household WTPs in the Chestnut and Rowe (1990) study among respondents who were
in-region-1 and out-of-region-1, respectively, nik denotes the annual average number of visitor-
days to the kth Class I area in California, and ni denotes the annual average number of visitor-
days to all Class I areas in California (that are included in the benefits analysis), then we assume
that
* WTP™ ,
and
6.A-14
-------
Yl
k
lk
Using WTPjin and WTPj°ut, either from the Chestnut and Rowe study (for j = 1, 2, and 3) or
derived by the benefits transfer method (for j = 4, 5, and 6), the same method is used to derive
Class I area-specific WTPs in each of the six regions.
While this is not a perfect allocation scheme, it is a reasonable scheme, given the
limitations of data. Visitors to national parks in the United States are not all from the United
States, and certainly not all from the region in which the park is located. A very large proportion
of the visitors to Yosemite National Park in California, for example, may come from outside the
United States. The above allocation scheme implicitly assumes that the relative frequencies of
visits to the parks in a region/rom everyone in the world is a reasonable index of the relative
WTP of an average household in that region (WTPj'n) or out of that region (but in the United
States) (WTPj°ut) for visibility improvements at these parks.10
A possible problem with this allocation scheme is that the relative frequency of visits is
an indicator of use value but not necessarily of nonuse value, which may be a substantial
component of the household's total WTP for a visibility improvement at Class I areas. If park A
is twice as popular (i.e., has twice as many visitors per year) as park B, this does not necessarily
imply that a household's WTP for an improvement in visibility at park A is twice its WTP for the
same improvement at park B. Although an allocation scheme based on relative visitation
frequencies has some obvious problems, however, it is still probably the best way to allocate a
collective WTP.
6.A.5.6 Derivation of Park- and Wilderness Area-Specific Parameters, Given Park- and
Wilderness-Specific WTP
Once the Class I area-specific WTPs have been estimated, we could derive the park- and
wilderness area-specific y's and 6's using the method used to derive region-specific y's and 6's.
Recall that that method involved (1) calibrating y and 6 to each of the three visibility
improvements in the Chestnut and Rowe study (producing three y's and three 6's),
(2) averaging the three y's and averaging the three 6's, and finally, (3) using these average y and
6 as starting points for a grid search to find the optimal y and the optimal 6— i.e., the y and 6
10 This might be thought of as two assumptions: (1) that the relative frequencies of visits to the parks in a region
from everyone in the world is a reasonable representation of the relative frequency of visits from people in the
United States—i.e., that the parks that are most popular (receive the most visitors per year) in general are also
the most popular among Americans; and (2) that the relative frequency with which Americans visit each of their
parks is a good index of their relative WTPs for visibility improvements at these parks.
6.A-15
-------
that would allow us to reproduce, as closely as possible, the three in-region and three out-of-
region WTPs in the study for the three visibility changes being valued.
Going through this procedure for each national park and each wilderness area
separately would be very time consuming, however. We therefore used a simpler approach,
which produces very close approximations to the y's and 6's produced using the above
approach. If:
WTPj"1 = the in-region WTP for the change in visibility from Qo to Qi in the jth region;
1 = the in-region WTP for the same visibility change (from Q0 to Qi) in the kth
Class I area in the jth region (= Sjk*WTPj'n, where Sjk is the kth area's share of
visitor-days in the jth region);
income;
the optimal value of v for the jth region; and
Yjk = the value of Yjk calibrated to WTPjk"1 and the change from Q0 to Qi;
then11:
(m- WTPfY - mp
7j *
and
(m- WTPmkY - m
r,k =
(<2op-0n
which implies that:
ajk *
where:
(m- WTPmkY - mp
a, =
Jk - - p
(m- WTPfY - m
11 Vj* is only approximately equal to the right-hand side because, although it is the optimal value designed to
reproduce as closely as possible all three of the WTPs corresponding to the three visibility changes in the
Chestnut and Rowe study, Vj* will n°t exactly reproduce any of these WTPs.
6.A-16
-------
We use the adjustment factor, ajk, to derive YJR from YJ*, for the kth Class I area in the jth
region. We use an analogous procedure to derive 6jk from 6j* for the kth Class I area in the jth
region (where, in this case, we use WTPj°ut and WTPjkout instead of WTPjin and WTPjkin).12
6.A.6 Estimating the Parameter for Visibility in Residential Areas: 6
In previous assessments, EPA used a study on residential visibility valuation conducted
in 1990 (McClelland et al., 1993). Consistent with advice from EPA's Science Advisory Board
(SAB), EPA designated the McClelland et al. study as significantly less reliable for regulatory
benefit-cost analysis, although it does provide useful estimates on the order of magnitude of
residential visibility benefits (U.S. EPA-SAB, 1999). 13 In order to estimate residential visibility
benefits in this analysis, we have replaced the previous methodology with a new benefits
transfer approach and incorporated additional valuation studies. This new approach was
developed for The Benefits and Costs of the Clean Air Act 1990 to 2020: EPA Report to Congress
(U.S. EPA, 2011) and reviewed by the SAB (U. S. EPA-SAB, 2010). To value residential visibility
improvements, the new approach draws upon information from the Brookshire et al. (1979),
Loehman et al. (1985) and Tolley et al. (1984) studies.14 These studies provide primary visibility
values for Atlanta, Boston, Chicago, Denver, Los Angeles, Mobile, San Francisco, and
Washington D.C.15
The estimation of 0 is a simpler procedure for residential visibility benefits, involving a
straightforward calibration using the study income and WTP for each study city:
p
(m - WTPy - m
^ '
12 This method uses a single in-region WTP and a single out-of-region WTP per region. Although the choice of WTP
will affect the resulting adjustment factors (the ajk's) and therefore the resulting vjk's and 6jk's, the effect is
negligible. We confirmed this by using each of the three in-region WTPs in California and comparing the
resulting three sets of Vjk's ar|d 6jk's, which were different from each other by about one one-hundredth of a
percent.
13 EPA's Advisory Council on Clean Air Compliance Analysis noted that the McClelland et al. (1993) study may not
incorporate two potentially important adjustments. First, their study does not account for the "warm glow"
effect, in which respondents may provide higher willingness to pay estimates simply because they favor "good
causes" such as environmental improvement. Second, while the study accounts for non-response bias, it may
not employ the best available methods. As a result of these concerns, the Council recommended that
residential visibility be omitted from the overall primary benefits estimate. (U.S. EPA-SAB, 1999)
14 Loehman et al. (1985) and Brookshire et al. (1979) were subsequently published in peer-reviewed journals (see
Loehman et al. (1994) and Brookshire et al. (1982). The Tolley et al. (1984) work was not published, but was
subject to peer review during study development.
15 Recognizing potential fundamental issues associated with data collected in Cincinnati and Miami (e.g., see
Chestnut et al. (1986) and Chestnut and Rowe (1990c), we do not include values for these cities in our analysis.
6.A-17
-------
where:
m = household income,
p = shape parameter (0.1),
0 = WTP parameter corresponding to the visibility at MSA,
Z0 = starting visibility, and
Zi = visibility after change.
Where studies provide multiple estimates for visual range improvements for a single
study city, we estimate one 0 as the simple average of the 0 calculated for each set of visual
range improvements.
6.A.7 Putting It All Together: The Household Utility and WTP Functions
Given an estimate of 0, derived as shown in Section 6.A.4, and estimates of the y's and
6's, derived as shown in Section 6.A.3, based on an assumed or estimated value of p, the utility
and WTP functions for a household in any region are fully specified. We could therefore
estimate the value to that household of visibility changes from any baseline level to any
alternative level in the household's residential area and/or at any or all of the Class I areas in
the United States, in a way that is consistent with economic theory. In particular, the WTP of a
household in the ith region and the nth residential area for any set of changes in the levels of
visibility at in-region Class I areas, out-of-region Class I areas, and the household's residential
area is:
-QfJ
The national benefits associated with any suite of visibility changes would be calculated
as the sum of these household WTPs for those changes. The benefit of any subset of visibility
changes (e.g., changes in visibility only at Class I areas in California) can be calculated by setting
all the other components of the WTP function to zero (that is, by assuming that all other
visibility changes that are not of interest are zero). This is effectively the same as assuming that
the subset of visibility changes of interest is the first or the only set of changes being valued by
households. Estimating benefit components in this way will yield slightly upward biased
estimates of benefits, because disposable income, m, is not being reduced by the WTPs for any
prior visibility improvements. That is, each visibility improvement (e.g., visibility at Class I areas
in the California) is assumed to be the first, and they cannot all be the first. The upward bias
6.A-18
-------
should be extremely small, however, because all of the WTPs for visibility changes are very
small relative to income.
Although we recognize that the approach described above is most consistent with
economic theory, we have chosen to not use this function with income constraints on overall
WTP. Instead, we simply add the total preference calibrated recreational visibility benefits to
the preference-calibrated residential visibility benefits. Again, because all of the WTPs for
visibility changes are very small relative to income, the upward bias should be extremely small.
6.A.8 Income Elasticity and Income Growth Adjustment for Visibility Benefits
Growth in real income over time is an important component of benefits analysis.
Economic theory argues that WTP for most goods (such as environmental protection) will
increase if real incomes increase. There is substantial empirical evidence that the income
elasticity16 of WTP for health risk reductions is positive, although there is uncertainty about its
exact value. Thus, as real income increases, the WTP for environmental improvements also
increases. Although many analyses assume that the income elasticity of WTP is unit elastic (i.e.,
a 10% higher real income level implies a 10% higher WTP to reduce risk changes), empirical
evidence suggests that income elasticity is substantially less than one and thus relatively
inelastic. As real income rises, the WTP value also rises but at a slower rate than real income.
The effects of real income changes on WTP estimates can influence benefits estimates
in two different ways: through real income growth between the year a WTP study was
conducted and the year for which benefits are estimated, and through differences in income
between study populations and the affected populations at a particular time. Empirical
evidence of the effect of real income on WTP gathered to date is based on studies examining
the former. The Environmental Economics Advisory Committee (EEAC) of the Science Advisory
Board (SAB) advised EPA to adjust WTP for increases in real income over time but not to adjust
WTP to account for cross-sectional income differences "because of the sensitivity of making
such distinctions, and because of insufficient evidence available at present" (U.S. EPA-SAB,
2000a). A recent advisory by another committee associated with the SAB, the Advisory Council
on Clean Air Compliance Analysis, has provided conflicting advice. While agreeing with "the
general principle that the willingness to pay to reduce mortality risks is likely to increase with
growth in real income (U.S. EPA-SAB, 2004)" and that "The same increase should be assumed
for the WTP for serious nonfatal health effects (U.S. EPA-SAB, 2004)," they note that "given the
16 Income elasticity is a common economic measure equal to the percentage change in WTP for a 1% change in
income.
6.A-19
-------
limitations and uncertainties in the available empirical evidence, the Council does not support
the use of the proposed adjustments for aggregate income growth as part of the primary
analysis (U.S. EPA-SAB, 2004)." Until these conflicting advisories have been reconciled, EPA will
continue to adjust valuation estimates to reflect income growth using the methods described
below, while providing sensitivity analyses for alternative income growth adjustment factors.
We assume that the WTP for improved visibility would increase with growth in real
income. The relative magnitude of the income elasticity of WTP for visibility compared with
those for health effects suggests that visibility is not as much of a necessity as health, thus, WTP
is more elastic with respect to income.
Details of the general procedure to account for projected growth in real U.S. income
between 1990 and 2020 can be found in Kleckner and Neumann (1999). Specifically, we use the
elasticity for visibility benefits provided in Chestnut (1997).
In addition to elasticity estimates, projections of real gross domestic product (GDP) and
populations from 1990 to 2020 are needed to adjust benefits to reflect real per capita income
growth. We used projections of real GDP provided in Kleckner and Neumann (1999) for the
years 1990 to 2010.17 We used projections of real GDP provided by Standard and Poor's (2000)
for the years 2010 to 2020.18 Visibility benefits are adjusted by multiplying the unadjusted
benefits by the appropriate adjustment factor.
6.A.9 Summary of Parameters
In Tables 6.A-3 through 6.A-6, we provide the parameters used to calculate recreational
and residential visibility benefits.
17 U.S. Bureau of Economic Analysis, Table 2A (available at http://www.bea.doc.gov/bea/dn/0897nip2/ tab2a.htm.
and U.S. Bureau of Economic Analysis, Economics and Budget Outlook. Note that projections for 2007 to 2010
are based on average GDP growth rates between 1999 and 2007.
18 In previous analyses, we used the Standard and Poor's projections of GDP directly. This led to an apparent
discontinuity in the adjustment factors between 2010 and 2011. We refined the method by applying the
relative growth rates for GDP derived from the Standard and Poor's projections to the 2010 projected GDP
based on the Bureau of Economic Analysis projections.
6.A-20
-------
Table 6.A-3. Mean Annual Household WTP for Changes in Visual Range for Recreational
Visibility (1990$)a
Region
California
Southwest
Southeast
WTP
In-region
$66.41
$80.19
$71.42
$50.12
$72.67
$61.40
$66.41
$82.70
$75.18
WTP
Out-of-region
$43.85
$53.88
$51.37
$45.11
$55.13
$48.87
$35.08
$53.88
$47.61
Starting Visual
Range (miles)
90
90
45
155
155
115
25
25
10
Ending Visual
Range (miles)
125
150
90
200
250
155
50
75
25
Study Household
Income
$48,759
$48,759
$48,759
a Based on Chestnut (1997) and adjusted for study sample income and currency year
Table 6.A-4. Region-Specific Parameters for Recreational Visibility Benefits3
Region
Optimal
Optimal 6
California
Southwest
Southeast
0.00517633
0.006402706
0.003552379
0.003629603
0.005092572
0.002163346
Northwest
Northern Rockies
Rest of U.S.
0.001172669
0.005263445
0.001211215
0.000823398
0.004176339
0.000738149
' Calculated using methodology described in sections 6.A.3 through 6.A.4
6.A-21
-------
Table 6.A-5. Mean Annual Household WTP for Changes in Visual Range for Residential
Visibility
City
Atlanta
(Tolley et al., 1984)
Boston
(Tolley et al., 1984)
Chicago
(Tolley et al., 1984)
Denver
(Tolley et al., 1984)
Los Angeles
(Brookshire et al.,
1979)
Mobile
(Tolley et al., 1984)
San Francisco
(Loehman et al., 1985)
Washington, DC
(Tolley et al., 1984)
WTP in
Original
Year's $
$188
$281
$82
$119
$139
$171
$202
$269
$121
$144
$115
$154
$43
$116
$71
$168
$197
$71
$238
$303
Starting
Visual
Range
(miles)
12
12
12
12
18
18
9
9
10
10
50
50
2
2
12
10
10
16.3
15
15
Ending
Visual
Range
(miles)
22
32
22
32
28
38
18
30
20
30
60
70
12
28
28
20
30
18.6
25
35
Study
Household
Income
$19,900a
$19,900a
$27,600d
$27,600d
$25,000a
$25,000a
$30,000b
$30,000b
$29,400d
$29,400d
$32,000°
$32,000°
$15,200d
$15,200d
$15,200d
$20,200a
$20,200a
$26,100°
$27,500a
$27,500a
Year of
Original
Data
1982
1982
1984
1984
1982
1982
1981
1981
1984
1984
1984
1984
1978
1978
1978
1982
1982
1980
1982
1982
0ifp = 0.1
(1990$,
1990
income)
0.033446
0.031661
0.010738
0.009417
0.026636
0.019049
0.022313
0.016696
0.013180
0.009732
0.038558
0.027803
0.003866
0.006716
0.011702
0.026078
0.018882
0.045307
0.036866
0.027804
0ifp =
0.1
(Simple
Average)
0.021316
0.022843
0.015480
0.033181
0.007428
0.022480
0.045307
0.032335
a See Chestnut et al. (1986), pages 5-5 through 5-10.
b See Tolley et al., (1984), page 127.
° See Loehman et al. (1985), page 38.
d Historical median income data by MSA from U.S. Census (1990).
6.A-22
-------
Table 6.A-6. Parameters for Income Growth Adjustment for Visibility Benefits
Adjustment Step Parameter Estimate
Central Estimate of Elasticitya 0.90
Adjustment Factor Used to Account for Projected Real Income Growth in 2020b 1.517
a Derivation of estimates can be found in Kleckner and Neumann (1999) and Chestnut (1997).
b Based on elasticity values reported in Table 5-3, U.S. Census population projections, and projections of real GDP
per capita.
6.A.10 References
Abt Associates, Inc. 2010. Environmental Benefits and Mapping Program (Version 4.0).
Bethesda, MD. Prepared for U.S. Environmental Protection Agency Office of Air Quality
Planning and Standards. Research Triangle Park, NC. Available on the Internet at
.
Brookshire, D.S., R.C. d'Arge, W.D. Schulze and M.A. Thayer. 1979. Methods Development for
Assessing Tradeoffs in Environmental Management, Vol. II: Experiments in Valuing Non-
Market Goods: A Case Study of Alternative Benefit Measures of Air Pollution Control in
the South Coast Air Basin of Southern California. Prepared for the U.S. Environmental
Protection Agency, Office of Research and Development. Available on the Internet at
.
Chestnut, L.G., and R.D. Rowe. 1990. "A New National Park Visibility Value Estimates." In
Visibility and Fine Particles, Transactions of an AWMA/EPA International Specialty
Conference, C.V. Mathai, ed. Air and Waste Management Association, Pittsburgh.
Chestnut, L.G., R.D. Rowe and J. Murdoch. 1986. Review of 'Establishing and Valuing the Effects
of Improved Visibility in Eastern United States.' Prepared for the U.S. Environmental
Protection Agency. October. Available on the Internet at
.
6.A-23
-------
Kleckner, N., and J. Neumann. 1999. Recommended Approach to Adjusting WTP Estimates to
Reflect Changes in Real Income. Memorandum to Jim Democker, U.S. EPA/OPAR. June.
Available on the Internet at .
Loehman, E.T., D. Boldt, K. Chaikin. 1985. Measuring the Benefits of Air Quality Improvements
in the San Francisco Bay Area. From Methods Development for Environmental Control
Benefits Assessment, Volume IV. Prepared for the U.S. Environmental Protection
Agency, Office of Policy, Planning and Evaluation, September. Grant #R805059-01-0
Available on the Internet at .
Loehman, E.T., S. Park, and D. Boldt. 1994. "Willingness to Pay for Gains and Losses in Visibility
and Health." Land Economics 70(4): 478-498.
McClelland, G., W. Schulze, D. Waldman, J. Irwin, D. Schenk, T. Stewart, L. Deck and M. Thayer.
1993. Valuing Eastern Visibility: A Field Test of the Contingent Valuation Method.
Prepared for U.S. Environmental Protection Agency, Office of Policy, Planning and
Evaluation. September. Available on the Internet at
.
Pitchford, M.L., and W.C. Malm. 1994. "Development and Applications of a Standard Visual
Index." Atmospheric Environment 28(5):1049-1054.
Sisler, J.F. 1996. Spatial and Seasonal Patterns and Long Term Variability of the Composition of
the Haze in the United States: An Analysis of Data from the IMPROVE Network. Colorado
State University, Cooperative Institute for Research in the Atmosphere (CIRA), ISSN
0737-5352-32. Fort Collins, CO. July. Available on the Internet at .
Smith, V.K., G. Van Houtven, and S.K. Pattanayak. 2002. "Benefit Transfer via Preference
Calibration." Land Economics 78:132-152.
Tolley, G., A. Randall, G. Blomquist, M. Brien, R. Fabian, G. Fishelson, A. Frankel, M. Grenchik, J.
Hoehn, A. Kelly, R. Krumm, E. Mensah, and T. Smith. 1984. Establishing and Valuing the
Effects of Improved Visibility in Eastern United States. Prepared for U.S. Environmental
Protection Agency, Office of Policy, Planning and Evaluation. March. U.S. Environmental
Protection Agency Grant #807768-01-0. Available on the Internet at
.
6.A-24
-------
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 1999. The Clean
Air Act Amendments (CAAA) Section 812 Prospective Study of Costs and Benefits (1999):
Advisory by the Advisory Council on Clean Air Compliance Analysis: Costs and Benefits of
the CAAA. EPA-SAB-COUNCIL-ADV-00-002. October. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2011b. The Benefits and Costs of the Clean Air
Act 1990 to 2020: EPA Report to Congress. Office of Air and Radiation, Washington, DC.
March. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA). 2000. Guidelines for Preparing Economic
Analyses. EPA 240-R-00-003. National Center for Environmental Economics, Office of
Policy Economics and Innovation. Washington, DC. September. Available on the Internet
at .
6.A-25
-------
CHAPTER 7
ENGINEERING COST ANALYSIS
7.1 Synopsis
This chapter summarizes the data sources and methodology used to estimate the
engineering costs of attaining the alternative, more stringent levels for the PM2.5 primary
standard analyzed in this RIA. This chapter provides the estimates of the engineering costs of
alternative annual standards of 13, 12, and 11 u.g/m3 in conjunction with retaining the 24-hour
standard of 35 u.g/m3, as well as one of the alternative, more stringent annual standards
(11 u.g/m3) in conjunction with an alternative, more stringent 24-hour standard of 30 u.g/m3
(referred to as 13/35,12/35, 11/35, and 11/30). This chapter also presents engineering cost
estimates for the control strategies outlined in Chapter 4. The cost discussion for known
controls in Section 7.2.2 is followed by a presentation of estimates for the engineering costs of
the additional (beyond known controls) tons of emission reductions that are needed to move to
full attainment of the alternative standards analyzed; this estimation approach, discussed in
Section 7.2.3, is referred to as extrapolated costs.
The engineering costs described in this chapter generally include the costs of
purchasing, installing, operating, and maintaining the referenced technologies. For a variety of
reasons, actual control costs may vary from the estimates EPA presents. As discussed
throughout this document, the technologies and control strategies selected for analysis are
illustrative of one way in which nonattainment areas could meet a revised standard. There are
numerous ways to construct and evaluate potential control programs that would bring areas
into attainment with alternative standards, and EPA anticipates that state and local
governments will consider programs that are best suited for local conditions. Furthermore,
based on past experience, EPA believes that it is reasonable to anticipate that the marginal cost
of control will decline over time due to technological improvements and more widespread
adoption of previously considered niche control technologies.1 Also, EPA recognizes the
extrapolated portion of the engineering cost estimates reflects substantial uncertainty about
which sectors, and which technologies, might become available for cost-effective application in
the future.
The engineering cost estimates are limited in their scope. This analysis focuses on the
emission reductions needed for attainment of a range of alternative revised standards, not
1 See Chapter 4, Section 4.3 for additional discussion of uncertainties associated with predicting technological
advancements that may occur between now and 2020.
7-1
-------
implementation of a final revised standard. EPA understands that some states will incur costs
both designing State Implementation Plans (SIPs) for and implementing new control strategies
to meet the revised standard. However, EPA does not know what specific actions states will
take to design their SIPs to meet the revised standards, therefore we do not present estimated
costs that government agencies may incur for managing the requirement, implementing these
(or other) control strategies, or for offering incentives that may be necessary to encourage the
implementation of specific technologies, especially for technologies that are not necessarily
market driven. This analysis does not assume specific control measures that would be required
in order to implement these technologies on a regional or local level.
7.2 PM2.s Engineering Costs
7.2.1 Data and Methods—Identified Control Costs (non-EGU Point and Area Sources)
After designing the hypothetical control strategy using the methodology discussed in
Chapter 4, EPA used the Control Strategy Tool2 (CoST) to estimate engineering control costs for
non-EGU point and area sources.3 CoST calculates engineering costs using one of three
different methods: (1) by multiplying an average annualized cost-per-ton estimate against the
total tons of a pollutant reduced to derive a total cost estimate; (2) by calculating cost using an
equation that incorporates key plant information; or (3) by using both cost-per-ton and cost
equations. Most control cost information within CoST was developed based on the cost-per-ton
approach because estimating engineering costs using an equation requires more data, and
parameters used in other non-cost-per-ton methods may not be readily available or broadly
representative across sources within the emissions inventory. The costing equations used in
CoST require either plant capacity or stack flow to determine annual, capital and/or operating
and maintenance (O&M) costs. Capital costs are converted to annual costs using the capital
recovery factor (CRF).4 Where possible, cost calculations are used to calculate total annual
control cost (TACC), which is a function of capital costs (CC) and O&M costs. The CRF
incorporates the interest rate and equipment life (in years) of the control equipment. Operating
costs are calculated as a function of annual O&M and other variable costs. The resulting TACC
equation is TACC = (CRF * CC) + O&M.
2 The Control Strategy Tool recently underwent peer review by an ad hoc panel of experts. Responses to the peer
review are currently under development and will be available by final promulgation of this rule.
3 Area sources are not necessarily non-urban sources.
4 The capital recovery factor formula is expressed as [r*(l+r)An/(l+r)An -1]. Where r is the real rate of interest and
n is the number of time periods. For more information on this cost methodology and the CoST, please refer to
the documentation at http://www.epa.gov/ttn/CoST, the EPA Air Pollution Control Cost Manual found at
http://epa.gov/ttn/catc/products.htmltfcccinfo, and EPA's Guidelines for Preparing Economic
Analyses, Chapter 6 found at http://yosemite.epa.gov/ee/epa/eed.nsf/webpages/Guidelines.htmlftdownload.
7-2
-------
Engineering costs will differ based upon quantity of emissions reduced, plant capacity,
or stack flow, which can vary by the emissions inventory year. Engineering costs will also differ
in nominal terms by the year the costs are calculated for (i.e., 1999$ versus 2006$).5 For capital
investment, in order to attain standards in 2020 we assume capital investment occurs at the
beginning of 2020. We make this simplifying assumption because we do not know what all firms
making capital investments will do and when they will do it. For 2020, our estimate of
annualized costs includes annualized capital and O&M costs for those controls included in our
identified control strategy analysis. Our engineering cost analysis uses the equivalent uniform
annual costs (EUAC) method, in which annualized costs are calculated based on the equipment
life for the control measure along with the interest rate incorporated into the CRF. Annualized
costs represent an equal stream of yearly costs over the period the control technology is
expected to operate. We make no presumption of additional capital investment in years
beyond 2020. The EUAC method is discussed in detail in the EPA Air Pollution Control Cost
Manual.6 Applied controls and their respective engineering costs are provided in the PM NAAQS
docket.
7.2.2 Identified Control Costs
In this section, we provide engineering cost estimates for the control strategies
identified in Chapter 4 that include control technologies on non-EGU point sources and area
sources. Engineering costs generally refer to the capital equipment expense, the site
preparation costs for the application, and annual operating and maintenance costs. Note that
the application of these control strategies results in some, but not all, geographic areas
reaching attainment for the alternative PM2.s standards.
Because we obtain control cost data from many sources, we are not always able to
obtain consistent data across original data sources.7 If disaggregated control cost data is
unavailable (i.e., where capital, equipment life value, and O&M costs are not separated out),
EPA typically assumes that the estimated control costs are annualized using a 7 percent
discount rate. When disaggregated control cost data is available (i.e., where capital, equipment
life value, and O&M costs are explicit) we can recalculate costs using a 3 percent discount rate.
For non-EGU point source controls, some disaggregated data is available and we were able to
calculate costs at both 3 and 7 percent discount rates for that control cost data. For the 12/35,
5 The engineering costs will not be any different in real (inflation-adjusted) terms if calculated in 2006 versus other-
year dollars, if the other-year dollars are properly adjusted. For this analysis, all costs are reported in real 2006
dollars.
http://epa.gov/ttn/catc/products.htmlffcccinfo
7 Data sources can include states and technical studies, which do not typically include the original data source.
7-3
-------
11/35, and 11/30 alternative standards approximately 31 percent, 33 percent, and 29 percent,
respectively, of known control costs are disaggregated at a level that could be discounted at 3
percent. Because we do not have disaggregated control cost data for any area source controls,
total annualized costs are assumed to be calculated using a 7 percent discount rate. See Table
7-1 for a summary of sectors, control costs, and discount rates used.
Table 7-1. Summary of Sectors, Control Costs, and Discount Rates for Known Control Costs
(millions of 2006$)a
Alternative
Standard
13/35
12/35
11/35
11/30
Emissions Sector
non-EGU Point Sources
Area Sources
Total
non-EGU Point Sources
Area Sources
Total
non-EGU Point Sources
Area Sources
Total
non-EGU Point Sources
Area Sources
Total
Known Control Costs (Millions of 2006$)
7 Percent
Discount Rateb
—
—
—
$0.098
$0.210
$0.31
$25
$28
$53
$46
$54
$100
Partial Control 3 Percent Discount Rate
Cost at 3 Percent (3 Percent & 7 Percent
Discount Rate c Discount Rates Combined) d
—
—
—
$0.057
—
$0.057
$16
—
$16
$25
—
$25
—
—
—
$0.061
$0.210
$0.27
$24
$28
$52
$42
$54
$96
All estimates rounded to two significant figures.
bAII non-EGU point source costs and all area source costs are included in this column and are assumed to be
calculated at a 7 percent discount rate.
c This column includes the non-EGU point source costs that we were able to calculate at a 3 percent rate and no
area source costs. The non-EGU point source costs calculated at a 3 percent rate are those for which we have
disaggregated control cost data.
d Our known control costs discounted at a 3 percent rate are a combination of area source costs and non-EGU
point source costs discounted at a 7 percent rate only and non-EGU point source costs discounted at 3 percent
when disaggregated control cost data is available.
The total annualized cost of control in each sector in the control scenario is summarized
by region in Table 7-2. Table 7-2 includes annualized control costs to allow for comparison
across regions and between costs and benefits. These numbers reflect the engineering costs
annualized at discount rates of 3 percent and 7 percent, consistent with the guidance provided
in the Office of Management and Budget's (OMB) (2003) Circular A-4. However, it is important
to note that it is not possible to estimate both 3 percent and 7 percent discount rates for each
individual facility.
7-4
-------
In this RIA, non-EGU point sources were the only sources with available data to perform
a sensitivity analysis of our annualized control costs to the choice of interest rate. As such, the 3
percent column in Table 7-2 reflects the sum of some non-EGU point sources at a 3 percent
discount rate, some non-EGU point sources at a 7 percent discount rate, and the area sources
at a 7 percent discount rate. The 3 percent column represents a slight overestimate because
some non-EGU point sources and area sources are included using a 7 percent discount rate.8
With the exception of the 3 percent Total Annualized Cost estimate in Table 7-2, engineering
cost estimates presented throughout this and subsequent chapters are based on a 7 percent
discount rate.
Table 7-2. Partial Attainment Known Control Costs in 2020 for Alternative Standards
Analyzed3 (millions of 2006$)a
Alternative
Standard
13/35°
12/35
11/35
11/30
Estimates are
Region
East
West
California
Total
East
West
California0
Total
East
West
California
Total
East
West
California
Total
rounded to two significant fij
3%b
—
—
—
—
$0.061
$0.21
—
$0.27
$46
$3.0
$3.0
$52
$46
$31
$19
$96
;ures, as such numbers may
Known Controls
7%
—
—
—
—
$0.098
$0.21
—
$0.31
$48
$3.0
$3.0
$53
$48
$33
$19
$100
not sum down columns.
' In these analyses, the discount rates refer to the rate at which capital costs are annualized. A higher discount, or
interest, rate results in a larger annualized cost of capital estimate.
7-5
-------
b Because we obtain control cost data from many sources, we are not always able to obtain consistent data across
original data sources. Where disaggregated control cost data is available (i.e., where capital, equipment life
value, and O&M costs are explicit) we can calculate costs using a 3 percent discount rate. Therefore the cost
estimate provided here is a summation of costs at 3 percent and 7 percent discount rates.
c All known controls were applied in the baseline.
The total annualized engineering costs associated with the application of known
controls, incremental to the baseline and using a 7 percent discount rate, are approximately
$310,000 for an alternative annual standard of 12/35, $53 million for an 11/35 alternative
standard, and $100 million for an 11/30 alternative standard. In addition, it is important to note
there is no partial attainment known control costs for a 13/35 alternative standard.9
7.2.3 Extrapolated Costs
This section presents the methodology and results of the extrapolated engineering cost
calculations for attainment of alternative PM2.5 standards of 13/35, 12/35,11/35, and 11/30. All
costs presented for the illustrative control strategies are calculated incrementally from the
current PM2.5 standard of 15/35, therefore, any additional emission reductions needed to attain
the current 24-hour standard of 35 u.g/m3 are part of the baseline analysis and not presented
here. Note that the extrapolated costs don't account for cost differences between reducing
additional tons by emissions source sector.
As mentioned earlier in this chapter, the application of the modeled control strategy
was not successful in reaching nationwide attainment for these alternative PM2.5 standards.
Because some areas remained in nonattainment, the engineering costs detailed in Section 7.2.2
represent the costs of partial attainment for PM2.5 standards of 13/35,12/35, 11/35, and 11/30.
For each alternative standard and geographic area that cannot reach attainment with known
controls, we estimated, in a least-cost way, the additional emission reductions needed for PM2.5
and its precursors, NOx and S02, to attain the standard. To generate estimates of the costs and
benefits of meeting alternative standards, EPA has assumed the application of unspecified
future controls that make possible the emission reductions needed for attainment in 2020. By
definition, no cost data currently exists for unidentified future technologies or innovative
strategies. EPA used two methodologies for estimating the costs of unspecified future controls:
a fixed-cost methodology and a hybrid methodology. The fixed-cost methodology is more
straight forward and transparent than the hybrid methodology. In the hybrid methodology, the
coefficient for the X2 term can be difficult to estimate and if it is zero, the functional form
9 Only one county (Riverside County, CA) exceeded the 13/35 alternative standard. All known controls were used in
the baseline analysis for this county; therefore there are no known control costs for this standard. Total costs
for 13/35 are represented by extrapolated costs only—see Table 7.3 below.
7-6
-------
becomes the same as the fixed-cost methodology's functional form. Additional discussion of
the functional form associated with the hybrid methodology is in Appendix 7.A. Both
approaches assume that innovative strategies and new control options make possible the
emissions reductions needed for attainment by 2020. The fixed-cost methodology uses a
$15,000/ton estimate for each ton of PM2.5, S02, and NOX reduced, and the hybrid approach is
similar to the hybrid approach used for the 2008 Ozone NAAQS RIA cost analysis. The
$15,000/ton amount is commensurate with that used in the 1997 Ozone Regulatory Impact
Analysis and is consistent with what an advisory committee to the Section 812 second
prospective analysis on the Clean Air Act Amendments suggested. In Section 7.A.2.1 we conduct
sensitivity analysis on the fixed-cost estimate of $15,000/ton. The fixed-cost methodology was
preferred by EPA's Science Advisory Board over two other options, including a marginal-cost-
based approach.
"When assigning costs to unidentified measures, the Council suggests that a simple,
transparent method that is sensitive to the degree of uncertainty about these costs is best.
Of the three approaches outlined, assuming a fixed cost/ton appears to be the simplest
and most straightforward. Uncertainty might be represented using alternative fixed costs
per ton of emissions avoided. "
EPA requests comments or suggestions on all aspects of the methodologies for estimating the
costs of unspecified future controls to provide illustrative estimates of NAAQS costs, including
choice of functional forms of the equations, initial parameter estimates, and the initial fixed-
cost estimate of $15,000/ton.
In Appendix 4.A we include estimates of the relationship between additional emission
reductions for each pollutant and air quality improvements. In this chapter we present
estimates of the costs for each additional emission reduction for each pollutant and geographic
area. The mix of pollutants varies by area, because each area has different amounts of known
controls, different additional air quality improvements required, and different amounts of
uncontrolled emissions remaining.
Estimating engineering costs for emission reductions needed beyond those from known
controls to reach attainment in 2020 is inherently a challenging exercise. As described later in
this chapter, our experience with Clean Air Act implementation shows that technological
advances and development of innovative strategies can reduce emissions and reduce the costs
of emerging technologies over time. Technological change may provide new possibilities for
controlling emissions as well as reducing the cost of known controls through technological
7-7
-------
improvements or higher control efficiencies. EPA requests comment on the likelihood that new
technologies that control direct PM2.5 and its precursors will become available between now
and 2020.
Because three different pollutants are involved, there are many different combinations
of pollutant reductions that would result in the required air quality improvements. Early in our
analysis we decided to use the hybrid-cost methodology to estimate the costs of the emission
reductions needed from unknown controls. However, the hybrid methodology still has a
number of important uncertainties, and its reliability for extrapolating costs has not been
evaluated. While we applied the methodology to generate emission reduction estimates
needed beyond known controls for the proposed PM2.5 standards and several alternatives, the
degree of extrapolation for emissions reductions caused us to reconsider applying the hybrid
methodology to obtain estimates of extrapolated costs. Consistent with an SAB
recommendation, we use the fixed-cost per ton methodology to generate the estimates of
extrapolated costs for emission reductions needed from unknown controls. We perform
sensitivity analyses using both the alternative fixed cost per ton and the hybrid methodologies
in Appendix 7.A
As discussed in Chapter 4, we developed the emission reduction estimates for each
alternative standard using the hybrid methodology. As a result, the emissions reductions that
form the basis of the primary cost and benefit estimates may include a different mix of PM2.5
and S02 emissions reductions than may have been identified as least cost had we employed the
fixed-cost methodology to develop the emission reduction estimates. Using the fixed-cost
methodology, the less expensive pollutant, for air quality improvements, to reduce will be
selected until there are no remaining tons to reduce. Using the hybrid methodology, the less
expensive pollutant to reduce will be selected until the marginal cost to reduce the next ton
exceeds the marginal cost to reduce the next ton of an alternate pollutant. At that point, the
methodology chooses a mix of pollutants to achieve the least-cost solution. Since the cost per
ton is held constant in the fixed-cost methodology, the least-cost solution would select all
available direct PM2.5 emissions reductions before selecting S02 emissions reductions.10
Therefore, the hybrid methodology estimates PM2.5 emissions reductions lower than or equal
10 Because the marginal cost equation for each pollutant is expected to be less accurate for the very last portion of
a pollutant in an area, and it is unlikely an area would reduce all anthropogenic emissions to zero on one
pollutant prior to controlling others, we included the constraint that no more than 90% of the remaining
emissions in an area for a given pollutant can be reduced from emission reductions beyond known control
measures.
7-8
-------
to the fixed-cost methodology and S02 emission reductions higher than or equal to the fixed-
cost methodology.
Because we used the hybrid methodology in selecting emissions reductions, the total
cost estimate is higher than if we had selected needed emissions reductions using the fixed-
cost methodology. This is because the total number of tons identified and summed across
pollutants under the hybrid methodology may be higher than the total number of tons needed
under the fixed-cost methodology. For example, the hybrid methodology may choose to reduce
direct PM2.5 by 15 tons of and S02 by 8 tons, whereas the fixed-cost methodology may choose
to reduce direct PM2.5 by 20 tons to obtain the same air quality improvement for an area.
Applying the fixed cost-per-ton to the total reductions, the hybrid methodology would result in
total costs of $345,000 (23 tons * $15,000/ton), and the fixed-cost methodology would result in
total costs of $300,000 (20 tons * $15,000/ton).
Extrapolated cost estimates are provided using a 7 percent discount rate because
known control measure information is available at 7 percent for all measures applied in this
analysis. Table 7-3 provides the extrapolated cost estimates using the fixed-cost methodology
described above, using a fixed cost-per-ton of $15,000/ton. The extrapolated costs estimate is
$69 million dollars (2006$) for the 12/35 alternative standard.
7-9
-------
Table 7-3. Fixed Costs by Alternative Standard Analyzed3 (millions of 2006$)a
Alternative Standard
13/35
12/35
11/35
11/30
Region
East
West
California
Total
East
West
California
Total
East
West
California
Total
East
West
California
Total
Fixed Cost Methodology
7%
—
—
$2.9
$2.9
—
$3.3
$65
$69
$1.3
$38
$180
$220
$21
$79
$190
$290
Estimates are rounded to two significant figures.
Of note is the geographic distribution of extrapolated costs. For all of the alternative
standards, the above costs indicate that California, as possibly expected, represents a
significant portion of the extrapolated costs. For the 11/30, 11/35, 12/35, and 13/35 alternative
standards, California represents 65 percent, 82 percent, 94 percent and 100 percent,
respectively, of the extrapolated cost estimates.
7.2.4 Total Cost Estimates
Table 7-4 presents a summary of the total national costs of attaining 13/35, 12/35,
11/35, and 11/30 alternative standards in 2020. This summary includes the engineering costs
presented above from the known controls analysis, as well as the extrapolated costs. As
discussed in Section 7.2.2, costs for known controls for non-EGU point sources where capital
7-10
-------
cost and equipment life information were available were calculated at a 3 percent discount
rate. Extrapolated costs were calculated at a 7 percent discount rate only.
To calculate total cost estimates at a 3 percent discount rate and to include the
extrapolated costs in those totals, we added the known control estimates at a 3 percent
discount rate to the extrapolated costs at a 7 percent discount rate. To more clearly present the
total cost calculations for both approaches, we include column labels in Table 7-4, e.g., A, B,
and C.
The costs associated with monitoring, reporting, and record keeping for affected
sources are not included in these annualized cost estimates. Based on preliminary estimates
prepared for the upcoming PM2.5 Implementation Rule Information Collection Request (ICR),
EPA believes these costs are minor compared to the control costs.
7-11
-------
Table 7-4. Total Costs by Alternative Standard Analyzed (millions of 2006$)a
Alternative Region
Standard
13/35 East
West
California
Total
12/35 East
West
California
Total
11/35 East
West
California
Total
11/30 East
West
California
Total
Known Controls Fixed-Cost
Methodology
3%b
A
—
—
—
—
$0.061
$0.21
—
$0.27
$46
$3.0
$3
$52
$46
$31
$19
$96
7%
B
—
—
—
—
$0.098
$0.21
—
$0.31
$48
$3.0
$3
$53
$48
$33
$19
$100
7%
C
—
—
$2.9
$2.9
—
$3.3
$65
$69
$1.3
$38
$180
$220
$21
$79
$190
$290
Total Costs
Fixed-Cost Methodology
3%
(A+C)
—
—
$2.9
$2.9
$0.061
$3.6
$65
$69
$47
$41
$180
$270
$67
$110
$210
$390
7%
(B+C)
—
—
$2.9
$2.9
$0.098
$3.6
$65
$69
$49
$41
$180
$270
$69
$110
$210
$390
a Estimates are rounded to two significant figures, as such numbers may not sum down columns.
bCost Estimates are not available at 3% for all control measures. Therefore the cost estimate provided here is a
summation of costs at 3% and 7% discount rates.
7.3 Changes in Regulatory Cost Estimates over Time
Our analyses focus on controls for non-EGU and area sources. Future technology
developments in sectors not analyzed here (e.g., EGUs) may be transferrable to non-EGU and
area sources, making these sources more viable for achieving future attainment at a lower cost.
These same future technology developments may also make the sectors not analyzed here
(e.g., EGUs) more viable for achieving future attainment at a lower cost. There are many
examples in which technological innovation and "learning by doing" have made it possible to
achieve greater emission reductions than had been feasible earlier, or have reduced the costs
7-12
-------
of emission control in relation to original estimates. Studies have concluded that costs of some
EPA programs have been less than originally estimated, due in part to EPA's inability to predict
and account for future technological innovation in regulatory impact analyses.11 Technological
change will affect baseline conditions for our analysis. This change may lead to potential
improvements in the efficiency with which firms produce goods and services; for example, firms
may use less energy to produce the same quantities of output.
Constantly increasing marginal abatement costs are likely to induce the type of
innovation that would result in lower costs than estimated in this chapter. By 2020,
breakthrough technologies in control equipment could result in a downward shift in the
marginal abatement cost curve for such equipment (Figure 7-1)12 as well as a decrease in its
slope, reducing marginal costs per unit of abatement. In addition, elevated abatement costs
may result in significant increases in the cost of production and would likely induce production
efficiencies, in particular those related to energy inputs, which would lower emissions from the
production side. EPA requests comment on how marginal control costs for specific technology
applications may have changed over the past 20 years.
Cost/
Ton
MCo
MCi
9o
Slope=/ Slope=>
MC
LONG
Induced Technology Shift
S02 Reductions
Figure 7-1. Technological Innovation Reflected by Marginal Cost Shift
Harrington et al. (2000) and previous studies cited by Harrington. Harrington, W., R.D. Morgenstern, and P.
Nelson. 2000. "On the Accuracy of Regulatory Cost Estimates." Journal of Policy Analysis and Management
19(2):297-322.
Figure 7-1 shows a linear marginal abatement cost curve. It is possible that the shape of the marginal abatement
cost curve is non-linear.
7-13
-------
7.3.1 Examples of Technological Advances in Pollution Control
There are numerous examples of low-emission technologies developed and/or
commercialized over the past 15 or 20 years, such as
• Selective catalytic reduction (SCR) and ultra-low NOX burners for NOX emissions
• Scrubbers, which achieve 95% and potentially greater S02 control on boilers
• Sophisticated new valve seals and leak detection equipment for refineries and
chemical plants
• Low- or zero-VOC paints, consumer products and cleaning processes
• Chlorofluorocarbon (CFC) free air conditioners, refrigerators, and solvents
• Water- and powder-based coatings to replace petroleum-based formulations
• Vehicles are much cleaner than believed possible in the late 1980s due to
improvements in evaporative controls, catalyst design and fuel control systems for
light-duty vehicles; and treatment devices and retrofit technologies for heavy-duty
engines
• Idle-reduction technologies for engines, including truck stop electrification efforts
• Market penetration of gas-electric hybrid vehicles, and clean fuels
• The development of retrofit technology to reduce emissions from in-use vehicles
and non-road equipment
These technologies were not commercially available 2 decades ago, and some did not
even exist. Yet today, all of these technologies are on the market, and many are widely
employed. Several are key components of major pollution control programs.
What is known as "learning by doing" or "learning curve impacts," which is a concept
distinct from technological innovation, have also made it possible to achieve greater emissions
reductions than had been feasible earlier, or have reduced the costs of emission control in
relation to original estimates. Learning curve impacts can be defined generally as the extent to
which variable costs (of production and/or pollution control) decline as firms gain experience
with a specific technology. Impacts such as these would manifest themselves as a lowering of
expected costs for operation of technologies in the future below what they may have been.
7-14
-------
The magnitude of learning curve impacts on pollution control costs has been estimated
for a variety of sectors as part of the cost analyses done for the Draft Direct Cost Report for the
second EPA Section 812 Prospective Analysis of the Clean Air Act Amendments of 1990.13 In
that report, learning curve adjustments were included for those sectors and technologies for
which learning curve data was available. A typical learning curve adjustment example is to
reduce either capital or O&M costs by a certain percentage given a doubling of output from
that sector or for that technology. In other words, capital or O&M costs will be reduced by
some percentage for every doubling of output for the given sector or technology.
T.P. Wright, in 1936, was the first to characterize the relationship between increased
productivity and cumulative production. He analyzed man-hours required to assemble
successive airplane bodies. He suggested the relationship is a log linear function, since he
observed a constant linear reduction in man-hours every time the total number of airplanes
assembled was doubled. The relationship he devised between number assembled and assembly
time is called Wright's Equation (Gumerman and Marnay, 2004).14 This equation, shown below,
has been shown to be widely applicable in manufacturing:
Wright's Equation: CN = C0 * Nb, (7.2)
where:
N = cumulative production
CN = cost to produce Nth unit of capacity
Co = cost to produce the first unit
b = learning parameter = In (l-LR)/ln(2), where
LR = learning by doing rate, or cost reduction per doubling of capacity or output.
The percentage adjustments to costs can range from 5 to 20 percent, depending on the
sector and technology. Learning curve adjustments were prepared in a memo by lEc supplied to
US EPA and applied for the mobile source sector (both onroad and nonroad) and for application
13 E.H. Pechan and Associates and Industrial Economics, Direct Cost Estimates for the Clean Air Act Second Section
812 Prospective Analysis: Draft Report, prepared for U.S. EPA, Office of Air and Radiation, February 2007.
Available at http://www.epa.gov/oar/sect812/mar07/direct_cost_draft.pdf.
14 Gumerman, Etan and Marnay, Chris. Learning and Cost Reductions for Generating Technologies in the National
Energy Modeling System (NEMS), Ernest Orlando Lawrence Berkeley National Laboratory, University of
California at Berkeley, Berkeley, CA. January 2004, LBNL-52559.
7-15
-------
of various ECU control technologies within the Draft Direct Cost Report.15 Advice received from
the SAB Advisory Council on Clean Air Compliance Analysis in June 2007 indicated an interest in
expanding the treatment of learning curves to those portions of the cost analysis for which no
learning curve impact data are currently available. Examples of these sectors are non-EGU point
sources and area sources. The memo by lEc outlined various approaches by which learning
curve impacts can be addressed for those sectors. The recommended learning curve impact
adjustment for virtually every sector considered in the Draft Direct Cost Report is a 10%
reduction in O&M costs for two doublings of cumulative output, with proxies such as
cumulative fuel sales or cumulative emission reductions being used when output data was
unavailable.
For this RIA, we do not have the necessary data for cumulative output, fuel sales, or
emission reductions for all sectors included in our analysis in order to properly generate control
costs that reflect learning curve impacts. Clearly, the effect of including these impacts would be
to lower our estimates of costs for our control strategies in 2020, but we are not able to include
such an analysis in this RIA.
7.3.2 Influence on Regulatory Cost Estimates
Studies indicate that it is not uncommon for pre-regulatory cost estimates to be higher
than later estimates, in part because of an inability to predict technological advances. Over
longer time horizons, the opportunity for technical advances is greater.
7.3.2.1 Multi-Rule Study
Harrington et al. of Resources for the Future (RFF)16 conducted an analysis of the
predicted and actual costs of 28 federal and state rules, including 21 issued by EPA and the
Occupational Safety and Health Administration (OSHA), and found a tendency for predicted
costs to overstate actual implementation costs. Costs were considered accurate if they fell
within the analysis error bounds or if they fall within 25 percent (greater or less than) of the
predicted amount. They found that predicted total costs were overestimated for 14 of the 28
rules, while total costs were underestimated for only three rules. Differences can result
because of quantity differences (e.g., overestimate of pollution reductions) or differences in
per-unit costs (e.g., cost per unit of pollution reduction). Per-unit costs of regulations were
15 Industrial Economics, Inc. Proposed Approach for Expanding the Treatment of Learning Curve Impacts for the
Second Section 812 Prospective Analysis: Memorandum, prepared for U.S. EPA, Office of Air and Radiation,
August 13, 2007.
16 Harrington, W., R.D. Morgenstern, and P. Nelson. 2000. "On the Accuracy of Regulatory Cost Estimates." Journal
of Policy Analysis and Management 19(2):297-322.
7-16
-------
overestimated in 14 cases, while they were underestimated in six cases. In the case of EPA
rules, the Agency overestimated per-unit costs for five regulations, underestimated them for
four regulations (three of these were relatively small pesticide rules), and accurately estimated
them for four. Based on examination of eight economic incentive rules, "for those rules that
employed economic incentive mechanisms, overestimation of per-unit costs seems to be the
norm," the study said. It is worth noting here that the controls applied for this NAAQS do not
use an economic incentive mechanism. In addition, Harrington et al. also states that
overestimation of total costs can be due to error in the quantity of emission reductions
achieved, which would also cause the benefits to be overestimated. A 2010 update to this study
by Harrington et al. of RFF showed that EPA and other regulatory agencies tend to overestimate
the total costs of regulations; their estimates of the cost per-unit of pollution eliminated by
regulations tend to be more accurate, however. Calculations of the total cost of regulation
include not only the "unit costs" multiplied by the number of units of pollution avoided, but
also estimates of the basic adjustment process and costs of change itself. Of the rules initially
examined, 14 projected inflated total costs, while pre-regulation estimates were too low for
only three rules. These exaggerated adjustment costs are often attributable to underestimates
of the potential that technological change could minimize pollution abatement costs.17
Based on the case study results and existing literature, the authors identified
technological innovation as one of five explanations of why predicted and actual regulatory cost
estimates differ: "Most regulatory cost estimates ignore the possibility of technological
innovation ... Technical change is, after all, notoriously difficult to forecast... In numerous case
studies actual compliance costs are lower than predicted because of unanticipated use of new
technology."
It should be noted that many (though not all) of the EPA rules examined by Harrington
et al. had compliance dates of several years, which allowed a limited period for technical
innovation.
17 Harrington, W, R Morgenstern, and P Nelson. "How Accurate Are Regulatory Cost Estimates?" Resources for the
Future, March 5, 2010. Available on the Internet at
http://www.rff.org/wv/Documents/HarringtonMorgensternNelson regulatory%20estimates
.pdf.
7-17
-------
7.3.2.2 Acid Rain SO2 Trading Program
Recent cost estimates of the Acid Rain S02 trading program by RFF and MIT have been
as much as 83 percent lower than originally projected by EPA.18 As noted in the RIA for the
Clean Air Interstate Rule, the ex ante numbers in 1989 were an overestimate in part because of
the limitation of economic modeling to predict technological improvement of pollution controls
and other compliance options, such as fuel switching. The fuel switching from high-sulfur to
low-sulfur coal was spurred by a reduction in rail transportation costs due to deregulation of
rail rates during the 1990's. Harrington et al. report that scrubbing turned out to be more
efficient (95% removal vs. 80-85% removal) and more reliable (95% vs. 85% reliability) than
expected, and that unanticipated opportunities arose to blend low- and high-sulfur coal in older
boilers up to a 40/60 mixture, compared with the 5/95 mixture originally estimated.
Table 7-5. Phase 2 Cost Estimates
Phase 2 Cost Estimates
Ex ante estimates $2.7 to $6.2 billion3
Ex post estimates $1.0 to $1.4 billion
3 2010 Phase II cost estimate in 1995$.
7.3.2.3 EPA Fuel Control Rules
A 2002 study by two economists with EPA's Office of Transportation and Air Quality19
examined EPA vehicle and fuels rules and found a general pattern that "all ex ante estimates
tended to exceed actual price impacts, with the EPA estimates exceeding actual prices by the
smallest amount." The paper notes that cost is not the same as price, but suggests that a
comparison nonetheless can be instructive.20 An example focusing on fuel rules is provided in
Table 7-6.
18 Carlson, Curtis, Dallas R. Burtraw, Maureen, Cropper, and Karen L Palmer. 2000. "Sulfur Dioxide Control by
Electric Utilities: What Are the Gains from Trade?" Journal of Political Economy 108(#6):1292-1326.
Ellerman, Denny. January 2003. Ex Post Evaluation of Tradable Permits: The U.S. SO2 Cap-and-Trade Program.
Massachusetts Institute of Technology Center for Energy and Environmental Policy Research.
19 Anderson, J.F., and Sherwood, T., 2002. "Comparison of EPA and Other Estimates of Mobile Source Rule Costs to
Actual Price Changes," Office of Transportation and Air Quality, U.S. Environmental Protection Agency.
Technical Paper published by the Society of Automotive Engineers. SAE 2002-01-1980.
20 The paper notes: "Cost is not the same as price. This simple statement reflects the fact that a lot happens
between a producer's determination of manufacturing cost and its decisions about what the market will bear in
terms of price change."
7-18
-------
Table 7-6. Comparison of Inflation-Adjusted Estimated Costs and Actual Price Changes of
EPA Fuel Control Rules3
Inflation-adjusted Cost Estimates (c/gal)
EPA
DOE
API
Other
Actual Price
Changes (c/gal)
Gasoline
Phase 2 RVP Control (7.8 RVP- 1.1
Summer) (1995$)
Reformulated Gasoline Phase 1 3.1-5.1
(1997$)
1.8
0.5
3.4^1.1 8.2-14.0 7.4 (CRA)
2.2
Reformulated Gasoline Phase 2 4.6-6.8
(Summer) (2000$)
30 ppm sulfur gasoline (Tier 2) 1.7-1.9
7.6-10.2 10.8-19.4 12
2.9-3.4 2.6 5.7 (NPRA),
3.1 (AIAM)
7.2(5.1, when
corrected to Syr
MTBE price)
N/A
Diesel
500 ppm sulfur highway diesel 1.9-2.4
fuel (1997$)
15 ppm sulfur highway diesel 4.5
fuel
4.2-6.0
3.3 (NPRA)
6.2
2.2
4.2-6.1
(NPRA)
N/A
a Anderson, J.F., and Sherwood, T., 2002. "Comparison of EPA and Other Estimates of Mobile Source Rule Costs to
Actual Price Changes," Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Technical
Paper published by the Society of Automotive Engineers. SAE 2002-01-1980.
Chlorofluorocarbon (CFC) Phase-Out: EPA used a combination of regulatory, market-
based (i.e., a cap-and-trade system among manufacturers), and voluntary approaches to phase
out the most harmful ozone depleting substances. This was done more efficiently than either
EPA or industry originally anticipated. The phase out for Class I substances was implemented
4-6 years faster, included 13 more chemicals, and cost 30 percent less than was predicted at
the time the 1990 Clean Air Act Amendments were enacted.21
The Harrington et al. study states, "When the original cost analysis was performed for
the CFC phase-out it was not anticipated that the hydrofluorocarbon HFC-134a could be
22
substituted for CFC-12 in refrigeration. However, as Hammit notes, 'since 1991 most new U.S.
Holmstead, Jeffrey, 2002. "Testimony of Jeffrey Holmstead, Assistant Administrator, Office of Air and Radiation,
U.S. Environmental Protection Agency, Before the Subcommittee on Energy and air Quality of the committee
on Energy and Commerce, U.S. House of Representatives, May 1, 2002, p. 10.
Hammit, J.K. (2000). "Are the costs of proposed environmental regulations overestimated? Evidence from the
CFC phase out." Environmental and Resource Economics, 16(#3): 281-302.
7-19
-------
automobile air conditioners have contained HFC-134a (a compound for which no commercial
production technology was available in 1986) instead of CFC-12" (p.13). He cites a similar story
for HCFRC-141b and 142b, which are currently substituting for CFC-11 in important foam-
blowing applications."
Additional examples of decreasing costs of emissions controls include: SCR catalyst costs
decreasing from $llk-$14k/m3 in 1998 to $3.5k-$5k/m3 in 2004, and improved low NOX burners
reduced emissions by 50% from 1993-2003 while the associated capital cost dropped from $25-
$38/kW to $15/kW.23 Also, FGD scrubber capital costs have been estimated to have decreased
by more than 50 percent from 1976 to 2005, and the O&M costs decreased by more than 50%
from 1982 to 2005. Many process improvements contributed to lowering the capital costs,
especially improved understanding and control of process chemistry, improved materials of
construction, simplified absorber designs, and other factors that improved reliability.24
We cannot estimate the precise interplay between EPA regulation and technology
improvement, but it is clear that a priori cost estimation often results in overestimation of costs
because changes in technology (whatever the cause) make less costly control possible.
7.3.3 Influence of Regulation on Technological Change
We cannot estimate the interplay between EPA regulation and technology improvement
but have reason to believe it may be significant. There is emerging research on technology-
forcing polices (i.e., where a regulator specifies a policy standard that cannot be met with
existing technology or met with existing technology but not at an acceptable cost, and over
time market demand will provide incentives for industry to develop the appropriate
technology). This is illustrated by Gerard and Lave (2005). Therein, they demonstrate through a
careful policy history that the 1970 CAA legislated dramatic improvements in the reduction of
emissions for 1975 and 1976 automobiles. Those mandated improvements went beyond the
capabilities of existing technologies. But the regulatory pressure "pulled" forth or "forced"
catalytic converting technology in 1975.
Work for EPA by RTI in 2011 studied the relationship between patents and the S02 cap-
and-trade program. This preliminary, non-peer reviewed study seems to indicate that patents
23ICF Consulting. October 2005. The Clean Air Act Amendment: Spurring Innovation and Growth While Cleaning
the Air. Washington, DC. Available at http://www.icfi.com/Markets/Environment/doc_files/caaa-success.pdf.
24 Yeh, Sonia and Rubin, Edward. February 2007. "Incorporating Technological Learning in the Coal Utility
Environmental Cost (CD ECost) Model: Estimating the Future Cost Trends of SO2, NOx, and Mercury Control
Technologies." Prepared for ARCADIS Geraghty and Miller, Research Triangle Park, NC 27711. Available at
http://steps.ucdavis.edu/People/slyeh/syeh-resources/Drft%20Fnl%20Rpt%20Lrng%20for%20CUECost_v3.pdf.
7-20
-------
may be related to S02 regulatory actions and drops in the long-term S02 allowance price. Popp
(2003) and Keohane (2002) have both provided empirical evidence that Title IV led to induced
technological change. Popp provides evidence that since Title IV there has been technological
innovations that have improved the removal efficiency of scrubbers. Keohane provides
evidence that fossil-fuel fired electric utilities that were subject to Title IV were, for a given
increase in the cost of switching to low sulfur coal, more likely to install a scrubber.
7.4 Uncertainties and Limitations
EPA bases its estimates of emission control costs on the best available information from
available engineering studies of air pollution controls and developed a reliable modeling
framework for analyzing the cost, emission changes, and other impacts of regulatory controls.
However, our cost analysis is subject to uncertainties and limitations, which we document on a
qualitative basis in Table 7-7 below. For additional discussion of how we assess uncertainty, see
Section 5.5.7.
Table 7-7. Summary of Qualitative Uncertainty for Modeling Elements of PM Engineering
Costs
Potential Source of Uncertainty
Degree of
Direction Magnitude of Confidence
of Impact on
Potential Monetized
Bias Costs3
in Our
Analytical
Approach
Ability to
Assess
Uncertaintyc
Uncertainties Associated with Engineering Costs
Engineering Cost Estimates
• Capital recovery factor estimates (7% and 3%)
• Estimates of private compliance cost
• Increased advancement in control
technologies as well as reduction in costs over
time
• Cost estimates for PM10
Both Medium-high Medium
Tier 2
Unquantified Costs
• Costs of federal and state administration of SIP
program, as well as permitting costs.
• Transactional costs
Low
Medium
Medium
Tierl
Extrapolated Costs
Both
High
Low
Tierl
Magnitude of Impact
High—If error could influence the total costs by more than 25%
Medium—If error could influence the total costs by 5% -25%
Low—If error could influence the total costs by less than 5%
7-21
-------
b Degree of Confidence in Our Analytic Approach
High—The current evidence is plentiful and strongly supports the selected approach
Medium—Some evidence exists to support the selected approach, but data gaps are present
Low—Limited data exists to support the selected approach
c Ability to Assess Uncertainty (using WHO Uncertainty Framework)
Tier 0—Screening level, generic qualitative characterization
Tier 1—Scenario-specific qualitative characterization
Tier 2—Scenario-specific sensitivity analysis
Tier 3—Scenario-specific probabilistic assessment of individual and combined uncertainty
7.5 References
Anderson, J.F., and Sherwood, T., 2002. "Comparison of EPA and Other Estimates of Mobile
Source Rule Costs to Actual Price Changes," Office of Transportation and Air Quality,
U.S. Environmental Protection Agency. Technical Paper published by the Society of
Automotive Engineers. SAE 2002-01-1980.
Carlson, Curtis, Dallas R. Burtraw, Maureen, Cropper, and Karen L. Palmer. 2000. "Sulfur Dioxide
Control by Electric Utilities: What Are the Gains from Trade?" Journal of Political
Economy 108(#6):1292-1326.
Ellerman, Denny. January 2003. Ex Post Evaluation of Tradable Permits: The U.S. S02 Cap-and-
Trade Program. Massachusetts Institute of Technology Center for Energy and
Environmental Policy Research.
Gerard, D. and L.B. Lave (2005). "Implementing Technology-Forcing Policies: The 1970 Clean Air
Act Amendments and the Introduction of Advanced Automotive Emissions Controls in
the United States," Technological Forecasting and Social Change, 72: 761-778.
Gumerman, Etan and Marnay, Chris. Learning and Cost Reductions for Generating Technologies
in the National Energy Modeling System (NEMS), Ernest Orlando Lawrence Berkeley
National Laboratory, University of California at Berkeley, Berkeley, CA. January 2004,
LBNL-52559.
Hammit, J.K. (2000). "Are the costs of proposed environmental regulations overestimated?
Evidence from the CFC phase out." Environmental and Resource Economics, 16(#3): 281-
302.
Harrington, W., R.D. Morgenstern, and P. Nelson. 2000. "On the Accuracy of Regulatory Cost
Estimates." Journal of Policy Analysis and Management 19(2):297-322.
Harrington, W., R.D. Morgenstern, and P. Nelson. 2010. "How Accurate Are Regulatory Cost
Estimates." Available at http://www.rff.org/wv/Documents/
HarringtonMorgensternNelson regulatory%20estimates.pdf.
7-22
-------
Holmstead, Jeffrey, 2002. "Testimony of Jeffrey Holmstead, Assistant Administrator, Office of
Air and Radiation, U.S. Environmental Protection Agency, Before the Subcommittee on
Energy and air Quality of the committee on Energy and Commerce, U.S. House of
Representatives, May 1, 2002, p. 10.
ICF Consulting. 2005. The Clean Air Act Amendment: Spurring Innovation and Growth While
Cleaning the Air. Washington, DC. Available at
http://www.icfi.com/Markets/Environment/doc files/caaa-success.pdf.
Industrial Economics, Inc. Proposed Approach for Expanding the Treatment of Learning Curve
Impacts for the Second Section 812 Prospective Analysis: Memorandum, prepared for
U.S. EPA, Office of Air and Radiation, August 13, 2007.
Keohane, Nathaniel. 2002. "Environmental Policy and the Choice of Abatement Technique"
Evidence from Coal-Fired Power Plants." February. Yale School of Management.
Manson, Nelson, and Neumann. Unpublished Paper,2002. "Assessing the Impact of Progress
and Learning Curves on Clean Air Act Compliance Costs."
Office of Management and Budget (OMB). 2003. Circular A-4. Available at
http://www.wh itehouse.gov/omb/circula rs_a004_a-4/
Popp, David. 2003. "Pollution Control Innovations and the Clean Air Act of 1990." Autumn.
Journal of Policy Analysis and Management, 22(4), 641-60.
E.H. Pechan and Associates and Industrial Economics, Direct Cost Estimates for the Clean Air Act
Second Section 812 Prospective Analysis: Draft Report, prepared for U.S. EPA, Office of
Air and Radiation, February 2007. Available at
http://www.epa.gov/oar/sect812/mar07/direct cost draft.pdf.
U.S. Environmental Protection Agency (U.S. EPA). 2002. EPA Air Pollution Control Cost
Manual—Sixth Edition (EPA452/B-02-001). Office of Air Quality Planning and Standards,
Research Triangle Park, NC. Also Available at http://epa.gOV/ttn/catc/dirl/c allchs.pdf.
U.S. Environmental Protection Agency (U.S. EPA). 2006. AirControlNET 4.1 Control Measures
Documentation Report. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. Also Available at http://www.epa.gov/ttnecasl/models/
DocumentationReport.pdf.
U.S. Environmental Protection Agency. June 2007. Advisory Council on Clean Air Compliance
Analysis (COUNCIL), Council Advisory on OAR's Direct Cost Report and Uncertainty
Analysis Plan. Washington, DC.
7-23
-------
U.S. Environmental Protection Agency (U.S. EPA). 2010. Guidelines for Preparing Economic
Analysis. Office of Policy, National Center for Environmental Economics. Also Available
at http://yosemite.epa.gov/ee/epa/eed.nsf/Webpages/Guidelines.html/
$file/Guidelines.pdf
Yeh, Sonia and Rubin, Edward. February 2007. "Incorporating Technological Learning in the Coal
Utility Environmental Cost (CUECost) Model: Estimating the Future Cost Trends of S02,
NOx, and Mercury Control Technologies." Prepared for ARCADIS Geraghty and Miller,
Research Triangle Park, NC 27711. Available at http://steps.ucdavis.edu/People/
slveh/sveh-resources/Drft%20Fnl%20Rpt%20Lrng%20for%20CUECost v3.pdf.
7-24
-------
APPENDIX 7.A
OTHER EXTRAPOLATED COST APPROACHES
7.A.1 Extrapolated Cost Equations
The hybrid methodology creates a total cost curve for each pollutant for unknown
future controls that might be used in order to move toward 2020 attainment. This approach
explicitly estimates the cost of removing emissions for each pollutant for each area, with a
higher cost-per-ton in areas needing a higher proportion of unknown controls relative to known
modeled controls. For each pollutant in each area, the cost begins with a national constant
cost-per-ton for that pollutant and increases as more of that pollutant is selected. The selection
is made so that costs are minimized. The incremental improvement in air quality for an
unknown control is determined, by pollutant, using an area-by-area ratio of air quality
improvement to air quality change multiplied by the emission reductions from unknown
controls for each pollutant and area. For each pollutant in an area, the per-unit costs of control
increase with each additional unit of that pollutant chosen. The choices are made to minimize
total cost.
The hybrid methodology has the advantage of using the information about how
significant the needed reductions from unspecified control technologies are relative to the
known control measures and matching that with expected increasing per-unit cost for applying
controls beyond those modeled. Under this approach, the relative costs of unknown controls in
different geographic areas reflect the expectation that average per-ton control costs are likely
to be higher in areas needing a higher ratio of emission reductions from unknown to known
controls. Because no cost data exists for unknown future strategies, it is unclear whether
approaches using hypothetical cost curves will be more accurate or less accurate in forecasting
total national costs of unknown controls than a fixed-cost approach that uses a range of
national cost-per-ton values. Extrapolated cost estimates are provided using a 7 percent
discount rate because known control measure information is available at 7 percent for all
measures applied in the analysis.
7.A.1.1 Theoretical Model for Hybrid Methodology
A model of increasing total costs was developed for each pollutant. The simplest form of
ax2 + bx + c was used where x is the tons of a particular pollutant to be reduced in a particular
area and a, b, and c are constants. For the hybrid methodology b is set to be a national, initial
cost-per-ton (N) for unknown controls for a pollutant, and c is zero because there is no cost to
7.A-1
-------
imposing no control. The hybrid methodology has a different a for each pollutant and
geographical area. For a particular geographic area and pollutant a is N/E where
N = national, initial cost/ton (b from above)
E = by geographic area and by pollutant, is the denominator1 and represents all
emission reductions estimated to be required (from applying known and
unknown controls to obtain the 15/35 baseline, as well as known controls to
achieve the alternative standard) prior to estimating needed emission reductions
from unknown controls to achieve the alternative standard.
The hybrid methodology attempts to consider the varying conditions in each geographic
area exceeding the alternative standard. Using the hybrid methodology, we develop marginal
control cost curves associated with unknown control costs for each geographic area. Because
the rate at which the unknown control costs increase varies across areas, the marginal cost will
vary. For example, applying unknown controls to obtain the needed emission reductions in a
geographic area with few emissions sources, very few known controls and many emission
reductions needed will have a higher marginal cost. Where in another circumstance, applying
unknown controls to obtain the needed emission reductions in a geographic area with many
emissions sources, many known controls and many emission reductions needed will have a
lower marginal cost.
The total cost equations presented below in Section 7.A.1.2 are used with the constraint
that
where
U = unknown emission reductions
Q = air quality change needed for area to reach attainment
R = air quality to emissions ratio for area.
This equation above uses the R developed in Section 3.3.1.3 (Estimating Full Attainment)
for each pollutant and area that is used to estimate the change in air quality for each unit of
emissions. In other words, R is a measure of how much the air quality concentration changes
1 The numerator differs by pollutant and is the national, initial cost-per-ton for unknown controls discussed below,
e.g., $17,000/ton for PM2.5.
7.A-2
-------
when the emissions change. The least-cost solution is found for each geographic area and for
each standard. See Section 4.2.2 for a discussion of the order in which controls were applied for
geographic areas and pollutants.
The cost-per-ton estimates were calculated by developing marginal cost curves for all
known controls applied for the alternative standards as well as the baseline. We reviewed data
on controls in approximately 120 counties with higher PM2.5 concentrations. We plotted, in
order of increasing marginal cost, the cost-per-ton and cumulative emission reductions
associated with these controls for PM2.5, S02, and NOX. For PM2.5, the data show that
approximately 96 percent of potential emission reductions associated with the known controls
in those 120 counties could be obtained for $17,000/ton (2006 dollars) or less. For S02, the data
show that approximately 94 percent of potential emission reductions associated with the
known controls in those 120 counties could be obtained for $5,400/ton (2006 dollars) or less.
For NOX, the data show that approximately 92 percent of potential emission reductions
associated with the known controls in those 120 counties could be obtained for $5,500/ton
(2006 dollars) or less. We conservatively defined a threshold of 90 percent for unknown
emission reductions for a particular pollutant in an area after reductions associated with known
controls are achieved. As such, the national, initial cost-per-ton for unknown controls for the
parameters NPMz5, NSO2i and JVyvorwere chosen at $17,000/ton (2006 dollars), $5,400/ton (2006
dollars), and $5,500/ton (2006 dollars) for PM2.5, S02; and NOX emission reductions,
respectively.
The results for the alternative hybrid methodology described in this Appendix and the
fixed-cost approach described in Chapter 7 are presented for comparison in Table 7.A-1.
Table 7.A-1. Extrapolated Costs by Alternative Standard Analyzed (millions of 2006$)a
Extrapolated Costs
Fixed-Cost Methodology
Alternative Hybrid
Methodology
Alternative Standard
13/35
12/35
Region
East
West
California
Total
East
West
7%
—
—
$2.9
$2.9
—
$3.3
7%
—
—
$3.6
$3.6
—
$24
7.A-3
-------
11/35
11/30
California
Total
East
West
California
Total
East
West
California
Total
$65
$69
$1.3
$38
$180
$220
$21
$79
$190
$290
$100
$130
$3
$310
$280
$590
$47
$470
$320
$830
a Estimates are rounded to two significant figures.
The hybrid model is presented more formally below. For the simple form of ax2 + bx + c,
where x is the tons of a particular pollutant to be reduced in a particular area and a, b, and c are
constants.
7.A.1.2 Framework Applied to This Analysis
Total Cost Equation for the three pollutants is
Minimize -f^ US022 + NS02US02 + c + ) [ F ' '"^ I UPM_.,,/ + NPM_.UPM_ _,J + d
\^o2/ Af V^
+ ^NOX^NOX + /
n
subject to the constraint ^ Qj ~ Rso2jUso2 ~ RPM2.5jUpM2.5j ~ RNOXJUNOX = 0
7=1
Where j isjth county up to the total number of nonattainment counties in each defined
geographic area
U.
SO2 '•
7 = 1
>0
US02<0.9(TS02-ES02)
7.A-4
-------
V UPM25. < = 0.9(7TM2.5y -
7=1
N ox
^ 0.9( 7~Wox ~
The constraints require the emissions changes to result in the required air quality
change. The emissions changes cannot be negative. The unknown emission reductions for a
particular pollutant in an area cannot exceed 90% of the remaining emissions after reductions
associated with known controls are achieved. Section 7.2.3.1 includes information on the
selection of the 90% threshold.
where
N = national constant cost/ton
E = known emission reductions
U = unknown emission reductions
Q = air quality change
R = air quality to emissions ratio
T = total emissions
This optimization can be executed in a number of ways (all resulting in the same
answer). For this analysis we employed the data solver add-in (solver) for Microsoft Excel. We
ran the solver for each area needing unknown controls for each standard. The detailed
spreadsheet will be in the PM NAAQS docket.
7.A.1.3 Cost Minimization Approach
The solver iterates to select how much of each pollutant to choose in a cost-minimizing
manner. In each geographic area with one or more counties that requires additional emission
reductions to reach attainment, to select a quantity of each pollutant to minimize the costs of
the needed emissions reductions, the solver
• looks simultaneously at PM2.5 emissions within any counties not meeting the
standard and S02 and N0xemissions from within the geographic area.
For example, for a three-county area where two counties are not estimated to meet the
alternative standard, the solver picks PM2.5 reductions in each of the two counties and then S02
7.A-5
-------
and NOX reductions for the entire geographic area such that both counties would reach
attainment in the least cost way.
In Chapter 3, we define ratios for each of the two counties associated with how much
air quality concentration improvement would result from 1,000 tons of PM2.5 reduction in each
respective county. Also in Chapter 3, we define ratios for each of the two counties associated
with how much air quality concentration improvement would result from 1,000 tons of S02
reduction in the defined geographic area (not just the respective counties). Lastly in Chapter 3,
we define ratios for each of the two counties associated with how much air quality
concentration improvement would result from 1,000 tons of NOX reduction in the defined
geographic area (not just the respective county). In Chapters 3 and 4, we define and present
geographic areas where S02and N0xemissions contribute to the air quality problems in a close
cluster of counties.
7.A.2 Sensitivity Analyses of Extrapolated Cost Approaches
Because of the uncertainties associated with estimating costs for the PM2.5 NAAQS and
because a significant portion of the estimated emissions reductions and related costs for
attaining the NAAQS come from unknown controls, it is important to test the sensitivity of the
assumptions applied to estimate unknown controls. The sensitivity analyses below are included
to help characterize the uncertainty for the cost estimates from unknown controls and the
responsiveness of the cost estimates to varying parameter estimates and assumptions. Note
that the tables below include cost estimates associated with unknown controls and not total
cost estimates.
While there are many approaches to sensitivity analysis, we selected analyses below,
keeping emissions estimates constant, to show variability in the cost estimates and remain
consistent with the benefits analysis. Note that the extrapolated cost estimates are provided
using a 7 percent discount rate because known control measure information is available at 7
percent for all measures applied in this analysis.
7.A.2.1 Sensitivity Analysis of Fixed-Cost Approach
Table 7.A.2 below presents the sensitivity analysis of the fixed-cost approach and
includes, by region and alternative standard, the primary cost estimate of $15,000/ton. The
Table also includes, by region and alternative standard, cost estimates using $10,000/ton and
$20,000/ton. For the 12/35 alternative standard, the total cost estimate associated with
unknown control costs ranges from $46 million to $92 million, depending on the fixed-cost-per-
ton assumed.
7.A-6
-------
Table 7.A-2. Sensitivity Analysis of Fixed-Cost Approach for Unknown Controls by Alternative
Standard Analyzed (millions of 2006$)a
Extrapolated Costs
Alternative
Standard
13/35
12/35
11/35
11/30
Region
East
West
California
Total
East
West
California
Total
East
West
California
Total
East
West
California
Total
$10,000/ton
7%
—
—
$1.9
$1.9
—
$2.2
$44
$46
$0.90
$26
$120
$150
$14
$53
$130
$190
$15,000/ton
7%
—
—
$2.9
$2.9
—
$3.3
$65
$69
$1.3
$38
$180
$220
$21
$79
$190
$290
$20,000/ton
7%
—
—
$3.9
$3.9
$4.5
$87
$92
$1.8
$51
$240
$290
$28
$110
$250
$390
a Estimates are rounded to two significant figures.
7.A.2.2 Sensitivity Analysis of Alternative Hybrid Approach
Table 7.A.3 below presents the sensitivity analysis of the alternative hybrid approach. To
be consistent with the sensitivity analysis of the fixed-cost approach, the table also includes, by
region and alternative standard, cost estimates using alternate parameter estimates for the
initial cost per ton. For the 12/35 alternative standard, the total cost estimate associated with
unknown control costs ranges from $85 million to $170 million.
7.A-7
-------
Table 7.A-3. Sensitivity Analysis of Alternative Hybrid Approach for Unknown Controls by
Alternative Standard Analyzed (millions of 2006$)a
Extrapolated Costs
Alternative
Standard
13/35
12/35
11/35
11/30
Region
East
West
California
Total
East
West
California
Total
East
West
California
Total
East
West
California
Total
30 Percent Lower
7%
—
—
$2.4
$2.4
—
$16
$69
$85
$1.8
$210
$190
$400
$32
$310
$210
$550
Estimate
w/Original Cost
of Initial Ton
7%
—
—
$3.6
$3.6
—
$24
$100
$130
$3
$310
$280
$590
$47
$470
$320
$830
30 Percent
Higher d
7%
—
—
$4.7
$4.7
—
$32
$140
$170
$3.6
$420
$370
$790
$63
$630
$420
$1,100
Estimates are rounded to two significant figures.
These estimates reflect national, initial cost-per ton estimates for the three parameters that are 30 percent
lower.
As discussed in Section 7.A. 1.1 above, the national, initial cost-per-ton for unknown controls for the parameters
NpM25, NS02 and NNOx were chosen at $17,000/ton (2006 dollars), $5,400/ton (2006 dollars), and $5,500/ton
(2006 dollars) for PM2.5, SO2, and NOX emission reductions, respectively. In addition, as discussed in Chapter 7,
Section 7.2.3, the hybrid cost methodology is used to estimate the needed emission reductions.
d These estimates reflect national, initial cost-per ton estimates for the three parameters that are 30 percent
higher.
7.A-8
-------
CHAPTER 8
COMPARISON OF BENEFITS AND COSTS
8.1 Synopsis
This chapter compares estimates of the benefits with economic costs and summarizes
the net benefits of alternative standards relative to a baseline that includes recently
promulgated national regulations (CSAPR, MATS, and others). We include a discussion of net
benefits for the case of full attainment and discuss selected limitations of the analyses.
8.2 Analysis
In the analysis, we estimate the net benefits of the proposed range of annual PM2.5
standards of 12/35 to 13/35. For 12/35, net benefits are estimated to be $2.3 billion to $5.9
billion at a 3% discount rate and $2.0 billion to $5.3 billion at a 7% discount rate in 2020 (2006
dollars).1 For 13/35, net benefits are estimated to be $85 million to $220 million at the 3%
discount rate and $76 million to $200 million at the 7% discount rate.
The RIA also analyzes the benefits and costs of two alternative primary PM2.5 standards
(11/35 and 11/30) that are more stringent than the proposed standard range of 12/35 to 13/35.
The EPA estimated the net benefits of the alternative annual PM2.5 standard of 11/35 to be $8.9
billion to $23 billion at a 3% discount rate and $8.0 billion to $21 billion at a 7% discount rate in
2020. The EPA estimated the net benefits of the alternative annual PM2.5 standard of 11/30 to
be $14 billion to $36 billion at a 3% discount rate and $13 billion to $33 billion at a 7% discount
rate in 2020. All estimates are in 2006$.2
The EPA determined that all counties would meet the 14/35 standard concurrently with
meeting the existing 15/35 standard at no additional cost. Consequently, there is no need to
present an analysis of 14/35 in this RIA.
We provide these results in Table 8-1. In Table 8-2, we provide the avoided health
incidences associated with these standard levels.
1 Using a 2010$ year increases estimated costs and benefits by approximately 8%. Because of data limitations, we
were unable to discount compliance costs for all sectors at 3%. As a result, the net benefit calculations at 3%
were computed by subtracting the monetized benefits at 3% minus the costs at 7%.
2 Using a 2010 $ year increases estimated costs and benefits by approximately 8%. Because of data limitations, we
were unable to discount compliance costs for all sectors at 3%. As a result, the net benefit calculations at 3%
were computed by subtracting the monetized benefits at 3% minus the costs at 7%.
8-1
-------
8.3 Conclusions of the Analysis
EPA's illustrative analysis has estimated the health and welfare benefits and costs
associated with proposed revised PM NAAQS. The results for 2020 suggest there will be
significant health and welfare benefits and these benefits will outweigh the costs associated
with the illustrative control strategies in 2020.
Table 8-1. Total Monetized Benefits, Total Costs, and Net Benefits in 2020 (millions of
2006$a)—Full Attainment
Alternative
Standard
13/35
12/35
11/35
11/30
Total
3% Discount
Ratec
$2.9
$69
$270
$390
Costs
7% Discount
Rate
$2.9
$69
$270
$390
Monetized
3% Discount
Rate
$88 to $220
$2,300 to
$5,900
$9,200 to
$23,000
$14,000 to
$36,000
Benefits b
7% Discount
Rate
$79 to $200
$2,100 to
$5,400
$8,300 to
$21,000
$13,000 to
$33,000
Net
3% Discount
Ratec
$85 to $220
$2,300 to
$5,900
$8,900 to
$23,000
$14,000 to
$36,000
Benefits b
7% Discount
Rate
$76 to $200
$2,000 to
$5,300
$8,000 to
$21,000
$13,000 to
$33,000
a Rounded to two significant figures. Using a 2010$ year increases estimated costs and benefits by approximately
8%.
bThe reduction in premature deaths each year accounts for over 90% of total monetized benefits. Mortality risk
valuation assumes discounting over the SAB-recommended 20-year segmented lag structure. Not all possible
benefits or disbenefits are quantified and monetized in this analysis. B is the sum of all unquantified benefits.
Data limitations prevented us from quantifying these endpoints, and as such, these benefits are inherently more
uncertain than those benefits that we were able to quantify.
c Due to data limitations, we were unable to discount compliance costs for all sectors at 3%. As a result, the net
benefit calculations at 3% were computed by subtracting the monetized benefits at 3% minus the costs at 7%.
For the lower end of the proposed standard range of 12/35, the EPA estimates that the
benefits of full attainment exceed the costs of full attainment by 34 to 86 times at a 3%
discount rate and 30 to 78 times at a 7% discount rate. For the upper end of the proposed
standard range of 13/35, the EPA estimates that the benefits of full attainment exceed the
costs of full attainment by 30 to 77 times at a 3% discount rate and 27 to 69 times at a 7%
discount rate. For the alternative standards, 11/35 and 11/30, the EPA estimates that the
benefits of full attainment exceed the costs of full attainment by 34 to 94 times at a 3%
discount rate and 30 to 85 times at a 7% discount rate.
8-2
-------
Table 8-2. Estimated Number of Avoided PM2.5 Health Impacts for Standard Alternatives-
Full Attainment3
Alternative Combination of Primary PM2.5 Standards
Health Effect
Adult Mortality
Krewski et al. (2009)
Laden et al. (2006) (adult)
Woodruff et al. (1997) (infant)
Non-fatal heart attacks (age >18)
Peters et al. (2001)
Pooled estimate of 4 studies
Hospital admissions— respiratory (all ages)
Hospital admissions— cardiovascular (age > 18)
Emergency department visits for asthma (age < 18)
Acute bronchitis (age 8-12)
Lower respiratory symptoms (age 7-14)
Upper respiratory symptoms (asthmatics age 9-11)
Asthma exacerbation (age 6-18)
Lost work days (age 18-65)
Minor restricted-activity days (age 18-65)
13/35
11
27
0
11
1
3
3
6
22
290
410
410
1,800
11,000
12/35
280
730
1
320
35
98
95
160
540
6,900
9,800
24,000
44,000
260,000
11/35
1,100
2,900
3
1,300
140
430
400
730
2,000
25,000
37,000
89,000
170,000
1,000,000
11/30
1,700
4,500
4
1,900
210
620
580
1,000
3,100
39,000
56,000
140,000
260,000
1,500,000
a Incidence estimates are rounded to whole numbers with no more than two significant figures.
8.4 Caveats and Limitations
EPA acknowledges several important limitations of the primary and secondary analysis.
These include:
8.4.1 Benefits Caveats
* PM2.5 mortality co-benefits represent a substantial proportion of total monetized
benefits (over 98%). To characterize the uncertainty in the relationship between
PM2.5 and premature mortality, we include a set of twelve estimates based on
results of the PM2.5 mortality expert elicitation study in addition to our core
estimates. Even these multiple characterizations omit the uncertainty in air quality
estimates, baseline incidence rates, populations exposed, and transferability of the
effect estimate to diverse locations. As a result, the reported confidence intervals
and range of estimates give an incomplete picture about the overall uncertainty in
the PM2.5 estimates. This information should be interpreted within the context of the
larger uncertainty surrounding the entire analysis.
• Most of the estimated avoided premature deaths occur at or above the lowest
measured PM2.5 concentration in the two studies used to estimate mortality
8-3
-------
benefits. In general, we have greater confidence in risk estimates based on PM2.5
concentrations where the bulk of the data reside and somewhat less confidence
where data density is lower.
• We analyzed full attainment in 2020, and projecting key variables introduces
uncertainty. Inherent in any analysis of future regulatory programs are uncertainties
in projecting atmospheric conditions and source-level emissions, as well as
population, health baselines, incomes, technology, and other factors.
• There are uncertainties related to the health impact functions used in the analysis.
These include within-study variability; pooling across studies; the application of C-R
functions nationwide and for all particle species; extrapolation of impact functions
across populations; and various uncertainties in the C-R function, including causality
and shape of the function at low concentrations. Therefore, benefits may be under-
or over-estimates.
• This analysis omits certain unquantified effects due to lack of data, time, and
resources. These unquantified endpoints include other health and ecosystem
effects. The EPA will continue to evaluate new methods and models and select those
most appropriate for estimating the benefits of reductions in air pollution.
8.4.2 Control Strategy and Cost Analysis Caveats and Limitations
Control Technology Data
* Technologies applied may not reflect emerging devices that may be available in
future years.
• Control efficiency data depend on equipment being well maintained.
• Area source controls assume a constant estimate of emission reductions, despite
variability in extent and scale of application.
Control Strategy Development
* States may develop different control strategies than the ones illustrated.
• Data on baseline controls from current SIPs are lacking.
• Timing of control strategies may be different than envisioned in the RIA.
• Controls are applied within the county with the violating monitor. It is possible that
additional known controls could be available in a wider geographical area.
• Unknown controls were needed to reach attainment in several counties. Costs
associated with these unknown controls were estimated using a fixed cost per ton
methodology as well as an extrapolated cost methodology.
• Emissions reductions from mobile sources, EGUs, other PM2.s precursors (i.e.,
ammonia and VOC), and voluntary programs are not reflected in the analyses.
8-4
-------
Technological Change
* Emission reductions do not reflect potential effects of technological change that may
be available in future years.
• Effects of "learning by doing" are not accounted for in the emission reduction
estimates.
• Future technology developments in sectors not analyzed here (e.g., EGUs) may be
transferrable to non-EGU and area sources, making these sources more viable for
achieving future attainment at a lower cost than the cost presented in this analysis.
Engineering Cost Estimates
* Because of data limitations, we were unable to discount compliance costs for all
sectors at 3%.
• Estimates of private compliance cost are used as a proxy for social cost.
Unquantified Costs
* A number of costs remain unquantified, including administration costs of federal and
state SIP programs, and transactional costs.
8-5
-------
CHAPTER 9
STATUTORY AND EXECUTIVE ORDER REVIEWS
9.1 Synopsis
This chapter summarizes the Statutory and Executive Order (EO) impact analyses
relevant for the PM NAAQS Regulatory Impact Analysis. For each EO and Statutory requirement
we describe both the requirements and the way in which our analysis addresses these
requirements.
9.2 Executive Order 12866: Regulatory Planning and Review
Under section 3(f)(l) of Executive Order 12866 (58 FR 51735, October 4, 1993), this
action is an "economically significant regulatory action" because it is likely to have an annual
effect on the economy of $100 million or more. Accordingly, the EPA submitted this action to
the Office of Management and Budget (OMB) for review under Executive Orders 12866 and
13563 (76 FR 3821, January 21, 2011), and any changes made in response to OMB
recommendations have been documented in the docket for this action.
9.3 Paperwork Reduction Act
This action does not impose an information collection burden under the provisions of
the Paperwork Reduction Act, 44 U.S.S. 3501 et seq. Burden is defined at 5 CFR 1320.3(b).
There are no information collection requirements directly associated with revisions to a NAAQS
under section 109 of the CAA.
9.4 Regulatory Flexibility Act
The Regulatory Flexibility Act (RFA) generally requires an agency to prepare a regulatory
flexibility analysis of any rule subject to notice and comment rulemaking requirements under
the Administrative Procedure Act or any other statute unless the agency certifies that the rule
will not have a significant economic impact on a substantial number of small entities. Small
entities include small businesses, small organizations, and small governmental jurisdictions.
For purposes of assessing the impacts of this rule on small entities, small entity is
defined as: (1) a small business that is a small industrial entity as defined by the Small Business
Administration's (SBA) regulations at 13 CFR 121.201; (2) a small governmental jurisdiction that
is a government of a city, county, town, school district or special district with a population of
less than 50,000; and (3) a small organization that is any not-for-profit enterprise which is
independently owned and operated and is not dominant in its field.
9-1
-------
After considering the economic impacts of this proposed rule on small entities, I certify
that this action will not have a significant economic impact on a substantial number of small
entities. This proposed rule will not impose any requirements on small entities. Rather, this rule
establishes national standards for allowable concentrations of particulate matter in ambient air
as required by section 109 of the CAA. See also American Trucking Associations v. EPA. 175 F.3d
at 1044-45 (NAAQS do not have significant impacts upon small entities because NAAQS
themselves impose no regulations upon small entities). Please refer to the preamble for
additional details.
9.5 Unfunded Mandates Reform Act
This action contains no Federal mandates under the provisions of Title II of the
Unfunded Mandates Reform Act of 1995 (UMRA), 2 U.S.C. 1531-1538 for state, local, or tribal
governments or the private sector. The action imposes no enforceable duty on any state, local
or tribal governments or the private sector. Therefore, this action is not subject to the
requirements of sections 202 or 205 of the UMRA.
This action is also not subject to the requirements section 205 of the UMRA because it
contains no regulatory requirements that might significantly or uniquely affect small
governments. This action imposes no new expenditure or enforceable duty on any state, local,
or tribal governments or the private sector, and the EPA has determined that this rule contains
no regulatory requirements that might significantly or uniquely affect small governments.
Furthermore, in setting a NAAQS, the EPA cannot consider the economic or
technological feasibility of attaining ambient air quality standards, although such factors may be
considered to a degree in the development of state plans to implement the standards. See also
American Trucking Associations v. EPA, 175 F. 3d at 1043 (noting that because the EPA is
precluded from considering costs of implementation in establishing NAAQS, preparation of a
Regulatory Impact Analysis pursuant to the Unfunded Mandates Reform Act would not furnish
any information which the court could consider in reviewing the NAAQS). The EPA
acknowledges, however, that any corresponding revisions to associated SIP requirements and
air quality surveillance requirements, 40 CFR part 51 and 40 CFR part 58, respectively, might
result in such effects. Accordingly, the EPA will address, as appropriate, unfunded mandates if
and when it proposes any revisions to 40 CFR parts 51 or 58.
9.6 Executive Order 13132: Federalism
This action does not have federalism implications. It will not have substantial direct
effects on the states, on the relationship between the national government and the states, or
9-2
-------
on the distribution of power and responsibilities among the various levels of government, as
specified in Executive Order 13132. The rule does not alter the relationship between the
Federal government and the states regarding the establishment and implementation of air
quality improvement programs as codified in the CAA. Under section 109 of the CAA, the EPA is
mandated to establish and review NAAQS; however, CAA section 116 preserves the rights of
states to establish more stringent requirements if deemed necessary by a state. Furthermore,
this proposed rule does not impact CAA section 107 which establishes that the states have
primary responsibility for implementation of the NAAQS. Finally, as noted in section D on UMRA
in the preamble, this rule does not impose significant costs on state, local, or Tribal
governments or the private sector. Thus, Executive Order 13132 does not apply to this action.
However, as also noted in section D on UMRA in the preamble, the EPA recognizes that
states will have a substantial interest in this rule and any corresponding revisions to associated
air quality surveillance requirements, 40 CFR part 58. Please refer to the preamble for
additional details on the Executive Order.
9.7 Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments
The action does not have tribal implications, as specified in Executive Order 13175 (65
FR 67249, November 9, 2000). It does not have a substantial direct effect on one or more Indian
Tribes, since Tribes are not obligated to adopt or implement any NAAQS. The Tribal Authority
Rule gives Tribes the opportunity to develop and implement CAA programs such as the PM
NAAQS, but it leaves to the discretion of the Tribe whether to develop these programs and
which programs, or appropriate elements of a program, they will adopt. Thus, Executive Order
13175 does not apply to this rule.
Although Executive Order 13175 does not apply to this rule, the EPA consulted with
tribal officials or other representatives of tribal governments in developing this action. Please
refer to the preamble for additional details on the Executive Order.
9.8 Executive Order 13045: Protection of Children from Environmental Health and Safety
Risks
This action is subject to Executive Order 13045 (62 FR 19885, April 23, 1997) because it
is an economically significant regulatory action as defined by Executive Order 12866, and the
EPA believes that the environmental health or safety risk addressed by this action may have a
disproportionate effect on children. Accordingly, we have evaluated the environmental health
or safety effects of PM exposures on children. The protection offered by these standards may
9-3
-------
be especially important for children because childhood represents a lifestage associated with
increased susceptibility to PM-related health effects. Because children have been identified as a
susceptible population, we have carefully evaluated the environmental health effects of
exposure to PM pollution among children. Discussions of the results of the evaluation of the
scientific evidence and policy considerations pertaining to children are contained in sections
III.B, III.D, IV.B, and IV.C of the preamble. A listing of documents that contain the evaluation of
scientific evidence and policy considerations that pertain to children is found in the section on
Children's Environmental Health in the Supplementary Information section of the preamble,
and a copy of all documents have been placed in the public docket for this action.
9.9 Executive Order 13211: Actions that Significantly Affect Energy Supply, Distribution or
Use
This action is not a "significant energy action" as defined in Executive Order 13211, (66
FR 28355, May 22, 2001) because it is not likely to have a significant adverse effect on the
supply, distribution, or use of energy. The purpose of this action concerns the review of the
NAAQS for PM. The action does not prescribe specific pollution control strategies by which
these ambient standards will be met. Such strategies are developed by states on a case-by-case
basis, and the EPA cannot predict whether the control options selected by states will include
regulations on energy suppliers, distributors, or users.
9.10 National Technology Transfer and Advancement Act
Section 12(d) of the National Technology Transfer and Advancement Act of 1995
(NTTAA), Public Law 104- 113, section 12(d) (15 U.S.C. 272 note) directs the EPA to use
voluntary consensus standards in its regulatory activities unless to do so would be inconsistent
with applicable law or otherwise impractical. Voluntary consensus standards are technical
standards (e.g., materials specifications, test methods, sampling procedures, and business
practices) that are developed or adopted by voluntary consensus standards bodies. The NTTAA
directs the EPA to provide Congress, through OMB, explanations when the Agency decides not
to use available and applicable voluntary consensus standards.
This proposed rulemaking involves technical standards for environmental monitoring
and measurement. Specifically, the EPA proposes to retain the indicators for fine (PM2.s) and
coarse (PM10) particles. The indicator for fine particles is measured using the Reference
Method for the Determination of Fine Particulate Matter as PM2.s in the Atmosphere (appendix
L to 40 CFR part 50), which is known as the PM2.s FRM, and the indicator for coarse particles is
measured using the Reference Method for the Determination of Particulate Matter as PM10 in
9-4
-------
the Atmosphere (appendix J to 40 CFR part 50), which is known as the PM10 FRM. The EPA also
proposes a separate secondary standard defined in terms of a calculated PM2.5 light extinction
indicator, which would use PM2.5 mass species and relative humidity data to calculate PM2.5
light extinction.
To the extent feasible, the EPA employs a Performance-Based Measurement System
(PBMS), which does not require the use of specific, prescribed analytic methods. The PBMS is
defined as a set of processes wherein the data quality needs, mandates or limitations of a
program or project are specified, and serve as criteria for selecting appropriate methods to
meet those needs in a cost-effective manner. It is intended to be more flexible and cost
effective for the regulated community; it is also intended to encourage innovation in analytical
technology and improved data quality. Though the FRM defines the particular specifications for
ambient monitors, there is some variability with regard to how monitors measure PM,
depending on the type and size of PM and environmental conditions. Therefore, it is not
practically possible to fully define the FRM in performance terms to account for this variability.
Nevertheless, our approach in the past has resulted in multiple brands of monitors being
approved as FRM for PM, and we expect this to continue. Also, the FRMs described in 40 CFR
part 50 and the equivalency criteria described in 40 CFR part 53, constitute a performance-
based measurement system for PM, since methods that meet the field testing and performance
criteria can be approved as FEMs. Since finalized in 2006 (71 FR, 61236, October 17, 2006) the
new field and performance criteria for approval of PM2.5 continuous FEMs has resulted in the
approval of six approved FEMs. In summary, for measurement of PM2.5 and PM10, the EPA
relies on both FRMs and FEMs, with FEMs relying on a PBMS approach for their approval. The
EPA is not precluding the use of any other method, whether it constitutes a voluntary
consensus standard or not, as long as it meets the specified performance criteria.
For the proposed distinct secondary standard defined in terms of a calculated PM2.5 light
extinction indicator, the EPA proposes to use existing monitoring technologies that are already
deployed in the CSN and IMPROVE monitoring programs as well as relative humidity data from
sensors already deployed at routine weather stations. The sampling and analysis protocols in
use in the CSN program are the result of substantial input and recommendations from CASAC
both during their initial deployment about ten years ago, and during the more recent transition
to carbon sampling that is consistent with IMPROVE protocols (Henderson 2005c). Monitoring
agencies also played a strong role in directing the sampling technologies used in the CSN.
During the first few years of implementing the CSN there were up to four different sampling
approaches used in the network. Over time as monitoring agencies shared their experiences
9-5
-------
and data with each other, several agencies shifted their network operations to the sampling
technology used today. By 2008, the EPA was working closely with all remaining monitoring
agencies to transition to the current CSN sampling for ions and elements. All carbon sampling
was fully transitioned to the current method by October of 2009 for consistency with the
IMPROVE program. Therefore, while the current CSN sampling methods were not developed or
adopted by a voluntary consensus standard body, they are the result of harmonizing the
network by monitoring agency users and EPA. The CSN network and methods are described in
more detail in the Policy Assessment (US EPA, 2011a, Appendix B, section B.I.3).
9.11 Executive Order 12898: Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations
Executive Order 12898 (59 FR 7629, February 16, 1994) establishes federal executive
policy on environmental justice. Its main provision directs federal agencies, to the greatest
extent practicable and permitted by law, to make environmental justice part of their mission by
identifying and addressing, as appropriate, disproportionately high and adverse human health
or environmental effects of their programs, policies, and activities on minority populations and
low-income populations in the United States.
The EPA maintains an ongoing commitment to ensure environmental justice for all
people, regardless of race, color, national origin, or income. Ensuring environmental justice
means not only protecting human health and the environment for everyone, but also ensuring
that all people are treated fairly and are given the opportunity to participate meaningfully in
the development, implementation, and enforcement of environmental laws, regulations, and
policies. The EPA has identified potential disproportionately high and adverse effects on
minority and/or low-income populations from this proposed rule.
The EPA has identified persons from lower socioeconomic strata as a susceptible
population for PM-related health effects. As a result, the EPA has carefully evaluated the
potential impacts on low-income and minority populations as discussed in section III.E.3.a of
the preamble. The Agency expects this proposed rule would lead to the establishment of
uniform NAAQS for PM. The Integrated Science Assessment and Policy Assessment contain the
evaluation of the scientific evidence and policy considerations that pertain to these
populations. These documents are available as described in the Supplementary Information
section of the preamble and copies of all documents have been placed in the public docket for
this action.
9-6
-------
CHAPTER 10
SECONDARY STANDARDS ANALYSIS
10.1 Introduction
As defined by section 109(b)(2) of the Clean Air Act (CAA), the purpose of a secondary
NAAQS standard is to protect public welfare against negative effects of criteria air pollutants,
including decreased visibility, climate effects, and damage to ecological systems and building
materials. Ambient PM has been associated with visibility impairment in diverse regions across
the United States and is considered adverse to the public welfare. EPA is proposing a national
visibility standard in conjunction with the Regional Haze Program as a means of achieving
appropriate levels of protection against PM-related visibility impairments in urban, non-urban
and Federal Class I areas across the country. EPA evaluated two proposed secondary standard
levels of 30 deciviews (dv) and 28 dv, along with a more stringent standard level of 25 dv, based
on 24-hour average speciated PM2.s measurements and with a 3-year average, 90th percentile
form.
10.2 The Secondary NAAQS Standard
The secondary PM2.s NAAQS standard consists of three parts: a level, averaging period,
and form. In the Urban-focused Visibility Assessment (U.S. EPA, 2010) and the Policy Assessment
for the Review of the PM NAAQS (U.S. EPA, 2011), several preference studies provide the
foundation for the secondary PM NAAQS.1 The three completed survey studies (all in the west)
included Denver, Colorado (Ely et al., 1991), one in the lower Fraser River valley near
Vancouver, British Columbia (BC), Canada (Pryor, 1996), and one in Phoenix, Arizona (BBC
Research & Consulting, 2003). A pilot focus group study was conducted in Washington, DC on
behalf of EPA to inform the 2006 PM NAAQS review (Abt Associates Inc., 2001). Using the
results of these studies, EPA determined that for a majority of individuals, the range of
acceptable urban visibility falls between 20 dv and 30 dv based on a 4-hour average indicator
(U.S. EPA, 2011). For this analysis, we consider the two proposed standard levels of 30 dv and
28 dv, both averaged over 24 hours, along with a more stringent 24-hour average standard
level.2 While a sub-daily (i.e., 4-hour) averaging period would capture the wide variations in
1 For more detail about these preference studies, including information about study designs and sampling
protocols, please see Section 2 of the Particulate Matter Urban-Focused Visibility Assessment (U.S. EPA, 2010b).
2ln order to provide generally equivalent protection, the level of a NAAQS based on a 24-hour average indicator
should include an adjustment compared to the level that would be applied to a NAAQS based on a daily
maximum daylight 4-hour average indicator. Using 15 study sites, EPA staff investigated five approaches to
making this adjustment, for 4-hour indicator NAAQS levels of 20, 25, and 30 dv. An approach thought by EPA
staff to be more appropriate for further consideration yielded adjusted NAAQS levels of 21, 25, and 28 dvas
10-1
-------
visibility conditions that occur over the course of a day, a 24-hour averaging period avoids data
quality uncertainties associated with instruments currently available to measure hourly PM2.5
mass (U.S. EPA, 2011).
EPA proposes using a 3-year average, 90th percentile form for the standard. Determining
attainment using this form requires comparing the level of the standard to the 3-year average
of the 90th percentile of the measured indicator. Using a multi-year percentile form for the
standard lessens the influence of unusual emissions values and provides a degree of stability for
implementation planning (U.S. EPA, 2011).
10.3 Visibility Benefits from PM Reduction3
Visibility directly impacts the quality of life in the places where people live, work, and
travel (U.S. EPA, 2009). Air pollution, including particulates, contributes to decreased visibility
by scattering and absorbing light, which reduces visual range and clarity. PM2.5 component
species that contribute to decreased visibility include sulfates, nitrates, organic carbon,
elemental carbon, and soil (Sisler, 1996). Visibility impairment is expressed in terms of light
extinction, measured in inverse megameters (Mm-1), or in terms of the deciview haze index,
which is calculated based on total light extinction (Pitchford and Malm, 1994). A change of one
dv is believed to be the smallest change in visible air quality perceptible by the human eye
(Pitchford and Malm, 1994).
Visibility conditions and sources of visibility impairment vary both by region and season.
Humidity increases visibility impairment because some particles, such as ammonium sulfate
and ammonium nitrate, absorb water and become larger when relative humidity increases,
resulting in increased visibility impairment (U.S. EPA, 2009). The eastern U.S. generally
experiences greater visibility impairment due to higher concentrations of particulates and
higher average humidity levels. Particulate sulfate is the dominant source of reduced visual air
quality in the eastern U.S. (>50% of the particulate light extinction) and an important
contributor to visibility impairment elsewhere in the country (>20% of particulate light
extinction) (U.S. EPA, 2009). Particulate nitrate contributes to decreased visibility in California
and the upper Midwest, particularly during the winter (U.S. EPA, 2009). In all regions, urban
particulate concentrations are higher than those in the surrounding non-urban area, but
the 24-hour PM2.5 light extinction indicator levels that are generally equivalent to levels of 20, 25, and 30 dv
applied to a daily maximum daylight 4-hour PM2.5 light extinction indicator (U.S. EPA, 2011).
Additional discussion of visibility benefits related to attainment of the primary PM NAAQS standard can be found
in Chapters of this RIA.
10-2
-------
western urban areas show greater differences from the surrounding non-urban areas than do
eastern urban areas (U.S. EPA, 2009).
10.4 Baseline Modeling Projection Data (2020)
EPA has proposed to use a calculated indicator of PM-related light extinction to
determine whether an area is in attainment for the secondary PM2.s NAAQS standard. The
IMPROVE4 algorithm uses the estimated impact of each PM component species and relative
humidity to calculate the amount of PM-related light extinction. In the equation, each PM
component species is multiplied by a factor related to its impact on light extinction. Component
species affected by the presence of water in the ambient air are also multiplied by a factor
representing the relative humidity. These factors are summed to determine the total light
extinction caused by PM. To calculate design values for this analysis, EPA is using a modified
version of this algorithm, which is explained in more detail in Chapter 3 of this RIA.
To estimate design values for this analysis, we apply the original IMPROVE algorithm to
24-hour, speciated PM2.s concentrations measured at 236 Chemical Speciation Network (CSN)
monitors across the country and climatological mean relative humidity data to calculate the
estimated light extinction in each location. The 207 counties with CSN monitors are identified in
Figure 10-1. We then compare the calculated light extinction from a monitor to the standard
level to determine whether the county where the monitor is located is in attainment (U.S. EPA,
2011).
Even before incorporating reductions to attain the current primary standard of 15 u.g/m3
annual and 35 u.g/m3 24-hour (denoted 15/35), no monitors are expected to exceed a
secondary standard level of 30 dv and only three monitors are expected to exceed a secondary
standard level of 28 dv in 2020. Because all three of these monitors also exceed the 15/35
primary standard,5 we would expect each would attain a secondary standard of 28 dv when
controlled at the primary standard level. Further emission reductions to meet a more stringent
primary standard would lead to additional improvement in visibility in all areas. Table 10-1
4The Interagency Monitoring of Protected Visual Environment (IMPROVE) program was established in 1985 to aid
in the creation of state and federal implementation plans for visibility in Class I areas as required in the 1977
amendments to the CAA.
5The monitors that are above 15/35 are monitor id numbers: 60658001 (located in Riverside, CA); 60290014
(located in Kern, CA); and 60990005 (located in Stanislaus, CA). The projected 2020 base case design values for
the primary standard for these monitors are 16.30/46.5 u.g/m3,14.18/44.0 u.g/m3, and 10.85/37.0 u.g/m3,
respectively. The projected 2020 base case design values for the secondary standard for these monitors are: 29
dv, 30 dv, and 29 dv, respectively. We believe that the emissions reductions needed to obtain the current
primary standard levels of 15/35 will be enough to lower the projected 2020 secondary standard design values
to 28 dv.
10-3
-------
shows the percentage of monitors projected to exceed 30, 28, or 25 dv in 2020, prior to full
attainment of the current primary standard.
'.1 r i LrfcM, w ti n&t Aon. tuw Timfn
A
Figure 10-1. Counties with Monitors Included in Analysis
Table 10-1. Percentage of Monitors Projected to Exceed Alternative Secondary Standards in
2020, Prior to Attainment of Primary Standard of 15/35
Level
Number Exceeding
Selected Level
% Exceeding Selected
Level
30
28
25
0
3
28
0.0%
1.3%
11.9%
Of the 236 monitors for which visibility design values are available, 208 (88%) attain a
secondary standard of 25 dv or better in 2020, prior to full attainment of the current primary
standard. Figure 10-2 shows the counties that would exceed the secondary standards in this
analysis. Visibility design values calculated from data for each monitor location included in this
analysis can be found in Appendix 10-A.
10-4
-------
€
^ \
m
a
- -J_ =
C3,
-CT —
1 4
D
a
Legend
hiit mctilnra tatit sisinry deagn
c
r^L_.:
\/
28
-------
methodology, described in Chapter 3, to estimate the small emissions reductions needed from
control measures to show attainment and to estimate the costs and benefits of attaining the
proposed alternative primary standards. It is not possible to apply this methodology to the
visibility design values.7 As a result, the only analysis available for the proposed alternative
secondary standards in 2020 is prior to full attainment of the current primary standard. All
monitors analyzed are projected to attain a secondary standard of 30 dv in the 2020 base case.
Given the 24-hr design value reductions that were included in simulating attainment of 15/35 in
the 2020 base case, it is likely that all monitors will also attain a secondary standard of 28 dv
when they attain the current primary standards.8
10.7 References
Abt Associates, Inc. 2001. Assessing Public Opinions on Visibility Impairment due to Air
Pollution: Summary Report. Prepared for EPA Office of Air Quality Planning and
Standards; funded under EPA Contract No. 68-D-98-001. Abt Associates Inc., Bethesda,
MD. Available online at .
BBC Research & Consulting. 2003. Phoenix Area Visibility Survey. Draft Report. Available online
at .
Ely, D.W., Leary, J.T., Stewart, T.R., Ross, D.M. 1991. "The Establishment of the Denver Visibility
Standard." Presented at Air and Waste Management Association 84th Annual Meeting,
June 16-21, Paper 91-48.4, 17pp. Available online at
.
Pitchford, M and W. Malm. 1994. "Development and Application of a Standard Visual Index."
Atmospheric Environment 28(5): 1049-1054.
7As described in Chapter 3, we apply a methodology of air quality ratios to estimate the emissions reductions
needed to meet the current and proposed alternative levels for the primary standard. While this methodology
can estimate how these emissions reductions will affect changes in the future-year annual design value and the
corresponding response of the future-year 24-hr design value to changes in the annual design value, it is unable
to estimate how each of the PM2.5 species will change with these emission reductions. Given that estimating
changes in future-year visibility is dependent on the IMPROVE equation and how the PM2.5 species are
projected to change in time, we are unable to estimate visibility design values for meeting the current and
proposed alternative levels for the primary standard.
8 The projected 2020 base case design values for the secondary standard for the following monitors with id
numbers 60658001 (located in Riverside, CA), 60290014 (located in Kern, CA), and 60990005 (located in
Stanislaus, CA) are 29 dv, 30 dv, and 29 dv, respectively. The emissions reductions selected for simulating
attainment of 15/35 in the 2020 base case resulted in the following reductions in the 24-hr design values for
these three monitors: 11.1 u.g/m3, 21.9 u.g/m3 and 5.3 ug/m3, respectively. Based on the trends presented in
Wayland, 2012, we believe that these emissions reductions and 24-hr design value changes for simulating the
current primary standard levels of 15/35 will be enough to lower the projected 2020 secondary standard design
values for these three monitors to 28 dv or lower.
10-6
-------
Pryor, S.C. 1996. "Assessing Public Perception of Visibility for Standard Setting Exercises."
Atmospheric Environment, vol. 30, no. 15, pp. 2705-2716.
Sisler, J.F. 1996. Spatial and seasonal patterns and long-term variability of the composition of
the haze in the United States: an analysis of data from the IMPROVE network. CIRA
Report, ISSN 0737-5352-32, Colorado State University.
U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for
Particulate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment—RTP Division. December. Available online at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010. Particulate Matter Urban-Focused
Visibility Assessment—Final Document (Corrected). Office of Air Quality Planning and
Standards, Research Triangle Park, NC. July. Available online at
.
U.S. Environmental Protection Agency (U.S. EPA). 2011. Policy Assessment for the Review of the
PM NAAQS—Final Document. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. April. Available online at
.
Wayland, 2012. Technical Analyses to Support Surrogacy Policy for Proposed Secondary PM2.5
NAAQS under NSR/PSD Program. EPA docket #EPA-HQ-OAR-2012-XXXX.
10-7
-------
APPENDIX 10.A
2017 MODELED DESIGN VALUES BY STATE, COUNTY, AND SITE
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site
State County
Alabama Barbour
Jefferson
Jefferson
Jefferson
Madison
Mobile
Montgomery
Morgan
Russell
Arizona Maricopa
Maricopa
Maricopa
Maricopa
Maricopa
Pima
Arkansas Ashley
Pulaski
White
California Butte
Fresno
Imperial
Kern
Los Angeles
Plumas
Riverside
Sacramento
Site
010050002
010730023
010732003
010731009
010890014
010970003
011011002
011030011
011130001
040130019
040139997
040139998
040137003
040137020
040191028
050030005
051190007
051450001
060070002
060190008
060250005
060290014
060371103
060631009
060658001
060670010
Design Value
20
26
23
20
22
21
22
21
23
20
19
18
15
15
14
21
22
20
24
28
23
30
27
22
29
28
(continued)
10.A-1
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State
California (continued)
Colorado
Connecticut
D.C.
Delaware
Florida
Georgia
County
Sacramento
San Diego
San Diego
Santa Clara
Stanislaus
Tulare
Ventura
Adams
El Paso
Mesa
Weld
New Haven
Washington
Washington
Kent
New Castle
Broward
Escambia
Hillsborough
Leon
Miami-Dade
Pinellas
Bibb
Chatham
Clarke
Coffee
De Kalb
Floyd
Site
060670006
060730003
060731002
060850005
060990005
061072002
061112002
080010006
080410011
080770017
081230008
090090027
110010042
110010043
100010003
100032004
120111002
120330004
120573002
120730012
120861016
121030026
130210007
130510017
130590001
130690002
130890002
131150005
Design Value
25
23
23
24
29
28
23
19
16
19
18
24
24
24
23
25
17
23
19
21
17
21
24
21
22
19
22
22
(continued)
10.A-2
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State County
Georgia (continued) Muscogee
Richmond
Walker
Idaho Ada
Canyon
Illinois Cook
Cook
Cook
Du Page
Macon
Madison
Indiana Allen
Dubois
Elkhart
Henry
Lake
Lake
Marion
St Joseph
Vanderburgh
Iowa Linn
Polk
Scott
Kansas Sedgwick
Wyandotte
Kentucky Boyd
Daviess
Daviess
Site
132150011
132450091
132950002
160010010
160270004
170310057
170310076
170314201
170434002
171150013
171192009
180030004
180372001
180390003
180650003
180890022
180892004
180970078
181411008
181630012
191130037
191530030
191630015
201730010
202090021
210190017
210590005
210590014
Design Value
22
23
23
23
23
27
25
25
27
25
25
25
26
27
24
27
27
26
26
24
24
22
25
21
23
24
26
23
(continued)
10.A-3
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State
Kentucky (continued)
Louisiana
Maryland
Massachusetts
Michigan
Minnesota
County
Fayette
Jefferson
Jefferson
Kenton
Laurel
McCracken
Perry
Warren
Bossier
East Baton Rouge
Anne Arundel
Baltimore
Prince Georges
Hampden
Suffolk
Allegan
Chippewa
Kalamazoo
Kent
Missaukee
Monroe
Washtenaw
Wayne
Wayne
Hennepin
Mille Lacs
Olmsted
Ramsey
Site
210670012
211110043
211110048
211170007
211250004
211451004
211930003
212270007
220150008
220330009
240030019
240053001
240330030
250130008
250250042
260050003
260330901
260770008
260810020
261130001
261150005
261610008
261630033
261630001
270530963
270953051
271095008
271230871
Design Value
24
24
24
23
22
23
21
24
20
24
23
25
23
21
22
25
22
25
25
21
26
26
27
25
23
19
23
22
(continued)
10.A-4
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
County
Forrest
Grenada
Harrison
Hinds
Jones
Clay
Cooper
Jefferson
St Louis City
Ste Genevieve
Lincoln
Missoula
Douglas
Clark
Washoe
Hillsborough
Rockingham
Camden
Middlesex
Morris
Union
Bernalillo
Bronx
Bronx
Erie
Essex
Monroe
New York
Site
280350004
280430001
280470008
280490018
280670002
290470005
290530001
290990012
295100085
291860005
300530018
300630031
310550019
320030561
320310016
330110020
330150014
340070003
340230006
340273001
340390004
350010023
360050110
360050083
360290005
360310003
360551007
360610062
Design Value
23
20
22
21
23
22
22
24
24
22
25
23
22
22
18
23
19
23
21
22
25
14
26
25
24
17
23
25
(continued)
10.A-5
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State County
New York (continued) Queens
Steuben
North Carolina Buncombe
Catawba
Cumberland
Davidson
Forsyth
Guilford
Lenoir
Mecklenburg
Rowan
Wake
North Dakota Burleigh
Cass
McKenzie
Ohio Butler
Cuyahoga
Cuyahoga
Franklin
Hamilton
Jefferson
Lawrence
Lorain
Lorain
Lucas
Mahoning
Montgomery
Stark
Site
360810124
361010003
370210034
370350004
370510009
370570002
370670022
370810013
371070004
371190041
371590021
371830014
380150003
380171004
380530002
390171004
390350038
390350060
390490081
390610040
390810017
390870010
390930016
390933002
390950026
390990014
391130031
391510017
Design Value
24
20
22
23
21
23
23
22
20
22
22
22
18
19
15
24
26
25
25
24
26
25
25
21
26
25
24
25
(continued)
10.A-6
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State
Ohio (continued)
Oklahoma
Oregon
Pennsylvania
Rhode Island
County
Stark
Summit
Ellis
Oklahoma
Tulsa
Jackson
Lane
Multnomah
Union
Adams
Allegheny
Allegheny
Centre
Chester
Dauphin
Delaware
Erie
Lackawanna
Lancaster
Northampton
Perry
Philadelphia
Philadelphia
Philadelphia
Washington
Westmoreland
York
Providence
Site
391510020
391530023
400450890
401091037
401431127
410290133
410390060
410510246
410610119
420010001
420030064
420030008
420270100
420290100
420430401
420450002
420490003
420692006
420710007
420950025
420990301
421010055
421010004
421010136
421255001
421290008
421330008
440070022
Design Value
24
24
17
20
21
24
19
22
19
23
28
24
24
25
26
25
23
22
27
24
21
26
25
23
20
23
25
22
(continued)
10.A-7
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
County
Charleston
Charleston
Chesterfield
Greenville
Richland
Minnehaha
Davidson
Hamilton
Knox
Lawrence
Shelby
Shelby
Sullivan
Sumner
Brewster
Dallas
El Paso
Harris
Davis
Salt Lake
Utah
Chittenden
Bristol City
Henrico
Page
Roanoke City
King
King
Site
450190049
450190046
450250001
450450009
450790019
460990006
470370023
470654002
470931020
470990002
471570024
471570047
471631007
471650007
480430101
481130069
481410044
482011039
490110004
490353006
490494001
500070012
515200006
510870014
511390004
517700014
530330024
530330057
Design Value
21
19
20
23
22
21
22
23
22
20
22
22
24
22
15
20
17
22
25
24
24
21
24
23
22
23
23
23
(continued)
10.A-8
-------
Table 10.A-1. 2017 Modeled Design Values by State, County, and Site (continued)
State
County
Site
Design Value
Washington (continued)
King
King
King
Pierce
Spokane
530330032
530330048
530330080
530530029
530630016
22
21
20
22
21
West Virginia
Kanawha
Kanawha
Marshall
540391005
540390011
540511002
24
22
22
Wisconsin
Dodge
Kenosha
Manitowoc
Milwaukee
Taylor
Waukesha
550270007
550590019
550710007
550790026
551198001
551330027
24
25
24
25
21
25
10.A-9
-------
CHAPTER 11
QUALITATIVE DISCUSSION OF EMPLOYMENT IMPACTS OF AIR QUALITY REGULATIONS
11.1 Introduction
Executive Order 13563 states that federal agencies should consider the effect of
regulations on employment. According to the Executive Order, "our regulatory system must
protect public health, welfare, safety, and our environment while promoting economic growth,
innovation, competitiveness, and job creation. It must be based on the best available science"
(Executive Order 13563, 2011). Although a stand-alone analysis of employment impacts is not
typically included in a standard cost-benefit analysis,1 employment impacts are currently of
particular concern due to recent economic conditions reflecting relatively high levels of
unemployment. This chapter is intended to provide context for considering the potential
influence of environmental regulation on growth and job shifts in the U.S. economy. Section
11.2 addresses the particular influence of this proposed rule on employment. Section 11.3
presents a descriptive overview of the peer-reviewed literature relevant to evaluating the
effect of air quality regulation on employment. Finally, in Section 11.4, we offer several
conclusions.
11.2 Influence of NAAQS Controls on Employment
Peer-reviewed econometric studies that estimate the impact of air quality regulation on
net overall employment and within the regulated sector converge on the finding that any net
employment effects, whether positive or negative, have been small. This finding holds for even
major nationwide environmental regulations. Therefore, given the overall small effect
environmental regulations have been shown to have on net employment in the regulated
sectors, we do not expect them to have a significant impact on the overall economy.
Other factors affecting U.S. employment include cyclical, technological, demographic,
and economic trends both in the United States and abroad. In this section, we focus on the
studies most directly applicable to EPA analyses.
Estimating specific employment impacts from a new NAAQS standard is particularly
challenging for two reasons. First, the NAAQS is a target level of public health protection that
individual areas have flexibility to meet in a variety of ways, and the primary regulatory activity
and implementation occur at the state or local level. Under these circumstances, states and
localities are given considerable flexibility in choosing which strategies to adopt to meet the
1 This is the case except to the extent that labor costs are part of total costs in a cost-benefit analysis.
11-1
-------
NAAQS target. State and local officials can consider employment impacts of various control
strategies, as well as other factors, when designing their state implementation plans (SIPs). This
makes it challenging to predict how specific sectors will be impacted and how those impacts
vary across regions of the country. Analyses in the RIA are based on a particular NAAQS
compliance scenario that reflects assumptions about control measures applied across all
sectors and locations, specific control strategies adopted by the states and associated
extrapolated costs. EPA believes this compliance scenario supports reasonably illustrative
quantitative estimates of the potential overall economic effects of the revised NAAQS.
However, EPA does not consider this illustrative, aggregate compliance scenario to be
sufficiently certain and precise to support quantitative projections of outcomes in particular
locations, sectors, or markets, including labor markets, in light of the scarcity of applicable
studies that can be used to generate such estimates. Therefore, this RIA does not include
quantitative projections of aggregate shifts in employment.
Second, we anticipate that national employment levels will be changing during the
period that the NAAQS is being implemented, a period that may be greater than 10 years for
some areas, following designations of nonattainment. Although current unemployment rates
remain high relative to historical averages largely due to the sharp increase in unemployment
that began in early 2008 (U.S. Department of Labor, Bureau of Labor Statistics, 2012a), current
data suggest unemployment rates have been declining in recent months (U.S. Department of
Labor, Bureau of Labor Statistics, 2012b). Policies to meet the NAAQS in all areas will not go
into effect for several years. By this time we anticipate the economy will have had a chance to
recover toward higher employment levels that more closely approximate full employment. In
addition, over a period of 10 years or longer, potentially significant changes in technology,
growth and distribution of economic activities, and other key determinants of local and national
labor market conditions further complicate projections of future employment and the potential
incremental effect of regulatory programs.
Although a quantitative assessment of employment consequences of today's proposed
revision to the national ambient PM standards remains beyond the reach of available data and
modeling tools, EPA is in the process of supporting the development of tools and research that
could assist in the future. In the interim, some insights on the potentially relevant
consequences of revising ambient air pollution standards can be gained by considering
currently available literature, including its limitations. In light of these challenges, Section 11.3
focuses on qualitative insights from currently available peer-reviewed literature on the impact
of air quality regulations in general.
11-2
-------
11.3 The Current State of Knowledge Based on the Peer-Reviewed Literature
There is limited peer-reviewed econometric literature estimating employment effects of
environmental regulations. We present an overview here, highlighting studies with particular
relevance for NAAQS. Determining the direction of employment effects in the regulated
industries is challenging because of competing effects. Complying with the new or more
stringent regulation requires additional inputs, including labor, and may alter the relative
proportions of labor and capital used by regulated firms in their production processes.
When the economy is at full employment, an environmental regulation is unlikely to
have a considerable impact on net employment in the long run. Instead, labor would primarily
be reallocated from one productive use to another (e.g., from producing electricity or steel to
producing pollution abatement equipment). Theory supports the argument that, in the case of
full employment, the net national employment effects from environmental regulation are likely
to be small and transitory (e.g., as workers move from one job to another). There is reason to
believe that when the economy is operating at less than full employment environmental
regulation could result in a short-run net increase in employment.2 Several empirical studies
suggest that net employment impacts may be positive but small even in the regulated sector.
Taken together, the peer- reviewed literature does not contain evidence that environmental
regulation would have a notable impact on net employment across the whole economy.
This discussion focuses on both short- and long-term employment impacts in the
regulated industries, as well as on the environmental protection sector for construction of
needed pollution control equipment prior to the compliance date of the regulation. EPA is
committed to using the best available science and the relevant theoretical and empirical
literature in this assessment and is pursuing efforts to support new research in this field.
11.3.1 Immediate and Short-Run Employment Impacts
Environmental regulations are typically phased in to allow firms time to invest in the
necessary technology and process changes to meet the new standards. Whatever effects a
regulation will have on employment in the regulated sector will typically occur only after a
regulation takes effect, or in the long term, as new technologies are introduced. However, the
environmental protection sector (pollution control equipment) often sees immediate
employment effects. When a regulation is promulgated, the first response of industry is to
order pollution control equipment and services to comply with the regulation when it becomes
effective. This can produce a short-term increase in labor demand for specialized workers
2 See Schmalansee and Stavin (2011)
11-3
-------
within the environmental protection sector related to design, construction, installation and
operation of the new pollution control equipment required by the regulation, (see Schmalansee
and Stavins, 2011; Bezdek, Wendling, and Diperna, 2008).
As the NAAQS are implemented, it is possible that the regulated sector will experience
short-run changes in employment. Because it is the states' responsibility to design their state
implementation plans (SIPs) over the next few years, we cannot assess the short-term effects of
those SIPs on the regulated sector with sufficient precision to quantify the resulting incremental
effects on employment. However, as previously noted, even in a full employment case, there
may be transitory effects as workers change jobs. Some workers may need to retrain or
relocate in anticipation of the new requirements or require time to search for new jobs, while
shortages in some sectors or regions could bid up wages to attract workers.
It is important to recognize that these adjustment costs can entail local labor
disruptions, and, although the net change in the national workforce might be small, gross
reductions in employment can still have negative impacts on individuals and communities. The
peer-reviewed literature that is currently available is focused on medium- and long-term
employment impacts and does not offer much insight into the short-term balance between
increased employment in the environmental protection sector and possible decreased
employment in some regulated sectors.
11.3.2 Long-Term Employment Impacts on the Regulated Industry
Determining the direction of net employment effects in regulated industries is
challenging because of competing effects. Morgenstern, Pizer, and Shih (2002) demonstrate
that environmental regulations can be understood as requiring regulated firms to add a new
output (environmental quality) to their product mix. Although legally compelled to produce this
new output, regulated firms have to finance this additional production input with the proceeds
of sales of their other (market) products. The current literature on employment impacts of air
quality regulations can be disaggregated into two types of approaches or models: 1) structural
and 2) reduced-form models. Two papers that present a formal structural model of the
underlying profit maximizing/cost minimizing problem of the firm are Berman and Bui (2001)
and Morgenstern, Pizer, and Shih (2002). Berman and Bui (2001) developed an innovative
approach to estimating the effect of environmental regulations designed to meet a NAAQS
(e.g., ozone and NOX) requirement in California on employment. Berman and Bui's model
allows environmental regulation to operate via two separate mechanisms: 1) the output
elasticity of labor demand and 2) the effect of pollution abatement activities on demand for
11-4
-------
variable factors, combined with the marginal rates of technical substitution between
abatement activity and variable factors, including labor. Berman and Bui demonstrate, using
economic theory, that the overall net effect of environmental regulation on employment,
predicted by this model, is ambiguous. Neoclassical economic theory predicts that the output
effect is, in most cases, negative, while the direction of the second, composite effect is
indeterminate making the overall net effect ambiguous.
Morgenstern, Pizer, and Shih (2002) developed a similar structural model to Berman and
Bui's (2001) model. Their model focuses on three mechanisms whereby environmental
regulation may impact employment in regulated industries. First, is the demand, or output,
effect, where new compliance costs increase the cost of production, raising prices and thereby
reducing consumer demand, which, in turn, reduces labor demand. Second, is the cost effect,
which increases the demand for inputs, including labor, as more inputs are now required to
produce the same amount of output. Finally, the factor-shift effect notes how regulated firms'
production technologies may be more or less labor intensive after complying with the
regulation (i.e., more/less labor is required relative to capital per dollar of output), implying an
ambiguous overall net effect on labor demand. Conceptually, this theoretical approach, which is
very similar to Berman and Bui's approach, could be applied to NAAQS. However, Morgenstern
et al.'s empirical approach uses pollution abatement expenditures for only four highly
polluting/regulated sectors (pulp and paper, plastics, steel, and petroleum refining) to estimate
effects on net employment; therefore, their empirical results are not directly applicable to the
full range of manufacturing and nonmanufacturing industries affected by NAAQS. Regardless,
their work represents one of the most rigorous attempts to quantify the net employment
impacts of regulation on the regulated sector. Morgenstern et al. conclude from their empirical
results that increased pollution abatement expenditures generally have not caused a significant
change in net employment in those four sectors. More specifically, their results show that, on
average across the industries studied, each additional $1 million ($1987) spent on pollution
abatement results in a (statistically insignificant) net increase of 1.5 jobs.
Berman and Bui (2001) use their model to empirically examine how an increase in local
air quality regulation that reduces NOX emissions as a precursor to ozone and PMi0 affects
manufacturing employment in the South Coast Air Quality Management District (SCAQMD),
which incorporates Los Angeles and its suburbs. During the time frame of their study, 1979 to
1992, the SCAQMD enacted some of the country's most stringent air quality regulations. Using
SCAQMD's local air quality regulations, which are more stringent than federal and state
regulations, Berman and Bui identify the effect of environmental regulations on net
11-5
-------
employment in the regulated sectors.3 They compare changes in employment in affected plants
to those in other plants in the same industries but in regions not subject to the local
regulations. The authors find that "while regulations do impose large costs, they have a limited
effect on employment"—even when exit and dissuaded entry effects are considered (Berman
and Bui, 2001, p. 269). Their conclusion is that local air quality regulation "probably increased
labor demand slightly" but that "the employment effects of both compliance and increased
stringency are fairly precisely estimated zeros [emphasis added], even when exit and dissuaded
entry effects are included" (Berman and Bui, 2001, p. 269). In their view, the limited effects
likely arose because 1) the regulations applied disproportionately to capital-intensive plants
with relatively little employment, 2) the plants sold to local markets where competitors were
subject to the same regulations (so that sales were relatively unaffected), and 3) abatement
inputs served as complements to employment. Although Berman and Bui focus on more sectors
than Morgenstern et al. and focus specifically on air regulations, the study only examined
impacts in Southern California and impacts may differ in other nonattainment areas.
Other studies, including Henderson (1996), Becker and Henderson (2000), Greenstone
(2002), and List et al. (2003), have taken a reduced-form approach to ask a related but quite
different question regarding the impact of environmental regulation on economic activity. All of
these studies examined the effect of attainment status, with respect to NAAQS, on various
forms of economic activity (e.g., employment growth, plant openings and closings, investment).
Polluting plants already located in and new polluting plants wanting to open in nonattainment
counties (counties not in compliance with one or more NAAQSs) are likely to face more
stringent air pollution regulations to help bring them into compliance. Thus, the stringency in
environmental regulations may vary spatially, which may affect the spatial distribution of
economic activity but not necessarily the overall level of economic activity. These studies find
limited evidence that employment grows more slowly, investment is lower, or fewer new
polluting plants open in nonattainment areas relative to attainment areas. However, this
evidence does not mean that there is less aggregate economic activity as a result of
environmental regulation nor does it provide evidence regarding absolute growth rates; it
simply suggests that the relative growth rate of some sectors may differ between attainment
and nonattainment areas. The approach used in all of these other studies is not capable of
estimating net employment effects as would be necessary for a national rulemaking, only
certain aspects of gross labor flows in selected areas.
3 Note, like Morgenstern, Pizer, and Shih (2002), this study does not estimate the number of jobs created in the
environmental protection sector.
11-6
-------
11.4 Conclusion
The long-term effects of a regulation on the environmental protection sector (which
provides goods and services to the regulated sector) are difficult to assess. Employment in the
industry supplying pollution control equipment is likely to increase with the increased demand
from the regulated industry for the equipment.4 According to U.S. Department of Commerce
(2010) data, by 2008, there were 119,000 environmental technology (ET) firms generating
approximately $300 billion in revenues domestically (2% of national gross domestic product
[GDP]), producing $43.8 billion in exports (2% of total exports), and supporting nearly 1.7
million jobs (0.93% of total jobs). Air pollution control accounted for 18% of the domestic ET
market and 16% of exports. Small and medium-size companies represent 99% of private ET
firms, producing 20% of total revenue. The remaining 1% of companies are large companies
supplying 49% of ET revenue (OEEI, 2010).5
As described above, deriving estimates of how regulations will impact economy-wide
net employment is a difficult task, especially in the case of setting a new NAAQS, given that
economic theory predicts that the net effect of an environmental regulation on regulated
sectors and the overall economy is indeterminate (not necessarily positive or negative). Peer-
reviewed econometric studies that use a structural approach, applicable to overall net effects in
the regulated sectors, converge on the finding that any net employment effects of
environmental regulation in general, whether positive or negative, have been small and have
not affected employment in the economy in a significant way.
11.5 References
Becker, R. and V. Henderson (2000). "Effects of Air Quality Regulations on Polluting Industries."
Journal of Political Economy 108(2): 379-421.
Berman, E. and L. T. M. Bui (2001). "Environmental Regulation and Labor Demand: Evidence
from the South Coast Air Basin." Journal of Public Economics 79(2): 265-295.
Bezdek, R. H., R. M. Wendling, and P. Diperna (2008). "Environmental Protection, the Economy,
and Jobs: National and Regional Analyses." Journal of Environmental Management
86(1): 63-79.
Executive Order 13563 (January 21, 2011). "Improving Regulation and Regulatory Review.
Section 1. General Principles of Regulation." Federal Register 76(14): 3821-3823.
See Bezdek et al. (2008), for example, and U.S. Department of Commerce (2010).
5 To calculate the percentages, total national 2008 GDP ($14,369.1 billion), exports ($1,842.68 billion), and
employment (181.75 million employees) were obtained from Bureau of Economic Analysis, U.S. Census Bureau
and Woods & Poole, respectively.
11-7
-------
Greenstone, M. (2002). "The Impacts of Environmental Regulations on Industrial Activity:
Evidence from the 1970 and 1977 Clean Air Act Amendments and the Census of
Manufactures." Journal of Political Economy 110(6): 1175-1219.
Henderson, V. (1996). "Effects of Air Quality Regulation." The American Economic Review 86(4):
789-813.
List, J. A., D. L. Millimet, P. G. Fredriksson, and W. W. McHone (2003). "Effects of Environmental
Regulations on Manufacturing Plant Births: Evidence from a Propensity Score Matching
Estimator." The Review of Economics and Statistics 85(4): 944-952.
Morgenstern, R. D., W. A. Pizer, and J-S. Shih (2002). "Jobs Versus the Environment: An
Industry-Level Perspective." Journal of Environmental Economics and Management
43(3): 412-436.
Schmalansee, R. and R. Stavins (2011). "A Guide to Economic and Policy Analysis for the
Transport Rule." White Paper. Boston, MA: Exelon Corp.
U.S. Department of Commerce, International Trade Administration (2010, April). Environmental
Industries Fact Sheet, using 2008 data from Environmental Business International, Inc.
http://web.ita.doc.gov/ete/eteinfo.nsf/068f3801d047f26e85256883006ffa54/4878b7e2
fc08ac6d85256883006c452c?OpenDocument.
U.S. Department of Labor, Bureau of Labor Statistics (2012a). Databases, Tables & Calculators
by Subject. http://data.bls.gov/timeseries/LNS14000000. Accessed April 16, 2012.
U.S. Department of Labor, Bureau of Labor Statistics (2012b, April 6). "The Employment
Situation—March 2012." News Release, http://www.bls.gov/news.release/pdf/empsit.pdf.
Accessed April 16, 2012.
11-8
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-452/R-12-003
Environmental Protection Health and Environmental Impacts Division June 2012
Agency Research Triangle Park, NC
------- |