&EPA
United Sfc*i6
Enviraimerilfll PiutmBmi
Agamy
Regulatory Impact Analysis (RIA) for Existing
Stationary Spark Ignition (SI) RICE
NESHAP
Final Report
-------
EPA452/R-10-010
August 2010
Regulatory Impact Analysis (RIA) for Existing Stationary Spark Ignition RICE
NESHAP
By:
Paramita Sinha
Fern Braun
Brooks Depro
RTI International
3040 Cornwallis Road
Research Triangle Park, NC 27709
Prepared for:
Larry Sorrels
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards (OAQPS)
Air Benefit and Cost Group (ABCG)
(MD-C439-02)
Research Triangle Park, NC 27711
Contract No. EP-D-06-003
Work Assignment No. 4-82
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Air Benefit and Cost Group
Research Triangle Park, North Carolina
-------
CONTENTS
Section Page
1 Executive Summary 1-1
2 Introduction 2-1
2.1 Organization of this Report 2-1
3 Industry Profile 3-1
3.1 Electric Power Generation, Transmission, and Distribution 3-1
3.1.1 Overview 3-1
3.1.2 Goods and Services Used 3-3
3.1.3 Business Statistics 3-4
3.2 Oil and Gas Extraction 3-11
3.2.1 Overview 3-11
3.2.2 Goods and Services Used 3-11
3.2.3 Business Statistics 3-13
3.2.4 Case Study: Marginal Wells 3-17
3.3 Pipeline Transportation of Natural Gas 3-19
3.3.1 Overview 3-19
3.3.2 Goods and Services Used 3-19
3.3.3 Business Statistics 3-21
4 Regulatory Alternatives, Costs, and Emission Impacts 4-1
4.1 Background 4-1
4.2 Summary of the Final Rule 4-1
4.2.1 What Is the Source Category Regulated by the Final Rule? 4-1
4.2.2 What Are the Pollutants Regulated by the Rule? 4-4
4.2.3 What Are the Final Requirements? 4-5
4.2.4 What Are the Operating Limitations? 4-9
4.2.5 What Are the Requirements for Demonstrating Compliance? 4-9
in
-------
4.2.6 What Are the Reporting and Recordkeeping Requirements? 4-11
4.3 Summary of Significant Changes Since Proposal 4-11
4.3.1 Applicability 4-11
4.3.2 Final Emission Standards 4-12
4.3.3 Management Practices 4-14
4.3.4 Startup, Shutdown, and Malfunction 4-15
4.3.5 Method 323 4-16
4.4 Cost Impacts 4-17
4.4.1 Introduction 4-17
4.4.2 Control Cost Methodology 4-17
4.4.3 Control Cost Equations 4-20
4.4.4 Summary 4-24
4.4.4 Caveats and Uncertainties in the Cost Estimates 4-35
4.5 Emissions and Emission Reductions 4-37
5 Economic Impact Analysis, Energy Impacts, and Social Costs 5-1
5.1 Compliance Costs of the Final Rule 5-1
5.2 How Might People and Firms Respond? A Partial Equilibrium Analysis 5-4
5.2.1 Changes in Market Prices and Quantities 5-5
5.2.2 Regulated Markets: The Electric Power Generation, Transmission,
and Distribution Sector 5-7
5.2.3 Partial Equilibrium Measures of Social Cost: Changes Consumer
and Producer Surplus 5-8
5.3 Social Cost Estimate 5-9
5.4 Energy Impacts 5-10
5.5 Unfunded Mandates 5-11
5.5.1 Future and Disproportionate Costs 5-11
5.5.2 Effects on the National Economy 5-11
6 Small Entity Screening Analysis 6-1
6.1 Small Entity Data Set 6-1
IV
-------
6.2 Small Entity Economic Impact Measures 6-2
6.2.1 Model Establishment Receipts and Annual Compliance Costs 6-2
6.3 Small Government Entities 6-10
7 Human Health Benefits of Emissions Reductions 7-1
7.1 Synopsis 7-1
7.2 Calculation of PM2.5 Human Health Co-Benefits 7-1
7.3 Unquantified Benefits 7-13
7.3.1 Carbon Monoxide Benefits 7-13
7.3.2 Other NOx Benefits 7-14
7.3.3 Ozone Co-Benefits 7-16
7.3.4 HAP Benefits 7-16
7.4 Characterization of Uncertainty in the Monetized PM2.5 Co-Benefits 7-26
7.5 Comparison of Co-Benefits and Costs 7-31
8 References 8-1
Appendices
A Summary of Expert Opinions on the Existence of a Threshold in the
Concentration-Response Function for PM2 5-related Mortality A-1
B Lowest Measured Level (LML) Assessment for Rules without Policy-Specific
Air Quality Data Available: Technical Support Document (TSD) B-l
-------
LIST OF FIGURES
Number Page
3-1 Industrial Production Index (NAICS 2211) 3-3
3-2 Internal Combustion Generators by State: 2006 3-4
3-3 2002 Regional Distribution of Establishments: Electric Power Generation,
Transmission, and Distribution Industry (NAICS 2211) 3-6
3-4 Industrial Production Index (NAICS 211) 3-12
3-5 2002 Regional Distribution of Establishments: Crude Petroleum and Natural
Gas Extraction Industry (NAICS 211111) 3-15
3-6 2002 Regional Distribution of Establishments: Natural Gas Liquid Extraction
Industry (NAICS 211112) 3-16
3-7 Trends in Marginal Oil and Gas Production: 1997 to 2006 3-19
3-8 Distribution of Establishments within Pipeline Transportation (NAICS 486) 3-20
3-9 Distribution of Revenue within Pipeline Transportation (NAICS 486) 3-21
3-10 2002 Regional Distribution of Establishments: Pipeline Transportation (NAICS
486) 3-23
3-11 Share of Establishments by Legal Form of Organization in the Pipeline
Transportation of Natural Gas Industry (NAICS 48621): 2002 3-23
5-1 Distribution of Annualized Direct Compliance Costs by Industry 5-2
5-2 Average Annualized Cost per Engine by Horsepower Group ($2009 5-3
5-3 Distribution of Engine Population by Horsepower Group 5-4
5-4 Market Demand and Supply Model: With and Without Regulation 5-6
5-5 Electricity Restructuring by State 5-9
6-1 Distribution of Engine Population by Size for All Industries 6-7
6-2 Distribution of Compliance Costs by Engine Size for All Industries 6-8
7-1. Breakdown of Monetized PM2.5 Health Co- Benefits using Mortality Function
from Pope et al. (2002) 7-7
7-2. Total Monetized PM2.5 Co-Benefits for the Final SI RICE NESHAP in 2013 7-12
7-3. Breakdown of Monetized Co-Benefits for the Final SI RICE NESHAP by
PM2.5 Precursor and Source 7-13
7-4. Estimated County Level Carcinogenic Risk (NATA, 2002) 7-18
7-5. Estimated County Level Noncancer (Respiratory) Risk (NATA, 2002) 7-18
7-6. Percentage of Adult Population by Annual Mean PM2.5 Exposure (pre- and
post-policy) 7-28
VI
-------
7-7. Cumulative Distribution of Adult Population at Annual Mean PM2.5 levels
(pre- and post-policy) 7-28
7-8. Net Benefits for the Final SI RICE NESHAP at 3% Discount Rate 7-33
7-9. Net Benefits for the Final SI RICE NESHAP at 7% Discount Rate 7-34
vn
-------
LIST OF TABLES
Number Page
1-1 Summary of the Monetized Co-Benefits, Social Costs, and Net benefits for the
Final SI RICE NESHAP in 2013 (millions of 2009$) 1-2
3-1 Key Statistics: Electric Power Generation, Transmission, and Distribution
(NAICS 2211) ($2007) 3-2
3-2 Direct Requirements for Electric Power Generation, Transmission, and
Distribution (NAICS 2211): 2002 3-5
3-3 Firm Concentration for Electric Power Generation, Transmission, and
Distribution (NAICS 2211): 2002 3-7
3-4 United States Retail Electricity Sales Statistics: 2008 3-8
3-5 FY 2007 Financial Data for 70 U.S. Shareholder-Owned Electric Utilities 3-9
3-6 Aggregate Tax Data for Accounting Period 7/07-6/08: NAICS 2211 3-9
3-7 Key Enterprise Statistics by Receipt Size for Electric Power Generation,
Transmission, and Distribution (NAICS 2211): 2002 3-10
3-8 Key Statistics: Crude Petroleum and Natural Gas Extraction (NAICS 211111):
($2007) 3-12
3-9 Key Statistics: Natural Gas Liquid Extraction (NAICS 211112) ($2007) 3-13
3-10 Direct Requirements for Oil and Gas Extraction (NAICS 211): 2002 3-13
3-11 Key Enterprise Statistics by Employment Size for Crude Petroleum and
Natural Gas Extraction (NAICS 211111): 2002 3-17
3-12 Key Enterprise Statistics by Employment Size for Crude Natural Gas Liquid
Extraction (NAICS 211112): 2002 3-17
3-13 Aggregate Tax Data for Accounting Period 7/07-6/08: NAICS 211 3-18
3-14 Reported Gross Revenue Estimates from Marginal Wells: 2007 3-18
3-15 Key Statistics: Pipeline Transportation of Natural Gas (NAICS 48621) ($2007) .... 3-20
3-16 Direct Requirements for Pipeline Transportation (NAICS 486): 2002 3-22
3-17 Firm Concentration for Pipeline Transportation of Natural Gas (NAICS
48621): 2002 3-24
3-18 Aggregate Tax Data for Accounting Period 7/07-6/08: NAICS 486 3-24
3-19 Key Enterprise Statistics by Receipt Size for Pipeline Transportation of Natural
Gas (NAICS 48621): 2002 3-25
4-1 Emission Standards for Existing Stationary SI RICE <500 HP Located at
Major Sources of HAP 4-5
4-2 Numerical Emission Standards for Existing Non-Emergency Stationary 4SLB
and 4SRB SI RICE >500 HP Located at Area Sources of HAP 4-7
Vlll
-------
4-3 Summary of Annual and Capital Costs Equations for Existing Stationary SI
Engines 4-24
4-4 Summary of Maj or Source and Area Source Costs for the SI RICE NESHAP 4-25
4-5 Summary of Major Source and Area Source NAICS Costs for the SI RICE
NESHAP 4-27
4-6 Summary of Major Source and Area Source NAICS Costs for the SI RICE
NESHAP, by Size 4-28
4-7 Summary of Major Source and Area Source NAICS Costs for the SI RICE
NESHAP, by Number of Engines 4-32
4-8 Summary of Major Source and Area Source Baseline for the SI RICE
NESHAP 4-37
4-9 Emissions Factors 4-37
4-10 Summary of Major Source and Area Source Emissions Reductions for the SI
RICE NESHAP 4-38
5-1 Selected Industry-Level Annualized Compliance Costs as a Fraction of Total
Industry Revenue: 2009 5-5
5-2 Hypothetical Price Increases for a 1% Increase in Unit Costs 5-7
5-3 Hypothetical Consumption Decreases for a 1% Increase in Unit Costs 5-8
5-4 U.S. Electric Powera Sector Energy Consumption (Quadrillion BTUs): 2013 5-10
6-1 Proposed NESHAP for Existing Stationary Reciprocating Internal Combustion
Engines (RICE): Affected Sectors and SB A Small Business Size Standards 6-3
6-2 Average Receipts for Affected Industry by Enterprise: 2002 ($2009
Million/Establishment) 6-4
6-3 Average Receipts for Affected Industry by Enterprise Receipt Range: 2002
($2009/Establishment) 6-4
6-4 Representative Establishment Costs Used for Small Entity Analysis ($2009) 6-6
7-1 Human Health and Welfare Effects of PM2.5 7-2
7-2. Summary of Monetized Co-Benefits Estimates for Final SI RICE NESHAP in
2013 (millions of 2009$) 7-9
7-3. Summary of Reductions in Health Incidences from PM2.5 Benefits for the Final
SI RICE NESHAP in 2013 7-10
7-4. All Monetized PM2.5 Co-Benefits from PM2.5 Benefits for the Final SI RICE
NESHAP in 2013 7-11
7-5. Summary of the Monetized Benefits, Social Costs, and Net Benefits for the
final SI RICE NESHAP in 2013 (millions of 2009$) 7-31
IX
-------
SECTION 1
EXECUTIVE SUMMARY
This final action promulgates NESHAP for existing stationary SI RICE with a site rating
of less than or equal to 500 HP located at major sources, and existing stationary SI RICE of any
site rating located at area sources. EPA is finalizing these standards to meet its statutory
obligation to address HAP emissions from these sources under sections 112(d), 112(c)(3) and
112(k) of the CAA. The final NESHAP for stationary RICE will be promulgated under 40 CFR
part 63, subpart ZZZZ, which already contains standards applicable to new and reconstructed
stationary RICE and some existing stationary RICE.
EPA estimates that complying with the final national emission standards for hazardous
air pollutants (NESHAP) for stationary spark-ignition (SI) reciprocating internal combustion
engines (RICE) will have an annualized cost of approximately $253 million per year (2009
dollars) in the year of full implementation of the rule (2013). Using these costs, EPA estimates in
its economic impact analysis that the NESHAP will have limited impacts on the industries
affected and their consumers. Using sales data obtained for affected small entities in an analysis
of the impacts of this rule on small entities, EPA expects that the NESHAP will not result in a
SISNOSE (significant economic impacts for a substantial number of small entities). EPA also
does not expect significant adverse energy impacts based on Executive Order 13211, an
Executive Order that requires analysis of energy impacts for rules such as this one that are
economically significant under Executive Order 12866.
In the year of full implementation (2013), EPA estimates that the total monetized benefits
of the final NESHAP are $510 million to $1.2 billion and $460 million to $1.1 billion, at 3% and
7% discount rates, respectively (Table 1-1). All estimates are in 2009 dollars for the year 2013.
Using alternate relationships between PM2.5 and premature mortality supplied by experts, higher
and lower benefits estimates are plausible, but most of the expert-based estimates fall between
these estimates. The benefits from reducing other air pollutants have not been monetized in this
analysis, including reducing 109,000 tons of carbon monoxide and 6,000 tons of hazardous air
pollutants (HAPs) each year. In addition, ecosystem benefits and visibility benefits have not been
monetized in this analysis.
In the year of full implementation (2013), EPA estimates the net benefits of the final
NESHAP are $250 million to $980 million and $210 million to $860 million, at 3% and 7%
discount rates, respectively (Table 1-1). All estimates are in 2009 dollars for the year 2013. The
final NESHAP is the MACT floor level of control for all major SI RICE non-emergency sources
1-1
-------
and the GACT level of control for area SI RICE sources. We also show results for an alternative
(referred to as Alternative 2) which is more stringent than the final NESHAP for major sources.
In Alternative 2, the MACT level of control is applied to all SI RICE major non-emergency
sources except for four-stroke rich-burn (4SRB) engines of 300-500 horsepower (HP), where the
required level of control is above the MACT floor, and the GACT level of control is applied for
area SI RICE sources.
It should be noted that there is a difference between the annualized social costs
estimated at 3% and 7%. We approximate the annualized social costs with the compliance costs
of the rule for the RIA, as we mention later in Section 5. The annualized compliance costs of the
rule are estimated to be $244 million (2009 dollars) using a 3% interest rate. Thus, the
annualized social costs for a 3% rate are also $244 million using our approximation, and this
estimate is very close to the annualized social cost estimate at a 7% rate.
1-2
-------
Table 1-1. Summary of the Monetized Co-Benefits, Social Costs, and Net benefits for the
Final SI RICE NESHAP in 2013 (millions of 2009S)1
3% Discount Rate
7% Discount Rate
Final NESHAP: Major
Total Monetized Benefits
Total Social Costs3
Net Benefits
Non-monetized Benefits
Non-monetized Benefits
Non-monetized Benefits
3.2
to
$20
$7.4 to $18
-$81
-$80 to
12,500 tons of carbon monoxide
1,300 tons of hazardous air pollutants (HAPs)
Ecosystem effects
Visibility impairment
to
-$70
Alternative 2: Major
Total Monetized Benefits2
Total Social Costs3
Net Benefits
$48
-$47
to $120
$95
to $22
$43
-$52
to $110
$95
to $11
17,800 tons of carbon monoxide
1,400 tons of hazardous air pollutants (HAPs)
Health effects from NO2 and ozone exposure
Ecosystem effects
Visibility impairment
Final NESHAP: Area5
Total Monetized Benefits2
Total Social Costs3
Net Benefits
$500
$330
to $1,200
$166
to $1,100
$450
$290
to $1,100
$166
to $930
97,000 tons of carbon monoxide
4,700 tons of hazardous air pollutants (HAPs)
Health effects from NO2 and ozone exposure
Ecosystem effects
Visibility impairment
Final Major and Area Source NESHAP
Total Monetized Benefits
Total Social Costs3
Net Benefits
Non-monetized Benefits
$510
$460
$210
to $1,200
$253
$250 to $980
109,000 tons of carbon monoxide
6,000 tons of hazardous air pollutants (HAPs)
Health effects from NO2 and ozone exposure
Ecosystem effects
Visibility impairment
to $1,100
$253
to $860
1 All estimates are for the implementation year (2013), and are rounded to two significant figures.
2 The total monetized co-benefits reflect the human health co-benefits associated with reducing exposure to PM2 5
through reductions of PM2 5 precursors such as NOx and VOC. It is important to note that the monetized co-
benefits include many but not all health effects associated with PM2 5 exposure. It is important to note that the
monetized benefits include many but not all health effects associated with PM2 5 exposure. Benefits are shown as a range
from Pope et al. (2002) to Laden et al. (2006). These models assume that all fine particles, regardless of their chemical
composition, are equally potent in causing premature mortality because there is no clear scientific evidence that would
support the development of differential effects estimates by particle type.
3 The annual compliance costs serve as a proxy for the annual social costs of this rule given the lack of difference
between the two.
The final NESHAP is the MACT floor level of control for all major SI RICE non-emergency sources, and the GACT
level of control for area SI RICE sources.
1-3
-------
' All of the benefits for area sources are attributable to reductions expected from 4SLB and 4SRB non-emergency
engines above 500 HP.
1-4
-------
SECTION 2
INTRODUCTION
EPA is promulgating NESHAP for existing stationary SI RICE that either are located at
area sources of hazardous air pollutant emissions or that have a site rating of less than or equal to
500 horsepower and are located at major sources of hazardous air pollutant emissions.
The rule is economically significant according to Executive Order 12866. As part of the
regulatory process of preparing these standards, EPA has prepared a regulatory impact analysis
(RIA). This analysis includes an analysis of impacts to small entities as part of compliance with
the Small Business Regulatory Enforcement Fairness Act (SBREFA) and an analysis of impacts
on energy consumption and production to comply with Executive Order 13211 (Statement of
Energy Effects).
2.1 Organization of this Report
The remainder of this report supports and details the methodology and the results of the
RIA:
• Section 3 presents a profile of the affected industries.
• Section 4 presents a summary of regulatory alternatives considered in the final rule,
and provides the compliance costs of the rule.
• Section 5 describes the estimated costs of the regulation and describes the economic
impact analysis (EIA) methodology and reports market, welfare, and energy impacts.
• Section 6 presents estimated impacts on small entities.
• Section 7 presents the benefits estimates.
• Appendices A and B present technical support documents related to the benefits
estimates
2-1
-------
SECTION 3
INDUSTRY PROFILE
This section provides an introduction to the industries affected by the rule, i.e., industries
in which the spark-ignition (SI) RICE being regulated are found. SI RICE generate electric
power, pump gas or other fluids, or compress air for machinery. The primary non-utility
application of internal combustion (1C) engines is in the natural gas industry to power
compressors used for pipeline transportation, field gathering (collecting gas from wells),
underground storage tanks, and in-gas processing plants. RICEs are separated into three design
classes: 2 cycle (stroke) lean burn, 4-stroke lean burn, and 4-stroke rich burn. Each of these has
design differences that affect both baseline emissions as well as the potential for emissions
control.
These industries include the following:
• electric power generation, transmission, and distribution (NAICS 2211),
• oil and gas extraction (including marginal wells) (NAICS 211), and
• pipeline transportation of natural gas (NAICS 48621).
These three industries incur over 80 percent of the annualized costs of the rule. The
purpose is to give the reader a general understanding of the economic aspects of the industry;
their relative size, relationships with other sectors in the economy, trends for the industries, and
financial statistics.
3.1 Electric Power Generation, Transmission, and Distribution
3.1.1 Overview
Electric power generation, transmission, and distribution (NAICS 2211) is an industry
group within the utilities sector (NAICS 22). It includes establishments that produce electrical
energy or facilitate its transmission to the final consumer.
From 2002 to 2007, revenues from electric power generation grew about 18% to over
$440 billion ($2007) (Table 3-1).1 At the same time, payroll rose about 7% and the number of
employees decreased by around 4%. The number of establishments rose by about 3%. Industrial
production within NAICS 2211 has increased 26% since 1997 (Figure 3-1).
1 We provide revenues from electric power generation for the years 2002 and 2007 for these are years of the
Economic Census. We reference data from these Economic Censuses frequently in this industry profile and
show revenues from this industry over this time frame due to availability of such data.
3-1
-------
Electric utility companies have traditionally been tightly regulated monopolies. Since
1978, several laws and orders have been passed to encourage competition within the electricity
market. In the late 1990s, many states began the process of restructuring their utility regulatory
framework to support a competitive market. Following market manipulation in the early 2000s,
however, several states have suspended their restructuring efforts. The majority (58%) of power
generators controlled by combined heat and power (CHP) or independent power producers are
located in states undergoing active restructuring (Figure 3-2).
Table 3-1. Key Statistics: Electric Power Generation, Transmission, and Distribution
(NAICS 2211) ($2007)
2002 2007
Revenue ($106) 373,309 440,355
Payroll ($106) 40,842 43,792
Employees 535,675 515,335
Establishments 9,394 9,642
Source: U.S. Census Bureau; American FactFinder; "Sector 22: EC0722I2: Utilities: Industry Series: Preliminary
Comparative Statistics for the United States (2002 NAICS Basis): 2007 and 2002." http://factfinder.census.gov
3-2
-------
115
110
o
3 105
(N
O
c
TO
OJ
100
95
_
TO
I 90
c
85
80
OO'HOOOO
O'HOOOO'HOOOO'HO
OOOO'HOOOO'H
cncncncncncncncnoooooooooooooooooooooooo
cncncncncncncncnoooooooooooooooooooooooo
Figure 3-1. Industrial Production Index (NAICS 2211)
Source: The Federal Reserve Board. "Industrial Production and Capacity Utilization: Industrial Production" Series
ID: G17/IP_MINING_AND_UTILITY_DETAIL/IP.G221 l.S .
(January 27, 2010).
3.1.2 Goods and Services Used
In Table 3-2, we use the latest detailed benchmark input-output data report by the Bureau
of Economic Analysis (BEA) (2002) to identify the goods and services used in electric power
generation. As shown, labor and tax requirements represent a significant share of the value of
power generation. Extraction, transportation, refining, and equipment requirements potentially
3-3
-------
Diesel and Natural Gas Internal
Combustion Generators By State
^J No Restructuring
^] Suspended Restructuring
| Active Restructuring
100
Figure 3-2. Internal Combustion Generators by State: 2006
Source: U.S. Department of Energy, Energy Information Administration. 2007. "2006 EIA-906/920 Monthly Time
Series."
associated with reciprocating internal combustion engines (oil and gas extraction, pipeline
transportation, petroleum refineries, and turbine manufacturing) represent around 10% of the
value of services.
3.1.3 Business Statistics
The U.S. Economic Census and Statistics of U.S. Businesses (SUSB) programs provide
national information on the distribution of economic variables by industry, location, and size of
business. Throughout this section and report, we use the following definitions:
• Establishment. An establishment is a single physical location where business is
conducted or where services or industrial operations are performed.
3-4
-------
Table 3-2. Direct Requirements for Electric Power Generation, Transmission, and
Distribution (NAICS 2211): 2002
Commodity
V00100
V00200
211000
212100
482000
230301
486000
722000
52AOOO
541100
Commodity Description
Compensation of employees
Taxes on production and imports, less subsidies
Oil and gas extraction
Coal mining
Rail transportation
Nonresidential maintenance and repair
Pipeline transportation
Food services and drinking places
Monetary authorities and depository credit intermediation
Legal services
Direct Requirements
Coefficients"
20.52%
13.71%
6.16%
5.86%
3.01%
2.83%
1.70%
1.40%
1.39%
1.13%
a These values show the amount of the commodity required to produce $1.00 of the industry's output. The values
are expressed in percentage terms (coefficient *100).
Source: U.S. Bureau of Economic Analysis. 2002. 2002 Benchmark Input-Output Accounts: Detailed Make Table,
Use Table and Direct Requirements Table. Tables 4 and 5.
• Receipts: Receipts (net of taxes) are defined as the revenue for goods produced,
distributed, or services provided, including revenue earned from premiums,
commissions and fees, rents, interest, dividends, and royalties. Receipts exclude all
revenue collected for local, state, and federal taxes.
• Firm: A firm is a business organization consisting of one or more domestic
establishments in the same state and industry that were specified under common
ownership or control. The firm and the establishment are the same for single-
establishment firms. For each multiestablishment firm, establishments in the same
industry within a state are counted as one firm; the firm employment and annual
payroll are summed from the associated establishments.
• Enterprise: An enterprise is a business organization consisting of one or more
domestic establishments that were specified under common ownership or control. The
enterprise and the establishment are the same for single-establishment firms. Each
multiestablishment company forms one enterprise; the enterprise employment and
annual payroll are summed from the associated establishments. Enterprise size
designations are determined by the summed employment of all associated
establishments.
3-5
-------
In 2002, Texas had almost 1,000 power establishments, while California, Georgia, and
Ohio all had between 400 and 500 (Figure 3-3). Hawaii, Nebraska, and Rhode Island all had
fewer than 20 establishments in their states.
o
&
Establishments by State
~\ Less than 100
f 100-199
f 200-349
f 350-500
^^1 More than 500
Figure 3-3. 2002 Regional Distribution of Establishments: Electric Power Generation,
Transmission, and Distribution Industry (NAICS 2211)
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 22: Utilities:
Geographic Area Series: Summary Statistics: 2002." ; (November 10, 2008)..
As shown in Table 3-3, the four largest firms owned over 1,200 establishments and
accounted for about 16% of total industry receipts/revenue. The 50 largest firms accounted for
almost 6,000 establishments and about 78% of total receipts/revenue.
Investor-owned energy providers accounted for only 2% of retail electricity sold in the
United States in 2008 (Table 3-4). In 2008, investor-owned energy provider companies with less
than 50% of their assets regulated were unprofitable overall, while other companies in this
category were profitable. (Table 3-5). In 2008, enterprises within NAICS 2211 had a pre-tax
profit margin of 8.1% (Table 3-6).
In 2002, about 82% of firms generating, transmitting, or distributing electric power had
receipts of under $50 million (Table 3-7). However, these firms accounted for only 11% of
employment, with 89% of employees working for firms with revenues in excess of $100 million.
3-6
-------
Table 3-3. Firm Concentration for Electric Power Generation, Transmission, and
Distribution (NAICS 2211): 2002
Receipts/Revenue
Commodity
All firms
4 largest firms
8 largest firms
20 largest firms
50 largest firms
Establishments
9,394
1,260
2,566
3,942
5,887
Amount ($106)
$325,028
$52,349
$95,223
$173,207
$253,015
Percentage
of Total
100.0%
16.1%
29.3%
53.3%
77.8%
Number of
Employees
535,675
68,432
151,575
271,393
408,021
Employees per
Establishment
57
54
59
69
69
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 22: Utilities:
Subject Series—Estab & Firm Size: Concentration by Largest Firms for the United States: 2002."
; (November 21, 2008).
3-7
-------
Table 3-4. United States Retail Electricity Sales Statistics: 2008
Full-Service Providers
Item
Number of entities
Number of retail customers
Retail sales (103 megawatthours)
Percentage of retail sales
Revenue from retail sales ($106)
Percentage of revenue
Average retail price (cents/kWh)
Investor-Owned
3
46,985
2,257
2
113
1.33
5.01
Public
62
2,160,220
70,303
67
5,934
69.65
8.44
Federal
1
36
9,625
9
473
5.55
4.91
Cooperative
25
940,697
21,868
21
1,994
23.41
9.12
Facility
1
1
117
0
6
0.07
5.25
Other Providers
Energy
NA
NA
NA
—
NA
—
NA
Delivery
NA
NA
NA
—
NA
—
NA
Total
92
3,147,939
104,170
100
8,520
100
8.18
oo
oo
Source: U.S. Department of Energy, Energy Information Administration. 2009. "State Electricity Profiles 2008." DOE/EIA-0348(01)/2. p. 260.
http://www.eia.doe.gov/cneaf/electricity/st_profiles/sep2008.pdf>.
-------
Table 3-5. FY 2007 Financial Data for 70 U.S. Shareholder-Owned Electric Utilities
Investor-Owned Utilities
Regulated3
Mostly regulatedb
Diversified0
Profit Margin
4.81%
7.25%
8.50%
-16.78%
Net Income
$20,677
$12,129
$17,704
-$9,156
Operating Revenues
$430,037
$167,194
$208,288
$54,554
a 80%+ of total assets are regulated.
b 50% to 80% of total assets are regulated.
0 Less than 50% of total assets are regulated.
Source: Edison Electric Institute. "Income Statement: Q4 2008 Financial Update. Quarterly Report of the U.S.
Shareholder-Owned Electric Utility Industry." .
Table 3-6. Aggregate Tax Data for Accounting Period 7/07-6/08: NAICS 2211
Number of enterprises3 1,187
Total receipts (103) $361,177,861
Net sales(103) $328,017,143
Profit margin before tax 8.1 %
Profit margin after tax 5.4%
a Includes corporations with and without net income.
Source: Internal Revenue Service, U.S. Department of Treasury. 2010. "Corporation Source Book: Data Files 2004-
2007." ; (May 2, 2010). 3.2.2 Goods and Services
Used.
3-9
-------
Table 3-7. Key Enterprise Statistics by Receipt Size for Electric Power Generation, Transmission, and Distribution (NAICS
2211): 2002
Owned by Enterprises with
Variable
Firms
Establishments
Employment
Receipts ($103)
Receipts/firm ($103)
Receipts/establishment
($103)
Receipts/employment
($)
All
Enterprises
1,756
9,493
515,769
$320,502,670
$182,519
$33,762
$621,407
0-99K
Receipts
129
129
429
$5,596
$43
$43
$13,044
100-
499.9K
Receipts
250
250
834
$63,339
$253
$253
$75,946
500-
999.9K
Receipts
80
85
3,139
$57,363
$717
$675
$18,274
1,000-
4,999.9K
Receipts
232
245
2,712
$627,414
$2,704
$2,561
$231,347
5,000,000-
9,999,999K
Receipts
205
262
5,620
$1,472,405
$7,182
$5,620
$261,994
<10,OOOK
Receipts
896
971
12,734
$2,226,117
$2,485
$2,293
$174,817
10,000-
49,999K
Receipts
538
978
31,573
$12,171,098
$22,623
$12,445
$385,491
50,000-
99,999K
Receipts
112
403
14,858
$7,607,166
$67,921
$18,876
$511,991
100,OOOK+
Receipts
210
7,141
456,604
$298,498,289
$1,421,420
$41,801
$653,736
Source: U.S. Census Bureau. 2008. "Firm Size Data from the Statistics of U.S.
.
Businesses: U.S. All Industries Tabulated by Receipt Size: 2002."
-------
3.2 Oil and Gas Extraction
3.2.1 Overview
Oil and gas extraction (NAICS 211) is an industry group within the mining sector
(NAICS 21). It includes establishments that operate or develop oil and gas field properties
through such activities as exploring for oil and gas, drilling and equipping wells, operating on-
site equipment, and conducting other activities up to the point of shipment from the property.
Oil and gas extraction consists of two industries: crude petroleum and natural gas
extraction (NAICS 211111) and natural gas liquid extraction (NAICS 211112). Crude petroleum
and natural gas extraction is the larger industry; in 2002, it accounted for 93% of establishments
and 75% of oil and gas extraction revenues.
Industrial production in this industry is particularly sensitive to hurricanes in the Gulf
Coast. In September of both 2005 and 2008, production dropped 14% from the previous month.
However, production is currently 3% higher than it was in 2002 (Figure 3-4).
From 2002 to 2007, revenues from crude petroleum and natural gas extraction (NAICS
211111) grew over 117% to almost $215 billion ($2007) (Table 3-8). At the same time, payroll
grew 55% and the number of employees grew by 48%. The number of establishments dropped
by over 17%; as a result, the average establishment revenue increased by 162%. Materials costs
were approximately 18% of revenue over the period.
From 2002 to 2007, revenue from natural gas liquid extraction (NAICS 211112) grew
over 26% to about $42 billion (Table 3-9). At the same time, payroll dropped 18% and the
number of employees dropped by 24%. The number of establishments dropped by 43%, resulting
in an increase of revenue per establishment of about 122%.
3.2.2 Goods and Services Used
The oil and gas extraction industry has similar labor and tax requirements as the electric
power generation sector. Extraction, support, power, and equipment requirements potentially
associated with RICE (oil and gas extraction, support activities, electric power generation,
machinery and equipment rental and leasing, and pipeline transportation) represent around 8% of
the value of services (Table 3-10).
3-11
-------
110
105
i 100
(N
o
o
| 90
re
x
01
"D oc
80
75
99
0101010101010101000000000000000000000000
O1O1O1O1O1O1O1O1OOOOOOOOOOOOOOOOOOOOOOOO
Figure 3-4. Industrial Production Index (NAICS 211)
Source: The Federal Reserve Board. "Industrial Production and Capacity Utilization: Industrial Production" Series
ID: G17/IP_MINING_AND_UTILITY_DETAIL/IP.G21 l.S .
(January 27, 2010).
Table 3-8. Key Statistics: Crude Petroleum and Natural Gas Extraction (NAICS 211111):
($2007)
2002 2007
Revenue ($106) $98,667 $214,198
Payroll ($106) $5,785 $8,980
Employees 94,886 140,160
Establishments 7,178 5,956
Sources: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21: Mining:
Industry Series: Historical Statistics for the Industry: 2002 and 1997." ;
(November 26, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21: EC0721I1: Mining:
Industry Series: Detailed Statistics by Industry for the United States: 2007 " ;
(April 27, 2010).
3-12
-------
Table 3-9. Key Statistics: Natural Gas Liquid Extraction (NAICS 211112) ($2007)
2002 2007
Revenue ($106) $33,579 $42,363
Payroll ($106) $607 $501
Employees 9,693 7,343
Establishments 511 291
Sources: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21: Mining:
Industry Series: Historical Statistics for the Industry: 2002 and 1997." ;
(November 26, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21: EC0721I1: Mining:
Industry Series: Detailed Statistics by Industry for the United States: 2007 " ;
(April 27, 2010).
Table 3-10. Direct Requirements for Oil and Gas Extraction (NAICS 211): 2002
Commodity
V00200
V00100
230301
211000
213112
221100
541300
532400
33291A
541511
Direct Requirements
Commodity Description Coefficients"
Taxes on production and imports, less subsidies
Compensation of employees
Nonresidential maintenance and repair
Oil and gas extraction
Support activities for oil and gas operations
Electric power generation, transmission, and distribution
Architectural, engineering, and related services
Commercial and industrial machinery and equipment rental and leasing
Valve and fittings other than plumbing
Custom computer programming services
8.93%
6.67%
6.36%
1.91%
1.51%
1.47%
1.24%
1.20%
1.10%
0.99%
a These values show the amount of the commodity required to produce $1.00 of the industry's output. The values
are expressed in percentage terms (coefficient *100).
Source: U.S. Bureau of Economic Analysis. 2002. 2002 Benchmark Input-Output Accounts: Detailed Make Table,
Use Table and Direct Requirements Table. Tables 4 and 5.
3.2.3 Business Statistics
The U.S. Economic Census and SUSB programs provide national information on the
distribution of economic variables by industry, location, and size of business. Throughout this
section and report, we use the following definitions:
• Establishment: An establishment is a single physical location where business is
conducted or where services or industrial operations are performed.
3-13
-------
• Receipts: Receipts (net of taxes) are defined as the revenue for goods produced,
distributed, or services provided, including revenue earned from premiums,
commissions and fees, rents, interest, dividends, and royalties. Receipts exclude all
revenue collected for local, state, and federal taxes.
• Firm: A firm is a business organization consisting of one or more domestic
establishments in the same state and industry that were specified under common
ownership or control. The firm and the establishment are the same for single-
establishment firms. For each multiestablishment firm, establishments in the same
industry within a state are counted as one firm; the firm employment and annual
payroll are summed from the associated establishments.
• Enterprise: An enterprise is a business organization consisting of one or more
domestic establishments that were specified under common ownership or control. The
enterprise and the establishment are the same for single-establishment firms. Each
multiestablishment company forms one enterprise; the enterprise employment and
annual payroll are summed from the associated establishments. Enterprise size
designations are determined by the summed employment of all associated
establishments.
In 2002, Texas had almost 2,500 crude petroleum and natural gas extraction
establishments, Oklahoma had about 900, and every other state had under 400 (Figure 3-5).
Twenty-two states had fewer than 10 establishments. Similarly, Texas had 830 natural gas liquid
extraction establishments, Oklahoma had 41, Louisiana had 37, and every other state had under
25 (Figure 3-6). Only seven states had 10 or more establishments, and 24 had no establishments.
According to the SUSB, 89% of crude petroleum and natural gas extraction firms had
fewer than 500 employees in 2002 (Table 3-11). Sixty-three percent of natural gas liquid
extraction firms had fewer than 500 employees in 2002 (Table 3-12).
Enterprises within this industry generated $193 billion in total receipts in 2008. Including
those enterprises without net income, the industry averaged an after-tax profit margin of 8.5%
(Table 3-13).
3-14
-------
o
&
Establishments by State
j Less than 100
] 100 - 249
] 250-499
| 500- 1,000
I More than 1,000
Figure 3-5. 2002 Regional Distribution of Establishments: Crude Petroleum and Natural
Gas Extraction Industry (NAICS 211111)
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21: Mining:
Geographic Area Series: Industry Statistics for the State or Offshore Areas: 2007." ;
(January 27, 2010).
3-15
-------
o
Establishments by State
^| Less than 5
• 5-9
^| 10-19
| 20-40
I More than 40
Figure 3-6. 2002 Regional Distribution of Establishments: Natural Gas Liquid
Extraction Industry (NAICS 211112)
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21: Mining:
Geographic Area Series: Industry Statistics for the State or Offshore Areas: 2007." ;
(January 27, 2010).
3-16
-------
Table 3-11. Key Enterprise Statistics by Employment Size for Crude Petroleum and
Natural Gas Extraction (NAICS 211111): 2002
Variable
Firms
Establishments
Employment
Receipts ($103)
Receipts/firm ($103)
Receipts/establishment
All
Enterprises
6,238
7,135
76,794
$88,388,300
$14,169
$12,388
1-20
Employees
5,130
5,185
5,825
$2,353,181
$459
$454
20-99
Employees
348
449
5,171
$2,559,239
$7,354
$5,700
Owned by
100-499
Employees
85
254
2,757
$2,051,860
$24,140
$8,078
Enterprises with
500-749
Employees
11
37
Not disclosed
Not disclosed
Not disclosed
Not disclosed
750-999
Employees
11
63
Not disclosed
Not disclosed
Not disclosed
Not disclosed
1,000-1,499
Employees
5
25
Not disclosed
Not disclosed
Not disclosed
Not disclosed
($103)
Receipts/employment ($) $1,150,979 $403,980 $494,921 $744,236 Not disclosed Not disclosed Not disclosed
Source: U.S. Census Bureau. 2008a. Firm Size Data from the Statistics of U.S. Businesses: U.S. Detail Employment
Sizes: 2002. http://www2.census.gov/csd/susb/2002/02us_detailed%20sizes_6digitnaics.txt.
Table 3-12. Key Enterprise Statistics by Employment Size for Crude Natural Gas Liquid
Extraction (NAICS 211112): 2002
Owned by Enterprises with
Variable
Firms
Establishments
Employment
Receipts ($103)
Receipts/firm ($103)
Receipts/establishment
($103)
Receipts/employment ($)
All
Enterprises
113
494
11,486
$72,490,930
$641,513
$146,743
$6,311,242
1-20
Employees
54
54
65
$13,862
$257
$257
$213,262
20-99
Employees
7
7
Not disclosed
Not disclosed
Not disclosed
Not disclosed
Not disclosed
100-499
Employees
10
38
241
$383,496
$38,350
$10,092
$1,591,270
500-749
Employees
2
23
Not disclosed
Not disclosed
Not disclosed
Not disclosed
Not disclosed
750-999
Employees
1
1
Not disclosed
Not disclosed
Not disclosed
Not disclosed
Not disclosed
1,000-1,499
Employees
2
6
Not disclosed
Not disclosed
Not disclosed
Not disclosed
Not disclosed
Source: U.S. Census Bureau. 2008a. Firm Size Data from the Statistics of U.S. Businesses: U.S. Detail Employment
Sizes: 2002. .
3.2.4 Case Study: Marginal Wells
To provide additional context for understanding energy sectors that use reciprocating
internal combustion engines, we examine one segment of the oil and gas sector: marginal wells.
This industry includes small-volume wells that are mature in age, are more difficult to extract oil
or natural gas from than other types of wells, and generally operate at very low levels of
3-17
-------
Table 3-13. Aggregate Tax Data for Accounting Period 7/07-6/08: NAICS 211
Number of enterprises3 19,441
Total receipts (103) $193,230,241
Net sales(103) $166,989,539
Profit margin before tax 12.9%
Profit margin after tax 8.5%
a Includes corporations with and without net income.
Source: Internal Revenue Service, U.S. Department of Treasury. 2010. "Corporation Source Book: Data Files 2004-
2007." ; (May 2, 2010).
profitability. As a result, well operations can be quite responsive to small changes in the benefits
and costs of their operation.
In 2007, there were approximately 400,000 marginal oil wells and 320,000 marginal gas
wells (Interstate Oil and Gas Compact Commission [IOGCC], 2008). These wells provide the
United States with 4% of all oil and 8% of all natural gas consumed (IOGCC, 2008). Data for
2007 show that revenue from the over 700,000 wells was approximately $30.6 billion
(Table 3-14).
Table 3-14. Reported Gross Revenue Estimates from Marginal Wells: 2007
Well Type
Oil
Natural gas
Total
Number of Wells
396,537
322,160
718,697
Production from
Marginal Wells
291. 067592 MMbbls
1763. 592746 MCF
Estimated Gross
Revenue (S109)
$18.6
$12.0
$30.6
Source: Interstate Oil & Gas Compact Commission. 2008. "Marginal Wells: Fuel for Economic Growth." Available
at .
Historical data show marginal oil production fluctuated between 1997 and 2007,
reflecting the industry's sensitivity to changes in economic conditions of fuel markets (see
Figure 3-7). In contrast, the number of marginal gas wells has continually increased during the
past decade; the IOGCC estimates that daily production levels from these wells reached a
10-year high in 2005. Although we have been unable to find data on what fraction of these
marginal wells are operated by small businesses, the IOGCC states that many are run by "mom
and pop operators" (IOGCC, 2007).
3-18
-------
•Marginal Oil Production
•Marginal Gas Production
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Figure 3-7. Trends in Marginal Oil and Gas Production: 1997 to 2006
Source: Interstate Oil & Gas Compact Commission. 2008. "Marginal Wells: Fuel for Economic Growth." Available
at .
3.3 Pipeline Transportation of Natural Gas
3.3.1 Overview
Pipeline transportation of natural gas (NAICS 48621) is an industry group within the
transportation and warehousing sector (NAICS 48-49), but more specifically in the pipeline
transportation subsector (486). It includes the transmission of natural gas as well as the
distribution of the gas through a local network to participating businesses.
From 2002 to 2007, natural gas transportation revenues fell by 29% to just over $16
billion ($2007) (Table 3-15). At the same time, payroll decreased by 14%, while the number of
paid employees decreased by nearly 25%. The number of establishments also fell by 8% from
1,701 establishments in 2002 to 1,560 in 2007.
3.3.2 Goods and Services Used
The BEA reports pipeline transportation of natural gas only for total pipeline
transportation (3-digit NAICS 486). In addition to pipeline transportation of natural gas (NAICS
4862), this industry includes pipeline transportation of crude oil (NAICS 4861) and other
pipeline transportation (NAICS 4869). However, the BEA data are likely representative of the
affected sector since pipeline transportation of natural gas accounts for 60% of NAICS 486
establishments and 66% of revenues (Figures 1-8 and 1-9).
3-19
-------
Table 3-15. Key Statistics: Pipeline Transportation of Natural Gas (NAICS 48621) ($2007)
Year 1997 2002
Revenue ($106) 22,964 16,368
Payroll ($106) 2,438 2,086
Employees 32,542 24,519
Establishments 1,701 1,560
Sources: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
Transportation and Warehousing: Industry Series: Comparative Statistics for the United States (1997 NAICS
Basis): 2002 and 1997. ; (December 12, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48: EC0748I1:
Transportation and Warehousing: Industry Series: Preliminary Summary Statistics for the United States: 2007."
http://factfinder.census.gov (January 27, 2010).
70%
60%
60%
50%
40%
30% 25%
20% 15%
10%
0%
4862 Pipeline 4869 Other Pipeline 4861 Pipeline
Transportation of Transportation Transportation of Crude
Natural Gas Oil
Figure 3-8. Distribution of Establishments within Pipeline Transportation (NAICS 486)
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
Transportation and Warehousing: Industry Series: Summary Statistics for the United States: 2002
; (January 27, 2010).
In Table 3-16, we use the latest detailed benchmark input-output data report by the BEA
(2002) to identify the goods and services used by pipeline transportation (NAICS 486). As
shown, labor, refineries, and maintenance requirements represent significant share of the cost
associated with pipeline transportation. Power and equipment requirements potentially associated
3-20
-------
with reciprocating internal combustion engines (electric power generation and distribution)
represent less than 2% of the value of services.
70% 66%
60%
50%
40%
30%
19%
20% 15%
10%
0%
4862 Pipeline 4869 Other Pipeline 4861 Pipeline
Transportation of Transportation Transportation of Crude
Natural Gas Oil
Figure 3-9. Distribution of Revenue within Pipeline Transportation (NAICS 486)
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
Transportation and Warehousing: Industry Series: Summary Statistics for the United States: 2002"
; (January 27, 2010).
3.3.3 Business Statistics
The pipeline transportation of natural gas is clearly concentrated in the two states closest
to the refineries in the Gulf of Mexico. In 2002, Texas and Louisiana contributed to 31% of all
pipeline transportation establishments in the United States (Figure 3-10) and 41% of all U.S.
revenues. Other larger contributors with over 50 establishments in their states include Oklahoma,
Pennsylvania, Kansas, Mississippi, and West Virginia.
According to 2002 U.S. Census data, about 86% of transportation of natural gas
establishments were owned by corporations and about 8% were owned by individual
proprietorships. About 6% were owned by partnerships (Figure 3-11). As shown in Table 3-17,
the four largest firms accounted for nearly half of the establishments, and just over half, 51%, of
total revenue. The 50 largest firms accounted for over 1,354 establishments and about 99% of
total revenue. The average number of employees per establishment was approximately 17 across
all groups of firms.
3-21
-------
Enterprises within pipeline transportation (NAICS 486) generated $11.1 billion in total
receipts in 2008. Including those enterprises without net income, the industry averaged an after-
tax profit margin of 9.6% (Table 3-18).
Table 3-16. Direct Requirements for Pipeline Transportation (NAICS 486): 2002
Commodity
V00100
324110
230301
211000
333415
561300
5416AO
541300
420000
332310
5419AO
524100
531000
52AOOO
V00200
541100
221100
Commodity Description
Compensation of employees
Petroleum refineries
Nonresidential maintenance and repair
Oil and gas extraction
Air conditioning, refrigeration, and warm air heating equipment
manufacturing
Employment services
Environmental and other technical consulting services
Architectural, engineering, and related services
Wholesale trade
Plate work and fabricated structural product manufacturing
All other miscellaneous professional, scientific, and technical services
Insurance carriers
Real estate
Monetary authorities and depository credit intermediation
Taxes on production and imports, less subsidies
Legal services
Electric power generation, transmission, and distribution
Direct
Requirements
Coefficients"
14.78%
13.55%
6.07%
4.94%
4.40%
4.26%
3.04%
3.04%
2.79%
2.72%
2.48%
2.38%
2.33%
1.76%
1.41%
1.19%
1.13%
a These values show the amount of the commodity required to produce $1.00 of the industry's output. The values
are expressed in percentage terms (coefficient *100).
Source: U.S. Bureau of Economic Analysis. 2002. 2002 Benchmark Input-Output Accounts: Detailed Make Table,
Use Table and Direct Requirements Table. Tables 4 and 5.
The 2002 SUSB shows that 47% of all firms in this industry made under $5 million in
revenue. Enterprises with revenue over $100 million provided an overwhelming share of
employment in this industry (98%) (Table 3-19).
3-22
-------
o
&
Establishments by State
^ Less than 15
15-39
40-79
I 80- 149
I More than 150
Figure 3-10. 2002 Regional Distribution of Establishments: Pipeline Transportation
(NAICS 486)
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48-49:
Geographic Distribution—Pipeline transportation of natural gas: 2002. ; (November
10, 2008).
uir/o
ono/
yUvo
ono/
oU/o
vno/
f\J7o
bU/o
cno/
OU/o
/i no/
4U /o
ono/
oU/o
ono/
zu /o
•i no/
1 U /o
no/_ .
86%
8% ROA
1 1
Corporations
Individual Proprietorships
Partnerships
Figure 3-11. Share of Establishments by Legal Form of Organization in the Pipeline
Transportation of Natural Gas Industry (NAICS 48621): 2002
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48-49:
Transportation and Warehousing: Subject Series—Estab & Firm Size: Legal Form of Organization for the United
States: 2002. ; (December 12, 2008).
3-23
-------
Table 3-17. Firm Concentration for Pipeline Transportation of Natural Gas (NAICS
48621): 2002
Receipts/Revenue
Commodity
All firms
4 largest firms
8 largest firms
20 largest firms
50 largest firms
Establishments
1,431
698
912
1,283
1,354
Amount ($106)
$14,797
$7,551
$10,059
$13,730
$14,718
Percentage of
Total
100%
51%
68%
93%
99%
Number of
Employees
23,677
11,814
15,296
21,792
23,346
Employees per
Establishment
16.5
16.9
16.8
17.0
17.2
Source: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
Transportation and Warehousing: Subject Series—Estab & Firm Size: Concentration by Largest Firms for the
United States: 2002" ; (December 12, 2008).
Table 3-18. Aggregate Tax Data for Accounting Period 7/07-6/08: NAICS 486
Number of enterprises3
Total receipts (103)
Net sales (103)
Profit margin before tax
Profit margin after tax
321
$11,062,608
$10,210,083
13.2%
9.6%
a Includes corporations with and without net income.
Source: Internal Revenue Service, U.S. Department of Treasury. 2010. "Corporation Source Book: Data Files 2004-
2007." ; (May 2, 2010).
3-24
-------
Table 3-19. Key Enterprise Statistics by Receipt Size for Pipeline Transportation of Natural Gas (NAICS 48621): 2002
Variable
Firms
Establishments
Employment
Receipts ($103)
Receipts/firm ($103)
Receipts/establishment
($103)
Receipts/employment
V° ($)
All
Enterprises
154
1,936
37,450
$35,896,535
$233,094
$18,542
$958,519
0-99K
Receipts
8
8
15
$524
$66
$66
$34,933
100-
499.9K
Receipts
32
32
58
$8,681
$271
$271
$149,672
500-999.9K
Receipts
10
10
69
$7,451
$745
$745
$107,986
Owned
1,000-
4,999.9K
Receipts
22
22
138
$46,429
$2,110
$2,110
$336,442
by Enterprises with
5,000,000-
9,999,999K
Receipts
6
7
88
$40,967
$6,828
$5,852
$465,534
<10,OOOK
Receipts
78
79
368
$104,052
$1,334
$1,317
$282,750
10,000-
49,999K
Receipts
11
21
216
$188,424
$17,129
$8,973
$872,333
50,000-
99,999K
Receipts
4
4
274
$154,384
$38,596
$38,596
$563,445
100,OOOK+
Receipts
61
1,832
36,592
$35,449,675
$581,142
$19,350
$968,782
Source: U.S. Census Bureau. 2008b. Firm Size Data from the Statistics of U.S. Businesses, U.S. All Industries Tabulated by Receipt Size: 2002.
http://www2.census.gov/csd/susb/2002/usalli_r02.xls.
-------
SECTION 4
REGULATORY ALTERNATIVES, COSTS, AND EMISSION IMPACTS
4.1 Background
This action promulgates NESHAP for existing stationary SI RICE with a site rating of
less than or equal to 500 HP located at major sources, and existing stationary SI RICE of any site
rating located at area sources. EPA is finalizing these standards to meet its statutory obligation to
address HAP emissions from these sources under sections 112(d), 112(c)(3) and 112(k) of the
CAA. The final NESHAP for stationary RICE will be promulgated under 40 CFR part 63,
subpart ZZZZ, which already contains standards applicable to new and reconstructed stationary
RICE and some existing stationary RICE.
EPA promulgated NESHAP for existing, new, and reconstructed stationary RICE greater
than 500 HP located at major sources on June 15, 2004 (69 FR 33474). EPA promulgated
NESHAP for new and reconstructed stationary RICE that are located at area sources of HAP
emissions and for new and reconstructed stationary RICE that have a site rating of less than or
equal to 500 HP that are located at major sources of HAP emissions on January 18, 2008 (73 FR
3568). On March 3, 2010, EPA promulgated NESHAP for existing stationary compression
ignition (CI) RICE with a site rating of less than or equal to 500 HP located at major sources,
existing non-emergency CI engines with a site rating greater than 500 HP at major sources, and
existing stationary CI RICE of any site rating located at area sources (75 FR 9674).
4.2 Summary of the Final Rule
4.2.1 What Is the Source Category Regulated by the Final Rule?
The final rule addresses emissions from existing stationary SI engines less than or equal
to 500 HP located at major sources and all existing stationary SI engines located at area sources.
A major source of HAP emissions is generally a stationary source that emits or has the potential
to emit 10 tons per year or more of any single HAP or 25 tons per year or more of any
combination of HAP. An area source of HAP emissions is a stationary source that is not a major
source.
This action revises the regulations at 40 CFR part 63, subpart ZZZZ. Through this action,
we are adding to 40 CFR part 63, subpart ZZZZ requirements for: existing SI stationary RICE
less than or equal to 500 HP located at major sources of HAP and existing SI stationary RICE
located at area sources of HAP.
4-1
-------
4.2.1.1 Existing Stationary SI RICE <500 HP at Major Sources of HAP
This action revises 40 CFR part 63, subpart ZZZZ, to address HAP emissions from
existing stationary SI RICE less than or equal to 500 HP located at major sources of HAP. For
stationary engines less than or equal to 500 HP at major sources, EPA must determine what is the
appropriate maximum achievable control technology (MACT) for those engines under sections
112(d)(2) and (d)(3) of the CAA.
EPA has divided stationary SI RICE less than or equal to 500 HP located at major
sources of HAP into the following subcategories:
• Non-emergency 2-stroke lean burn (2SLB) stationary SI RICE 100-500 HP;
• Non-emergency 4-stroke lean burn (4SLB) stationary SI RICE 100-500 HP;
• Non-emergency 4-stroke rich burn (4SRB) stationary SI RICE 100-500 HP;
• Non-emergency landfill and digester gas stationary SI RICE 100-500 HP;
• Non-emergency stationary SI RICE <100 HP; and
• Emergency stationary SI RICE.
4.2.1.2 Existing Stationary SI RICE at Area Sources of HAP
This action revises 40 CFR part 63, subpart ZZZZ, in order to address HAP emissions
from existing stationary SI RICE located at area sources of HAP. Section 112(d) of the CAA
requires EPA to establish NESHAP for both major and area sources of HAP that are listed for
regulation under CAA section 112(c). As noted above, an area source is a stationary source that
is not a major source.
Section 112(k)(3)(B) of the CAA calls for EPA to identify at least 30 HAP that, as a
result of emissions of area sources, pose the greatest threat to public health in the largest number
of urban areas. EPA implemented this provision in 1999 in the Integrated Urban Air Toxics
Strategy (64 FR 38715, July 19, 1999). Specifically, in the Strategy, EPA identified 30 HAP that
pose the greatest potential health threat in urban areas, and these HAP are referred to as the "30
urban HAP." Section 112(c)(3) of the CAA requires EPA to list sufficient categories or
subcategories of area sources to ensure that area sources representing 90 percent of the emissions
of the 30 urban HAP are subject to regulation. EPA implemented these requirements through the
Integrated Urban Air Toxics Strategy (64 FR 38715, July 19, 1999). The area source stationary
engine source category was one of the listed categories. A primary goal of the Strategy is to
4-2
-------
achieve a 75 percent reduction in cancer incidence attributable to HAP emitted from stationary
sources.
Under CAA section 112(d)(5), EPA may elect to promulgate standards or requirements
for area sources "which provide for the use of generally available control technologies or
management practices by such sources to reduce emissions of hazardous air pollutants."
Additional information on generally available control technologies (GACT) and management
practices is found in the Senate report on the legislation (Senate report Number 101-228,
December 20, 1989), which describes GACT as:
. . . methods, practices and techniques which are commercially available and appropriate
for application by the sources in the category considering economic impacts and the
technical capabilities of the firms to operate and maintain the emissions control systems.
Consistent with the legislative history, EPA can consider costs and economic impacts in
determining GACT, which is particularly important when developing regulations for source
categories, like this one, that have many small businesses.
Determining what constitutes GACT involves considering the control technologies and
management practices that are generally available to the area sources in the source category.
EPA also considers the standards applicable to major sources in the same industrial sector to
determine if the control technologies and management practices are transferable and generally
available to area sources. In appropriate circumstances, EPA may also consider technologies and
practices at area and major sources in similar categories to determine whether such technologies
and practices could be considered generally available for the area source category at issue.
Finally, as EPA has already noted, in determining GACT for a particular area source category,
EPA considers the costs and economic impacts of available control technologies and
management practices on that category.
The urban HAP that must be regulated from stationary SI RICE to achieve the CAA
section 112(c)(3) requirement to regulate categories accounting for 90 percent of the urban HAP
are: 7 polycyclic aromatic hydrocarbons (PAH), formaldehyde, and acetaldehyde.
Similar to existing stationary SI RICE at major sources, EPA has also divided the
existing stationary SI RICE at area sources into subcategories in order to properly take into
account the differences between these engines. The subcategories for stationary SI RICE at area
sources are as follows:
• Non-emergency 2SLB stationary SI RICE
4-3
-------
• Non-emergency 4SLB stationary SI RICE
- <500HP
- >500HP
• Non-emergency 4SRB stationary SI RICE
- <500HP
- >500HP
• Non-emergency landfill and digester gas stationary SI RICE
• Emergency stationary SI RICE.
4.2.2 What Are the Pollutants Regulated by the Rule?
The final rule regulates emissions of HAP. Available emissions data show that several
HAP, which are formed during the combustion process or which are contained within the fuel
burned, are emitted from stationary engines. The HAP which have been measured in emission
tests conducted on SI stationary RICE include: formaldehyde, acetaldehyde, acrolein, methanol,
benzene, toluene, 1,3-butadiene, 2,2,4-trimethylpentane, hexane, xylene, naphthalene, PAH,
methylene chloride, and ethylbenzene. EPA described the health effects of these HAP and other
HAP emitted from the operation of stationary RICE in the preamble to 40 CFR part 63, subpart
ZZZZ, published on June 15, 2004 (69 FR 33474). These HAP emissions are known to cause, or
contribute significantly to air pollution, which may reasonably be anticipated to endanger public
health or welfare.
For the standards being finalized in this action, EPA believes that previous
determinations regarding the appropriateness of using formaldehyde and carbon monoxide (CO)
both in concentration (parts per million [ppm]) levels as surrogates for HAP for stationary RICE
are still valid. Consequently, EPA is promulgating CO or formaldehyde standards in order to
regulate HAP emissions.
In addition to reducing HAP, the emission control technologies that will be installed on
stationary RICE to reduce HAP will also reduce CO and VOC, and for rich burn engines will
also reduce NOx.
4-4
-------
4.2.3 What Are the Final Requirements ?
4.2.3.1 Existing Stationary SI RICE <500 HP at Major Sources
The numerical emission standards that are being finalized for existing stationary non-
emergency SI RICE less than or equal to 500 HP located at major sources of HAP are shown in
Table 4-1. The emission standards are in units of ppm by volume, dry basis (ppmvd).
Table 4-1. Emission Standards for Existing Stationary SI RICE <500 HP Located at
Major Sources of HAP
Subcategory Except during Periods of Startup
2SLB Non-Emergency 100
-------
an oil analysis program as discussed below and none of the condemning limits are
exceeded;
• Inspect spark plugs every 1,440 hours of operation or annually, whichever comes
first, and replace as necessary; and
• Inspect all hoses and belts every 1,440 hours of operation or annually, whichever
comes first, and replace as necessary.
Existing 2SLB stationary SI RICE less than 100 HP located at major sources of HAP are subject
to the following work practices:
• Change oil and filter every 4,320 hours of operation or annually, whichever comes
first, except that sources can extend the period for changing the oil if the oil is part of
an oil analysis program as discussed below and none of the condemning limits are
exceeded;
• Inspect spark plugs every 4,320 hours of operation or annually, whichever comes
first, and replace as necessary; and
• Inspect all hoses and belts every 4,320 hours of operation or annually, whichever
comes first, and replace as necessary.
Sources also have the option to use an oil change analysis program to extend the oil
change frequencies specified above. The analysis program must at a minimum analyze the
following three parameters: Total Acid Number, viscosity, and percent water content. The
analysis must be conducted at the same frequencies specified for changing the engine oil. If the
condemning limits provided below are not exceeded, the engine owner or operator is not
required to change the oil. If any of the condemning limits are exceeded, the engine owner or
operator must change the oil within two days of receiving the results of the analysis; if the engine
is not in operation when the results of the analysis are received, the engine owner or operator
must change the oil within two days or before commencing operation, whichever is later. The
condemning limits are as follows:
• Total Acid Number increases by more than 3.0 milligrams (mg) potassium hydroxide
per gram (KOH/g) from Total Acid Number of the oil when new; or
• viscosity of the oil changes by more than 20 percent from the viscosity of the oil
when new; or
• percent water content (by volume) is greater than 0.5.
Pursuant to the provisions of 40 CFR 63.6(g), sources can also request that the
Administrator approve alternative work practices.
4-6
-------
4.2.3.2 Existing Stationary SI RICE at Area Sources of HAP.
The numerical emission standards that EPA is finalizing for non-emergency 4SLB
stationary SI RICE and non-emergency 4SRB stationary SI RICE located at area sources of HAP
are shown in Table 4-2.
Table 4-2. Numerical Emission Standards for Existing Non-Emergency Stationary 4SLB
and 4SRB SI RICE >500 HP Located at Area Sources of HAP
Subcategory Except during Periods of Startup
4SLB Non-Emergency >500 HP that operate more than 47 ppmvd CO at 15% O2 or 93% CO reduction
24 hours per calendar year
4SRB Non-Emergency >500 HP that operate more than 2.7 ppmvd formaldehyde at 15% O2 or 76%
24 hours per calendar year formaldehyde reduction
EPA is finalizing management practices for existing non-emergency 4SLB stationary SI
RICE less than or equal to 500 HP located at area sources of HAP, existing non-emergency
4SLB stationary SI RICE greater than 500 HP located at area sources of HAP that operate 24
hours or less per calendar year, existing non-emergency 4SRB stationary SI RICE less than or
equal to 500 HP located at area sources of HAP, existing non-emergency 4SRB stationary SI
RICE greater than 500 HP located at area sources of HAP that operate 24 hours or less per
calendar year, existing 2SLB non-emergency stationary SI RICE located at area sources of HAP,
existing non-emergency landfill and digester gas stationary RICE located at area sources of
HAP, and existing emergency stationary SI RICE located at area sources of HAP.
Existing non-emergency 4SLB and 4SRB stationary SI RICE less than or equal to 500
HP located at area sources of HAP and existing landfill or digester gas non-emergency stationary
SI RICE located at area sources of HAP are subject to the following management practices:
• Change oil and filter every 1,440 hours of operation or annually, whichever comes
first, except that sources can extend the period for changing the oil if the oil is part of
an oil analysis program as discussed below and none of the condemning limits are
exceeded;
• Inspect spark plugs every 1,440 hours of operation or annually, whichever comes
first, and replace as necessary; and
• Inspect all hoses and belts every 1,440 hours of operation or annually, whichever
comes first, and replace as necessary.
4-7
-------
Existing stationary 2SLB non-emergency engines located at area sources of HAP are
subject to the following work practices:
• Change oil and filter every 4,320 hours of operation or annually, whichever comes
first, except that sources can extend the period for changing the oil if the oil is part of
an oil analysis program as discussed below and none of the condemning limits are
exceeded;
• Inspect spark plugs every 4,320 hours of operation or annually, whichever comes
first, and replace as necessary; and
• Inspect all hoses and belts every 4,320 hours of operation or annually, whichever
comes first, and replace as necessary.
Existing stationary emergency SI RICE located at area sources of HAP and existing non-
emergency 4SLB and 4SRB stationary SI RICE greater than 500 HP located at area sources of
HAP that operate 24 hours or less per calendar year are subject to the following work practices:
• Change oil and filter every 500 hours of operation or annually, whichever comes first,
except that sources can extend the period for changing the oil if the oil is part of an oil
analysis program as discussed below and none of the condemning limits are
exceeded;
• Inspect spark plugs every 1,000 hours of operation or annually, whichever comes
first, and replace as necessary; and
• Inspect all hoses and belts every 500 hours of operation or annually, whichever comes
first, and replace as necessary.
As discussed above for major sources, these sources may utilize an oil analysis program
in order to extend the specified oil change requirement specified above. Also, sources have the
option to work with state permitting authorities pursuant to EPA's regulations at 40 CFR subpart
E ("Approval of State Programs and Delegation of Federal Authorities") for approval of
alternative management practices. 40 CFR subpart E implements section 112(1) of the CAA,
which authorizes EPA to approve alternative state/local/tribal HAP standards or programs when
such requirements are demonstrated to be no less stringent than EPA promulgated standards.
4.2.3.3 Startup Requirements
Existing stationary SI RICE less than or equal to 500 HP located at major sources of HAP
and existing stationary SI RICE located at area sources of HAP must meet specific operational
standards during engine startup. Engine startup is defined as the time from initial start until
applied load and engine and associated equipment reaches steady state or normal operation. For
4-8
-------
stationary engines with catalytic controls, engine startup means the time from initial start until
applied load and engine and associated equipment reaches steady state, or normal operation,
including the catalyst. Owners and operators must minimize the engine's time spent at idle and
minimize the engine's startup to a period needed for appropriate and safe loading of the engine,
not to exceed 30 minutes, after which time the engine must meet the otherwise applicable
emission standards. These requirements will limit the HAP emissions during periods of engine
startup. Pursuant to the provisions of 40 CFR 63.6(g), engines at major sources may petition the
Administrator for an alternative work practice. An owner or operator of an engine at an area
source can work with its State permitting authority pursuant to EPA's regulations at 40 CFR
subpart E for approval of an alternative management practice. See 40 CFR subpart E (setting
forth requirements for, among other things, equivalency by permit, rule substitution).
4.2.4 What Are the Operating Limitations ?
In addition to the standards discussed above, EPA is finalizing operating limitations for
stationary non-emergency 4SLB and 4SRB RICE that are greater than 500 HP and located at an
area source of HAP and operated more than 24 hours per calendar year. Owners and operators of
engines that are equipped with oxidation catalyst or non-selective catalytic reduction (NSCR)
must maintain the catalyst so that the pressure drop across the catalyst does not change by more
than 2 inches of water from the pressure drop across the catalyst that was measured during the
initial performance test. If the engine is equipped with oxidation catalyst, owners and operators
must also maintain the temperature of the stationary RICE exhaust so that the catalyst inlet
temperature is between 450 and 1,350 degrees Fahrenheit (°F). If the engine is equipped with
NSCR, owners and operators must maintain the temperature of the stationary RICE exhaust so
that the NSCR inlet temperature is between 750 and 1,250 °F. Owners and operators may
petition for a different temperature range; the petition must demonstrate why it is operationally
necessary and appropriate to operate below the temperature range specified in the final rule (see
40 CFR 63.8(f)). Owners and operators of engines that are not using oxidation catalyst or NSCR
must comply with any operating limitations approved by the Administrator.
4.2.5 What Are the Requirements for Demonstrating Compliance?
The following sections describe the requirements for demonstrating compliance under the
final rule.
4.2.5.1 Existing Stationary SI RICE <500 at Major Sources of HAP.
Owners and operators of existing stationary non-emergency SI RICE located at major
sources that are less than 100 HP and existing stationary emergency SI RICE located at major
4-9
-------
sources must operate and maintain their stationary RICE and aftertreatment control device (if
any) according to the manufacturer's emission-related written instructions or develop their own
maintenance plan. The maintenance plan must specify how the work practices will be met and
provide to the extent practicable for the maintenance and operation of the engine in a manner
consistent with good air pollution control practices for minimizing emissions. Owners and
operators of existing stationary non-emergency SI RICE located at major sources that are less
than 100 HP and existing stationary emergency SI RICE located at major sources do not have to
conduct any performance testing because they are not subject to numerical emission standards.
Owners and operators of existing stationary non-emergency SI RICE located at major
sources that are greater than or equal to 100 HP and less than or equal to 500 HP must conduct
an initial performance test to demonstrate that they are achieving the required emission
standards.
4.2.5.2 Existing Stationary SI RICE at Area Sources of HAP
Owners and operators of existing stationary RICE located at area sources of HAP that are
subject to management practices do not have to conduct any performance testing; they must
develop a maintenance plan that specifies how the management practices will be met and
provides to the extent practicable for the maintenance and operation of the engine in a manner
consistent with good air pollution control practices for minimizing emissions. Owners and
operators of existing 4SLB and 4SRB non-emergency stationary SI RICE that are greater than
500 HP and located at an area source of HAP , and operated more than 24 hours per calendar
year must conduct an initial performance test to demonstrate compliance with the applicable
emission limitations and must conduct subsequent performance testing every 8,760 hours of
operation or 3 years, whichever comes first. Owners and operators of existing 4SLB and 4SRB
non-emergency stationary SI RICE that are greater than 500 HP and located at an area source of
HAP , and operated more than 24 hours per calendar year must continuously monitor and record
the inlet temperature of the oxidation catalyst or NSCR and also take monthly measurements of
the pressure drop across the oxidation catalyst or NSCR. If an oxidation catalyst or NSCR is not
being used on the engine, the owner or operator must continuously monitor and record the
operating parameters (if any) approved by the Administrator. As discussed in the March 3, 2010,
final NESHAP for existing stationary CI RICE (75 FR 9648) and in section V.E. of the
preamble, EPA is finalizing performance specification requirements in 40 CFR part 63, subpart
ZZZZ for the continuous parametric monitoring systems used for continuous catalyst inlet
temperature monitoring.
4-10
-------
4.2.6 What Are the Reporting and Recordkeeping Requirements ?
The following sections describe the reporting and recordkeeping requirements that are
required under the final rule.
Owners and operators of existing stationary emergency SI RICE that do not meet the
requirements for non-emergency engines are required to keep records of their hours of operation.
Owners and operators of existing stationary emergency SI RICE must install a non-resettable
hour meter on their engines to record the hours of operation of the engine.
Owners and operators of existing stationary SI RICE located at major sources that are
subject to work practices and existing stationary SI RICE located at area sources that are subject
to management practices are required to keep records that show that the work or management
practices that are required are being met. These records must include, at a minimum: oil and
filter change dates and engine hours of operation; inspection and replacement dates for spark
plugs, hoses, and belts; and records of other emission-related repairs and maintenance
performed.
In terms of reporting requirements, owners and operators of existing non-emergency
stationary SI RICE greater than or equal to 100 HP and less than or equal to 500 HP located at
major sources of HAP and existing non-emergency 4SLB and 4SRB stationary RICE greater
than 500 HP located at area sources of HAP that are operated more than 24 hours per calendar
year must submit the notifications required in Table 8 of 40 CFR part 63, subpart ZZZZ, which
lists the NESHAP General Provisions applicable to this rule. (40 CFR part 63, subpart A) These
notifications include an initial notification, notification of performance test, and a notification of
compliance for each stationary RICE which must comply with the specified emission limitations.
Owners and operators of existing stationary non-emergency SI RICE greater than or equal to 100
HP and less than or equal to 500 HP located at major sources of HAP and existing stationary
4SLB and 4SRB non-emergency SI RICE greater than 500 HP located at area sources of HAP
that are operated more than 24 hours per calendar year must submit semiannual compliance
reports.
4.3 Summary of Significant Changes Since Proposal
4.3.1 Applicability
A change from the proposal is that the final rule is not applicable to existing stationary
emergency engines at area sources that are located at residential, commercial, or institutional
facilities. These engines are not subject to any requirements under the final rule because they are
4-11
-------
not part of the regulated source category. EPA has found that existing stationary emergency
engines located at residential, commercial, and institutional facilities that are area sources were
not included in the original Urban Air Toxics Strategy inventory and were not included in the
listing of urban area sources. More information on this issue can be found in the memorandum
titled, "Analysis of the Types of Engines Used to Estimate the CAA Section 112(k) Area Source
Inventory for Stationary Reciprocating Internal Combustion Engines," available from the
rulemaking docket. In the March 3, 2010, final NESHAP for existing stationary CI RICE (75 FR
9648), EPA included a definition for residential/commercial/institutional emergency stationary
RICE. After the final rule was promulgated, EPA received numerous questions regarding the
definition and whether certain types of facilities would meet the definition. In the final rule, EPA
is separating the definition into individual definitions for residential emergency stationary RICE,
commercial emergency stationary RICE, and institutional emergency stationary RICE, and is
also providing additional examples of the types of facilities that would be included under those
categories in the definitions. EPA has also prepared a memorandum to provide further guidance
regarding the types of facilities that would or would not be considered residential, commercial,
or institutional facilities. The memorandum is titled, "Guidance Regarding Definition of
Residential, Commercial, and Institutional Emergency Stationary RICE in the NESHAP for
Stationary RICE," and is available in the rulemaking docket.
4.3.2 Final Emission Standards
4.3.2.1 Existing Stationary SI Engines <500 HP Located at Major Sources of HAP
EPA is revising the emission standards that it proposed for the subcategories of stationary
SI engines less than or equal to 500 HP located at major sources. As discussed in section V.B. of
the preamble, numerous commenters indicated that EPA's dataset used to establish the proposed
emission limits was insufficient and urged EPA to gather more data to obtain a more complete
representation of emissions from existing stationary SI engines. Commenters also questioned the
emission standard setting approach that EPA used at proposal and claimed that the proposed
standards did not take into account emissions variability. For the final rule, EPA has obtained
additional test data for existing stationary SI engines and has included this additional data in the
MACT floor analysis. EPA is also using an approach that better considers emissions variability,
as discussed below. EPA is also not using the Population Database to determine a percentage of
engines that have emission controls installed, as it did at proposal. The Population Database has
not been updated since 2000. It contains information regarding whether or not an engine has
emission controls, but does not generally contain other types of emission-related information,
like engine-out emissions or operational controls, and it does not include any emissions
4-12
-------
concentration data, which is necessary to determine the MACT floor. EPA determined that it
would be more appropriate and more defensible to base the MACT floor analysis directly on the
emissions data that EPA has for stationary SI engines.
For 2SLB non-emergency engines, EPA proposed a limit of 85 ppmvd CO for engines
from 50 to 249 HP and 8 ppmvd CO or 90 percent CO reduction for engines greater than or
equal to 250 HP. EPA is finalizing an emission limit of 225 ppmvd CO for 2SLB non-emergency
engines from 100 to 500 HP. For 4SLB non-emergency engines, EPA proposed a limit of 95
ppmvd CO for engines from 50 to 249 HP and 9 ppmvd CO or 90 percent CO reduction for
engines greater than or equal to 250 HP. EPA is finalizing an emission limit of 47 ppmvd CO for
4SLB non-emergency engines from 100 to 500 HP. For 4SRB non-emergency engines from 50
to 500 HP, EPA proposed an emission limit of 200 ppbvd (parts per billion by volume, dry basis)
formaldehyde or 90 percent formaldehyde reduction. EPA is finalizing an emission limit of 10.3
ppmvd formaldehyde for 4SRB non-emergency engines from 100 to 500 HP. For landfill and
digester gas engines, EPA proposed an emission limit of 177 ppmvd CO; EPA is finalizing an
emission limit of 177 ppmvd CO.
For the proposed rule, EPA required existing stationary engines less than 50 HP that are
located at major sources to meet a formaldehyde emission standard. As discussed in the final rule
published on March 3, 2010, for existing stationary CI RICE (75 FR 9674), EPA is not finalizing
a formaldehyde emission standard for stationary SI engines less than 50 HP, but is instead
requiring compliance with work practices. In addition, in light of several comments asserting that
the level at which EPA subcategorized small engines at major sources was inappropriate, EPA is
finalizing a work practice standard for engines less than 100 HP. These work practices are
described in section III.C. of the preamble to the final rule. EPA believes that work practices are
appropriate and justified for this group of stationary engines because the application of
measurement methodology is not practicable due to technological and economic limitations.
Further information on EPA's decision can be found in the memorandum titled, "MACT Floor
and MACT Determination for Existing Stationary Non-Emergency SI RICE <100 HP and
Existing Stationary Emergency SI RICE Located at Major Sources and GACT for Existing
Stationary SI RICE Located at Area Sources," which is available from the rulemaking docket.
For existing stationary emergency engines located at major sources, EPA proposed that
these engines be subject to a 2 ppmvd formaldehyde emission standard. In the final rule, existing
stationary emergency SI engines located at major sources of HAP must meet work practices.
These work practices are described in section III.C. of the preamble to the final rule. EPA
believes that work practices are appropriate and justified for this group of stationary engines
4-13
-------
because the application of measurement methodology is not practicable due to technological and
economic limitations. Further information on EPA's decision can be found in the memorandum
titled, "MACT Floor and MACT Determination for Existing Stationary Non-Emergency SI
RICE <100 HP and Existing Stationary Emergency SI RICE Located at Major Sources and
GACT for Existing Stationary SI RICE Located at Area Sources," which is available from the
rulemaking docket.
4.3.2.2 Existing Stationary SI Engines Located at Area Sources of HAP
EPA proposed numerical emission standards for the following stationary SI engines
located at area sources of HAP: non-emergency 2SLB and 4SLB greater than or equal to 250
HP, non-emergency 4SRB greater than or equal to 50 HP, landfill and digester gas fired greater
than 500 HP, and emergency greater than 500 HP. For the remaining engines at area sources,
EPA proposed management practice standards.
In the final rule, EPA is promulgating numerical emission standards for non-emergency
4SLB and 4SRB stationary SI RICE larger than 500 HP located at area sources of HAP
emissions that operate more than 24 hours per calendar year. For non-emergency 4SLB engines
greater than 500 HP located at area sources of HAP, EPA proposed an emission limit of 9 ppmvd
CO or 90 percent CO reduction; EPA is finalizing an emission limit of 47 ppmvd CO or 93
percent CO reduction. For non-emergency 4SRB engines greater than 500 HP located at area
sources of HAP, EPA proposed an emission limit of 200 ppbvd formaldehyde or 90 percent
formaldehyde reduction and is finalizing an emission limit of 2.7 ppmvd formaldehyde or 76
percent formaldehyde reduction. For stationary SI RICE located at area sources of HAP that are
non-emergency 2SLB stationary SI RICE greater than or equal to 250 HP, non-emergency 4SLB
stationary SI RICE between 250 and 500 HP, non-emergency 4SRB stationary SI RICE between
50 and 500 HP, landfill/digester gas stationary SI RICE greater than 500 HP, or emergency
stationary SI RICE greater than 500 HP, EPA is finalizing management practices rather than
numeric emission limitations as proposed. EPA is also finalizing management practices for non-
emergency 4SLB and 4SRB stationary SI RICE that are greater than 500 HP, located at area
sources of HAP, and operated 24 hours or less per calendar year.
4.3.3 Management Practices
EPA proposed management practices for several subcategories of engines located at area
sources. EPA explained that the proposed management practices would be expected to ensure
that emission control systems are working properly and would help minimize HAP emissions
from the engines. EPA proposed specific maintenance practices and asked for comments on the
4-14
-------
need and appropriateness for those procedures. Based on feedback received during the public
comment period, which included information submitted in comment letters and additional
information EPA received following the close of the comment period from different industry
groups, EPA is finalizing management practices for existing stationary 2SLB non-emergency SI
engines located at area sources of HAP, existing stationary 4SLB and 4SRB non-emergency SI
engines less than or equal to 500 HP located at area sources of HAP; existing stationary landfill
and digester gas non-emergency engines located at area sources of HAP; and all existing
emergency stationary SI engines located at area sources of HAP.
Based on the comments on the proposal and additional information received from
stakeholders, EPA made changes to the intervals for the management practices from the
proposal. EPA is also adding an option for sources to use an oil change analysis program to
extend the oil change frequencies specified above. The analysis program must at a minimum
analyze the following three parameters: Total Acid Number, viscosity, and percent water
content. If the condemning limits for these parameters are not exceeded, the engine owner or
operator is not required to change the oil. If any of the limits are exceeded, the engine owner or
operator must change the oil within two days of receiving the results of the analysis; if the engine
is not in operation when the results of the analysis are received, the engine owner or operator
must change the oil within two days or before commencing operation, whichever is later. Owners
and operators of all engines subject to management practices also have the option to work with
State permitting authorities pursuant to EPA's regulations at 40 CFR subpart E for alternative
management practices to be used instead of the specific management practices promulgated in
the final rule. The management practices must be at least as stringent as those specified in the
final rule.
4.3.4 Startup, Shutdown, and Malfunction
EPA proposed formaldehyde and CO emission standards for existing stationary engines
at major sources to apply during periods of startup and malfunction. EPA also proposed certain
standards for existing stationary engines at area sources that would apply during startup and
malfunction. EPA did not propose distinct standards for periods of shutdown. EPA proposed that
engines would be subject to the same standards during shutdown as are applicable during other
periods of operation.
Based on various comments and concerns with the proposed emission standards for
periods of startup, EPA has determined that it is not feasible to finalize numerical emission
standards that would apply during startup because the application of measurement methodology
4-15
-------
to this operation is not practicable due to technological and economic limitations. This issue is
discussed in detail in the final rule published on March 3, 2010 (75 FR 9674), and as discussed in
the Response to Comments for this rule, the analysis is the same for the engines regulated in this
final rule.
As a result, EPA is extending the operational standards during startup it promulgated in
the March 3, 2010, final rule (75 FR 9674), which specify that owners and operators must limit
the engine startup time to no more than 30 minutes and must minimize the engine's time spent at
idle during startup, to the engines newly subject to regulation in this rule.
With respect to malfunctions, EPA proposed two options for subcategories where the
proposed emission standard was based on the use of catalytic controls. The first proposed option
was to have the same standards apply during normal operation and malfunctions. The second
proposed option was that standards during malfunctions be based on emissions expected from
the best controlled sources prior to the full warm-up of the catalytic control. For subcategories
where the proposed emission standard was not based on the use of catalytic controls, we
proposed the same emission limitations apply during malfunctions and periods of normal
operations. EPA is finalizing the first option described above, which is that the same standards
apply during normal operation and malfunctions. In the proposed rule, EPA expressed the view
that there are different modes of operation for any stationary source, and that these modes
generally include startup, normal operations, shutdown, and malfunctions. However, as
discussed in detail in the final rule published on March 3, 2010 (75 FR 9674), and as discussed in
the Response to Comments for this rule, after considering the issue of malfunctions more
carefully, EPA has determined that malfunctions should not be viewed as a distinct operating
mode and, therefore, any emissions that occur at such times do not need to be factored into
development of CAA section 112(d) standards, which, once promulgated, apply at all times. In
addition, as discussed in detail in the final rule published on March 3, 2010 (75 FR 9674), and as
discussed in the Response to Comments for this rule, EPA believes that malfunctions will not
cause stationary engines to violate the standard that applies during normal operations. Therefore,
the standards that apply during normal operation also apply during malfunction.
4.3.5 Method 323
EPA proposed to remove Method 323 as an option for determining compliance with
formaldehyde emission limitations in 40 CFR part 63, subpart ZZZZ. EPA Method 323 was first
proposed as part of the NESHAP for Stationary Combustion Turbines published January 14,
2003, (68 FR 1888) for measuring formaldehyde emissions from natural gas-fired sources.
4-16
-------
However, the method was not included in the final Stationary Combustion Turbines NESHAP
due to reliability concerns and EPA never promulgated EPA Method 323 as a final standard in
40 CFR part 63, appendix A. Due to unresolved technical issues with the method affecting
engine test results, EPA found it appropriate to propose to remove the method from 40 CFR part
63, subpart ZZZZ. As discussed in greater detail in section V.D. of the preamble, after EPA
proposed to remove Method 323 as a compliance test Method, the Agency received test data
comparing Method 323 to EPA Method 320. The results of this comparison testing showed good
agreement between the two methods and there was no evidence of bias in the results from
Method 323. Therefore, EPA has determined that it is appropriate to promulgate Method 323 and
to allow it as an option for measuring formaldehyde in 40 CFR part 63 subpart ZZZZ.
4.4 Cost Impacts
4.4.1 Introduction
EPA has determined that oxidation catalysts for two-stroke lean burn (2SLB) and four-
stroke lean burn (4SLB) engines, and non-selective catalytic reduction (NSCR) for four-stroke
rich burn (4SRB) engines are applicable controls for HAP reduction from existing stationary SI
RICE. To determine the capital and annual costs for these control technologies, equipment cost
information was obtained from industry groups2 and vendors and manufacturers of SI engine
control technology. In some cases, the industry groups provided a breakdown of the capital and
annual cost components for each of the retrofit options. Using this cost data, annualized cost and
capital cost equations for oxidation catalysts and NSCR were developed.
4.4.2 Control Cost Methodology
The following sections describe the methodology used to derive the total capital and total
annual costs for each of the control technology options. These methodologies were used to
calculate total capital and total annual costs when only purchased equipment costs were available
(e.g., vendor equipment costs). The methodologies were not used for cost data provided by
industry groups because they included a breakdown of the actual total capital and total annual
costs. A summary of the methodologies, equations, and assumptions used to estimate the total
capital and total annual costs for some of the cost data are described in the following sections.
! Reciprocating Internal Combustion Engine National Emission Standards for Hazardous Air Pollutants (RICE
NESHAP) Proposed Revisions - Emission Control Costs Analysis Background for "Above the Floor" Emission
Controls for Natural Gas-Fired RICE, Innovative Environmental Solutions Inc., October 2009. (EPA-HQ-OAR-
2008-0708-0279).
4-17
-------
4.4.2.1 Total Capital Costs
The total capital cost includes the direct and indirect costs of purchasing and installing
the control equipment. The direct cost includes the cost of purchasing the equipment and
instrumentation, cost of shipping, and the cost of installing the control equipment. The indirect
cost includes the costs for engineering, contractor fees, testing costs, and also includes costs for
contingencies, such as additional modifications, or delays in startup. The total capital cost
equation can be summarized as follows:
Total Capital Cost (TCC) = Direct Costs (DC) + Indirect Costs (1C)
The direct costs include the costs of purchasing and installing the control equipment and can be
summarized using the following equation;
DC = Purchased Equipment Cost (PEC) + Direct Installation Costs (DIC).
A summary of the cost assumptions for PEC includes the following:
• Control Device and Auxiliary Equipment (EC);
• Instrumentation (10% of EC);
• Sales Tax (3% of EC);
• Freight (5% of EC);
and can be summarized as:
PEC= 118% EC.
A summary of the cost assumptions for DIC includes the following:
• Foundations and Supports (8% of PEC);
• Handling and Erection (14% of PEC);
• Electrical (4% of PEC);
• Piping (2% of PEC);
• Insulation for Ductwork (1 % of PEC);
• Painting (1% of PEC);
and can be summarized as:
4-18
-------
DIG = 30% PEC = 0.3 PEC.
Therefore, the direct costs can be simplified using the following equation:
DC = PEC + 0.3 PEC = 1.3 PEC.
The indirect costs include the costs of engineering and contractor fees and contingencies and can
be summarized using the following equation:
1C = Indirect Installation Costs (ICC) + Contingencies (C).
A summary of the cost assumptions for ICC includes the following:
• Engineering (10% of PEC);
• Construction and Field Expenses (5% of PEC);
• Contractor Fees (10% of PEC);
• Startup (2% of PEC);
• Performance Test (1 % of PEC);
and can be summarized as:
IIC = 28% PEC = 0.28 PEC.
A summary of the cost assumptions for C includes the following:
• Equipment Redesign and Modifications;
• Cost Escalations;
• Delays in Startup;
and is assumed to be:
C = 3% PEC = 0.03 PEC.
Therefore, the 1C can be summarized using the following equation:
1C = 0.28 PEC + 0.03 PEC = 0.31 PEC,
and the simplified TCC equation can be expressed as:
4-19
-------
TCC = 1.3 PEC+ 0.31 PEC = 1.61 PEC = 1.61 (1.18EC)= 1.9 EC
4.4.2.2 Total Annual Costs
The total annual cost includes the direct and indirect annual costs of operating and
maintaining the control equipment. The direct annual cost includes the cost of the utilities,
operating labor, and control device cleaning and maintenance. The indirect annual cost includes
the overhead costs such as spare parts for the control equipment, administrative charges, and the
capital recovery of the control technology. The total annual cost equation can be summarized as
follows:
Total Annual Cost (TAC) = Direct Annual Costs (DAC) + Indirect Annual Costs (IAC).
The DAC includes the following parameters:
• Utilities;
• Operating Labor;
• Maintenance;
• Annual Compliance Test;
• Catalyst Cleaning;
• Catalyst Replacement;
• Catalyst Disposal.
The IAC includes the following parameters:
• Overhead;
• Fuel Penalty;
• Property Tax;
• Insurance;
• Administrative Charges;
• Capital Recovery = (1(1+!)"/((!+I)n-l)*TCC} where I is the interest rate, and n is the
equipment life.
To calculate DAC, the costs were broken up into three separate costs: operation and
maintenance materials cost, operation and maintenance labor cost, and the cost for annual
4-20
-------
performance testing or downtime or allowance for catalyst washing. Actual annual cost data
from the industry groups were used to estimate the DAC for each of the control technologies.
The I AC was broken up into three separate costs: administrative, fuel penalty, and capital
recovery. Again, cost data from the industry groups was used to estimate these costs for each of
the control technologies. No fuel penalty was estimated for the oxidation catalyst control
technologies, because this control technology does not increase the fuel usage of the SI engine.
4.4.3 Control Cost Equations
Control cost equations were developed for 2SLB oxidation catalyst, 4SLB oxidation
catalyst, and a NSCR for 4SRB engines using the total capital cost and total annual cost data for
each control technology. Control cost equations for 2SLB and 4SLB oxidation catalysts were
developed separately because the 2SLB oxidation catalyst requires a premium catalyst to reduce
the HAP compounds because of the low exhaust temperature of 2SLB engines.
4.4.3.1 2SLB Oxidation Catalyst
The 2SLB oxidation catalyst is an effective control technology that reduces HAP
emissions from a 2SLB SI engine by oxidizing organic compounds using a catalyst. The
oxidation catalyst unit contains a honeycomb-like structure or substrate with a large surface area
that is coated with a premium active catalyst layer such as platinum or palladium. The oxidation
catalyst works by oxidizing carbon monoxide (CO) and gaseous hydrocarbons (HAP) in the
exhaust gas to carbon dioxide (CO2) and water. The reduction of CO and HAP varies depending
on the type of catalyst used and the exhaust temperature of the pollutant stream.
The cost of retrofitting an oxidation catalyst to an existing 2SLB engine was estimated
using cost data obtained from vendors and industry groups covering engines ranging from 58
horsepower (HP) to 4,670 HP. An equipment life of 10 years and an interest rate of 7 percent
were used to estimate the capital recovery of the control technology and the fuel penalty was
assumed to be negligible. The cost equations are presented in 2009 dollars.
The total annualized cost equation for retrofitting an oxidation catalyst on a 2SLB engine
was estimated to be:
2SLB Oxidation Catalyst Total Annual Cost = $11.4 x HP + $13,928
where
HP = engine size in HP.
4-21
-------
The linear equation has a correlation coefficient of 0.8046, which shows the data fits the
equation closely. Therefore, this equation was used to estimate annualized cost for an oxidation
catalyst on a 2SLB engine.
The total capital cost equation for retrofitting an oxidation catalyst on a 2SLB engine was
estimated to be:
2SLB Oxidation Catalyst Total Capital Cost = $47.1 x HP + $41,603
where
HP = engine size in HP.
4.4.3.2 4SLB Oxidation Catalyst
The 4SLB oxidation catalyst is an effective control technology that reduces HAP
emissions from a 4SLB SI engine by oxidizing organic compounds using a catalyst. The
oxidation catalyst unit contains a honeycomb-like structure or substrate with a large surface area
that is coated with a premium active catalyst layer such as platinum or palladium. The oxidation
catalyst works by oxidizing CO and gaseous hydrocarbons (HAP) in the exhaust gas to CO2 and
water. The reductions of CO and HAP vary depending on the type of catalyst used and the
exhaust temperature of the pollutant stream.
The cost of retrofitting an oxidation catalyst to an existing 4SLB engine was estimated
using cost data obtained from vendors and industry groups covering engines ranging from 400
HP to 8,000 HP. Again, an equipment life of 10 years and an interest rate of 7 percent were used
to estimate the capital recovery of the control technology and the fuel penalty was assumed to be
negligible. The cost equations are presented in 2009 dollars.
The total annualized cost equation for retrofitting an oxidation catalyst on a 4SLB engine
was estimated to be:
4SLB Oxidation Catalyst Total Annual Cost = $1.81 x HP + $3,442
where
HP = engine size in HP.
4-22
-------
The linear equation has a correlation coefficient of 0.9779, which shows the data fits the
equation very closely. Therefore, this equation was used to estimate annualized cost for an
oxidation catalyst on a 4SLB engine.
The total capital cost equation for retrofitting an oxidation catalyst on a 4SLB SI engine was
estimated to be:
4SLB Oxidation Catalyst Total Capital Cost = $12.8 x HP + $3,069
where
HP = engine size in HP.
A summary of the cost calculations, regression analyses, and graphical representations of the
annual and capital cost data are presented in Appendix A of the cost memo that is the basis for
the cost data presented in this RIA.3
4.4.3.3 Non-Selective Catalytic Reduction
The NSCR or three-way catalyst is used to control HAP emissions from 4SRB engines.
In addition to HAP reductions, NSCR also reduces the emissions of nitrogen oxides (NOx), CO,
and other hydrocarbons (HC). The reduction of HAP and CO takes place through an oxidation
reaction that converts HAP to CO2 and water and converts CO to CO2. The conversion of NOx
takes place through a reduction of the NOx to nitrogen gas and oxygen.
The cost of retrofitting an NSCR on an existing 4SRB engine was estimated based on
cost data received from vendors and industry groups. A linear regression analysis was done on
the data set and the linear equation for annualized cost was;
NSCR Annual Cost = $4.77 x HP + $5,679
where
HP = engine size in HP.
The linear equation has a correlation coefficient of 0.7987, which shows an acceptable
representation of the cost data. Therefore, this equation was used to estimate annualized cost for
retrofitting the NSCR control technology on 4SRB engines.
3 Memorandum from Bradley Nelson, EC/R to Melanie King, EPA. OAQPS/SPPD/ESG. Impacts Associated with
NESHAP for Existing Stationary SI RICE. June 29, 2010.
4-23
-------
The capital cost equation for retrofitting an air-to-fuel ratio (APR) controller and NSCR on a
4SRB engine was estimated to be:
NSCR Capital Cost = $24.9 x HP + $13,118
where
HP = engine size in HP.
4.4.4 Summary
Table 4-3 presents a summary of the annual and capital control costs as a function of
engine size for the control technologies applicable to existing stationary SI engines, as discussed
in this memorandum.
Table 4-3. Summary of Annual and Capital Costs Equations for Existing Stationary SI
Engines
HAP Control Device
Annual Cost ($2009)
Capital Cost ($2009)
2SLB Oxidation Catalyst
4SLB Oxidation Catalyst
NSCR
$11.4 x HP+ $13,928
$1.Six HP+ $3,442
$4.77 x HP + $5,679
$47. Ix HP+ $41,603
$12.8 x HP+ $3,069
$24.9xHP + $13,118
A summary of the annual and capital costs associated with the rule and obtained using the
methodology described above are presented in Tables 4-4 to 4-7 below.4 These costs are used in
the economic impact as well as the small entity analysis.
4 Memorandum from Bradley Nelson, EC/R to Melanie King, EPA. OAQPS/SPPD/ESG. Impacts Associated with
NESHAP for Existing Stationary SI RICE. June 29, 2010.
4-24
-------
Table 4-4. Summary of Major Source and Area Source Costs for the SI RICE NESHAP
,a,b
Size Range (HP)
Major Sources
25-50
50-100
100-175
175-300
300-500
500-600
600-750
>750
Total
•{^ Area Sources
to
^ 25-50
50-100
100-175
175-300
300-500
500-600
600-750
>750
Total
Grand Total
Total
Capital
Control Cost
$0
$0
$48,502,361
$13,225,919
$10,934,795
$0
$0
$0
$72,663,076
$0
$0
$0
$0
$0
$75,474,331
$15,222,363
$210,754,181
$301,450,875
$374,113,951
Annual
Control Cost
$0
$0
$37,071,061
$8,382,568
$5,562,872
$0
$0
$0
$51,016,500
$0
$0
$0
$0
$0
$26,628,053
$5,052,207
$62,143,967
$93,824,227
$144,840,727
Initial Test
$0
$0
$15,971,384
$3,442,648
$2,123,326
$0
$0
$0
$21,537,358
$0
$0
$0
$0
$0
$3,655,719
$652,400
$6,951,011
$11,259,129
$32,796,487
Record-
keeping
$4,060,795
$1,087,540
$1,721,899
$371,157
$228,919
$0
$0
$0
$7,470,310
$6,668,944
$2,868,511
$3,529,711
$1,264,799
$908,913
$454,493
$77,882
$829,795
$16,603,048
$24,073,358
Reporting
$0
$0
$5,725,314
$1,234,097
$761,155
$0
$0
$0
$7,720,566
$0
$0
$0
$0
$0
$1,264,260
$225,620
$2,403,874
$3,893,754
$11,614,321
Monitoring —
Capital Cost
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$2,821,013
$503,438
$5,363,896
$8,688,347
$8,688,347
Monitoring —
Annual Cost
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$13,003,822
$2,320,662
$24,725,562
$40,050,046
$40,050,046
Total Annual
Costs
$4,060,795
$1,087,540
$60,489,657
$13,430,470
$8,676,262
$0
$0
$0
$87,744,734
$6,668,944
$2,868,511
$3,529,711
$1,264,799
$908,913
$45,006,347
$8,328,771
$97,054,209
$165,630,205
$253,374,939
Total Capital
Costs
$0
$0
$48,502,361
$13,225,919
$10,934,795
$0
$0
$0
$72,663,076
$0
$0
$0
$0
$0
$78,295,345
$15,725,801
$216,118,077
$310,139,222
$382,802,298
Costs are presented in 2009 dollars.
-------
b For some HP ranges, the annual compliance cost is greater than the capital compliance cost because not all of the engines in those HP ranges are expected to
incur capital costs for controls, but all of the engines in those HP ranges are expected to incur annual costs of testing, monitoring, recordkeeping, and reporting.
to
-------
Table 4-5. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP"
,b
to
Major Source
NAICS
2211
48621
211111
211112
92811
335312
335312
333992
Electric Power Generation
Natural Gas Transmission
Crude Petroleum & NG
Production
Natural Gas Liquid Producers
National Security
Hydro Power Units
Irrigation Sets
Welders
Total
Capital Cost
$52,905,258
$1,484,494
$4,561,236
$4,561,236
$5,878,362
$0
$3,025,050
$247,440
$72,663,076
Annual Cost
$63,062,494
$1,462,530
$6,138,383
$6,138,383
$7,006,944
$25,248
$3,230,856
$679,896
$87,744,734
Area Source
Capital Cost
$120,301,416
$140,977,276
$732,943
$732,943
$13,366,824
$0
$34,027,819
$0
$310,139,222
Annual Cost
$65,334,028
$67,467,484
$1,258,072
$1,258,072
$7,259,336
$37,872
$22,445,211
$570,130
$165,630,205
Total (Major + Area)
Capital Cost Annual Cost
$173,206,675
$142,461,771
$5,294,179
$5,294,179
$19,245,186
$0
$37,052,869
$247,440
$382,802,298
$128,396,522
$68,930,015
$7,396,454
$7,396,454
$14,266,280
$63,120
$25,676,067
$1,250,027
$253,374,939
a Costs are presented in 2009 dollars.
b For some HP ranges, the annual compliance cost is greater than the capital compliance cost because not all of the engines in those HP ranges are expected to
incur capital costs for controls, but all of the engines in those HP ranges are expected to incur annual costs of testing, monitoring, recordkeeping, and reporting.
-------
Table 4-6. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Size
a,b
to
Major Source
Area Source
Total (Major + Area)
NAICS
Electric Power Generation (2211)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Electric Power Generation 221 1
Natural Gas Transmission (48621)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Natural Gas Transmission (48621)
Crude Petroleum & NG Production (211111)
25-50 hp
50-100 hp
100-175 hp
Capital Cost
$0
$0
$33,868,173
$10,603,849
$8,433,236
$0
$0
$0
$52,905,258
$0
$0
$301,721
$643,157
$539,617
$0
$0
$0
$1,484,494
$0
$0
$4,549,775
Annual Cost
$2,758,459
$606,144
$42,238,648
$10,767,847
$6,691,397
$0
$0
$0
$63,062,494
$102
$4,872
$376,291
$653,104
$428,162
$0
$0
$0
$1,462,530
$388,115
$66,698
$5,674,246
Capital Cost
$0
$0
$0
$0
$0
$26,390,293
$5,325,628
$88,585,495
$120,301,416
$0
$0
$0
$0
$0
$19,975,323
$9,808,436
$111,193,518
$140,977,276
$0
$0
$0
Annual Cost
$4,137,688
$909,215
$1,803,548
$446,361
$264,820
$15,169,876
$2,820,584
$39,781,934
$65,334,028
$1,934
$92,571
$203,518
$342,928
$214,637
$11,482,372
$5,194,789
$49,934,735
$67,467,484
$582,173
$100,047
$242,285
Capital Cost
$0
$0
$33,868,173
$10,603,849
$8,433,236
$26,390,293
$5,325,628
$88,585,495
$173,206,675
$0
$0
$301,721
$643,157
$539,617
$19,975,323
$9,808,436
$111,193,518
$142,461,771
$0
$0
$4,549,775
Annual Cost
$6,896,147
$1,515,359
$44,042,196
$11,214,209
$6,956,217
$15,169,876
$2,820,584
$39,781,934
$128,396,522
$2,036
$97,443
$579,809
$996,032
$642,799
$11,482,372
$5,194,789
$49,934,735
$68,930,015
$970,288
$166,744
$5,916,531
(continued)
-------
a,b
Table 4-6. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Size ' (continued)
to
VO
NAICS
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Crude Petroleum & NG Production
(211111)
Natural Gas Liquid Producers (211112)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Natural Gas Liquid Producers (211112)
National Security (92811)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
Major
Capital Cost
$1,037
$10,424
$0
$0
$0
$4,561,236
$0
$0
$4,549,775
$1,037
$10,424
$0
$0
$0
$4,561,236
$0
$0
$3,763,130
$1,178,205
$937,026
$0
Source
Annual Cost
$1,053
$8,271
$0
$0
$0
$6,138,383
$388,115
$66,698
$5,674,246
$1,053
$8,271
$0
$0
$0
$6,138,383
$306,495
$67,349
$4,693,183
$1,196,427
$743,489
$0
Area Source
Capital Cost
$0
$0
$32,184
$0
$700,760
$732,943
$0
$0
$0
$0
$0
$32,184
$0
$700,760
$732,943
$0
$0
$0
$0
$0
$2,932,255
Annual Cost
$44
$327
$18,500
$0
$314,697
$1,258,072
$582,173
$100,047
$242,285
$44
$327
$18,500
$0
$314,697
$1,258,072
$459,743
$101,024
$200,394
$49,596
$29,424
$1,685,542
Total (Major + Area)
Capital Cost
$1,037
$10,424
$32,184
$0
$700,760
$5,294,179
$0
$0
$4,549,775
$1,037
$10,424
$32,184
$0
$700,760
$5,294,179
$0
$0
$3,763,130
$1,178,205
$937,026
$2,932,255
Annual Cost
$1,096
$8,598
$18,500
$0
$314,697
$7,396,454
$970,288
$166,744
$5,916,531
$1,096
$8,598
$18,500
$0
$314,697
$7,396,454
$766,239
$168,373
$4,893,577
$1,246,023
$772,913
$1,685,542
(continued)
-------
a,b
Table 4-6. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Size ' (continued)
-^
o
NAICS
600-750 hp
>750 hp
Total Natural Gas Liquid Producers (2111 12)
Hydro Power Units (335312)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Hydro Power Units (3353 12)
Irrigation Sets (335312)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Irrigation Sets (335312)
Major
Capital Cost
$0
$0
$5,878,362
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$1,222,348
$798,634
$1,004,068
$0
$0
$0
$3,025,050
Source
Annual Cost
$0
$0
$7,006,944
$22,688
$2,560
$0
$0
$0
$0
$0
$0
$25,248
$32,913
$65,825
$1,524,449
$810,986
$796,683
$0
$0
$0
$3,230,856
Area Source
Capital Cost
$591,736
$9,842,833
$13,366,824
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$28,933,107
$0
$5,094,712
$34,027,819
Annual Cost
$313,398
$4,420,215
$7,259,336
$34,032
$3,840
$0
$0
$0
$0
$0
$0
$37,872
$625,338
$1,250,677
$824,505
$425,827
$399,376
$16,631,556
$0
$2,287,931
$22,445,211
Total (Major + Area)
Capital Cost
$591,736
$9,842,833
$19,245,186
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$1,222,348
$798,634
$1,004,068
$28,933,107
$0
$5,094,712
$37,052,869
Annual Cost
$313,398
$4,420,215
$14,266,280
$56,721
$6,399
$0
$0
$0
$0
$0
$0
$63,120
$658,251
$1,316,502
$2,348,954
$1,236,813
$1,196,060
$16,631,556
$0
$2,287,931
$25,676,067
(continued)
-------
a,b
Table 4-6. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Size ' (continued)
NAICS
Welders (333992)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Welders (333992)
Total
Major
Capital Cost
$0
$0
$247,440
$0
$0
$0
$0
$0
$247,440
$72,663,076
Source
Annual Cost
$163,908
$207,394
$308,594
$0
$0
$0
$0
$0
$679,896
$87,744,734
Area Source
Capital Cost
$0
$0
$0
$0
$0
$0
$0
$0
$0
$310,139,222
Annual Cost
$245,862
$311,091
$13,177
$0
$0
$0
$0
$0
$570,130
$165,630,205
Total (Major + Area)
Capital Cost
$0
$0
$247,440
$0
$0
$0
$0
$0
$247,440
$382,802,298
Annual Cost
$409,771
$518,485
$321,771
$0
$0
$0
$0
$0
$1,250,027
$253,374,939
Costs are presented in 2009 dollars.
For some HP ranges, the annual compliance cost is greater than the capital compliance cost because not all of the engines in those HP ranges are expected to
incur capital costs for controls, but all of the engines in those HP ranges are expected to incur annual costs of testing, monitoring, recordkeeping, and reporting.
-------
Table 4-7. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Number of Engines
a,b
-^
to
Number of Engines
NAICS
Electric Power Generation (2211)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Electric Power Generation 221 1
Natural Gas Transmission (48621)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Natural Gas Transmission (48621)
Crude Petroleum & NG Production (211111)
25-50 hp
50-100 hp
100-175 hp
Major
37,933
8,336
16,534
4,092
2,428
0
0
0
69,323
1
67
147
248
155
0
0
0
619
5,337
917
2,221
Area
56,900
12,503
24,802
6,138
3,642
2,107
363
4,677
111,132
27
1,273
2,799
4,716
2,952
1,595
668
5,871
19,899
8,006
1,376
3,332
Total
94,833
20,839
41,336
10,230
6,070
2,107
363
4,677
180,455
28
1,340
2,946
4,964
3,107
1,595
668
5,871
20,519
13,343
2,293
5,553
Total (Major + Area)
Capital Cost
$0
$0
$33,868,173
$10,603,849
$8,433,236
$26,390,293
$5,325,628
$88,585,495
$173,206,675
$0
$0
$301,721
$643,157
$539,617
$19,975,323
$9,808,436
$111,193,518
$142,461,771
$0
$0
$4,549,775
Annual Cost
$6,896,147
$1,515,359
$44,042,196
$11,214,209
$6,956,217
$15,169,876
$2,820,584
$39,781,934
$128,396,522
$2,036
$97,443
$579,809
$996,032
$642,799
$11,482,372
$5,194,789
$49,934,735
$68,930,015
$970,288
$166,744
$5,916,531
(continued)
-------
Table 4-7. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Number of Engines"
(continued)
Number of Engines
NAICS
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Grade Petroleum & NG Production (211111)
Natural Gas Liquid Producers (211112)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Natural Gas Liquid Producers (2111 12)
National Security (92811)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
Major
0
3
0
0
0
8,479
5,337
917
2,221
0
3
0
0
0
8,479
4,215
926
1,837
455
270
0
Area
1
5
3
0
37
12,758
8,006
1,376
3,332
1
5
3
0
37
12,758
6,322
1,389
2,756
682
404
234
Total
1
8
3
0
37
21,237
13,343
2,293
5,553
1
8
o
6
0
37
21,237
10,537
2,315
4,593
1,137
674
234
Total (Major + Area)
Capital Cost
$1,037
$10,424
$32,184
$0
$700,760
$5,294,179
$0
$0
$4,549,775
$1,037
$10,424
$32,184
$0
$700,760
$5,294,179
$0
$0
$3,763,130
$1,178,205
$937,026
$2,932,255
Annual Cost
$1,096
$8,598
$18,500
$0
$314,697
$7,396,454
$970,288
$166,744
$5,916,531
$1,096
$8,598
$18,500
$0
$314,697
$7,396,454
$766,239
$168,373
$4,893,577
$1,246,023
$772,913
$1,685,542
(continued)
-------
Table 4-7. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Number of Engines"
(continued)
Number of Engines
NAICS
600-750 hp
>750 hp
Total Natural Gas Liquid Producers (211112)
Hydro Power Units (335312)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Hydro Power Units (335312)
Irrigation Sets (335312)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Irrigation Sets (335312)
Major
0
0
7,702
312
35
0
0
0
0
0
0
347
453
905
597
308
289
0
0
0
2,552
Area
40
520
12,347
468
53
0
0
0
0
0
0
521
8,599
17,199
11,338
5,856
5,492
2,310
0
269
51,063
Total
40
520
20,050
780
88
0
0
0
0
0
0
868
9,052
18,104
11,935
6,164
5,781
2,310
0
269
53,615
Total (Major + Area)
Capital Cost
$591,736
$9,842,833
$19,245,186
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$1,222,348
$798,634
$1,004,068
$28,933,107
$0
$5,094,712
$37,052,869
Annual Cost
$313,398
$4,420,215
$14,266,280
$56,721
$6,399
$0
$0
$0
$0
$0
$0
$63,120
$658,251
$1,316,502
$2,348,954
$1,236,813
$1,196,060
$16,631,556
$0
$2,287,931
$25,676,067
(continued)
-------
Table 4-7. Summary of Major Source and Area Source NAICS Costs for the SI RICE NESHAP, by Number of Engines"
(continued)
Number of Engines
NAICS
Welders (333992)
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total Welders (333992)
Total
Major
2,254
2,852
121
0
0
0
0
0
5,227
102,729
Area
3,381
4,278
181
0
0
0
0
0
7,840
228,319
Total
5,635
7,130
302
0
0
0
0
0
13,067
331,047
Total (Major + Area)
Capital Cost
$0
$0
$247,440
$0
$0
$0
$0
$0
$247,440
$382,802,298
Annual Cost
$409,771
$518,485
$321,771
$0
$0
$0
$0
$0
$1,250,027
$253,374,939
a Costs are presented in 2009 dollars.
bFor some HP ranges, the annual compliance cost is greater than the capital compliance cost because not all of the engines in those HP ranges are expected
to incur capital costs for controls, but all of the engines in those HP ranges are expected to incur annual costs of testing, monitoring, recordkeeping, and
reporting.
-------
4.4.5 Caveats and Uncertainties in the Cost Estimates
* Current knowledge about NOx control techniques and costs is applied in this
study. Advances such as alternative catalyst formulations may occur between
now and when sources comply with this rulemaking that may lower costs. Scale
economies can also lower per unit production costs as the market for these NOx
control techniques expands.
* The alternative control techniques and corresponding emission reductions and
costs may not apply to every unit within the source category. Many factors
influence the performance and cost of any control technique. Because control
technology references typically evaluate average retrofit situations, costs may
be underestimated for the fraction of the source population with difficult to
retrofit conditions. Difficult to retrofit conditions may be less of an issue for
RICEs than for other point sources, however.
* NOx control efficiency and cost estimates associated with source category-control
strategy combinations are represented as point estimates. In practice, control
effectiveness and costs will vary by engine.
4-36
-------
4.5 Emissions and Emission Reductions
The baseline emissions, emissions factors and emissions reductions associated with the
final rule are provided in the tables below. Emissions are in tons per year.
Table 4-8. Summary of Major Source and Area Source Baseline Emissions for the SI
RICE NESHAP
Size Range (HP)
Major Sources
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total
Area Sources
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total
Major + Area
Total
Table 4-9. Emissions
HAP
Engine (Ib/hp-hr)
2SLB 5.96xlO'4
4SLB 5.41xlO"4
4SRB 2.43xlO'4
HAP
1,107
593
1,721
641
666
0
0
0
4,728
1,818
1,564
3,529
2,184
2,643
1,830
383
6,041
19,993
24,721
Factors
CO
(Ib/hp-hr)
1.06xlO'2
3.92xlO"3
1.93xlO'2
Baseline
CO
28,557
15,296
44,399
16,530
17,171
0
0
0
121,953
46,898
40,344
91,013
56,331
68,178
47,273
9,876
155,890
515,803
637,756
NOx
(Ib/hp-hr)
4.18xlO'2
1.15xlO"2
1.47xlO'2
Emissions (TPY)
NOx
41,751
22,363
64,913
24,168
25,105
0
0
0
178,301
68,566
58,985
133,065
82,359
99,679
69,094
14,438
227,890
754,077
932,378
voc
(Ib/hp-hr)
3.07xlO'3
2.78xlO"3
1.25xlO'3
VOC
5,696
3,051
8,855
3,297
3,425
0
0
0
24,323
9,354
8,047
18,153
11,235
13,598
9,415
1,969
31,076
102,846
127,169
Formaldehyde
(Ib/hp-hr)
4.29xlO'4
3.96xlO"4
1.75xlO'4
4-37
-------
Table 4-10. Summary of Major Source and Area Source Emissions Reductions for the SI
RICE NESHAP
Emission Reductions (TPY)
Size Range (HP)
Major Sources
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total
Area Sources
25-50 hp
50-100 hp
100-175 hp
175-300 hp
300-500 hp
500-600 hp
600-750 hp
>750 hp
Total
Major + Area Total
HAP
0
0
744
277
288
0
0
0
1,308
0
0
0
0
0
1,005
220
3,475
4,700
6,008
CO
0
0
7,124
2,653
2,755
0
0
0
12,532
0
0
0
0
0
20,698
4,533
71,557
96,789
109,321
NOx
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20,632
4,519
71,328
96,479
96,479
voc
0
0
3,826
1,424
1,480
0
0
0
6,730
0
0
0
0
0
5,170
1,132
17,874,
24,177
30,907
4-38
-------
SECTION 5
ECONOMIC IMPACT ANALYSIS, ENERGY IMPACTS, AND SOCIAL COSTS
The EIA provides decision makers with social cost estimates and enhances understanding
of how the costs may be distributed across stakeholders (EPA, 2000). Although several
economic frameworks can be used to estimate social costs for regulations of this size and sector
scope, OAQPS has typically used partial equilibrium market models. However, the current data
do not provide sufficient details to develop a market model; the data that are available have little
or no sector/firm detail and are reported at the national level. In addition, some sectors have
unique market characteristics that make developing partial equilibrium models difficult. Given
these constraints, we used the direct compliance costs as a measure of total social costs. In
addition, we also provide a qualitative analysis of the final rule's impact on stakeholder
decisions, a qualitative discussion on if unfunded mandates occur as a result of this final rule,
and the potential distribution of social costs between consumers and producers.
5.1 Compliance Costs of the Final Rule
EPA's engineering cost analysis estimates the total annualized costs of the final rule are
$253 million (in 2009 dollars) (Nelson, 2010).
As shown in Figure 5-1, the majority of the costs fall on the electric power sector (51%),
followed by natural gas transmission (27%). The remaining industries each account for less than
15% of the total annualized cost. The industrial classification for each engine is taken from the
Power Systems Research (PSR) database, which is the major source of data for the engines
affected by the final rule. The PSR database used as a basis for the analyses in this RIA contains
information on both mobile and stationary engines, among other data, and does so not only for
the U.S. but worldwide. PSR has collected such data for more than 30 years. The Office of
Transportation and Air Quality (OTAQ) uses this database frequently in the development of their
mobile source rules.
The annualized compliance costs per engine vary by the engine size (see Figure 5-2). For
500 hp engines or less, the annualized per-engine costs are below $1,200 per engine. Per-engine
costs for higher horsepower (hp) engines range between $7,200 and $8,500.
The final rule will affect approximately 331,000 existing stationary SI engines. As shown
in Figure 5-3, most of the affected engines fall within the 25 to 50 hp category (45%). The next
highest categories are 100 to 175 hp (22%) and 50 to 100 hp (16%).
5-1
-------
100%
51%
27%
6% I0%
3% 3%
0% 0%
C2-
C2-
Q_
O
o
a.
p
O
g)
O
Figure 5-1. Distribution of Annualized Direct Compliance Costs by Industry
5-2
-------
$9,000
$8,000
$7,000
$6,000
$5,000
$4,000
$3,000
$2,000
$1,000
$0
$73
$8,505
$7.777
$7,201
$1,100
$653
$73
25-50 hp 50-100 hp 100-175 hp 175-300 hp 300-500 hp 500-600 hp 600-750 hp >750 hp
Figure 5-2. Average Annualized Cost per Engine by Horsepower Group ($2009)
5-3
-------
100%
90%
80%
70%
60%
50%
45%
40%
30%
22%
20%
16%
7% ~
3%
0%
0%
25-50 hp 50-100 hp 100-175 hp 175-300 hp 300-500 hp 500-600 hp 600-750 hp >750 hp
Figure 5-3. Distribution of Engine Population by Horsepower Group
To assess the size of the compliance relative to the value of the goods and services for
industries using affected engines, we collected Census data for selected industries. At the
industry level, the annualized costs represent a very small fraction of revenue (less than 1%), for
all affected industries. Results for affected industries can be found in Table 5-1. These industry
level cost-to-sales ratios can be interpreted as an average impact on potentially affected firms in
these industries. Based on the cost-to-sales ratios, we can conclude that the annualized cost of
this rule should be no higher than 1% of the sales on average for a firm in each of these
industries, excluding natural gas transmission and natural gas liquid producers, which face
slightly higher costs to sales ratios.
5.2 How Might People and Firms Respond? A Partial Equilibrium Analysis
Markets are composed of people as consumers and producers trying to do the best they
can given their economic circumstances. One way economists illustrate behavioral responses to
pollution control costs is by using market supply and demand diagrams. The market supply curve
describes how much of a good or service firms are willing and able to sell to people at a
5-4
-------
Table 5-1. Selected Industry-Level Annualized Compliance Costs as a Fraction of Total
Industry Revenue: 2009
Industry
(NAICS)
2211
48621
211111
211112
92811
333992
111 and 112
Industry Name
Electric Power Generation
Natural Gas Transmission
Crude Petroleum & NG Production
Natural Gas Liquid Producers
National Security
Welders
Agriculture using irrigation systems3
Total
Annualized
Costs
($ million)3
$128.4
$68.9
$7.4
$7.4
$14.3
$1.3
$25.7
Sales, Shipments, Receipt,
or Revenue ($ Billion)
($2007)
$440.4
$16.4
$214.2
$42.4
#N/A
$5.2
$27.9
($2009)
$453.7
$16.9
$220.5
$43.6
#N/A
$5.5
$28.8
Cost-to-
Sales Ratio
0.004%
0.41%
0.001%
0.005%
#N/A
0.025%
0.09%
a Irrigation engine costs assumed to be passed on to agricultural sectors that use irrigation systems.
N/A: receipts are Not Available for National Security
Sources: U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 00: All sectors:
Geographic Area Series: Economy-Wide Key Statistics: 2007" ; (July 7th,
2010).
U.S. Department of Agriculture (USDA), National Agricultural Statistics Service (NASS). 2009. "2008
Farm and Ranch Irrigation Survey." Washington, DC: USDA-NASS.
Costs from Existing SI RICE NESHAP Impacts 6-24-2010.xls received from EPA 6/24/10
particular price; we often draw this curve as upward sloping because some production resources
are fixed. As a result, the cost of producing an additional unit typically rises as more units are
made. The market demand curve describes how much of a good or service consumers are willing
and able to buy at some price. Holding other factors constant, the quantity demand is assumed to
fall when prices rise. In a perfectly competitive market, equilibrium price (Po) and quantity (Qo)
is determined by the intersection of the supply and demand curves (see Figure 5-4).
5.2.1 Changes in Market Prices and Quantities
To qualitatively assess how the regulation may influence the equilibrium price and
quantity in the affected markets, we assumed the market supply function shifts up by the
additional cost of producing the good or service; the unit cost increase is typically calculated by
dividing the annual compliance cost estimate by the baseline quantity (Qo) (see Figure 5-4). As
shown, this model makes two predictions: the price of the affected goods and services are likely
to rise and the consumption/production levels are likely to fall.
5-5
-------
Price
Increase
Si: With Regulation
Unit Cost Increase
S0: Without Regulation
Output
consumer surplus = -[fghd + dhc]
producer surplus = [fghd - aehb] - bdc
total surplus = consumer surplus + producer surplus
-[aehb + dhc + bdc]
Figure 5-4. Market Demand and Supply Model: With and Without Regulation
The size of these changes depends on two factors: the size of the unit cost increase
(supply shift) and differences in how each side of the market (supply and demand) responds to
changes in price. Economists measure responses using the concept of price elasticity, which
represents the percentage change in quantity divided by the percentage change in price. This
dependence has been expressed in the following formula:1
Share ofper-unit cost =
Price Elasticity of Supply
(Price Elasticity of Supply - Price Elasticity of Demand))
As a general rule, a higher share of the per-unit cost increases will be passed on to
consumers in markets where
• goods and services are necessities and people do not have good substitutes that they
can switch to easily (demand is inelastic) and
JFor examples of similar mathematical models in the public finance literature, see Nicholson (1998), pages 444-447,
or Fullerton and Metcalf (2002).
5-6
-------
• suppliers have excess capacity and can easily adjust production levels at minimal
costs, or the time period of analysis is long enough that suppliers can change their
fixed resources; supply is more elastic over longer periods.
Short-run demand elasticities for energy goods (electricity and natural gas), agricultural
products, and construction are often inelastic. Specific estimates of short-run demand elasticities
for these products can be obtained from existing literature. For the short-run demand of energy
products, the National Energy Modeling System (NEMS) buildings module uses values between
0.1 and 0.3; a 1% increase in price leads to a 0.1 to 0.3% decrease in energy demand (Wade,
2003). For the short-run demand of agriculture and construction, the EPA has estimated
elasticities to be 0.2 for agriculture and approximately 1 for construction (U.S. EPA, 2004). As a
result, a 1% increase in the prices of agriculture products would lead to a 0.2% decrease in
demand for those products, while a 1% increase in construction prices would lead to
approximately a 1% decrease in demand for construction. Given these demand elasticity
scenarios (shaded in gray), approximately a 1% increase unit costs would result in a price
increase of 0.1 to 1% (Table 5-2). As a result, 10 to 100% of the unit cost increase could be
passed on to consumers in the form of higher goods/services prices. This price increase would
correspond to a 0.1 to 0.8% decline in consumption in these markets (Table 5-3).
Table 5-2. Hypothetical Price Increases for a 1% Increase in Unit Costs
Market Demand
Elasticity
-0.1
-0.3
-0.5
-0.7
-1.0
-1.5
-3.0
Market Supply Elasticity
0.1
0.5%
0.3%
0.2%
0.1%
0.1%
0.1%
0.0%
0.3
0.8%
0.5%
0.4%
0.3%
0.2%
0.2%
0.1%
0.5
0.8%
0.6%
0.5%
0.4%
0.3%
0.3%
0.1%
0.7
0.9%
0.7%
0.6%
0.5%
0.4%
0.3%
0.2%
1
0.9%
0.8%
0.7%
0.6%
0.5%
0.4%
0.3%
1.5
0.9%
0.8%
0.8%
0.7%
0.6%
0.5%
0.3%
3
1.0%
0.9%
0.9%
0.8%
0.8%
0.7%
0.5%
5.2.2 Regulated Markets: The Electric Power Generation, Transmission, and Distribution
Sector
Given that the electric power sector bears majority of the estimated compliance costs
(Figure 5-1) and the industry is also among the last major regulated energy industries in the
United States (EIA, 2000), the competitive model is not necessarily applicable for this industry.
5-7
-------
Table 5-3. Hypothetical Consumption Decreases for a 1% Increase in Unit Costs
Market Demand
Elasticity
-0.1
-0.3
-0.5
-0.7
-1.0
-1.5
-3.0
Market Supply Elasticity
0.1
-0.1%
-0.1%
-0.1%
-0.1%
-0.1%
-0.1%
-0.1%
0.3
-0.1%
-0.2%
-0.2%
-0.2%
-0.2%
-0.3%
-0.3%
0.5
-0.1%
-0.2%
-0.3%
-0.3%
-0.3%
-0.4%
-0.4%
0.7
-0.1%
-0.2%
-0.3%
-0.4%
-0.4%
-0.5%
-0.6%
1
-0.1%
-0.2%
-0.3%
-0.4%
-0.5%
-0.6%
-0.8%
1.5
-0.1%
-0.3%
-0.4%
-0.5%
-0.6%
-0.8%
-1.0%
3
-0.1%
-0.3%
-0.4%
-0.6%
-0.8%
-1.0%
-1.5%
Although the electricity industry continues to go through a process of restructuring, whereby the
industry is moving toward a more competitive framework (see Figure 5-5 for the status of
restructuring by state),2 in many states, electricity prices continue to be fully regulated by Public
Service Commissions. As a result, the rules and processes outlined by these agencies would
ultimately determine how these additional regulatory costs would be recovered by affected
entities.
5.2.3 Partial Equilibrium Measures of Social Cost: Changes Consumer and Producer
Surplus
In partial equilibrium analysis, the social costs are estimated by measuring the changes in
consumer and producer surplus, and these values can be determined using the market supply and
demand model (Figure 5-4). The change in consumer surplus is measured as follows:
ACS = -[AQ] x zip]+ [0.5 xAQxAp]. (5.1)
Higher market prices and lower quantities lead to consumer welfare losses. Similarly, the change
in producer surplus is measured as follows:
APS = [AQ} x Ap] - [AQ} x t] - [0.5 x AQ x (Ap - f}}. (5.2)
Higher unit costs and lower production level reduce producer surplus because the net
price change (Ap -1) is negative. However, these losses are mitigated because market prices tend
to rise.
2http://tonto.eia.doe.gov/energy_in_brief/print_pages/electricity.pdf.
5-8
-------
Electricity Restructuring by State
Figure 5-5. Electricity Restructuring by State
Source. U.S. Energy Information Administration. 2008a.
. Last updated September
2008.
5.3 Social Cost Estimate
As shown in Table 5-1 the compliance costs are only a small fraction of the affected
product value; this suggests that shift of the supply curve may also be small and result in small
changes in market prices and consumption. EPA believes the national annualized compliance
cost estimates provide a reasonable approximation of the social cost of this final rule. EPA
believes this approximation is better for industries whose markets are well characterized as
perfectly competitive. This approximation is less well understood for industries where the
characterization of markets is not always perfectly competitive such as electric power generation
whose legal incidence of this rule is approximately 50 percent of the annualized compliance cost.
However, given the data limitation noted earlier, EPA believes the accounting for compliance
cost is a reasonable approximation to inform policy discussion in this rulemaking. To shed more
light on this issue, EPA ran hypothetical analyses and the results are in Tables 5-2 and 5-3.
5-9
-------
5.4 Energy Impacts
Executive Order 13211 (66 FR 28355, May 22, 2001) provides that agencies will prepare
and submit to the Administrator of the Office of Information and Regulatory Affairs, Office of
Management and Budget, a Statement of Energy Effects for certain actions identified as
"significant energy actions." Section 4(b) of Executive Order 13211 defines "significant energy
actions" as any action by an agency (normally published in the Federal Register) that
promulgates or is expected to lead to the promulgation of a final rule or regulation, including
notices of inquiry, advance notices of proposed rulemaking, and notices of proposed rulemaking:
(1) (i) that is a significant regulatory action under Executive Order 12866 or any successor order,
and (ii) is likely to have a significant adverse effect on the supply, distribution, or use of energy;
or (2) that is designated by the Administrator of the Office of Information and Regulatory Affairs
as a significant energy action.
This rule is not a significant energy action as designated by the Administrator of the
Office of Information and Regulatory Affairs because it is not likely to have a significant adverse
impact on the supply, distribution, or use of energy. EPA has prepared an analysis of energy
impacts that explains this conclusion as follows below.
With respect to energy supply and prices, the analysis in Table 5-1 suggests at the
industry level, the annualized costs represent a very small fraction of revenue (all industries are
impacted under 1%). As a result, we can conclude supply and price impacts should be small.
To enhance understanding regarding the regulation's influence on energy consumption,
we examined publicly available data describing energy consumption for the electric power sector
that will be affected by this rule. The Annual Energy Outlook 2010 (EIA, 2009) provides energy
consumption data. As shown in Table 5-4, this industry account for less than 0.5% of the U.S.
total liquid fuels and less than 5.2% of natural gas. As a result, any energy consumption changes
attributable to the regulatory program should not significantly influence the supply, distribution,
or use of energy.
Table 5-4. U.S. Electric Power" Sector Energy Consumption (Quadrillion BTUs): 2013
Quantity Share of Total Energy Use
Distillate fuel oil
Residual fuel oil
Liquid fuels subtotal
Natural gas
0.12
0.34
0.45
5.17
0.1%
0.3%
0.5%
5.1%
5-10
-------
Steam coal 20.69 20.6%
Nuclear power 8.59 8.5%
Renewable energyb 6.06 6.0%
Electricity Imports 0.09 0.1%
Total Electric Power Energy Consumption0 41.18 40.9%
Delivered Energy Use 72.41 72.0%
Total Energy Use 100.59 100.0%
"Includes consumption of energy by electricity-only and combined heat and power plants whose primary business is
to sell electricity, or electricity and heat, to the public. Includes small power producers and exempt wholesale
generators.
blncludes conventional hydroelectric, geothermal, wood and wood waste, biogenic municipal solid waste, other
biomass, petroleum coke, wind, photovoltaic and solar thermal sources. Excludes net electricity imports.
Includes non-biogenic municipal waste not included above.
Source: U.S. Energy Information Administration. 2009a. Supplemental Tables to the Annual Energy Outlook 2010.
Table 2. Available at: .
5.5 Unfunded Mandates
The UMRA requires that we estimate, where accurate estimation is reasonably feasible,
future compliance costs imposed by the rule and any disproportionate budgetary effects. Our
estimates of the future compliance costs of the final rule are discussed previously in Chapter 4 of
this RIA. We do not believe that there will be any disproportionate budgetary effects of the final
rule on any particular areas of the country, State or local governments, types of communities
(e.g., urban, rural), or particular industry segments.
5.5.1 Future and Disproportionate Costs
The UMRA requires that we estimate, where accurate estimation is reasonably feasible,
future compliance costs imposed by the rule and any disproportionate budgetary effects. Our
estimates of the future compliance costs of the final rule are discussed previously in Chapter 4 of
this RIA. We do not believe that there will be any disproportionate budgetary effects of the final
rule on any particular areas of the country, State or local governments, types of communities
(e.g., urban, rural), or particular industry segments.
5.5.2 Effects on the National Economy
The UMRA requires that we estimate the effect of the final rule on the national economy.
To the extent feasible, we must estimate the effect on productivity, economic growth, full
employment, creation of productive jobs, and international competitiveness of the U.S. goods
and services if we determine that accurate estimates are reasonably feasible and that such effect
is relevant and material. The nationwide economic impact of the final rule is presented earlier in
5-11
-------
this RIA chapter. This analysis provides estimates of the effect of the final rule on most of the
categories mentioned above, and these estimates are presented earlier in this RIA chapter. In
addition, we have determined that the final rule contains no regulatory requirements that might
significantly or uniquely affect small governments. Therefore, today's rule is not subject to the
requirements of section 203 of the UMRA.
5-12
-------
SECTION 6
SMALL ENTITY SCREENING ANALYSIS
The Regulatory Flexibility Act as amended by the Small Business Regulatory
Enforcement Fairness Act (SBREFA) 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 governmental jurisdictions, and small not-for-profit enterprises.
After considering the economic impact of the final rule on small entities, the screening
analysis indicates that this final rule will not have a significant economic impact on a substantial
number of small entities (or "SISNOSE"). Under the analyses EPA considered, sales and
revenue tests for establishments owned by model small entities are less than 1% except electric
power generation (NAICS 2211 with receipts less than $100,000 per year) and crop and animal
production (NAICS 111 and 112 with receipts less than $25,000 per year).
6.1 Small Entity Data Set
The industry sectors covered by the final rule were identified during the development of
the cost analysis (Nelson, 2010). The SUSB provides national information on the distribution of
economic variables by industry and enterprise size (U.S. Census, 2006a, b).1 The Census Bureau
and the Office of Advocacy of the SB A supported and developed these files for use in a broad
range of economic analyses.2 Statistics include the total number of establishments and receipts
for all entities in an industry; however, many of these entities may not necessarily be covered by
the final rule. SUSB also provides statistics by enterprise employment and receipt size.
The Census Bureau's definitions used in the SUSB are as follows:
• Establishmeni: An establishment is a single physical location where business is
conducted or where services or industrial operations are performed.
• Receipts: Receipts (net of taxes) are defined as the revenue for goods produced,
distributed, or services provided, including revenue earned from premiums,
commissions and fees, rents, interest, dividends, and royalties. Receipts exclude all
revenue collected for local, state, and federal taxes.
lrThe SUSB data do not provide establishment information for the national security NAICS code (92811) or irrigated
farms. Since most national security installations are owned by the federal government (e.g., military bases), EPA
assumes these entities would not be considered small. For irrigated farms, we relied on receipt data provided in
the 2008 Farm and Irrigation Survey (USDA, 2009).
2See http://www.census.gov/csd/susb/ and http://www.sba.gov/advo/research/data.html for additional details.
6-1
-------
• Enterprise: An enterprise is a business organization consisting of one or more
domestic establishments that were specified under common ownership or control. The
enterprise and the establishment are the same for single-establishment firms. Each
multiestablishment company forms one enterprise—the enterprise employment and
annual payroll are summed from the associated establishments. Enterprise size
designations are determined by the summed employment of all associated
establishments.
Because the SBA's business size definitions (SBA, 2008) apply to an establishment's
"ultimate parent company," we assumed in this analysis that the "enterprise" definition above is
consistent with the concept of ultimate parent company that is typically used for SBREFA
screening analyses and the terms are used interchangeably.
6.2 Small Entity Economic Impact Measures
The analysis generated a set of establishment sales tests (represented as cost-to-receipt
ratios)3 for NAICS codes associated with sectors listed in Table 6-1. Although the appropriate
SBA size definition should be applied at the parent company (enterprise) level, we can only
compute and compare ratios for a model establishment owned by an enterprise within an SUSB
size range (employment or receipts). Using the SUSB size range helps us account for receipt
differences between establishments owned by large and small enterprises and also allows us to
consider the variation in small business definitions across affected industries. Using
establishment receipts is also a conservative approach, because an establishment's parent
company (the "enterprise") may have other economic resources that could be used to cover the
costs of the final rule.
6.2.1 Model Establishment Receipts and Annual Compliance Costs
The sales test compares a representative establishment's total annual engine costs to the
average establishment receipts for enterprises in several size categories.4 For industries with SBA
employment size standards, we calculated average establishment receipts for each enterprise
employment range (Table 6-2).5 For industries with SBA receipt size standards, we calculated
3The following metrics for other small entity economic impact measures (if applicable) would potentially include
small governments (if applicable): "revenue" test; annualized compliance cost as a percentage of annual
government revenues and
small nonprofits (if applicable): "expenditure" test; annualized compliance cost as a percentage of annual
operating expenses,
4For the 1 to 20 employee category, we excluded SUSB data for enterprises with zero employees. These enterprises
did not operate the entire year.
5 We use 2002 Economic Census data in estimating number of establishments by industry instead of using 2007
Economic Census since this data was not available in time for use in our analysis. The release schedules for
different types of 2007 Economic Census data are at
http://www.census.gov/econ/census07/pdf/EconCensusScheduleByDate.pdf.
6-2
-------
Table 6-1. SI NESHAP for Existing Stationary Reciprocating Internal Combustion
Engines (RICE): Affected Sectors and SBA Small Business Size Standards
Industry Description
Electric Power Generation
Natural Gas Transmission
Crude Petroleum & NG Production
Natural Gas Liquid Producers
National Security
Hydro Power Units
Irrigation Sets
Corresponding
NAICS
2211
48621
211111
211112
92811
See NAICS 22 11
Affects NAICS 111 and
SBA Size Standard for
Businesses (August 22nd,
2008)
a
$7.0 million in annual receipts
500 employees
500 employees
NA
1,000 employees
Generally $750,000 or less in
Type of Small
Entity
Business and
government
Business
Business
Business
Government
Business and
government
Business
Welders
112
Affects industries that
use heavy equipment
such as construction,
mining, farming
annual receipts
Varies by 6-digit NAICS code;
Example industry:
NAICS 238 = $14 million in
annual receipts
Business
aNAICS codes 221111, 221112, 221113, 221119, 221121, 221122: A firm is small if, including its affiliates, it is
primarily engaged in the generation, transmission, and/or distribution of electric energy for sale and its total
electric output for the preceding fiscal year did not exceed 4 million megawatt hours.
average establishment receipts for each enterprise receipt range (Table 6-3). We included the
utility sector in the second group, although the SBA size standard for this industry is defined in
terms of physical units (megawatt hours) versus receipts. Crop and animal production (NAICS
111 and 112) also have an SBA receipt size standard that defines a small business as receiving
$750,000 or less in receipts per year. However, SUSB data were not available for these
industries. Therefore, we conducted the sales test using the following range of establishment
receipts: farms with annual receipts of $25,000 or less, farms with annual receipts of $100,000 or
less, farms with annual receipts of $500,000 or less, and farms with annual receipts of $750,000
or less.
6-3
-------
Table 6-2. Average Receipts for Affected Industry by Enterprise: 2002 ($2009 Million/Establishment)
NAICS
211111
211112
335312
333992
SBA Size Standard
for Businesses All 1-20
NAICS Description (effective August 22, 2008) Enterprises employees
Crude Petroleum & NG Production 500
Natural Gas Liquid Producers 500
Motor & generator mfg
employees
employees
1,000 employees
Welding & soldering equipment mfg 500
employees
$14.76 $0.54
$174.86 $0.31
$18.80 $1.38
$18.73 $1.58
Owned Bj
20 to 99
employees
$6.79
NA
$6.22
$6.67
Enterprises with Employee Range:
1,000 to
100 to 499 500 to 749 750 to 999 1,499
employees employees employees employees
$9.63
$12.03
$16.15
$33.64
NA NA NA
NA NA NA
$29.82 NA NA
NA NA $115.91
NA = Not available.
Table
NAICS
2211
48621
92811
6-3. Average Receipts for Affected Industry
NAICS Description
Electric Power
Generation
Natural Gas
Transmission
National Security
SBA Size
Standard
for Businesses
(effective August
22nd, 2008)
a
$7.0 million in
annual receipts
NA
All
Enterprises
$40.23
$22.09
NA
by Enterprise Receipt Range:
Owned
100- 500- 1,000-
2002 ($2009/Establishment)
By Enterprises with Receipt Range:
5,000,000-
0-99K 499.9K 999.9K 4,999.9K 9,999,999K
Receipts Receipts Receipts Receipts
$0.1 $0.3 $0.8 $3.1
$0.08 $0.32 $0.89 $2.51
NA NA NA NA
Receipts
$6.7
$6.97
NA
<10,000
K
Receipts
$2.7
$1.57
NA
10,000- 50,000-
49,999K 99,999K 100,OOOK
Receipts Receipts + Receipts
$14.8 $22.5 $49.8
$10.69 $45.99 $23.06
NA NA NA
a NAICS codes 221111, 221112, 221113, 221119, 221121, 221122 - A firm is small if, including its affiliates, it is primarily engaged in the generation,
transmission, and/or distribution of electric energy for sale and its total electric output for the preceding fiscal year did not exceed 4 million megawatt hours.
NA = Not available. SUSB did not report this data disclosure or other reasons.
-------
Annual entity compliance costs vary depending on the size of the SI engines used at the
affected establishment. Absent facility-specific information, we computed per-entity compliance
costs based for three different cases based on representative establishments—Cases 1, 2, and 3
(see Table 6-4). Each representative establishment differs based on the size and number of SI
engines being used. Compliance costs are calculated by summing the total annualized
compliance costs for the relevant engine categories, dividing the sum by the total existing
population of those engines, and multiplying the average engine cost by the number of engines
assumed to be at the establishment. Since NAICS 2211 and 48621 are fundamentally different
than other industries considered in this analysis due to the number of engines affected and
amount of cost incurred resulting from this final rule, we used different assumptions about what
constitutes the representative establishment and report these assumptions separately.
• Case 1: The representative establishment for all industries uses three 750+ hp engines
with an average compliance cost of $8,500 per engine, resulting in a total annualized
compliance cost of approximately $25,500 for this representative establishment.
• Case 2: The representative establishment in NACIS 2211 and 48621 uses two 25 to
750+ hp engines with an average compliance cost of $1,013 per engine, resulting in a
total annualized compliance cost of $2,026 for this representative establishment. For
all other industries, the representative establishment uses two 25 to 300 hp engines
with an average compliance cost of $245 per engine, resulting in a total compliance
cost of $490 for this representative establishment.
• Case 3: The representative establishment for all industries uses two 50 to 100 hp
engines with an average compliance cost of $73 per engine, resulting in a total
compliance cost of $145 for this representative establishment.
EPA believes that small entities are most likely to face costs similar to Case 2 (columns
shaded in gray in Table 6-4) because most of the engines to be affected by this final rule in
NAICS 335312, 333992, 211111, and 211112 are under 300 hp capacity, and most small entities
in these industries will own engines of this size or smaller. This is corroborated by Figure 6-1
and 6-2 which shows the distribution of engine population and compliance costs by engine size
for all industries. However, it is difficult to make a similar claim for NAICS 2211 and 48621
based on the existing distribution of engines in these industries.6
6This claim also cannot be made for NAICS 92811: National Security. However, since most national security
installations are owned by the federal government (e.g., military bases), EPA assumes these entities would not be
considered small.
6-5
-------
For the sales test, we divided the representative establishment compliance costs reported
in Table 6-4 by the representative establishment receipts reported in Tables 6-2 and 6-3. This is
known as the cost-to-receipt (i.e., sales) ratio, or the "sales test." The "sales test" is the impact
Table 6-4. Representative Establishment Costs Used for Small Entity Analysis ($2009)
Casel
NAICS All Other
2211,48621 NAICS
(+750 hp (+750 hp
only) only)
Case 2
All Other
NAICS NAICS
2211,48621 (25-300
(25-750+ hp) hp)
Case 3
NAICS
2211,
48621
(25-100
hp only)
All Other
NAICS
(25-100
hp only)
Total Annualized Costs ($)
Engine Population
Average Engine Cost
($/engine)
Assumed Engines Per
Establishment
Total Annualized Costs per
Establishment
$89,716,669 $7,337,540
10,548 863
$203,582,405 $28,057,197
200,974 114,517
$8,510,985 $6,174,805
117,040 84,913
$8,506
$8,502
$25,517 $25,507
$1,013
$2,026
$245
$490
$73
$145
$73
$145
6-6
-------
100%
90%
70%
60%
40%
30%
0%
>750hp
600-750 hp
500-600 hp
300-500 hp
1175-300 hp
100-175 hp
I 50-100 hp
I 25-50 hp
Electric Power Natural Gas Crude Petroleum Natural Gas National Security Hydro Power Irrigation Sets Welders
Generation Transmission & NG Production Liquid Producers (92811) Units (335312) (335312) (333992)
(2211) (48621) (211111) (211112)
Figure 6-1. Distribution of Engine Population by Size for All Industries
6-7
-------
100%
70%
60%
30%
20%
>750hp
600-750 hp
500-600 hp
300-500 hp
1175-300 hp
100-175 hp
I 50-100 hp
I 25-50 hp
Electric Power Natural Gas Crude Petroleum Natural Gas National Security Hydro Power Irrigation Sets Welders
Generation Transmission & NG Production Liquid Producers (92811) Units (335312) (335312) (333992)
(2211) (48621) (211111) (211112)
Figure 6-2. Distribution of Compliance Costs by Engine Size for All Industries
methodology EPA employs in analyzing small entity impacts as opposed to a "profits test," in
which annualized compliance costs are calculated as a share of profits.
This is because revenues or sales data are commonly available data for entities normally
impacted by EPA regulations and profits data normally made available are often not the true
profit earned by firms because of accounting and tax considerations. Revenues as typically
published are usually correct figures and are more reliably reported when compared to profit
data. The use of a "sales test" for estimating small business impacts for a rulemaking such as this
one is consistent with guidance offered by EPA on compliance with SBREFA7 and is consistent
with guidance published by the U.S. SBA's Office of Advocacy that suggests that cost as a
percentage of total revenues is a metric for evaluating cost increases on small entities in relation
to increases on large entities.8
7The SBREFA compliance guidance to EPA rulewriters regarding the types of small business analysis that should be
considered can be found at http://www.epa.gov/sbrefa/documents/rfafinalguidance06.pdf, pp. 24-25.
8U.S. SBA, Office of Advocacy. A Guide for Government Agencies, How to Comply with the Regulatory
Flexibility Act, Implementing the President's Small Business Agenda and Executive Order 13272, May 2003.
6-8
-------
If the cost-to-receipt ratio is less than 1%, then we consider the final rule to not have a
significant impact on the establishment company in question. We summarize the industries with
cost-to-receipt ratios exceeding 1% below:
Primary Analysis:
• Case 2: NAICS 2211 with receipts less than $100,000 per year and NAICS 111 and
112 with receipts less than $25,000 per year
• Case 3: No industries
Sensitivity Analysis (unlikely):
• Case 1: NAICS 2211 with receipts less than $100,000 per year
In the Case 2 primary analysis, only establishments in NAICS 2211 with receipts less
than $100,000 per year (less than 5 percent of the total), and establishments in NAICS 111 and
112 with receipts less than $25,000 per year (around 30 percent of the total) have cost-to-receipt
ratios above 1%. However, establishments earning this level of receipts are likely to be using
smaller engines than those assumed in Case 2, such as 25 to 300 hp engines. The results of our
Case 3 analysis demonstrate that these establishments are not significantly impacted when taking
this engine size into account.
After considering the economic impacts of this final rule on small entities, we certify that
this action will not have a significant economic impact on a substantial number of small entities.
This certification is based on the economic impact of this final action to all affected small entities
across all industries affected. We estimate that all small entities will have annualized costs of less
than 1 percent of their sales in all industries except NAICS 2211 (electric power generation,
transmission, and distribution) and NAICS 111 and 112 (Crop and Animal Production). The
number of small entities in NAICS 2211 having annualized costs of greater than 1 percent of
their sales is less than 5 percent, and the number of small entities in NAICS 111 and 112 having
annualized costs of greater than 1 percent of their sales (but less than 2 percent of sales) is 30
percent. We thus conclude that there is no significant economic impact on a substantial number
of small entities (SISNOSE) for this rule.
Although the final rule would not have a significant economic impact on a substantial
number of small entities, EPA nonetheless tried to reduce the impact of the final rule on small
entities. When developing the revised standards, EPA took special steps to ensure that the
burdens imposed on small entities were minimal. EPA conducted several meetings with industry
6-9
-------
trade associations to discuss regulatory options and the corresponding burden on industry, such
as recordkeeping and reporting. In this rule, we are applying the minimum level of control (i.e.,
the MACT floor) to small non-emergency engines (below 300 HP) and emergency engines
located at major HAP sources and the minimum level of testing, monitoring, recordkeeping, and
reporting to affected RICE sources, both major and area, allowed by the CAA. Other alternatives
considered that provided more than the minimum level of control were deemed as not technically
feasible or cost-effective for EPA to implement for small non-emergency engines and emergency
engines as explained earlier.
6.3 Small Government Entities
The rule also covers sectors that include entities owned by small and large governments.
However, given the uncertainty and data limitations associated with identifying and
appropriately classifying these entities, we computed a "revenue" test for a model small
government, where the annualized compliance cost is a percentage of annual government
revenues (U.S. Census, 2005a, b). The use of a "revenue test" for estimating impacts to small
governments for a rulemaking such as this one is consistent with guidance offered by EPA on
compliance with SBREFA,9 and is consistent with guidance published by the US SBA's Office
of Advocacy.10 For example, from the 2002 Census (in 2008 dollars), the average revenue for
small governments (counties and municipalities) with populations fewer than 10,000 is $3
million per entity, and the average revenue for local governments with populations fewer than
50,000 is $8 million per entity (U.S. Census Bureau, 2005a; U.S. Census Bureau, 2005b). For the
smallest group of local governments (<10,000 people), the cost-to-revenue ratio would be 0.2%
or less under each case. For the larger group of governments (<50,000 people), the cost-to-
revenue ratio is 0.1% or less under all cases.
9The SBREFA compliance guidance to EPA rule writers regarding the types of small business analysis that should
be considered can be found at http://www.epa.gov/sbrefa/documents/rfafinalguidance06.pdf. pp. 24-25.
10U.S. SBA, Office of Advocacy. A Guide for Government Agencies, How to Comply with the Regulatory
Flexibility Act, Implementing the President's Small Business Agenda and Executive Order 13272, May 2003.
6-10
-------
SECTION 7
HUMAN HEALTH BENEFITS OF EMISSIONS REDUCTIONS
7.1 Synopsis
In this section, we provide an estimate of the monetized co-benefits associated with
reducing particulate matter (PM) for the final NESHAP for spark ignition reciprocating internal
combustion engines (SI RICE). Specifically, we calculated the co-benefits of this rule in terms of
the co-benefits associated with reducing PM rather than calculating the co-benefits associated
with reducing hazardous air pollutants (HAPs). These PM reductions are a consequence of the
technologies installed to reduce HAP emissions from SI RICE. These estimates reflect the
monetized human health co-benefits of reducing cases of morbidity and premature mortality
among populations exposed to the PM2.5 precursors reduced by this rulemaking. Using a 3%
discount rate, we estimate the total monetized co-benefits of the final NESHAP to be $510
million to $1.2 billion in the implementation year (2013). Using a 7% discount rate, we estimate
the total monetized co-benefits of the final NESHAP to be $460 million to $1.1 billion in the
implementation year. All estimates are in 2009$.
These estimates reflect EPA's most current interpretation of the scientific literature.
Higher or lower estimates of benefits are possible using other assumptions; examples of this are
provided in Figure 7-2. Data, resource, and methodological limitations prevented EPA from
monetizing the benefits from several important benefit categories, including benefits from
reducing hazardous air pollutants, ecosystem effects, and visibility impairment. The benefits
from reducing other air pollutants have not been monetized in this analysis, including reducing
109,000 tons of carbon monoxide and 6,000 tons of HAPs each year.
7.2 Calculation of PM2.5 Human Health Co-Benefits
This rulemaking would reduce emissions of NOx and VOCs. Because NOx and VOCs
are precursors to PM2.5; reducing these emissions would also reduce PM2.5 formation, human
exposure, and the incidence of PM2.s-related health effects. These PM reductions are a
consequence of the technologies installed to reduce HAP emissions from SI RICE. Due to
analytical limitations, it was not possible to provide a comprehensive estimate of PM2.s-related
co-benefits. Instead, we used the "benefit-per-ton" approach to estimate these co-benefits based
on the methodology described in Fann, Fulcher, and Hubbell (2009). The key assumptions are
described in detail below. These PM2.5 benefit-per-ton estimates provide the total monetized
human health co-benefits (the sum of premature mortality and premature morbidity) of reducing
one ton of PM2.5 from a specified source. EPA has used the benefit per-ton technique in several
7-1
-------
previous RIAs, including the recent NO2 NAAQS RIA (U.S. EPA, 2010b). Table 7-1 shows the
quantified and unquantified co-benefits captured in those benefit-per-ton estimates.
Table 7-1. Human Health and Welfare Effects of PM2.5
Pollutant / Quantified and Monetized in Primary
Effect Estimates
Unquantified Effects Changes in:
Adult premature mortality
Bronchitis: chronic and acute
Hospital admissions: respiratory and
cardiovascular
Emergency room visits for asthma
Nonfatal heart attacks (myocardial infarction)
Lower and upper respiratory illness
Minor restricted-activity days
Work loss days
Asthma exacerbations (asthmatic population)
Infant mortality
Subchronic bronchitis cases
Low birth weight
Pulmonary function
Chronic respiratory diseases other than chronic
bronchitis
Non-asthma respiratory emergency room visits
Visibility
Household soiling
Consistent with the Portland Cement NESHAP (U.S. EPA, 2009a), the PM2.5 co-benefits
estimates utilize the concentration-response functions as reported in the epidemiology literature,
as well as the 12 functions obtained in EPA's expert elicitation study as a sensitivity analysis.
• One estimate is based on the concentration-response (C-R) function developed from
the extended analysis of American Cancer Society (ACS) cohort, as reported in Pope
et al. (2002), a study that EPA has previously used to generate its primary benefits
estimate. When calculating the estimate, EPA applied the effect coefficient as
reported in the study without an adjustment for assumed concentration threshold of 10
|ig/m3 as was done in recent (2006-2009) Office of Air and Radiation RIAs.
• One estimate is based on the C-R function developed from the extended analysis of
the Harvard Six Cities cohort, as reported by Laden et al. (2006). This study,
published after the completion of the Staff Paper for the 2006 PM2.5 NAAQS, has
been used as an alternative estimate in the PM2.5 NAAQS RIA and PM2.5 benefits
estimates in RIAs completed since the PM2.5 NAAQS. When calculating the estimate,
EPA applied the effect coefficient as reported in the study without an adjustment for
assumed concentration threshold of 10 |ig/m3 as was done in recent (2006-2009)
RIAs.
• Twelve estimates are based on the C-R functions from EPA's expert elicitation study
(lEc, 2006; Roman et al., 2008) on the PM2 5 -mortality relationship and interpreted
for benefits analysis in EPA's final RIA for the PM2.5 NAAQS. For that study, twelve
7-2
-------
experts (labeled A through L) provided independent estimates of the PM2.5 -mortality
concentration-response function. EPA practice has been to develop independent
estimates of PM2.5 -mortality estimates corresponding to the concentration-response
function provided by each of the twelve experts, to better characterize the degree of
variability in the expert responses.
The effect coefficients are drawn from epidemiology studies examining two large
population cohorts: the American Cancer Society cohort (Pope et al., 2002) and the Harvard Six
Cities cohort (Laden et al., 2006).J These are logical choices for anchor points in our presentation
because, while both studies are well designed and peer reviewed, there are strengths and
weaknesses inherent in each, which we believe argues for using both studies to generate co-
benefits estimates. Previously, EPA had calculated co-benefits based on these two empirical
studies, but derived the range of co-benefits, including the minimum and maximum results, from
an expert elicitation of the relationship between exposure to PM2.5 and premature mortality
(Roman et al., 2008).2 Within this assessment, we include the co-benefits estimates derived from
the concentration-response function provided by each of the twelve experts to better characterize
the uncertainty in the concentration-response function for mortality and the degree of variability
in the expert responses. Because the experts used these cohort studies to inform their
concentration-response functions, co-benefits estimates using these functions generally fall
between results using these epidemiology studies (see Figure 7-2). In general, the expert
elicitation results support the conclusion that the co-benefits of PM2.5 control are very likely to be
substantial.
Readers interested in reviewing the methodology for creating the benefit-per-ton
estimates used in this analysis should consult Fann, Fulcher, and Hubbell (2009). As described in
the documentation for the benefit per-ton estimates cited above, national per-ton estimates are
developed for selected pollutant/source category combinations. The per-ton values calculated
therefore apply only to tons reduced from those specific pollutant/source combinations (e.g.,
NOx emitted from electric generating units; NC>2 emitted from mobile sources). Our estimate of
PM2.5 control co-benefits is therefore based on the total NOx and VOC emissions controlled by
sector and multiplied by this per-ton value.
These models assume that all fine particles, regardless of their chemical composition, are
equally potent in causing premature mortality because there is no clear scientific evidence that
would support the development of differential effects estimates by particle type. NOx and VOCs
1 These two studies specify multi-pollutant models that control for NOx, among other pollutants.
2 Please see the Section 5.2 of the Portland Cement proposal RIA in Appendix 5 A for more information regarding
the change in the presentation of co-benefits estimates.
7-3
-------
are the primary PM2.5 precursors affected by this rule. Even though we assume that all fine
particles have equivalent health effects, the benefit-per-ton estimates vary between precursors
because each ton of precursor reduced has a different propensity to form PM2.5. For example,
NOx has a lower benefit-per-ton estimate than direct PM2.5 because it does not form as much
PM2.5, thus the exposure would be lower, and the monetized health co-benefits would be lower.
The benefit-per-ton coefficients in this analysis were derived using modified versions of
the health impact functions used in the PM NAAQS Regulatory Impact Analysis. Specifically,
this analysis uses the benefit-per-ton method first applied in the Portland Cement NESHAP RIA
(U.S. EPA, 2009a), which incorporated three updates: a new population dataset, an expanded
geographic scope of the benefit-per-ton calculation, and the functions directly from the
epidemiology studies without an adjustment for an assumed threshold.3 Removing the threshold
assumption is a key difference between the method used in this analysis of PM co-benefits and
the methods used in RIAs prior to Portland Cement, and we now calculate incremental co-
benefits down to the lowest modeled PM2.5 air quality levels.
EPA strives to use the best available science to support our benefits analyses, and we
recognize that interpretation of the science regarding air pollution and health is dynamic and
evolving. Based on our review of the current body of scientific literature, EPA now estimates
PM-related mortality without applying an assumed concentration threshold. EPA's Integrated
Science Assessment for Particulate Matter (U.S. EPA, 2009b), which was recently 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
recognizing potential uncertainty about the exact shape of the concentration-response function.
Since then, the Health Effects Subcommittee (U.S. EPA-SAB, 2010) of EPA's Council
concluded, "The HES fully supports EPA's decision to use a no-threshold model to estimate
mortality reductions. This 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. Therefore, there is no evidence to support a truncation of the CRF."
In conjunction with the underlying scientific literature, this document provided a basis for
reconsidering the application of thresholds in PM2.5 concentration-response functions used in
EPA's RIAs. For a summary of these scientific review statements and the panel members
3 The benefit-per-ton estimates have also been updated since the Cement RIA to incorporate a revised VSL, as
discussed on the next page.
7-4
-------
commenting on thresholds since 2002, please consult the Technical Support Document (TSD)
Summary of Expert Opinions on the Existence of a Threshold (U.S. EPA, 2010c), which is
provided as an appendix to this RIA.
Consistent with this recent scientific advice, we are replacing the previous threshold
sensitivity analysis with a new "Lowest Measured Level" (LML) assessment. This information
allows readers to determine the portion of population exposed to annual mean PM2.5 levels at or
above the LML of each study; in general, our confidence in the estimated PM mortality
decreases as we consider air quality levels further below the LML in major cohort studies that
estimate PM-related mortality. While an LML assessment provides some insight into the level of
uncertainty in the estimated PM mortality benefits, EPA does not view the LML as a threshold
and continues to quantify PM-related mortality impacts using a full range of modeled air quality
concentrations. It is important to emphasize that we have high confidence in PM2.s-related effects
down to the lowest LML of the major cohort studies. Just because we have greater confidence in
the benefits above the LML, this does not mean that we have no confidence that benefits occur
below the LML.
For this analysis, policy-specific air quality data is not available due to time or resource
limitations. For these rules, we are unable to estimate the percentage of premature mortality
associated with this specific rule's emission reductions at each PM2.5 level. However, we believe
that it is still important to characterize the distribution of exposure to baseline air quality levels.
As a surrogate measure of mortality impacts, we provide the percentage of the population
exposed at each PM2.5 level using the most recent modeling available from the recently proposed
Transport Rule (U.S. EPA, 2010e). It is important to note that baseline exposure is only one
parameter in the health impact function, along with baseline incidence rates population, and
change in air quality. In other words, the percentage of the population exposed to air pollution
below the LML is not the same as the percentage of the population experiencing health impacts
as a result of a specific emission reduction policy. The most important aspect, which we are
unable to quantify for rules without air quality modeling, is the shift in exposure associated with
this specific rule. Therefore, caution is warranted when interpreting the LML assessment. For
more information on the data and conclusions in the LML assessment for rules without policy-
specific air quality modeling, please consult the LML TSD (U.S. EPA, 2010d), which is
provided as an appendix to this RIA. The results of this analysis are provided in Section 7.4.
As is the nature of Regulatory Impact Analyses (RIAs), the assumptions and methods
used to estimate air quality co-benefits evolve over time to reflect the Agency's most current
interpretation of the scientific and economic literature. For a period of time (2004-2008), the
7-5
-------
Office of Air and Radiation (OAR) valued mortality risk reductions using a value of statistical
life (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 $5.5 million (2000$)4 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)5 while the Agency continues its efforts to
update its guidance on this issue. This approach calculates a mean value across VSL estimates
derived from 26 labor market and contingent valuation studies published between 1974 and
1991. The mean VSL across these studies is $6.3 million (2000$).6 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.
4. After adjusting the VSL for a different currency year (2009$) and to account for income growth to 2015 to the
$5.5 million value, the VSL is $7.9 million.
5 In the (draft) update of the Economic Guidelines (U.S. EPA, 2008), EPA retained the VSL endorsed by the SAB
with the understanding that further updates to the mortality risk valuation guidance would be forthcoming in the
near future. Therefore, this report does not represent final agency policy.
6 In this analysis, we adjust the VSL to account for a different currency year (2009$) and to account for income
growth to 2015. After applying these adjustments to the $6.3 million value, the VSL is $9.1 million.
7-6
-------
Figure 7-1 illustrates the relative breakdown of the monetized PM2.5 health co-benefits by
health endpoint.
Acute Respiratory Symptoms
0.5%
,fant Mortality 0.4%
Work Loss Days 0.2%
Hospital Admissions,
0.2%
Hhlospital Admissions, Resp
0.04%
Asthma Exacerbation 0.01%
Acute Bronchitis 0.01%
Upper Resp Symp 0.00%
Lower Resp Symp 0.00%
ER Visits, Resp 0.00%
Figure 7-1. Breakdown of Monetized PM2.s Health Co-Benefits using Mortality Function
from Pope et al. (2002)a
a This pie chart breakdown is illustrative, using the results based on Pope et al. (2002) as an example. Using the
Laden et al. (2006) function for premature mortality, the percentage of total monetized co-benefits due to adult
mortality would be 97%. This chart shows the breakdown using a 3% discount rate, and the results would be
similar if a 7% discount rate was used.
Table 7-2 provides a general summary of the monetized co-benefits results by pollutant,
including the emission reductions and benefits-per-ton estimates at discount rates of 3% and 7%.7
Table 7-3 provides a summary of the reductions in health incidences anticipated as a result of the
pollution reductions. In Table 7-4, we provide the monetized co-benefits using our anchor points
of Pope et al. and Laden et al. as well as the results from the expert elicitation on PM mortality.
To comply with Circular A-4, EPA provides monetized co-benefits using discount rates of 3% and 7% (OMB,
2003). These co-benefits are estimated for a specific analysis year (i.e., 2013), and most of the PM co-benefits
occur within that year with two exceptions: acute myocardial infarctions (AMIs) and premature mortality. For
AMIs, we assume 5 years of follow-up medical costs and lost wages. For premature mortality, we assume that
there is a "cessation" lag between PM exposures and the total realization of changes in health effects. 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, 2004). Changes in the lag assumptions do not change
the total number of estimated deaths but rather the timing of those deaths. Therefore, discounting only affects the
AMI costs after the analysis year and the valuation of premature mortalities that occur after the analysis year. As
such, the monetized co-benefits using a 7% discount rate are only approximately 10% less than the monetized
co-benefits using a 3% discount rate.
7-7
-------
Figures 7-2 and 7-3 provide a visual representation of the range of monetized co-benefits
estimates and the pollutant breakdown of the monetized co-benefits of the proposed option. The
final NESHAP is the MACT floor level of control for all major SI RICE sources except for four-
stroke rich-burn (4SRB) engines of 300-500 horsepower (HP), where the required level of
control is above the MACT floor, and the GACT level of control for area SI RICE sources. We
also show results for an alternative in which only the MACT level of control is applied to all SI
RICE major sources.
7-8
-------
Table 7-2. Summary of Monetized Co-Benefits Estimates for the Final Spark Ignition
NESHAP in 2013 (2009$)a
„ . . Benefit Benefit Benefit Benefit
Emissions ,
T. 11 ^ ^ T> ., +• per ton per ton per ton per ton
Pollutant Reductions * ' , ,n ' ,
, , (Pope, (Laden, (Pope, (Laden,
(tons) 3%)
3%)
7%)
7%)
Total Total
Monetized Monetized
Benefits Benefits
(millions 2008$ (millions
at 3%) 2008$ at 7%)
Final NESHAP:
Majorb
Alternative 1:
Major
si Va
at
< °
..
-------
Table 7-3. Summary of Reductions in Health Incidences from PMi.s Co-Benefits for Final
SI RICE NESHAP in 2013a
Final NESHAP: Alternative 2:
Majorb Major
Final: Final: Major
Area Source and Area
onlyc Sources TOTAL
Avoided Premature Mortality
Pope et al.
Laden et al.
Avoided Morbidity
Chronic Bronchitis
Acute Myocardial Infarction
Hospital Admissions, Respiratory
Hospital Admissions, Cardiovascular
Emergency Room Visits, Respiratory
Acute Bronchitis
Work Loss Days
Asthma Exacerbation
Minor Restricted Activity Days
Lower Respiratory Symptoms
Upper Respiratory Symptoms
1
2
1
2
0
0
1
2
130
16
740
18
14
2
5
2
4
1
1
2
3
280
37
1,700
41
31
16
42
12
31
4
8
12
27
2,200
300
13,000
320
240
17
44
12
33
4
9
13
29
2,400
310
14,000
340
260
a All estimates are for the analysis year (2013) and are rounded to whole numbers with two significant figures. All
fine particles are assumed to have equivalent health effects, but each PM2 5 precursor pollutant has a different
propensity to form PM25. Confidence intervals are unavailable for this analysis because of the benefit-per-ton
methodology.
bThe final NESHAP is the MACT floor level of control for all major SI RICE sources and the GACT level of
control for area SI RICE sources. We also show results for an alternative (referred to as Alternative 2) in which
the MACT level of control is applied to all SI RICE major sources except for four-stroke rich-burn (4SRB)
engines of 300-500 horsepower (HP), where the level of control is above the MACT floor, and the GACT level of
control is applied to all area SI RICE sources.
0 All of the benefits for area sources are attributable to reductions expected from 4SLB and 4SRB non-emergency
engines above 500 HP.
7-10
-------
Table 7-4. All Monetized PM2.5 Co-Benefits Estimates for the Final SI RICE NESHAP at
discount rates of 3% and 7% in 2013 (in millions of 2009$)a
Final NESHAP:
Majorb
Benefit-per-ton
Pope et al.
Laden et al.
Benefit-per-ton
Expert A
Expert B
Expert C
Expert D
Expert E
Expert F
Expert G
Expert H
Expert I
Expert J
Expert K
Expert L
3%
Coefficients
$8.2
$20
Coefficients
$21
$16
$16
$11
$26
$15
$10
$12
$16
$13
$3.3
$12
7%
derived
$7.4
$18
Derived
$19
$15
$15
$10
$24
$13
$9
$11
$14
$12
$3.0
$11
Alternative 2:
Major
3%
7%
from Epidemiology
$48
$120
from Expert
$124
$95
$95
$67
$153
$86
$57
$71
$94
$76
$19
$70
$43
$110
Final: Area Source only0
3%
Literature
$500
$1,200
7%
$450
$1,100
Final: Major and Area
Sources Total
3%
$510
$1,200
7%
$460
$1,100
Elicitation
$112
$86
$85
$61
$139
$78
$51
$65
$85
$69
$18
$63
$1,300
$1,000
$1,000
$700
$1,600
$900
$590
$700
$1,000
$800
$200
$700
$1,160
$900
$900
$600
$1,400
$800
$530
$700
$900
$700
$190
$700
$1,300
$1,000
$1,000
$710
$1,600
$910
$600
$750
$990
$810
$210
$740
$1,200
$910
$900
$640
$1,500
$820
$540
$680
$890
$730
$190
$670
a All estimates are rounded to two significant figures. Estimates do not include confidence intervals because they
were derived through the benefit-per-ton technique described above. The co-benefits estimates from the Expert
Elicitation are provided as a reasonable characterization of the uncertainty in the mortality estimates associated
with the concentration-response function. Confidence intervals are unavailable for this analysis because of the
benefit-per-ton methodology.
b The final NESHAP is the MACT floor level of control for all major SI RICE and the GACT level of control for
area SI RICE sources. We also show results for an alternative (referred to as Alternative 2) in which the MACT
level of control is applied to all SI RICE major sources except for four-stroke rich-burn (4SRB) engines of 300-
500 horsepower (HP), where the required level of control is above the MACT floor, and the GACT level of
controls is applied to all area SI RICE sources.
0 All of the benefits for area sources are attributable to reductions expected from 4SLB and 4SRB non-emergency
engines above 500 HP.
7-11
-------
$1,600
$1,400
$1,200
$1,000
$800
$600
$400
$200
$0
Benefits estimates derived from 2 epidemiology functions and 12 expert functions
Figure 7-2. Total Monetized PM2.5 Co-Benefits for the Final SI RICE NESHAP in 2013
a This graph shows the estimated co-benefits at discount rates of 3% and 7% using effect coefficients derived from
the Pope et al. study and the Laden et al. study, as well as 12 effect coefficients derived from EPA's expert
elicitation on PM mortality. The results shown are not the direct results from the studies or expert elicitation;
rather, the estimates are based in part on the concentration-response function provided in those studies.
7-12
-------
NOx
(major)
0%
Figure 7-3. Breakdown of Monetized Co-Benefits for the Final SI RICE NESHAP by
PM2.s Precursor Pollutant and Source
7.3 Unquantified Benefits
The monetized co-benefits estimated in this RIA only reflect the portion of co-benefits
attributable to the health effect reductions associated with ambient fine particles. Data, resource,
and methodological limitations prevented EPA from quantifying or monetizing the benefits from
several important benefit categories, including benefits from reducing toxic emissions,
ecosystem effects, and visibility impairment. The health co-benefits from reducing hazardous air
pollutants (HAPs) and carbon monoxide each year have not been monetized in this analysis. In
addition to being a PM2.5 precursor, NOx emissions also contribute to adverse effects from acidic
deposition in aquatic and terrestrial ecosystems, increase mercury methylation, as well as
visibility impairment.
7.3.1 Carbon Monoxide Benefits
Carbon monoxide (CO) exposure is associated with a variety of health effects. Without
knowing the location of the emission reductions and the resulting ambient concentrations using
fine-scale air quality modeling, we were unable to estimate the exposure to CO for nearby
populations. Due to data, resource, and methodological limitations, we were unable to estimate
7-13
-------
the benefits associated with the reductions in CO emissions that would occur as a result of this
rule.
Carbon monoxide in ambient air is formed primarily by the incomplete combustion of
carbon-containing fuels and photochemical reactions in the atmosphere. The amount of CO
emitted from these reactions, relative to carbon dioxide (CO2), is sensitive to conditions in the
combustion zone, such as fuel oxygen content, burn temperature, or mixing time. Upon
inhalation, CO diffuses through the respiratory system to the blood, which can cause hypoxia
(reduced oxygen availability). Carbon monoxide can elicit a broad range of effects in multiple
tissues and organ systems that are dependent upon concentration and duration of exposure.
The Integrated Science Assessment for Carbon Monoxide (U.S. EPA, 2010a) concluded
that short-term exposure to CO is "likely to have a causal relationship" with cardiovascular
morbidity, particularly in individuals with coronary heart disease. Epidemiologic studies
associate short-term CO exposure with increased risk of emergency department visits and
hospital admissions. Coronary heart disease includes those who have angina pectoris (cardiac
chest pain), as well as those who have experienced a heart attack. Other subpopulations
potentially at risk include individuals with diseases such as chronic obstructive pulmonary
disease (COPD), anemia, or diabetes, and individuals in very early or late life stages, such as
older adults or the developing young. The evidence is suggestive of a causal relationship
between short-term exposure to CO and respiratory morbidity and mortality. The evidence is also
suggestive of a causal relationship for birth outcomes and developmental effects following long-
term exposure to CO, and for central nervous system effects linked to short- and long-term
exposure to CO.
7.3.2 OtherNOx Benefits
In addition to being a precursor to PM2.5, NOx emissions are also associated with a
variety of respiratory health effects. Unfortunately, we were unable to estimate the health
benefits associated with reduced NOx exposure in this analysis because we do not have air
quality modeling data available. Without knowing the location of the emission reductions and the
resulting ambient concentrations, we were unable to estimate the exposure to NOx for nearby
populations. Therefore, this analysis only quantifies and monetizes the PM2 5 co-benefits
associated with the reductions in NOx emissions.
Following an extensive evaluation of health evidence from epidemiologic and laboratory
studies, the Integrated Science Assessment (ISA) for Nitrogen Dioxide concluded that there is a
likely causal relationship between respiratory health effects and short-term exposure to NO2
7-14
-------
(U.S. EPA, 2008a). Persons with preexisting respiratory disease, children, and older adults may
be more susceptible to the effects of NC>2 exposure. Based on our review of this information, we
identified four short-term morbidity endpoints that the NC>2 ISA identified as a "likely causal
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 NC>2 ISA. The NC>2 ISA also
concluded that the relationship between short-term NC>2 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 NC>2 alone. Although the NC>2 ISA stated that studies consistently
reported a relationship between NC>2 exposure and mortality, the effect was generally smaller
than that for other pollutants such as PM.
NOx emissions also contribute to adverse welfare effects from acidic deposition, nutrient
enrichment, and visibility impairment. Deposition of nitrogen causes acidification, which can
cause a loss of biodiversity of fishes, zooplankton, and macro invertebrates in aquatic
ecosystems, as well as a decline in sensitive tree species, such as red spruce (Picea rubens) and
sugar maple (Acer saccharum) in terrestrial ecosystems. In the northeastern United States, the
surface waters affected by acidification are a source of food for some recreational and
subsistence fishermen and for other consumers and support several cultural services, including
aesthetic and educational services and recreational fishing. Biological effects of acidification in
terrestrial ecosystems are generally linked to aluminum toxicity, which can cause reduced root
growth, which restricts the ability of the plant to take up water and nutrients. These direct effects
can, in turn, increase the sensitivity of these plants to stresses, such as droughts, cold
temperatures, insect pests, and disease leading to increased mortality of canopy trees. Terrestrial
acidification affects several important ecological services, 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). (U.S. EPA, 2008d)
Deposition of is also associated with aquatic and terrestrial nutrient enrichment. In
estuarine waters, excess nutrient enrichment can lead to eutrophication. Eutrophication of
estuaries can disrupt an important source of food production, particularly fish and shellfish
production, and a variety of cultural ecosystem services, including water-based recreational and
aesthetic services. Terrestrial nutrient enrichment is associated with changes in the types and
number of species and biodiversity in terrestrial systems. Excessive nitrogen deposition upsets
the balance between native and nonnative plants, changing the ability of an area to support
7-15
-------
biodiversity. When the composition of species changes, then fire frequency and intensity can
also change, as nonnative grasses fuel more frequent and more intense wildfires. (U.S. EPA,
2008d)
Reducing NOx emissions and the secondary formation of PM2.5 would improve the level
of visibility throughout the United States. Fine particles with significant light-extinction
efficiencies include sulfates, nitrates, organic carbon, elemental carbon, and soil (Sisler, 1996).
These suspended particles and gases degrade visibility by scattering and absorbing light. Higher
visibility impairment levels in the East are due to generally higher concentrations of fine
particles, particularly sulfates, and higher average relative humidity levels. Visibility has direct
significance to people's enjoyment of daily activities and their overall sense of wellbeing. Good
visibility increases the quality of life where individuals live and work, and where they engage in
recreational activities.
7.3.3 Ozone Co-Benefits
In the presence of sunlight, NOx and VOCs can undergo a chemical reaction in the
atmosphere to form ozone. Reducing ambient ozone concentrations is associated with significant
human health benefits, including mortality and respiratory morbidity (U.S. EPA, 2008).
Epidemiological researchers have associated ozone exposure with adverse health effects in
numerous lexicological, clinical and epidemiological studies (U.S. EPA, 2006c). These health
effects include respiratory morbidity such as fewer asthma attacks, hospital and ER visits, school
loss days, as well as premature mortality.
7.3.4 HAP Benefits
Americans are exposed to ambient concentrations of air toxics at levels which have the
potential to cause adverse health effects.8 The levels of air toxics to which people are exposed
vary depending on where people live and work and the kinds of activities in which they engage.
In order to identify and prioritize air toxics, emission source types and locations which are of
greatest potential concern, U.S. EPA conducts the National-Scale Air Toxics Assessment
(NATA). The most recent NATA was conducted for calendar year 2002, and was released in
June 2009.9 NATA for 2002 includes four steps:
1) Compiling a national emissions inventory of air toxics emissions from outdoor sources
8 U.S. EPA. (2009) 2002 National-Scale Air Toxics Assessment, http://www.epa.gov/ttn/atw/nata2002/
9 U.S. EPA. (2009) 2002 National-Scale Air Toxics Assessment, http://www.epa.gov/ttn/atw/nata2002/
7-16
-------
2) Estimating ambient concentrations of air toxics across the United States
3) Estimating population exposures across the United States
4) Characterizing potential public health risk due to inhalation of air toxics including both
cancer and noncancer effects
Noncancer health effects can result from chronic,10 subchronic,11 or acute12 inhalation
exposures to air toxics, and include neurological, cardiovascular, liver, kidney, and respiratory
effects as well as effects on the immune and reproductive systems. According to the 2002
NAT A, nearly the entire U.S. population was exposed to an average concentration of air toxics
that has the potential for adverse noncancer respiratory health effects.13 Figures 7-4 and 7-5
depict estimated county-level carcinogenic risk and noncancer respiratory hazard from the
assessment. The respiratory hazard is dominated by a single pollutant, acrolein.
10 Chronic exposure is defined in the glossary of the Integrated Risk Information (IRIS) database
(http://www.epa.gov/iris) as repeated exposure by the oral, dermal, or inhalation route for more than
approximately 10% of the life span in humans (more than approximately 90 days to 2 years in typically used
laboratory animal species).
11 Defined in the IRIS database as exposure to a substance spanning approximately 10% of the lifetime of an
organism.
12 Defined in the IRIS database as exposure by the oral, dermal, or inhalation route for 24 hours or less.
13 The NAT A modeling framework has a number of limitations which prevent its use as the sole basis for setting
regulatory standards. These limitations and uncertainties are discussed on the 2002 NATA website. Even so, this
modeling framework is very useful in identifying air toxic pollutants and sources of greatest concern, setting
regulatory priorities, and informing the decision making process. U.S. EPA. (2009) 2002 National-Scale Air
Toxics Assessment, http://www.epa.gov/ttn/atw/nata2002/
7-17
-------
Average Risk Level
I <1 in a Million
^] 1 - 25 in a Million
^] 25 - 50 in a Million
| 50 - 75 in a Million
| 75-100 in a Million
• > 100 in a Million
Figure 7-4. Estimated County Level Carcinogenic Risk from HAP exposure from
outdoor sources (from 2002 NATA)
7-18
-------
Average Risk Level
Hazard Index
Figure 7-5. Estimated County Level Noncancer (Respiratory) Risk from HAP exposure
from outdoor sources (from 2002 NATA)
Due to data, resource, and methodology limitations, we were unable to estimate the
benefits associated with the thousands tons of hazardous air pollutants that would be reduced as a
result of this rule. Available emissions data show that several different HAPs are emitted from SI
RICE, either contained within the fuel burned or formed during the combustion process.
Although numerous HAPs may be emitted from SI RICE, a few HAPs account for over
90% of the total mass of HAPs emissions emitted. These HAPs are formaldehyde (72%),
acetaldehyde (8%), acrolein (7%), methanol (3%), and benzene (3%). Although we do not have
estimates of emission reductions for each HAP, this rule is anticipated to reduce 6,000 tons of
HAPs each year. Below we describe the health effects associated with the top 5 HAPs by mass
emitted from SI RICE.
7.3.4.1 Formaldehyde
Since 1987, EPA has classified formaldehyde as a probable human carcinogen based on
evidence in humans and in rats, mice, hamsters, and monkeys.14 EPA is currently reviewing
14U.S. EPA. 1987. Assessment of Health Risks to Garment Workers and CertainHome Residents fromExposure to
Formaldehyde, Office of Pesticides and Toxic Substances, April 1987.
7-19
-------
recently published epidemiological data. For instance, research conducted by the National
Cancer Institute (NCI) found an increased risk of nasopharyngeal cancer and
lymphohematopoietic malignancies such as leukemia among workers exposed to
formaldehyde.15'16 In an analysis of the lymphohematopoietic cancer mortality from an extended
follow-up of these workers, NCI confirmed an association between lymphohematopoietic cancer
risk and peak exposures.17 A recent National Institute of Occupational Safety and Health
(NIOSH) study of garment workers also found increased risk of death due to leukemia among
workers exposed to formaldehyde.18 Extended follow-up of a cohort of British chemical workers
did not find evidence of an increase in nasopharyngeal or lymphohematopoietic cancers, but a
continuing statistically significant excess in lung cancers was reported.19
In the past 15 years there has been substantial research on the inhalation dosimetry for
formaldehyde in rodents and primates by the CUT Centers for Health Research (formerly the
Chemical Industry Institute of Toxicology), with a focus on use of rodent data for refinement of
the quantitative cancer dose-response assessment.20'21'22 CIIT's risk assessment of formaldehyde
incorporated mechanistic and dosimetric information on formaldehyde. However, it should be
noted that recent research published by EPA indicates that when two-stage modeling
assumptions are varied, resulting dose-response estimates can vary by several orders of
magnitude.23'24'25'26 These findings are not supportive of interpreting the CUT model results as
15Hauptmann, M; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2003. Mortality from lymphohematopoetic
malignancies among workers in formaldehyde industries. Journal of the National Cancer Institute 95: 1615-1623.
16Hauptmann, M.; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Blair, A. 2004. Mortality from solid cancers among
workers in formaldehyde industries. American Journal of Epidemiology 159: 1117-1130.
17 Beane Freeman, L. E.; Blair, A.; Lubin, J. H.; Stewart, P. A.; Hayes, R. B.; Hoover, R. N.; Hauptmann, M. 2009.
Mortality from lymphohematopoietic malignancies among workers in formaldehyde industries: The National
Cancer Institute cohort. J. National Cancer Inst. 101: 751-761.
18 Pinkerton, L. E. 2004. Mortality among a cohort of garment workers exposed to formaldehyde: an update.
Occup. Environ. Med. 61: 193-200.
19 Coggon, D, EC Harris, J Poole, KT Palmer. 2003. Extended follow-up of a cohort of British chemical workers
exposed to formaldehyde. J National Cancer Inst. 95:1608-1615.
20 Conolly, RB, JS Kimbell, D Janszen, PM Schlosser, D Kalisak, J Preston, and FJ Miller. 2003. Biologically
motivated computational modeling of formaldehyde carcinogenicity in the F344 rat. Tox Sci 75: 432-447.
21 Conolly, RB, JS Kimbell, D Janszen, PM Schlosser, D Kalisak, J Preston, and FJ Miller. 2004. Human respiratory
tract cancer risks of inhaled formaldehyde: Dose-response predictions derived from biologically-motivated
computational modeling of a combined rodent and human dataset. Tox Sci 82: 279-296.
22 Chemical Industry Institute of Toxicology (CUT). 1999. Formaldehyde: Hazard characterization and dose-response
assessment for carcinogenicity by the route of inhalation. CUT, September 28, 1999. Research Triangle Park, NC.
23 U.S. EPA. Analysis of the Sensitivity and Uncertainty in 2-Stage Clonal Growth Models for Formaldehyde with
Relevance to Other Biologically-Based Dose Response (BBDR) Models. U.S. Environmental Protection Agency,
Washington, D.C., EPA/600/R-08/103, 2008
24 Subramaniam, R; Chen, C; Crump, K; .et .al. (2008) Uncertainties in biologically-based modeling of
formaldehyde-induced cancer risk: identification of key issues. Risk Anal 28(4):907-923.
25 Subramaniam, R; Chen, C; Crump, K; .et .al. (2007). Uncertainties in the CUT 2-stage model for formaldehyde-
induced nasal cancer in the F344 rat: a limited sensitivity analysis-I. Risk Anal 27:1237
7-20
-------
providing a conservative (health protective) estimate of human risk.27 EPA research also
examined the contribution of the two-stage modeling for formaldehyde towards characterizing
the relative weights of key events in the mode-of-action of a carcinogen. For example, the
model-based inference in the published CUT study that formaldehyde's direct mutagenic action
is not relevant to the compound's tumorigenicity was found not to hold under variations of
modeling assumptions.28
Based on the developments of the last decade, in 2004, the working group of the IARC
concluded that formaldehyde is carcinogenic to humans (Group 1), on the basis of sufficient
evidence in humans and sufficient evidence in experimental animals - a higher classification than
previous IARC evaluations. After reviewing the currently available epidemiological evidence,
the IARC (2006) characterized the human evidence for formaldehyde carcinogenicity as
"sufficient," based upon the data on nasopharyngeal cancers; the epidemiologic evidence on
leukemia was characterized as "strong."29EPA is reviewing the recent work cited above from the
NCI and NIOSH, as well as the analysis by the CUT Centers for Health Research and other
studies, as part of a reassessment of the human hazard and dose-response associated with
formaldehyde.
Formaldehyde exposure also causes a range of noncancer health effects, including
irritation of the eyes (burning and watering of the eyes), nose and throat. Effects from repeated
exposure in humans include respiratory tract irritation, chronic bronchitis and nasal epithelial
lesions such as metaplasia and loss of cilia. Animal studies suggest that formaldehyde may also
cause airway inflammation - including eosinophil infiltration into the airways. There are several
studies that suggest that formaldehyde may increase the risk of asthma - particularly in the
young.30'31
26 Cramp, K; Chen, C; Fox, J; .et .al. (2008) Sensitivity analysis of biologically motivated model for formaldehyde-
induced respiratory cancer in humans. Ann Occup Hyg 52:481-495.
27 Cramp, K; Chen, C; Fox, J; .et .al. (2008) Sensitivity analysis of biologically motivated model for formaldehyde-
induced respiratory cancer in humans. Ann Occup Hyg 52:481-495.
28 Subramaniam, R; Chen, C; Cramp, K; .et .al. (2007). Uncertainties in the CUT 2-stage model for formaldehyde-
induced nasal cancer in the F344 rat: a limited sensitivity analysis-I. Risk Anal 27:1237
29 International Agency for Research on Cancer (2006) Formaldehyde, 2-Butoxyethanol and l-tert-Butoxypropan-2-
ol. Monographs Volume 88. World Health Organization, Lyon, France.
30 Agency for Toxic Substances and Disease Registry (ATSDR). 1999. Toxicological profile for Formaldehyde.
Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service.
http://www.atsdr.cdc.gov/toxprofiles/tplll.html
31 WHO (2002) Concise International Chemical Assessment Document 40: Formaldehyde. Published under the joint
sponsorship of the United Nations Environment Programme, the International Labour Organization, and the
World Health Organization, and produced within the framework of the Inter-Organization Programme for the
Sound Management of Chemicals. Geneva.
7-21
-------
7.3.4.2 Acetaldehyde
Acetaldehyde is classified in EPA's IRIS database as a probable human carcinogen,
based on nasal tumors in rats, and is considered toxic by the inhalation, oral, and intravenous
routes.32 Acetaldehyde is reasonably anticipated to be a human carcinogen by the U.S. DHHS in
the 11th Report on Carcinogens and is classified as possibly carcinogenic to humans (Group 2B)
by the IARC.33'34 EPA is currently conducting a reassessment of cancer risk from inhalation
exposure to acetaldehyde.
The primary noncancer effects of exposure to acetaldehyde vapors include irritation of
the eyes, skin, and respiratory tract.35 In short-term (4 week) rat studies, degeneration of olfactory
epithelium was observed at various concentration levels of acetaldehyde exposure.36'37 Data from
these studies were used by EPA to develop an inhalation reference concentration. Some
asthmatics have been shown to be a sensitive subpopulation to decrements in functional
expiratory volume (FEV1 test) and bronchoconstriction upon acetaldehyde inhalation.38 The
agency is currently conducting a reassessment of the health hazards from inhalation exposure to
acetaldehyde.
7.3.4.3 Acrolein
EPA determined in 2003 that the human carcinogenic potential of acrolein could not be
determined because the available data were inadequate. No information was available on the
carcinogenic effects of acrolein in humans and the animal data provided inadequate evidence of
32U.S. Environmental Protection Agency (U.S. EPA). 1991. Integrated Risk Information System File of
Acetaldehyde. Research and Development, National Center for Environmental Assessment, Washington, DC.
This material is available electronically at http://www.epa.gov/iris/subst/0290.htm.
33 U.S. Department of Health and Human Services National Toxicology Program 11th Report on Carcinogens
available at: http://ntp.niehs.nih.gov/go/16183.
34 International Agency for Research on Cancer (IARC). 1999. Re-evaluation of some organic chemicals, hydrazine,
and hydrogen peroxide. IARC Monographs on the Evaluation of Carcinogenic Risk of Chemical to Humans, Vol
71. Lyon, France.
35 U.S. Environmental Protection Agency (U.S. EPA). 1991. Integrated Risk Information System File of
Acetaldehyde. Research and Development, National Center for Environmental Assessment, Washington, DC.
This material is available electronically at http://www.epa.gov/iris/subst/0290.htm.
36 U.S. Environmental Protection Agency (U.S. EPA). 1991. Integrated Risk Information System File of
Acetaldehyde. Research and Development, National Center for Environmental Assessment, Washington, DC.
This material is available electronically at http://www.epa.gov/iris/subst/0290.htm.
37 Appleman, L.M., R.A. Woutersen, and V.J. Feron. (1982). Inhalation toxicity of acetaldehyde in rats. I. Acute and
subacute studies. Toxicology. 23: 293-297.
38 Myou, S.; Fujimura, M; Nishi K.; Ohka, T.; and Matsuda, T. (1993) Aerosolized acetaldehyde induces
histamine-mediated bronchoconstriction in asthmatics. Am. Rev. Respir.Dis. 148(4 Pt 1): 940-943.
7-22
-------
carcinogen!city.39 The IARC determined in 1995 that acrolein was not classifiable as to its
carcinogenicity in humans.40
Acrolein is extremely acrid and irritating to humans when inhaled, with acute exposure
resulting in upper respiratory tract irritation, mucus hypersecretion and congestion. The intense
irritancy of this carbonyl has been demonstrated during controlled tests in human subjects, who
suffer intolerable eye and nasal mucosal sensory reactions within minutes of exposure.41 These
data and additional studies regarding acute effects of human exposure to acrolein are
summarized in EPA's 2003 IRIS Human Health Assessment for acrolein.42 Evidence available
from studies in humans indicate that levels as low as 0.09 ppm (0.21 mg/m3) for five minutes
may elicit subjective complaints of eye irritation with increasing concentrations leading to more
extensive eye, nose and respiratory symptoms.43 Lesions to the lungs and upper respiratory tract
of rats, rabbits, and hamsters have been observed after subchronic exposure to acrolein.44 Acute
exposure effects in animal studies report bronchial hyper-responsiveness.45 In a recent study, the
acute respiratory irritant effects of exposure to 1.1 ppm acrolein were more pronounced in mice
with allergic airway disease by comparison to non-diseased mice which also showed decreases in
respiratory rate.46 Based on these animal data and demonstration of similar effects in humans
(i.e., reduction in respiratory rate), individuals with compromised respiratory function (e.g.,
emphysema, asthma) are expected to be at increased risk of developing adverse responses to
strong respiratory irritants such as acrolein.
39 U.S. Environmental Protection Agency (U.S. EPA). 2003. Integrated Risk Information System File of Acrolein.
Research and Development, National Center for Environmental Assessment, Washington, DC. This material is
available at http://www.epa.gov/iris/toxreviews/0364tr.pdf.
40 International Agency for Research on Cancer (IARC). 1995. Monographs on the evaluation of carcinogenic risk
of chemicals to humans, Volume 63, Dry cleaning, some chlorinated solvents and other industrial chemicals,
World Health Organization, Lyon, France.
41 U.S. Environmental Protection Agency (U.S. EPA). 2003. Integrated Risk Information System File of Acrolein.
EPA/635/R-03/003. p. 10. Research and Development, National Center for Environmental Assessment,
Washington, DC. This material is available at http://www.epa.gov/iris/toxreviews/0364tr.pdf.
42 U.S. Environmental Protection Agency (U.S. EPA). 2003. Integrated Risk Information System File of Acrolein.
2003. Research and Development, National Center for Environmental Assessment, Washington, DC.
EPA/635/R-03/003. This material is available at http://www.epa.gov/iris/toxreviews/0364tr.pdf.
43 U.S. Environmental Protection Agency (U.S. EPA). 2003. Integrated Risk Information System File of Acrolein.
Research and Development, National Center for Environmental Assessment, Washington, DC. EPA/635/R-
03/003. p. 11. This material is available at http://www.epa.gov/iris/toxreviews/0364tr.pdf.
44 U.S. Environmental Protection Agency (U.S. EPA). 2003. Integrated Risk Information System File of Acrolein.
Research and Development, National Center for Environmental Assessment, Washington, DC. EPA/635/R-03/003.
This material is available at http://www.epa.gov/iris/toxreviews/0364tr.pdf.
45 U.S. Environmental Protection Agency (U.S. EPA). 2003. Integrated Risk Information System File of Acrolein.
Research and Development, National Center for Environmental Assessment, Washington, DC. EPA/635/R-03/003.
This material is available at http://www.epa.gov/iris/toxreviews/0364tr.pdf.
46 Morris JB, Symanowicz PT, Olsen JE, et al. 2003. Immediate sensory nerve-mediated respiratory responses to
irritants in healthy and allergic airway-diseased mice. J Appl Physiol 94(4): 1563-1571.
7-23
-------
7.3.4.4Methanol
Exposure of humans to methanol by inhalation or ingestion may result in central nervous
system depression and degenerative changes in the brain and visual systems. After inhaled or
ingested, methanol is converted to formate, a highly toxic metabolite that within the course of a
few hours can cause narcosis, metabolic acidosis, headaches, severe abdominal and leg pain and
visual degeneration that can lead to blindness.47
Methanol has been demonstrated to cause developmental toxicity in rats and mice, and
reproductive and developmental toxicity in monkeys. A number of studies have reported adverse
effects in the offspring of rats and mice exposed to methanol by inhalation including reduced
weight of brain pituitary gland, thymus, thyroid, reduced overall fetal body weight and increased
incidence of extra ribs and cleft palate.48'49'50 Methanol inhalation studies using rhesus monkeys
have reported a decrease in the length of pregnancy, and limited evidence of impaired learning
ability in offspring.51'52'53'54 EPA has not classified methanol with respect to its carcinogenicity.
7.3.4.5 Benzene
The EPA's IRIS database lists benzene as a known human carcinogen (causing leukemia)
by all routes of exposure, and concludes that exposure is associated with additional health
effects, including genetic changes in both humans and animals and increased proliferation of
bone marrow cells in mice.55'56'57 EPA states in its IRIS database that data indicate a causal
47 Rowe, VK and McCollister, SB. 1981. Alcohols. In: Patty's Industrial Hygiene and Toxicology, 3rd ed. Vol. 2C,
GD Clayton, FE Clayton, Eds. John Wiley & Sons, New York, pp. 4528-4541.
48 New Energy Development Organization (NEDO). 1987. Toxicological research of methanol as a fuel for power
station: summary report on tests with monkeys, rats and mice. Tokyo, Japan.
49 Nelson, BK; Brightwell, WS; MacKenzie, DR; Khan, A; Burg, JR; Weigel, WW; Goad, PT. 1985. Teratological
assessment of methanol and ethanol at high inhalation levels in rats. Toxicol Sci, 5: 727-736.
50 Rogers, JM; Barbee, BD; Rehnberg, BF. 1993. Critical periods of sensitivity for the developmental toxicity of
inhaled methanol. Teratology, 47: 395.
51 Burbacher, T; Grant, K; Shen, D; Damian, D; Ellis, S; Liberate, N. 1999. Reproductive and offspring
developmental effects following maternal inhalation exposure to methanol in nonhuman primates Part II:
developmental effects in infants exposed prenatally to methanol. Health Effects Institute. Cambridge, MA.
52 Burbacher, T; Shen, D; Grant, K; Sheppard, L; Damian, D; Ellis, S; Liberate, N. 1999. Reproductive and
offspring developmental effects following maternal inhalation exposure to methanol in nonhuman primates Part I:
methanol disposition and reproductive toxicity in adult females. Health Effects Institute. Cambridge, MA.
53 Burbacher, TM; Grant, KS; Shen, DD; Sheppard, L; Damian, D; Ellis, S; Liberate, N. 2004. Chronic maternal
methanol inhalation in nonhuman primates (Macaca fascicularis): reproductive performance and birth outcome.
Neurotoxicol Teratol, 26: 639-650.
54 Burbacher, TM; Shen, DD; Lalovic, B; Grant, KS; Sheppard, L; Damian, D; Ellis, S; Liberate, N. 2004. Chronic
maternal methanol inhalation in nonhuman primates (Macaca fascicularis): exposure and toxicokinetics prior to
and during pregnancy. Neurotoxicol Teratol, 26: 201-221.
55 U.S. Environmental Protection Agency (U.S. EPA). 2000. Integrated Risk Information System File for Benzene.
Research and Development, National Center for Environmental Assessment, Washington, DC. This material is
7-24
-------
relationship between benzene exposure and acute lymphocytic leukemia and suggest a
relationship between benzene exposure and chronic non-lymphocytic leukemia and chronic
lymphocytic leukemia. The International Agency for Research on Carcinogens (IARC) has
determined that benzene is a human carcinogen and the U.S. Department of Health and Human
Services (DHHS) has characterized benzene as a known human carcinogen.58'59
A number of adverse noncancer health effects including blood disorders, such as preleukemia
and aplastic anemia, have also been associated with long-term exposure to benzene.60'61 The
most sensitive noncancer effect observed in humans, based on current data, is the depression of
the absolute lymphocyte count in blood.62'63 In addition, recent work, including studies sponsored
by the Health Effects Institute (HEI), provides evidence that biochemical responses are occurring
at lower levels of benzene exposure than previously known.64'65'66'67 EPA's IRIS program has not
yet evaluated these new data.
available electronically at: http://www.epa.gov/iris/subst/0276.htm.
56 International Agency for Research on Cancer, IARC monographs on the evaluation of carcinogenic risk of
chemicals to humans, Volume 29, Some industrial chemicals and dyestuffs, International Agency for Research on
Cancer, World Health Organization, Lyon, France, p. 345-389, 1982.
57 Irons, R.D.; Stillman, W.S.; Colagiovanni, D.B.; Henry, V.A. (1992) Synergistic action of the benzene metabolite
hydroquinone on myelopoietic stimulating activity of granulocyte/macrophage colony-stimulating factor in vitro,
Proc. Natl. Acad. Sci. 89:3691-3695.
58 International Agency for Research on Cancer (IARC). 1987. Monographs on the evaluation of carcinogenic risk
of chemicals to humans, Volume 29, Supplement 7, Some industrial chemicals and dyestuffs, World Health
Organization, Lyon, France.
59 U.S. Department of Health and Human Services National Toxicology Program 11th Report on Carcinogens
available at: http://ntp.niehs.nih.gov/go/16183.
60Aksoy, M. (1989). Hematotoxicity and carcinogenicity of benzene. Environ. Health Perspect. 82:193-197.
61 Goldstein, B.D. (1988). Benzene toxicity. Occupational medicine. State of the Art Reviews. 3:541-554.
62Rothman, N., G.L. Li, M. Dosemeci, W.E. Bechtold, G.E. Marti, Y.Z. Wang, M. Linet, L.Q. Xi, W. Lu, M.T.
Smith, N. Titenko-Holland, L.P. Zhang, W. Blot, S.N. Yin, and R.B. Hayes (1996) Hematotoxicity among
Chinese workers heavily exposed to benzene. Am. J. Ind. Med. 29: 236-246.
63 U.S. Environmental Protection Agency (U.S. EPA). 2000. Integrated Risk Information System File for Benzene
(Noncancer Effects). Research and Development, National Center for Environmental Assessment, Washington,
DC. This material is available electronically at: http://www.epa.gov/iris/subst/0276.htm.
64Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen, B.; Melikian, A.; Eastmond, D.; Rappaport, S.; Li, H.; Rupa,
D.; Suramaya, R.; Songnian, W.; Huifant, Y.; Meng, M.; Winnik, M.; Kwok, E.; Li, Y.; Mu, R.; Xu, B.;
Zhang, X.; Li, K. (2003). HEI Report 115, Validation & Evaluation of Biomarkers in Workers Exposed to
Benzene in China.
65 Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, B. Cohen, et al. (2002). Hematological changes among Chinese
workers with a broad range of benzene exposures. Am. J. Industr. Med. 42: 275-285.
66Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. (2004). Hematotoxically in Workers Exposed to Low Levels
of Benzene. Science 306: 1774-1776.
67 Turtletaub, K.W. and Mani, C. (2003). Benzene metabolism in rodents at doses relevant to human exposure from
Urban Air. Research Reports Health Effect Inst. Report No. 113.
7-25
-------
7.3.4.6 Other Air Toxics
In addition to the compounds described above, other compounds from SI RICE would be
affected by this rule. Information regarding the health effects of these compounds can be found
in EPA's IRIS database.68
7.4 Characterization of Uncertainty in the Monetized PM2.s Co-Benefits
In any complex analysis, there are likely to be many sources of uncertainty. Many inputs
are used to derive the final estimate of economic co-benefits, including emission inventories, air
quality models (with their associated parameters and inputs), epidemiological estimates of
concentration-response (C-R) functions, estimates of values, population estimates, income
estimates, and estimates of the future state of the world (i.e., regulations, technology, and human
behavior). For some parameters or inputs it may be possible to provide a statistical representation
of the underlying uncertainty distribution. For other parameters or inputs, the necessary
information is not available. Because we used the benefit-per-ton approach for this analysis,
confidence intervals are unavailable.
The annual benefit estimates presented in this analysis are also inherently variable due to
the processes that govern pollutant emissions and ambient air quality in a given year. Factors
such as hours of equipment use and weather are constantly variable, regardless of our ability to
measure them accurately. As discussed in the PM2.5 NAAQS RIA (Table 5.5) (U.S. EPA,
2006a), there are a variety of uncertainties associated with these PM co-benefits. Therefore, the
estimates of annual co-benefits should be viewed as representative of the magnitude of co-
benefits expected, rather than the actual co-benefits that would occur every year.
It is important to note that the monetized benefit-per-ton estimates used here reflect
specific geographic patterns of emissions reductions and specific air quality and co-benefits
modeling assumptions. For example, these estimates do not reflect local variability in population
density, meteorology, exposure, baseline health incidence rates, or other local factors. Use of
these $/ton values to estimate co-benefits associated with different emission control programs
(e.g., for reducing emissions from large stationary sources like EGUs) may lead to higher or
lower benefit estimates than if co-benefits were calculated based on direct air quality modeling.
Great care should be taken in applying these estimates to emission reductions occurring in any
specific location, as these are all based on national or broad regional emission reduction
programs and therefore represent average co-benefits-per-ton over the entire United States. The
68 U.S. EPA Integrated Risk Information System (IRIS) database is available at: www.epa.gov/iris
7-26
-------
co-benefits- per-ton for emission reductions in specific locations may be very different than the
estimates presented here.
Understanding the transport of pollutants is a critical component for estimating exposure
and the associated human health benefits from reducing air pollution emissions. The underlying
emissions modeling and air quality modeling supporting the PM2.5 co-benefits analysis accounts
for the current distribution of emissions sources, including both urban and rural sources. In
addition, the air quality modeling included 14 vertical layers to simulate the differences between
ground-level emissions and higher stack emissions (U.S. EPA, 2006b). The distance that HAPs
travel away from the emission source depends on several factors. HAPs such as formaldehyde,
acetaldehyde, acrolein, methanol, and benzene are emitted as gases. Regional photochemical
model simulations, examining particular scenarios, have shown that gaseous HAPs like
formaldehyde and acetaldehyde can be transported hundreds of kilometers from their emissions
source in distinct plumes (U.S. EPA, 2010f). Further, these emissions can contribute to regional
airmasses with elevated concentrations of gaseous HAPs. These polluted airmasses can be
transported thousands of kilometers and affect locations well distant from the original emissions
source. For the SI RICE examined in this rule, EPA does not have enough information to
determine the extent of transport specific to the HAPs reduced.
PM2.5 mortality benefits are the largest benefit category that we monetized in this
analysis. To better characterize the uncertainty associated with mortality impacts that are
estimated to occur in areas with low baseline levels of PM2.5, we included the LML assessment.
Without policy-specific air quality modeling, we are unable to quantify the shift in exposure
associated with this specific rule. For this rule, as a surrogate measure of mortality impacts, we
provide the percentage of the population exposed at each PM2.5 level using the most recent
modeling available from the recently proposed Transport Rule (U.S. EPA, 2010e). A very large
proportion of the population is exposed at or above the lowest LML of the cohort studies
(Figures 7-6 and 7-7), increasing our confidence in the PM mortality analysis. Figure 7-6 shows
a bar chart of the percentage of the population exposed to various air quality levels in the pre-
and post-policy policy. Figure 7-7 shows a cumulative distribution function of the same data.
Both figures identify the LML for each of the major cohort studies. As the policy shifts the
distribution of air quality levels, fewer people are exposed to PM2.5 levels at or above the LML.
Using the Pope et al. (2002) study, the 85% of the population is exposed to annual mean PM2.5
levels at or above the LML of 7.5 |ig/m3. Using the Laden et al. (2006) study, 40% of the
population is exposed above the LML of 10 |ig/m3. As we model mortality impacts among
populations exposed to levels of PM2.5 that are successively lower than the LML of the lowest
7-27
-------
cohort study, our confidence in the results diminishes. However, the analysis above confirms that
the great majority of the impacts occur at or above the lowest cohort study's LML. It is important
to emphasize that we have high confidence in PM2.s-related effects down to the lowest LML of
the major cohort studies. Just because we have greater confidence in the benefits above the LML,
this does not mean that we have no confidence that benefits occur below the LML.
7-28
-------
25%
20%
15%
10%
5%
0%
Pope et al. 2002
Laden et al. 2006
The control strategy lowers PM2 5
levels substantially, particularly
among highly exposed
populations. In the baseline, 85%
of the population lived in areas
where PM2 5 levels above the
lowest measured levels of the
Pope study, increasing our
confidence in the estimated
mortality reductions for this rule.
1 2 3 4 5 5.8 6 7 7.5 8 9 10 11 12 13 14 15 16 17 18 19 20
• Post-control • Baseline
Figure 7-6. Percentage of Adult Population by Annual Mean
Exposure (pre- and post-policy policy)
100%
Pope etal, 2002 Laden etal. 2006
70%
50%
The control strategy lowers PM2S levels
substantial \j, parti cularr/ among high I/
exposed populations, In the baseline, 89% of
the population lived in areas where PM25
levels above the lowest measured levels of
the Pope stud/, increasing our confidence in
the estimated mortality reductions for this
rule,
1 2 34 5 5.8 6 7 7.5 S 9 10 11 12 13 14 15 16 17 13 19 20
Post-control Baseline
Figure 7-7. Cumulative Distribution of Adult Population at Annual Mean PM2.slevels
(pre- and post-policy policy)
7-29
-------
Above we present the estimates of the total monetized co-benefits, based on our
interpretation of the best available scientific literature and methods and supported by the SAB-
HES and the NAS (NRC, 2002). The co-benefits estimates are subject to a number of
assumptions and uncertainties. For example, for key assumptions underlying the estimates for
premature mortality, which typically account for at least 90% of the total monetized co-benefits,
we were able to quantify include the following:
1. PM2.5 co-benefits were derived through benefit per-ton estimates, which do not reflect
local variability in population density, meteorology, exposure, baseline health
incidence rates, or other local factors that might lead to an over-estimate or under-
estimate of the actual co-benefits of controlling directly emitted fine particulates.
2. 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 via transported precursors emitted from EGUs may differ
significantly from direct PM2.5 released from diesel engines and other industrial
sources, but no clear scientific grounds exist for supporting differential effects
estimates by particle type.
3. We assume that the health impact function for fine particles is linear down to the
lowest air quality levels modeled in this analysis. Thus, the estimates include health
co-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.
4. To characterize the uncertainty in the relationship between PM2.5 and premature
mortality (which typically accounts for 85% to 95% of total monetized co-benefits),
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. For more
information on the uncertainties associated with PM2.5 co-benefits, please consult the
PM2.5NAAQS RIA (Table 5-5).
This RIA does not include the type of detailed uncertainty assessment found in the PM
NAAQS RIA because we lack the necessary air quality input and monitoring data to run the co-
benefits model. Moreover, it was not possible to develop benefit-per-ton metrics and associated
estimates of uncertainty using the co-benefits estimates from the PM RIA because of the
significant differences between the sources affected in that rule and those regulated here.
However, the results of the Monte Carlo analyses of the health and welfare co-benefits presented
7-30
-------
in Chapter 5 of the PM NAAQS RIA can provide some evidence of the uncertainty surrounding
the co-benefits results presented in this analysis.
7.5 Comparison of Co-Benefits and Costs
Using a 3% discount rate, we estimate the total combined monetized co-benefits of the
final SI RICE NESHAP to be $510 million to $1.2 billion in the implementation year (2013).
Using a 7% discount rate, we estimate the total monetized co-benefits of the final rule to be $460
million to $1.1 billion. The annualized social costs of the final NESHAP are $253 million at a
7% interest rate.69 Thus, the net benefits are $250 million to $980 million at a 3% discount rate
and $210 million to $860 million at a 7% discount rate. All estimates are in 2009$ for the year
2013.
Table 7-5 shows a summary of the monetized co-benefits, social costs, and net benefits
for the final SI RICE NESHAP, respectively. Figures 7-8 and 7-9 show the full range of net
benefits estimates (i.e., annual co-benefits minus annualized costs) utilizing the 14 different
PM2.5 mortality functions at discount rates of 3% and 7%. In addition, the benefits from reducing
109,000 tons of carbon monoxide and 6,000 tons of HAPs each year from SI RICE have not
been included in these estimates. EPA believes that the co-benefits are likely to exceed the costs
under this rulemaking even when taking into account uncertainties in the cost and benefit
estimates. As mentioned earlier in this RIA, the final NESHAP is the MACT floor level of
control for all SI RICE major sources and the GACT level of control for all SI RICE area
sources. We show results in Table 7-5 for an alternative (referred to as "Alternative 2") which is
a more stringent alternative than the final NESHAP for major sources. For this alternative, the
MACT floor level of control is applied to all SI RICE major sources except for four-stroke rich-
burn (4SRB) engines of 300-500 horsepower (HP), where the level of control is above the
MACT floor, and the GACT level of control is applied to all area SI RICE sources.
69 For more information on the annualized social costs, please refer to Section 5 of this RIA.
7-31
-------
Table 7-5. Summary of the Monetized Co-Benefits, Social Costs, and Net benefits for the
Final SI RICE NESHAP in 2013 (millions of 2009S)1
3% Discount Rate 7% Discount Rate
Final NESHAP: Major4
Total Monetized Benefits7 $8.2 to $20 $7.4 to $18
Total Social Costs3 $88 $88
Net Benefits -$80 to -$68 -$81 to -$70
12,500 tons of carbon monoxide
XT .. j „ c. 2.000 tons of hazardous air pollutants (HAPs)
Non-monetized Benefits r . „• . r
Ecosystem effects
Visibility impairment
Alternative 2: Major
Total Monetized Benefits2
Total Social Costs3
Net Benefits
$48
-$47
to $120
$95
to $22
$43
-$52
to $110
$95
to $11
17,800 tons of carbon monoxide
1,400 tons of hazardous air pollutants (HAPs)
Non-monetized Benefits Health effects from NO2 and ozone exposure
Ecosystem effects
Visibility impairment
Final NESHAP: Area5
Total Monetized Benefits2
Total Social Costs3
Net Benefits
$500
$330
to $1,200
$166
to $1,100
$450
$290
to $1,100
$166
to $930
97,000 tons of carbon monoxide
4,700 tons of hazardous air pollutants (HAPs)
Non-monetized Benefits Health effects from NO2 and ozone exposure
Ecosystem effects
Visibility impairment
Final Major and Area Source NESHAP
Total Monetized Benefits2 $510 to $1,200 $460 to $1,100
Total Social Costs3 $253 $253
Net Benefits $250 to $980 $210 to $860
109,000 tons of carbon monoxide
6,000 tons of hazardous air pollutants (HAPs)
Non-monetized Benefits Health effects from NO2 and ozone exposure
Ecosystem effects
Visibility impairment
1 All estimates are for the implementation year (2013), and are rounded to two significant figures.
2 The total monetized co-benefits reflect the human health co-benefits associated with reducing exposure to PM2 5
through reductions of PM25 precursors such as NOx and VOC. It is important to note that the monetized co-
benefits include many but not all health effects associated with PM2 5 exposure. It is important to note that the
monetized benefits include many but not all health effects associated with PM2 5 exposure. Benefits are shown as a
range from Pope et al. (2002) to Laden et al. (2006). These models assume that all fine particles, regardless of
their chemical composition, are equally potent in causing premature mortality because there is no clear scientific
evidence that would support the development of differential effects estimates by particle type.
3
The annual compliance costs serve as a proxy for the annual social costs of this rule given the lack of difference
between the two.
7-32
-------
4The final NESHAP is the MACT floor level of control for all major SI RICE non-emergency sources. We also
show results for Alternative 2, an more stringent alternative than the final NESHAP for major sources. In this
alternative, the MACT level of control is applied to all SI RICE major non-emergency sources except for four-
stroke rich-burn (4SRB) engines of 300-500 horsepower (HP), where the level of control is above the MACT
floor, and the GACT level of control is applied to all area SI RICE sources.
5 All of the benefits for area sources are attributable to reductions expected from 4SLB and 4SRB non-emergency
engines above 500 HP.
7-33
-------
$1,400
$1,200
Laden et al.
$1,000
_ $800
•CO-
§ $600
c
g
'^ $400
Pope et al.
$200
-$200
Cost estimate combined with total monetized benefits estimates derived from 2
epidemiology functions and 12 expert functions
Figure 7-8. Net Benefits for the Final SI RICE NESHAP at 3% Discount Rate a
a Net Benefits are quantified in terms of PM2 5 co-benefits for implementation year (2013). This graph shows 14
benefits estimates combined with the cost estimate. All combinations are treated as independent and equally
probable. All fine particles are assumed to have equivalent health effects, but the benefit per ton estimates vary
because each ton of precursor reduced has a different propensity to become PM25. The monetized co-benefits
incorporate the conversion from precursor emissions to ambient fine particles.
7-34
-------
$0
-$200
Cost estimate combined with total monetized benefits estimates derived from 2
epidemiology functions and 12 expert functions
Figure 7-9. Net Benefits for the Final SI RICE NESHAP at 7% Discount Rate a
a Net benefits are quantified in terms of PM2 5 co-benefits for implementation year (2013). This graph shows 14 co-
benefits estimates combined with the cost estimate. All combinations are treated as independent and equally
probable. All fine particles are assumed to have equivalent health effects, but the benefit per ton estimates vary
because each ton of precursor reduced has a different propensity to become PM25. The monetized co-benefits
incorporate the conversion from precursor emissions to ambient fine particles.
7-35
-------
SECTION 8
REFERENCES
Anadarko Petroleum Corporation Comments on the Proposed Revisions to the National
Emission Standard for Hazardous Air Pollutants for Reciprocating Internal Combustion
Engines, June 2, 2009, Case Study (EPA-HQ-OAR-2008-0708-0186).
Edison Electric Institute. "Income Statement: Q4 2008 Financial Update. Quarterly Report of the
U.S. Shareholder-Owned Electric Utility Industry." .
Email from Nick Huff, Miratech to Jennifer Synder, AGTI, SCR Questions for RICE MACT,
October 23, 2003 (EPA-HQ-OAR-2005-0029-0038).
Email from Mike Leonard, Miratech to Brad Nelson, AGTI, Information Request, July 21, 2005
(EPA-HQ-OAR-2005-0030-0087).
Email from Bruce Chrisman, Cameron's Compression Systems to Tanya Parise, EC/R, Subject:
Existing RICE NESHAP - Information for EPA for 2SLB Engines, October 16, 2009.
Email from James Harrison, Exterran to Melanie King, EPA, Subject: 2SLB - Cameron
oxidation catalyst pricing, October 20, 2009.
Email from Mike Leonard, Miratech Corporation to Brenda Riddle, AGTI, RE: Clarification of
SCR Cost Information for EPA, July 20, 2005. (EPA-HQ-OAR-2005-0030-0086).
Email from Antonio Santos, MECA to Tanya Parise, EC/R, Subject: EPA Proposed Existing
RICE NESHAP - Cost of Aftertreatment, October 2, 2009. Response #2.
Fann, N., C.M. Fulcher, B.J. Hubbell. 2009. The influence of location, source, and emission type
in estimates of the human health benefits of reducing a ton of air pollution. Air Qual
Atmos Health 2:169-176.
Federal Reserve Board. "Industrial Production and Capacity Utilization: Industrial Production"
Series ID: G17/IP_MINING_AND_UTILITY_DETAIL/IP.G2211.S
. (January 27, 2010).
Four Corners Air Quality Task Force Report of Mitigation Options, November 1, 2007.
Mitigation Option: Use of NSCR/3-Way Catalysts and Air/Fuel Ratio Controllers on
Rich Burn Stoichiometric Engines (EPA-HQ-OAR-2008-0708-0009).
Fullerton, D., and G. Metcalf 2002. "Tax Incidence." In A. Auerbach and M. Feldstein, eds.,
Handbook of Public Economics., Vol.4, Amsterdam: Elsevier.
Internal Revenue Service, U.S. Department of Treasury. 2010. "Corporation Source Book: Data
Files 2004-2007." ; (May 2,
2010).
8-1
-------
Interstate Oil & Gas Compact Commission. 2007. "Marginal Wells: Fuel for Economic Growth."
Table 3.B. .
Interstate Oil & Gas Compact Commission (IOGCC). 2008. "Marginal Wells: Fuel for Economic
Growth." Available at < http://iogcc.publishpath.com/Websites/iogcc/pdfs/2008-
Marginal-Well-Report.pdf>.
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.
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.
Lincoln Electric Holdings. 2006. Form 10-K. Filed February 24, 2006.
Memorandum from Tom McGrath, IES to Brad Nelson, EC/R, Request for Additional Cost
Detail for Gas-Fired Engines Emission Controls, April 19, 2010.
Mrozek, J.R., and L.O. Taylor. 2002. "What Determines the Value of Life? A Meta-Analysis."
Journal of Policy Analysis and Management 21(2):253-270.
National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed Air
Pollution Regulations. Washington, DC: The National Academies Press.
Nelson, B., EC/R Inc. June 29, 2010. Memorandum to Melanie King, U.S. Environmental
Protection Agency.Control Costs for Existing Stationary SI RICE.
Nicholson, Walter. \99%.Microeconomic Theory. Orlando: The Dry den Press.
Office of Management and Budget (OMB), 2003. Circular A-4: Regulatory Analysis.
Washington, DC. Available on the internet at
.
Pope, C.A., III, R.T. Burnett, MJ. 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.
Price quote from Charles Ball, Emissions & Silencer Technology for an oxidation catalyst for a
500 HP 4SLB engine.
Roman, H.A., K.D. Walker, T.L. Walsh, L. Conner, H.M. Richmond, J. Hubbell, and P.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.
8-2
-------
Technical Report: RICE NESHAP Control Costs Background for "Above the Floor Analysis",
October 2009, Attachment N (EPA-HQ-OAR-2008-0708-0279).
Technical Report: RICE NESHAP Control Costs Background for "Above the Floor Analysis",
October 2009, Attachment D (EPA-HQ-OAR-2008-0708-0279).
U.S. Bureau of Economic Analysis. 2002. 2002 Benchmark Input-Output Accounts: Detailed
Make Table, Use Table and Direct Requirements Table. Tables 4 and 5.
U.S. Bureau of Economic Analysis (BEA). 2010. Table 1.1.9. Implicit Price Deflators for Gross
Domestic Product. Last Revised June 25, 2010.
U.S. Census Bureau. 2008. "Firm Size Data from the Statistics of U.S. Businesses: U.S. All
Industries Tabulated by Receipt Size: 2002."
http://www.census.gov/csd/susb/susb02.htm.
U.S. Census Bureau. 2008a. Firm Size Data from the Statistics of U.S. Businesses: U.S. Detail
Employment Sizes: 2002.
http://www2.census.gov/csd/susb/2002/02us_detailed%20sizes_6digi tnaics.txt.
U.S. Census Bureau. 2008b. Firm Size Data from the Statistics of U.S. Businesses, U.S. All
Industries Tabulated by Receipt Size: 2002.
http://www2.census.gov/csd/susb/2002/usalli_r02.xls.
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21:
Mining: Industry Series: Historical Statistics for the Industry: 2002 and 1997."
; (November 26, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21:
EC0721I1: Mining: Industry Series: Detailed Statistics by Industry for the United States:
2007 " ; (April 27, 2010).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 21:
Mining: Geographic Area Series: Industry Statistics for the State or Offshore Areas:
2007." ; (January 27, 2010).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 22:
Utilities: Geographic Area Series: Summary Statistics: 2002."
; (November 10, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 22:
Utilities: Subject Series—Estab & Firm Size: Concentration by Largest Firms for the
United States: 2002." ; (November 21, 2008).
U.S. Census Bureau; American FactFinder; "Sector 22: EC0722I2: Utilities: Industry Series:
Preliminary Comparative Statistics for the United States (2002 NAICS Basis): 2007 and
2002."
8-3
-------
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 23:
Geographic Distribution—Construction: 2007."
(January 27, 2010).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 31:
EC0731I1: Manufacturing: Industry Series: Detailed Statistics by Industry for the United
States: 2007"
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 31:
Manufacturing: Industry Series: Historical Statistics for the Industry: 2002 and Earlier
Years" ; (November 25, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
Transportation and Warehousing: Industry Series: Comparative Statistics for the United
States (1997NAICS Basis): 2002 and 1997" ;
(December 12, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
Transportation and Warehousing: Subject Series—Estab & Firm Size: Concentration by
Largest Firms for the United States: 2002" ; (December 12,
2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
Transportation and Warehousing: Industry Series: Summary Statistics for the United
States: 2002" ; (January 27, 2010).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48:
EC0748I1: Transportation and Warehousing: Industry Series: Preliminary Summary
Statistics for the United States: 2007." http://factfmder.census.gov (January 27, 2010).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48-49:
Geographic Distribution—Pipeline transportation of natural gas: 2002."
; (November 10, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 48-49:
Transportation and Warehousing: Subject Series—Estab & Firm Size: Legal Form of
Organization for the United States: 2002" ; (December 12,
2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 62:
Health Care and Social Assistance: Geographic Area Series: Summary Statistics: 2002."
; (November 10, 2008).
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 62:
Health Care and Social Assistance: Subject Series—Estab & Firm Size: Legal Form of
Organization for the United States: 2002" ; (November 21,
2008).
8-4
-------
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 62:
Health Care and Social Assistance: Geographic Area Series: 2002 and 2007."
; (January 27, 2010).
U.S. Census Bureau. 2005a. 2002 Census of Governments, Volume 4, Number 3, Finances of
County Governments: 2002 GC02(4)-3. U.S. Government Printing Office, Washington,
DC. Table 12.
U.S. Census Bureau. 2005b. 2002 Census of Governments, Volume 4, Number 4, Finances of
Municipal and Township Governments: 2002 GC02(4)-4. U.S. Government Printing
Office, Washington, DC. Table 13.
U.S. Census Bureau; generated by RTI International; using American FactFinder; "Sector 00: All
sectors: Geographic Area Series: Economy-Wide Key Statistics: 2007"
; (July 7, 2010).
U.S. Department of Agriculture, National Agricultural Statistics Service. 2009. "2008 Farm and
Ranch Irrigation Survey." Washington, DC: USDA-NASS. Table 1.
U.S. Department of Agriculture, National Agricultural Statistics Service. 2009. "2008 Farm and
Ranch Irrigation Survey." Washington, DC: USDA-NASS. Table 20.
U.S. Department of Agriculture (USD A), National Agricultural Statistics Service (NASS). 2008.
"2008 Farm and Ranch Irrigation Survey." Washington, DC: USDA-NASS. Table 34.
U.S. Department of Agriculture (USDA). 2008. "Agricultural Projections to 2017."
.
U.S. Department of Energy, Energy Information Administration. 2007. "2006 EIA-906/920
Monthly Time Series."
U.S. Department of Energy, Energy Information Administration. 2009. "State Electricity Profiles
2008." DOE/EIA-0348(01)/2. p. 260. <
http://www.eia. doe. gov/cneaf/electricity/st_profiles/sep2008.pdf>.
U.S. Energy Information Administration. 2008a. . Last updated September 2008.
U.S. Energy Information Administration. 2009. Supplemental Tables to the Annual Energy
Outlook 2010. Table 2. Available at: .
U.S. Environmental Protection Agency (U.S. EPA). May 2004. Final Regulatory Analysis:
Control of Emissions from Nonroad Diesel Engines. EPA420-R-04-007. Washington,
DC: EPA. < http://www.epa.gov/nonroad-diesel/2004fr/420r04007.pdf>.
8-5
-------
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 at
.
U.S. Environmental Protection Agency (U.S. EPA). 2006a. Regulatory Impact Analysis, 2006
National Ambient Air Quality Standards for Particulate Matter, Chapter 5. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2006b. Technical Support Document for the
ProposedPMNAAQS Rule: Response Surface Modeling. Office of Air Quality Planning
and Standards, Research Triangle Park, NC. February. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2006c. Air Quality Criteria for Ozone and
Related Photochemical Oxidants (Final). EPA/600/R-05/004aF-cF. Washington, DC:
U.S. EPA. February. Available 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. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2008a. Integrated Science Assessment for
Oxides of Nitrogen - Health Criteria (Final Report). National Center for Environmental
Assessment, Research Triangle Park, NC. July. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2008c. Guidelines for Preparing Economic
Analyses: External Review Draft. National Center for Environmental Economics, Office
of Policy Economics and Innovation. Washington, DC. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2008d. 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 at .
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 at
.
8-6
-------
U.S. Environmental Protection Agency (U.S. EPA). 2009b. Integrated Science Assessment for
Paniculate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment—RTF Division. December. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2009c. 2002 National-Scale Air Toxics
Assessment. Office of Air Quality Planning and Standards. June. Available at
U.S. Environmental Protection Agency (U.S. EPA). 2010a. Integrated Science Assessment for
Carbon Monoxide. National Center for Environmental Assessment, Research Triangle
Park, NC. EPA/600/R-09/019F. January. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010b. Final Regulatory Impact Analysis
(RIA)for the NO2 National Ambient Air Quality Standards (NAAQS). Office of Air
Quality Planning and Standards, Research Triangle Park, NC. January. Available at <
http://www.epa.gov/ttn/ecas/regdata/RIAs/FinalNO2RIAfulldocument.pdf>.
U.S. Environmental Protection Agency (U.S. EPA). 2010c. Summary of Expert Opinions on the
Existence of a Threshold in the Concentration-Response Function for PM2.s-related
Mortality: Technical Support Document. Compiled by Office of Air Quality Planning and
Standards, Research Triangle Park, NC. July. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010d. Lowest Measured Level (LML)
Assessment for Rules without Policy-Specific Air Quality Data Available: Technical
Support Document. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. July. Available at .
U.S. Environmental Protection Agency (U.S. EPA). 2010e. Regulatory Impact Analysis for the
Transport Rule. Office of Air Quality Planning and Standards, Research Triangle Park, NC.
July. Available at .
U.S. Environmental Protection Agency (U.S. EPA). 2010f. Air Quality Modeling Technical
Support Document: Changes to the Renewable Fuel Standard Program., Office of Air
Quality Planning and Standards, EPA 454/R-10-001. February. Available at
.
U.S. Environmental Protection Agency - Science Advisory Board (U.S. 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 at
.
8-7
-------
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 at
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2009a.
Review of EPA 's Integrated Science Assessment for Paniculate Matter (First External
Review Draft, December 2008). EPA-COUNCIL-09-008. May. Available at
.
U.S. Small Business Administration (SBA). 2008. "Table of Small Business Size Standards
Matched to North American Industry Classification System Codes." Effective August 22,
2008. .
Viscusi, V.K., and I.E. Aldy. 2003. "The Value of a Statistical Life: A Critical Review of Market
Estimates throughout the World." Journal of Risk and Uncertainty 27(l):5-76.
Wade, S.H. 2003. "Price Responsiveness in the AEO2003 NEMS Residential and Commercial
Buildings Sector Models."
.
-------
APPENDIX A
SUMMARY OF EXPERT OPINIONS ON THE EXISTENCE OF A THRESHOLD IN
THE CONCENTRATION-RESPONSE FUNCTION FOR PM2.5-RELATED
MORTALITY
-------
Summary of Expert Opinions on the Existence of a Threshold in the
Concentration-Response Function for PM2.s-related Mortality
Technical Support Document (TSD)
June 2010
Compiled by:
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impact Division
Air Benefit-Cost Group
Research Triangle Park, North Carolina
Contents:
A. HES comments on 812 Analysis (2010)
B. American Heart Association Scientific Statement (2010)
C. Integrated Science Assessment for Particulate Matter (2009)
D. CASAC comments on PM ISA and REA (2009)
E. Krewski et al. (2009)
F. Schwartz et al. (2008)
G. Expert Elicitation on PM Mortality (2006, 2008)
H. CASAC comments on PM Staff Paper (2005)
I. HES comments on 812 Analysis (2004)
J. NRC (2002)
A-l
-------
A. HES Comments on 812 Analysis (2010)
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
.
Pg 2: "The HES generally agrees with other decisions made by the EPA project team with
respect to PM, in particular, the PM mortality effect threshold model, the cessation lag model,
the inclusion of infant mortality estimation, and differential toxicity of PM."
Pg 2: "Further, the HES fully supports EPA's use of a no-threshold model to estimate the
mortality reductions associated with reduced PM exposure."
Pg 6: "The HES also supports the Agency's choice of a no-threshold model for PM-related
effects."
Pg 13: "The HES fully supports EPA's decision to use a no-threshold model to estimate mortality
reductions. This 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. Therefore, there is
no evidence to support a truncation of the CRF."
HES Panel Members
Dr. John Bailar, Chair of the Health Effects Subcommittee, Scholar in Residence, The National
Academies, Washington, DC
Dr. Michelle Bell, Associate Professor, School of Forestry and Environmental Studies, Yale
University, New Haven, CT
Dr. James K. Hammitt, Professor, Department of Health Policy and Management, Harvard
School of Public Health, Boston, MA
Dr. Jonathan Levy, Associate Professor, Department of Environmental Health, Harvard School
of Public Health, Boston, MA
Dr. C. Arden Pope, III Professor, Department of Economics, Brigham Young University,
Provo, UT
Mr. John Fintan Hurley, Research Director, Institute of Occupational Medicine (IOM),
Edinburgh, United Kingdom, UK
Dr. Patrick Kinney, Professor, Department of Environmental Health Sciences, Mailman School
of Public Health, Columbia University, New York, NY
Dr. Michael T. Kleinman, Professor, Department of Medicine, Division of Occupational and
Environmental Medicine, University of California, Irvine, Irvine, C A
A-2
-------
Dr. Bart Ostro, Chief, Air Pollution Epidemiology Unit, Office of Environmental Health
Hazard Assessment, California Environmental Protection Agency, Oakland, CA
Dr. Rebecca Parkin, Professor and Associate Dean, Environmental and Occupational Health,
School of Public Health and Health Services, The George Washington University Medical
Center, Washington, DC
A-3
-------
B. Scientific Statement from American Heart Association (2010)
Brook RD, Rajagopalan S, Pope CA 3rd, Brook JR, Bhatnagar A, Diez-Roux AV, Holguin
F, Hong Y, Luepker RV, Mittleman MA, Peters A, Siscovick D, Smith SC Jr, Whitsel L,
Kaufman JD; on behalf of the American Heart Association Council on Epidemiology and
Prevention, Council on the Kidney in Cardiovascular Disease, and Council on Nutrition,
Physical Activity and Metabolism. (2010). "Particulate matter air pollution and
cardiovascular disease: an update to the scientific statement from the American Heart
Association." Circulation. 121: 2331-2378.
Pg 2338: "Finally, there appeared to be no lower-limit threshold below which PMio was not
associated with excess mortality across all regions."
Pg 2350: "There also appears to be a monotonic (e.g., linear or log-linear) concentration-
response relationship between PM2.5 and mortality risk observed in cohort studies that extends
below present-day regulations of 15 |ig/m3 for mean annual levels, without a discernable "safe"
threshold." (cites Pope 2004, Krewski 2009, and Schwartz 2008)
Pg 2364: "The PM2.5 concentration- cardiovascular risk relationships for both short- and long-
term exposures appear to be monotonic, extending below 15 |ig/m3 (the 2006 annual NAAQS
level) without a discernable "safe" threshold."
Pg 2365: "This updated review by the AHA writing group corroborates and strengthens the
conclusions of the initial scientific statement. In this context, we agree with the concept and
continue to support measures based on scientific evidence, such as the US EPA NAAQS, that
seek to control PM levels to protect the public health. Because the evidence reviewed supports
that there is no safe threshold, it appears that public health benefits would accrue from lowering
PM2.5 concentrations even below present-day annual (15 |ig/m3) and 24-hour (35 |ig/m3)
NAAQS, if feasible, to optimally protect the most susceptible populations."
Pg 2366: "Although numerous insights have greatly enhanced our understanding of the PM-
cardiovascular relationship since the first AHA statement was published, the following list
represents broad strategic avenues for future investigation: ... Determine whether any "safe" PM
threshold concentration exists that eliminates both acute and chronic cardiovascular effects in
healthy and susceptible individuals and at a population level."
Scientific Statement Authors
Dr. Robert D. Brook, MD
Dr. Sanjay Rajagopalan, MD
Dr. C. Arden Pope, PhD
Dr. Jeffrey R. Brook, PhD
Dr. Aruni Bhatnagar, PhD, FAHA
Dr. Ana V. Diez-Roux, MD, PhD, MPH
A-4
-------
Dr. Fernando Holguin, MD
Dr. Yuling Hong, MD, PhD, FAHA
Dr. Russell V. Luepker, MD, MS, FAHA
Dr. Murray A. Mittleman, MD, DrPH, FAHA
Dr. Annette Peters, PhD
Dr. David Siscovick, MD, MPH, FAHA
Dr. Sidney C. Smith, Jr, MD, FAHA
Dr. Laurie Whitsel, PhD
Dr. Joel D. Kaufman, MD, MPH
A-5
-------
C. Integrated Science Assessment for Particulate Matter (2009)
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
.
Pg 1-22: "An important consideration in characterizing the public health impacts associated with
exposure to a pollutant is whether the concentration-response relationship is linear across the full
concentration range encountered, or if nonlinear relationships exist along any part of this range.
Of particular interest is the shape of the concentration-response curve at and below the level of
the current standards. The shape of the concentration-response curve varies, depending on the
type of health outcome, underlying biological mechanisms and dose. At the human population
level, however, various sources of variability and uncertainty tend to smooth and "linearize" the
concentration-response function (such as the low data density in the lower concentration range,
possible influence of measurement error, and individual differences in susceptibility to air
pollution health effects). In addition, many chemicals and agents may act by perturbing naturally
occurring background processes that lead to disease, which also linearizes population
concentration-response relationships (Clewell and Crump, 2005, 156359; Crump et al., 1976,
003192; Hoel, 1980, 156555). These attributes of population dose-response may explain why the
available human data at ambient concentrations for some environmental pollutants (e.g., PM, Os,
lead [Pb], ETS, radiation) do not exhibit evident thresholds for health effects, even though likely
mechanisms include nonlinear processes for some key events. These attributes of human
population dose-response relationships have been extensively discussed in the broader
epidemiologic literature (Rothman and Greenland, 1998, 086599)."
Pg 2-16: "In addition, cardiovascular hospital admission and mortality studies that examined the
PMio concentration-response relationship found evidence of a log-linear no-threshold
relationship between PM exposure and cardiovascular-related morbidity (Section 6.2) and
mortality (Section 6.5)."
Pg 2-25: "2.4.3. PM Concentration-Response Relationship
An important consideration in characterizing the PM-morbidity and mortality association is
whether the concentration-response relationship is linear across the full concentration range that
is encountered or if there are concentration ranges where there are departures from linearity (i.e.,
nonlinearity). In this ISA studies have been identified that attempt to characterize the shape of
the concentration-response curve along with possible PM "thresholds" (i.e., levels which PM
concentrations must exceed in order to elicit a health response). The epidemiologic studies
evaluated that examined the shape of the concentration-response curve and the potential presence
of a threshold have focused on cardiovascular hospital admissions and ED visits and mortality
associated with short-term exposure to PMio and mortality associated with long-term exposure to
PM2.5.
"A limited number of studies have been identified that examined the shape of the PM
cardiovascular hospital admission and ED visit concentration-response relationship. Of these
A-6
-------
studies, some conducted an exploratory analysis during model selection to determine if a linear
curve most adequately represented the concentration-response relationship; whereas, only one
study conducted an extensive analysis to examine the shape of the concentration-response curve
at different concentrations (Section 6.2.10.10). Overall, the limited evidence from the studies
evaluated supports the use of a no-threshold, log-linear model, which is consistent with the
observations made in studies that examined the PM-mortality relationship.
"Although multiple studies have previously examined the PM-mortality concentration-response
relationship and whether a threshold exists, more complex statistical analyses continue to be
developed to analyze this association. Using a variety of methods and models, most of the
studies evaluated support the use of a no-threshold, log-linear model; however, one study did
observe heterogeneity in the shape of the concentration-response curve across cities (Section
6.5). Overall, the studies evaluated further support the use of a no-threshold log-linear model, but
additional issues such as the influence of heterogeneity in estimates between cities, and the effect
of seasonal and regional differences in PM on the concentration-response relationship still
require further investigation.
"In addition to examining the concentration-response relationship between short-term exposure
to PM and mortality, Schwartz et al. (2008, 156963) conducted an analysis of the shape of the
concentration-response relationship associated with long-term exposure to PM. Using a variety
of statistical methods, the concentration-response curve was found to be indistinguishable from
linear, and, therefore, little evidence was observed to suggest that a threshold exists in the
association between long-term exposure to PM2.5 and the risk of death (Section 7.6)."
Pg 6-75: "6.2.10.10. Concentration Response
The concentration-response relationship has been extensively analyzed primarily through studies
that examined the relationship between PM and mortality. These studies, which have focused on
short- and long-term exposures to PM have consistently found no evidence for deviations from
linearity or a safe threshold (Daniels et al., 2004, 087343; Samoli et al., 2005, 087436; Schwartz,
2004, 078998; Schwartz et al., 2008, 156963) (Sections 6.5.2.7 and 7.1.4). Although on a more
limited basis, studies that have examined PM effects on cardiovascular hospital admissions and
ED visits have also analyzed the PM concentration-response relationship, and contributed to the
overall body of evidence which suggests a log-linear, no-threshold PM concentration-response
relationship.
"The results from the three multicity studies discussed above support no-threshold log-linear
models, but issues such as the possible influence of exposure error and heterogeneity of shapes
across cities remain to be resolved. Also, given the pattern of seasonal and regional differences
in PM risk estimates depicted in recent multicity study results (e.g., Peng et al., 2005, 087463),
the very concept of a concentration-response relationship estimated across cities and for all-year
data may not be very informative."
Pg 6-197: "6.5.2.7. Investigation of Concentration-Response Relationship
The results from large multicity studies reviewed in the 2004 PM AQCD (U.S. EPA, 2004,
056905) suggested that strong evidence did not exist for a clear threshold for PM mortality
effects. However, as discussed in the 2004 PM AQCD (U.S. EPA, 2004, 056905), there are
A-7
-------
several challenges in determining and interpreting the shape of PM-mortality concentration-
response functions and the presence of a threshold, including: (1) limited range of available
concentration levels (i.e., sparse data at the low and high end); (2) heterogeneity of susceptible
populations; and (3) investigate the PM-mortality concentration-response relationship.
"Daniels et al. (2004, 087343) evaluated three concentration-response models: (1) log-linear
models (i.e., the most commonly used approach, from which the majority of risk estimates are
derived); (2) spline models that allow data to fit possibly non-linear relationship; and (3)
threshold models, using PMio data in 20 cities from the 1987-1994 NMMAPS data. They
reported that the spline model, combined across the cities, showed a linear relation without
indicating a threshold for the relative risks of death for all-causes and for cardiovascular-
respiratory causes in relation to PMi0, but "the other cause" deaths (i.e., all cause minus
cardiovascular-respiratory) showed an apparent threshold at around 50 ug/m3 PM10, as shown in
Figure 6-35. For all-cause and cardio-respiratory deaths, based on the Akaike's Information
Criterion (AIC), a log-linear model without threshold was preferred to the threshold model and
to the spline model.
"The HEI review committee commented that interpretation of these results required caution,
because (1) the measurement error could obscure any threshold; (2) the city-specific
concentration-response curves exhibited a variety of shapes; and (3) the use of AIC to choose
among the models might not be appropriate due to the fact it was not designed to assess scientific
theories of etiology. Note, however, that there has been no etiologically credible reason
suggested thus far to choose one model over others for aggregate outcomes. Thus, at least
statistically, the result of Daniels et al. (2004, 087343) suggests that the log-linear model is
appropriate in describing the relationship between PMIO and mortality.
"The Schwartz (2004, 078998) analysis of PMio and mortality in 14 U.S. cities, described in
Section 6.5.2.1, also examined the shape of the concentration-response relationship by including
indicator variables for days when concentrations were between 15 and 25 ug/m3, between 25 and
34 ug/m3, between 35 and 44 ug/m3, and 45 ug/m3 and above. In the model, days with
concentrations below 15 ug/m3 served as the reference level. This model was fit using the single
stage method, combining strata across all cities in the case-crossover design. Figure 6-36 shows
the resulting relationship, which does not provide sufficient evidence to suggest that a threshold
exists. The authors did not examine city-to-city variation in the concentration-response
relationship in this study.
"PMio and mortality in 22 European cities (and BS in 15 of the cities) participating in the
APHEA project. In nine of the 22 cities, PMIO levels were estimated using a regression model
relating co-located PMIO to BS or TSP. They used regression spline models with two knots (30
and 50 ug/m3) and then combined the individual city estimates of the splines across cities. The
investigators concluded that the association between PM and mortality in these cities could be
adequately estimated using the log-linear model. However, in an ancillary analysis of the
concentration-response curves for the largest cities in each of the three distinct geographic areas
(western, southern, and eastern European cities): London, England; Athens, Greece; and Cracow,
Poland, Samoli et al. (2005, 087436) observed a difference in the shape of the concentration-
response curve across cities. Thus, while the combined curves (Figure 6-37) appear to support
A-8
-------
no-threshold relationships between PMio and mortality, the heterogeneity of the shapes across
cities makes it difficult to interpret the biological relevance of the shape of the combined curves.
"The results from the three multicity studies discussed above support no-threshold log-linear
models, but issues such as the possible influence of exposure error and heterogeneity of shapes
across cities remain to be resolved. Also, given the pattern of seasonal and regional differences
in PM risk estimates depicted in recent multicity study results (e.g., Peng et al., 2005, 087463),
the very concept of a concentration-response relationship estimated across cities and for all-year
data may not be very informative."
Authors of ISA
Dr. Lindsay Wichers Stanek (PM Team Leader)—National Center for Environmental
Assessment (NCEA), U.S. Environmental Protection Agency (U.S. EPA), Research Triangle
Park, NC
Dr. Jeffrey Arnold—NCEA, U.S. EPA, Research Triangle Park, NC (now at Institute for Water
Resources, U.S. Army Corps of Engineers, Washington, D.C)
Dr. Christal Bowman—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. James S. Brown—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Barbara Buckley—NCEA, U.S. EPA, Research Triangle Park, NC
Mr. Allen Davis—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Jean-Jacques Dubois—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Steven J. Dutton—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Tara Greaver—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Erin Hines—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Douglas Johns—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Ellen Kirrane—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Dennis Kotchmar—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Thomas Long—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Thomas Luben—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Qingyu Meng—Oak Ridge Institute for Science and Education, Postdoctoral Research
Fellow to NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Kristopher Novak—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Joseph Pinto—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Jennifer Richmond-Bryant—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Mary Ross—NCEA, U.S. EPA, Research Triangle Park, NC
Mr. Jason Sacks—NCEA, U.S. EPA, Research Triangle Park, NC
A-9
-------
Dr. Timothy J. Sullivan—E&S Environmental Chemistry, Inc., Corvallis, OR
Dr. David Svendsgaard—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Lisa Vinikoor—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. William Wilson—NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Lori White— NCEA, U.S. EPA, Research Triangle Park, NC (now at National Institute for
Environmental Health Sciences, Research Triangle Park, NC)
Dr. Christy Avery—University of North Carolina, Chapel Hill, NC
Dr. Kathleen Belanger —Center for Perinatal, Pediatric and Environmental Epidemiology,
Yale University, New Haven, CT
Dr. Michelle Bell—School of Forestry & Environmental Studies, Yale University, New Haven,
CT
Dr. William D. Bennett—Center for Environmental Medicine, Asthma and Lung Biology,
University of North Carolina, Chapel Hill, NC
Dr. Matthew J. Campen—Lovelace Respiratory Research Institute, Albuquerque, NM
Dr. Leland B. Deck— Stratus Consulting, Inc., Washington, DC
Dr. Janneane F. Gent—Center for Perinatal, Pediatric and Environmental Epidemiology, Yale
University, New Haven, CT
Dr. Yuh-Chin Tony Huang—Department of Medicine, Division of Pulmonary Medicine, Duke
University Medical Center, Durham, NC
Dr. Kazuhiko Ito—Nelson Institute of Environmental Medicine, NYU School of Medicine,
Tuxedo, NY
Mr. Marc Jackson—Integrated Laboratory Systems, Inc., Research Triangle Park, NC
Dr. Michael Kleinman—Department of Community and Environmental Medicine, University
of California, Irvine
Dr. Sergey Napelenok—National Exposure Research Laboratory, U.S. EPA, Research Triangle
Park, NC
Dr. Marc Pitchford—National Oceanic and Atmospheric Administration, Las Vegas, NV
Dr. Les Recio—Genetic Toxicology Division, Integrated Laboratory Systems, Inc., Research
Triangle Park, NC
Dr. David Quincy Rich—Department of Epidemiology, University of Medicine and Dentistry
of New Jersey, Piscataway, NJ
Dr. Timothy Sullivan— E&S Environmental Chemistry, Inc., Corvallis, OR
Dr. George Thurston—Department of Environmental Medicine, NYU, Tuxedo, NY
Dr. Gregory Wellenius—Cardiovascular Epidemiology Research Unit, Beth Israel Deaconess
Medical Center, Boston, MA
A-10
-------
Dr. Eric Whitsel—Departments of Epidemiology and Medicine, University of North Carolina,
Chapel Hill, NC
Peer Reviewers
Dr. Sara Dubowsky Adar, Department of Epidemiology, University of Washington, Seattle,
WA
Mr. Chad Bailey, Office of Transportation and Air Quality, Ann Arbor, MI
Mr. Richard Baldauf, Office of Transportation and Air Quality, Ann Arbor, MI
Dr. Prakash Bhave, National Exposure Research Laboratory, U.S. EPA, Research Triangle
Park, NC
Mr. George Bowker, Office of Atmospheric Programs, U.S. EPA, Washington, D.C.
Dr. Judith Chow, Division of Atmospheric Sciences, Desert Research Institute, Reno, NV
Dr. Dan Costa, U.S. EPA, Research Triangle Park, NC
Dr. Ila Cote, NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Robert Devlin, National Health and Environmental Effects Research Laboratory, U.S. EPA,
Research Triangle Park, NC
Dr. David DeMarini, National Health and Environmental Effects Research Laboratory, U.S.
EPA, Research Triangle Park, NC
Dr. Neil Donahue, Department of Chemical Engineering, Carnegie Mellon University,
Pittsburgh, PA
Dr. Aimen Farraj, National Health and Environmental Effects Research Laboratory, U.S. EPA,
Research Triangle Park, NC
Dr. Mark Frampton, Department of Environmental Medicine, University of Rochester Medical
Center, Rochester, NY
Mr. Neil Frank, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Mr. Tyler Fox, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Dr. Jim Gauderman, Department of Environmental Medicine, Department of Preventive
Medicine, University of Southern California, Los Angeles, CA
Dr. Barbara Glenn, National Center for Environmental Research, U.S. EPA, Washington, D.C.
Dr. Terry Gordon, School of Medicine, New York University, Tuxedo, NY
Mr. Tim Hanley, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Dr. Jack Harkema, Department of Pathobiology and Diagnostic Investigation, Michigan State
University, East Lansing, MI
Ms. Beth Hassett-Sipple, Office of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC
A-ll
-------
Dr. Amy Herring, Department of Biostatistics, University of North Carolina, Chapel Hill, NC
Dr. Israel Jirak, Department of Meteorology, Embry-Riddle Aeronautical University, Prescott,
AZ
Dr. Mike Kleeman, Department of Civil and Environmental Engineering, University of
California, Davis, CA
Dr. Petros Koutrakis, Exposure, Epidemiology and Risk Program, Harvard School of Public
Health, Boston, MA
Dr. Sagar Krupa, Department of Plant Pathology, University of Minnesota, St. Paul, MN
Mr. John Langstaff, Office of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC
Dr. Meredith Lassiter, Office of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC
Mr. Phil Lorang, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Dr. Karen Martin, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Ms. Connie Meacham, NCEA, U.S. EPA, Research Triangle Park, NC
Mr. Tom Pace, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Dr. Jennifer Peel, Department of Environmental and Radiological Health Sciences, College of
Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO
Dr. Zackary Pekar, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Mr. Rob Pinder, National Exposure Research Laboratory, U.S. EPA, Research Triangle Park,
NC
Mr. Norm Possiel, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Dr. Sanjay Rajagopalan, Division of Cardiovascular Medicine, Ohio State University,
Columbus, OH
Dr. Pradeep Rajan, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Mr. Venkatesh Rao, Office of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC
Ms. Joann Rice, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Mr. Harvey Richmond, Office of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC
A-12
-------
Ms. Victoria Sandiford, Office of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC
Dr. Stefanie Sarnat, Department of Environmental and Occupational Health, Emory University,
Atlanta, GA
Dr. Frances Silverman, Gage Occupational and Environmental Health, University of Toronto,
Toronto, ON
Mr. Steven Silverman, Office of General Council, U.S. EPA, Washington, D.C.
Dr. Barbara Turpin, Department of Environmental Sciences, Rutgers University, New
Brunswick, NJ
Dr. Robert Vanderpool, National Exposure Research Laboratory, U.S. EPA, Research Triangle
Park, NC
Dr. John Vandenberg (Director)—NCEA-RTP Division, U.S. EPA, Research Triangle Park,
NC
Dr. Alan Vette, National Exposure Research Laboratory, U.S. EPA, Research Triangle Park,
NC
Ms. Debra Walsh (Deputy Director)—NCEA-RTP Division, U.S. EPA, Research Triangle
Park, NC
Mr. Tim Watkins, National Exposure Research Laboratory, U.S. EPA, Research Triangle Park,
NC
Dr. Christopher Weaver, NCEA, U.S. EPA, Research Triangle Park, NC
Mr. Lewis Weinstock, Office of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC
Ms. Karen Wesson, Office of Air Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC
Dr. Jason West, Department of Environmental Sciences and Engineering, University of North
Carolina, Chapel Hill, NC
Mr. Ronald Williams, National Exposure Research Laboratory, U.S. EPA, Research Triangle
Park, NC
Dr. George Woodall, NCEA, U.S. EPA, Research Triangle Park, NC
Dr. Antonella Zanobetti, Department of Environmental Health, Harvard School of Public
Health, Boston, MA
A-13
-------
D. CASAC comments on PM ISA and REA (2009)
U.S. Environmental Protection Agency - Science Advisory Board (U.S. EPA-SAB). 2009.
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
.
Pg 9: "There is an appropriate discussion of the time-series studies, but this section needs to have
an explicit finding that the evidence supports a relationship between PM and mortality that is
seen in these studies. This conclusion should be followed by the discussion of statistical
methodology and the identification of any threshold that may exist."
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2009.
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
.
Pg 6: "On the issue of cut-points raised on 3-18, the authors should be prepared to offer a
scientifically cogent reason for selection of a specific cut-point, and not simply try different cut-
points to see what effect this has on the analysis. The draft ISA was clear that there is little
evidence for a population threshold in the C-R function."
U.S. Environmental Protection Agency - Science Advisory Board (U.S. EPA-SAB). 2009. Review of
Integrated Science Assessment for Particulate Matter (Second External Review Draft, July 2009).
EPA-CASAC-10-001. November. Available on the Internet at
.
Pg 2: "The paragraph on lines 22-30 of page 2-37 is not clearly written. Twice in succession it
states that the use of a no-threshold log-linear model is supported, but then cites other studies
that suggest otherwise. It would be good to revise this paragraph to more clearly state - well, I'm
not sure what. Probably that more research is needed."
CASAC Panel Members
Dr. Jonathan M. Samet, Professor and Chair, Department of Preventive Medicine, University of
Southern California, Los Angeles, CA
Dr. Joseph Brain, Philip Drinker Professor of Environmental Physiology, Department of Environmental
Health, Harvard School of Public Health, Harvard University, Boston, MA
Dr. Ellis B. Cowling, University Distinguished Professor At-Large Emeritus, Colleges of Natural
Resources and Agriculture and Life Sciences, North Carolina State University, Raleigh, NC
A-14
-------
Dr. James Crapo, Professor of Medicine, Department of Medicine, National Jewish Medical and
Research Center, Denver, CO
Dr. H. Christopher Frey, Professor, Department of Civil, Construction and Environmental Engineering,
College of Engineering, North Carolina State University, Raleigh, NC
Dr. Armistead (Ted) Russell, Professor, Department of Civil and Environmental Engineering, Georgia
Institute of Technology, Atlanta, GA
Dr. Lowell Ashbaugh, Associate Research Ecologist, Crocker Nuclear Lab, University of California,
Davis, Davis, CA
Prof. Ed Avol, Professor, Preventive Medicine, Keck School of Medicine, University of Southern
California, Los Angeles, CA
Dr. Wayne Cascio, Professor, Medicine, Cardiology, Brody School of Medicine at East Carolina
University, Greenville, NC
Dr. David Grantz, Director, Botany and Plant Sciences and Air Pollution Research Center, Riverside
Campus and Kearney Agricultural Center, University of California, Parlier, CA
Dr. Joseph Helble, Dean and Professor, Thayer School of Engineering, Dartmouth College, Hanover,
NH
Dr. Rogene Henderson, Senior Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM
Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical Engineering,
Clarkson University, Potsdam, NY
Dr. Morton Lippmann, Professor, Nelson Institute of Environmental Medicine, New York University
School of Medicine, Tuxedo, NY
Dr. Helen Suh Macintosh, Associate Professor, Environmental Health, School of Public Health,
Harvard University, Boston, MA
Dr. William Malm, Research Physicist, National Park Service Air Resources Division, Cooperative
Institute for Research in the Atmosphere, Colorado State University, Fort Collins, CO
Mr. Charles Thomas (Tom) Moore, Jr., Air Quality Program Manager, Western Governors'
Association, Cooperative Institute for Research in the Atmosphere, Colorado State University, Fort
Collins, CO
Dr. Robert F. Phalen, Professor, Department of Community & Environmental Medicine; Director, Air
Pollution Health Effects Laboratory; Professor of Occupational & Environmental Health, Center for
Occupation & Environment Health, College of Medicine, University of California Irvine, Irvine, CA
Dr. Kent Pinkerton, Professor, Regents of the University of California, Center for Health and the
Environment, University of California, Davis, CA
Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Frank Speizer, Edward Kass Professor of Medicine, Channing Laboratory, Harvard Medical School,
Boston, MA
Dr. Sverre Vedal, Professor, Department of Environmental and Occupational Health Sciences, School of
Public Health and Community Medicine, University of Washington, Seattle, WA
Dr. Donna Kenski, Data Analysis Director, Lake Michigan Air Directors Consortium, Rosemont, IL
A-15
-------
Dr. Kathy Weathers, Senior Scientist, Gary Institute of Ecosystem Studies, Millbrook, NY
A-16
-------
E. Krewski et al. (2009)
Krewski, Daniel, Michael Jerrett, Richard T. Burnett, Renjun Ma, Edward Hughes, Yuanli
Shi, Michelle C. Turner, C. Arden Pope III, George Thurston, Eugenia E. Calle, and
Michael J. Thun with Bernie Beckerman, Pat DeLuca, Norm Finkelstein, Kaz Ito, D.K.
Moore, K. Bruce Newbold, Tim Ramsay, Zev Ross, Hwashin Shin, and Barbara
Tempalski. (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.
Pg 119: [About Pope et al. (2002)] "Each 10-ug/m3 increase in long-term average ambient PM2.5
concentrations was associated with approximately a 4%, 6%, or 8% increase in risk of death
from all causes, cardiopulmonary disease, and lung cancer, respectively. There was no evidence
of a threshold exposure level within the range of observed PM2.5 concentrations."
Krewski (2009). Letter from Dr. Daniel Krewski to HEI's Dr. Kate Adams (dated July 7,
2009) regarding "EPA queries regarding HEI Report 140". Dr. Adams then forwarded
the letter on July 10, 2009 to EPA's Beth Hassett-Sipple. (letter placed in docket #EPA-
HQ-OAR-2007-0492).
Pg4: "6. The Health Review Committee commented that the Updated Analysis completed by
Pope et al. 2002 reported "no evidence of a threshold exposure level within the range of
observed PM2.5 concentrations" (p. 119). In the Extended Follow-Up study, did the analyses
provide continued support for a no-threshold response or was there evidence of a threshold?
"Response: As noted above, the HEI Health Review Committee commented on the lack of
evidence for a threshold exposure level in Pope et al. (2002) with follow-up through the year
1998. The present report, which included follow-up through the year 2000, also does not appear
to demonstrate the existence of a threshold in the exposure-response function within the range of
observed PM2.5 concentrations."
HEI Health Review Committee Members
Dr. Homer A. Boushey, MD, Chair, Professor of Medicine, Department of Medicine,
University of California-San Francisco
Dr. Ben Armstrong, Reader, in Epidemiological Statistics, Department of Public Health and
Policy, London School of Hygiene and Tropical Medicine, United Kingdom
Dr. Michael Brauer, ScD, Professor, School of Environmental Health, University of British
Columbia, Canada
Dr. Bert Brunekreef, PhD, Professor of Environmental Epidemiology, Institute of Risk
Assessment Sciences, University of Utrecht, The Netherlands
Dr. Mark W. Frampton, MD, Professor of Medicine & Environmental Medicine, University of
Rochester Medical Center, Rochester, NY
A-17
-------
Dr. Stephanie London, MD, PhD, Senior Investigator, Epidemiology Branch, National Institute
of Environmental Health Sciences
Dr. William N. Rom, MD, MPH, Sol and Judith Bergstein Professor of Medicine and
Environmental Medicine and Director of Pulmonary and Critical Care Medicine, New York
University Medical Center
Dr. Armistead Russell, Georgia Power Distinguished Professor of Environmental Engineering,
School of Civil and Environmental Engineering, Georgia Institute of Technology
Dr. Lianne Sheppard, PhD, Professor, Department of Biostatistics, University of Washington
A-18
-------
F. Schwartz et al. (2008)
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.
Pg 67: "A key finding of this study is that there is little evidence for a threshold in the
association between exposure to fine particles and the risk of death on follow-up, which
continues well below the U.S. EPA standard of 15 ug/m3."
Pg 68: "In conclusion, penalized spline smoothing and model averaging represent reasonable,
feasible approaches to addressing questions of the shape of the exposure-response curve, and can
provide valuable information to decisionmakers. In this example, both approaches are consistent,
and suggest that the association of particles with mortality has no threshold down to close to
background levels."
A-19
-------
G. Expert Elicitation on PM-Mortality (2006, 2008)
Industrial Economics, Inc., 2006. Expanded Expert Judgment Assessment of the
Concentration-Response Relationship Between PM2.s Exposure and Mortality. Prepared for
the U.S.EPA, Office of Air Quality Planning and Standards, September. Available on the
Internet at .
Pg v: "Each expert was given the option to integrate their judgments about the likelihood of a
causal relationship and/or threshold in the C-R function into his distribution or to provide a
distribution "conditional on" one or both of these factors."
Pg vii: "Only one of 12 experts explicitly incorporated a threshold into his C-R function.3 The
rest believed there was a lack of empirical and/or theoretical support for a population threshold.
However, three other experts gave differing effect estimate distributions above and below some
cut-off concentration. The adjustments these experts made to median estimates and/or
uncertainty at lower PM2 5 concentrations were modest."
"3 Expert K indicated that he was 50 percent sure that a threshold existed. If there
were a threshold, he thought that there was an 80 percent chance that it would be
less than or equal to 5 ug/m3, and a 20 percent chance that it would fall between 5
and 10 ug/m3."
Pg ix: "Compared to the pilot study, experts in this study were in general more confident in a
causal relationship, less likely to incorporate thresholds, and reported higher mortality effect
estimates. The differences in results compared with the pilot appear to reflect the influence of
new research on the interpretation of the key epidemiological studies that were the focus of both
elicitation studies, more than the influence of changes to the structure of the protocol."
Pg 3-25: "3.1.8 THRESHOLDS
The protocol asked experts for their judgments regarding whether a threshold exists in the PM2.5
mortality C-R function. The protocol focused on assessing expert judgments regarding theory
and evidential support for a population threshold (i.e., the concentration below which no member
of the study population would experience an increased risk of death).32 If an expert wished to
incorporate a threshold in his characterization of the concentration-response relationship, the
team then asked the expert to specify the threshold PM2.5 concentration probabilistically,
incorporating his uncertainty about the true threshold level.
"From a theoretical and conceptual standpoint, all experts generally believed that individuals
exhibit thresholds for PM-related mortality. However, 11 of them discounted the idea of a
population threshold in the C-R function on a theoretical and/or empirical basis. Seven of these
experts noted that theoretically one would be unlikely to observe a population threshold due to
the variation in susceptibility at any given time in the study population resulting from
combinations of genetic, environmental, and socioeconomic factors.33 All 11 thought that there
was insufficient empirical support for a population threshold in the C-R function. In addition,
two experts (E and L) cited analyses of the ACS cohort data in Pope et al. (2002) and another (J)
cited Krewski et al. (2000a & b) as supportive of a linear relationship in the study range.
A-20
-------
"Seven of the experts favored epidemiological studies as ideally the best means of addressing the
population threshold issue, because they are best able to evaluate the full range of susceptible
individuals at environmentally relevant exposure levels. However, those who favored
epidemiologic studies generally acknowledged that definitive studies addressing thresholds
would be difficult or impossible to conduct, because they would need to include a very large and
diverse population with wide variation in exposure and a long follow-up period. Furthermore,
two experts (B and I) cited studies documenting difficulties in detecting a threshold using
epidemiological studies (Cakmak et al. 1999, and Brauer et al., 2002, respectively). The experts
generally thought that clinical and toxicological studies are best suited for researching
mechanisms and for addressing thresholds in very narrowly defined groups. One expert, B,
thought that a better understanding of the detailed biological mechanism is critical to addressing
the question of a threshold.
"One expert, K, believed it was possible to make a conceptual argument for a population
threshold. He drew an analogy with smoking, indicating that among heavy smokers, only a
proportion of them gets lung cancer or demonstrates an accelerated decline in lung function. He
thought that the idea that there is no level that is biologically safe is fundamentally at odds with
toxicological theory. He did not think that a population threshold was detectable in the currently
available epidemiologic studies. He indicated that some of the cohort studies showed greater
uncertainty in the shape of the C-R function at lower levels, which could be indicative of a
threshold.
"Expert K chose to incorporate a threshold into his C-R function. He indicated that he was 50
percent sure that a threshold existed. If there were a threshold, he thought that there was an 80
percent chance that it would be less than or equal to 5 ug/m3, and a 20 percent chance that it
would fall between 5 and 10 ug/m3."
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.
Pg 2271: "Eight experts thought the true C-R function relating mortality to changes in annual
average PM2.5 was log-linear across the entire study range (In(mortality) ) P x PM). Four experts
(B, F, K, and L) specified a "piecewise" log-linear function, with different P coefficients for PM
concentrations above and below an expert-specified break point. This approach allowed them to
express increased uncertainty in mortality effects seen at lower concentrations in major
epidemiological studies. Expert K thought the relationship would be log-linear above a
threshold."
Pg 2271: "Expert K also applied a threshold, T, to his function, which he described
probabilistically. He specified P(T > 0) = 0.5. Given T > 0, he indicated P(T < 5 ug/m3) = 0.8
and P(5 ug/m3 < T < 10 ug/m3) = 0.2. Figure 3 does not include the impact of applying expert
K's threshold, as the size of the reduction in benefits will depend on the distribution of baseline
PM levels in a benefits analysis."
A-21
-------
Experts:
Dr. Doug W. Dockery, Harvard School of Public Health
Dr. Kazuhiko Ito, Nelson Institute of Environmental Medicine, NYU School of Medicine,
Tuxedo, NY
Dr. Dan Krewski, University of Ottawa
Dr. Nino Kiinzli, University of Southern California Keck School of Medicine
Dr. Morton Lippmann, Professor, Nelson Institute of Environmental Medicine, New York University
School of Medicine, Tuxedo, NY
Dr. Joe Mauderly, Lovelace Respiratory Research Institute
Dr. Bart Ostro, Chief, Air Pollution Epidemiology Unit, Office of Environmental Health
Hazard Assessment, California Environmental Protection Agency, Oakland, CA
Dr. Arden Pope, Professor, Department of Economics, Brigham Young University, Provo, UT
Dr. Richard Schlesinger, Pace University
Dr. Joel Schwartz, Harvard School of Public Health
Dr. George Thurston—Department of Environmental Medicine, NYU, Tuxedo, NY
Dr. Mark Utell, University of Rochester School of Medicine and Dentistry
A-22
-------
H. CASAC comments on PM Staff Paper (2005)
U.S. Environmental Protection Agency - Science Advisory Board (U.S. EPA-SAB). 2005.
EPA's Review of the National Ambient Air Quality Standards for Particulate Matter
(Second Draft PM Staff Paper, January 2005). EPA-SAB-CASAC-05-007. June. Available
on the Internet at
-------
Dr. Petros Koutrakis, Professor of Environmental Science, Environmental Health , School of
Public Health, Harvard University (HSPH), Boston, MA
Dr. Allan Legge, President, Biosphere Solutions, Calgary, Alberta
Dr. Paul J. Lioy, Associate Director and Professor, Environmental and Occupational Health
Sciences Institute, UMDNJ - Robert Wood Johnson Medical School, NJ
Dr. Morton Lippmann, Professor, Nelson Institute of Environmental Medicine, New York
University School of Medicine, Tuxedo, NY
Dr. Joe Mauderly, Vice President, Senior Scientist, and Director, National Environmental
Respiratory Center, Lovelace Respiratory Research Institute, Albuquerque, NM
Dr. Roger O. McClellan, Consultant, Albuquerque, NM
Dr. Frederick J. Miller, Consultant, Gary, NC
Dr. Gunter Oberdorster, Professor of Toxicology, Department of Environmental Medicine,
School of Medicine and Dentistry, University of Rochester, Rochester, NY
Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Robert D. Rowe, President, Stratus Consulting, Inc., Boulder, CO
Dr. Jonathan M. Samet, Professor and Chair, Department of Epidemiology, Bloomberg School
of Public Health, Johns Hopkins University, Baltimore, MD
Dr. Frank Speizer, Edward Kass Professor of Medicine, Channing Laboratory, Harvard
Medical School, Boston, MA
Dr. Sverre Vedal, Professor of Medicine, School of Public Health and Community
Medicine University of Washington, Seattle, WA
Mr. Ronald White, Research Scientist, Epidemiology, Bloomberg School of Public Health,
Johns Hopkins University, Baltimore, MD
Dr. Warren H. White, Visiting Professor, Crocker Nuclear Laboratory, University of California
-Davis, Davis, CA
Dr. George T. Wolff, Principal Scientist, General Motors Corporation, Detroit, MI
Dr. Barbara Zielinska, Research Professor, Division of Atmospheric Science, Desert Research
Institute, Reno, NV
A-24
-------
I. HES Comments on 812 Analysis (2004)
U.S. Environmental Protection Agency - Science Advisory Board (U.S. 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
.
Pg 20: "The Subcommittee agrees that the whole range of uncertainties, such as the questions of
causality, shape of C-R functions and thresholds, relative toxicity, years of life lost, cessation lag
structure, cause of death, biologic pathways, or susceptibilities may be viewed differently for
acute effects versus long-term effects.
"For the studies of long-term exposure, the HES notes that Krewski et al. (2000) have conducted
the most careful work on this issue. They report that the associations between PM2 5 and both all-
cause and cardiopulmonary mortality were near linear within the relevant ranges, with no
apparent threshold. Graphical analyses of these studies (Dockery et al., 1993, Figure 3 and
Krewski et al., 2000, page 162) also suggest a continuum of effects down to lower levels.
Therefore, it is reasonable for EPA to assume a no threshold model down to, at least, the low end
of the concentrations reported in the studies."
HES Panel Members
Dr. Bart Ostro, California Office of Environmental Health Hazard Assessment (OEHHA),
Oakland, CA
Mr. John Fintan Hurley, Institute of Occupational Medicine (IOM), Edinburgh, Scotland
Dr. Patrick Kinney, Columbia University, New York, NY
Dr. Michael Kleinman, University of California, Irvine, C A
Dr. Nino Kiinzli, University of Southern California, Los Angeles, CA
Dr. Morton Lippmann, New York University School of Medicine, Tuxedo, NY Dr. Rebecca
Parkin, The George Washington University, Washington, DC
Dr. Trudy Cameron, University of Oregon, Eugene, OR
Dr. David T. Allen, University of Texas, Austin, TX
Ms. Lauraine Chestnut, Stratus Consulting Inc., Boulder, CO
Dr. Lawrence Goulder, Stanford University, Stanford, CA
Dr. James Hammitt, Harvard University, Boston, MA
Dr. F. Reed Johnson, Research Triangle Institute, Research Triangle Park, NC
Dr. Charles Kolstad, University of California, Santa Barbara, C A
A-25
-------
Dr. Lester B. Lave, Carnegie Mellon University, Pittsburgh, PA
Dr. Virginia McConnell, Resources for the Future, Washington, DC
Dr. V. Kerry Smith, North Carolina State University, Raleigh, NC
Other Panel Members
Dr. John Evans, Harvard University, Portsmouth, NH Dr. Dale Hattis, Clark University,
Worcester, MA Dr. D. Warner North, NorthWorks Inc., Belmont, CA Dr. Thomas S. Wallsten,
University of Maryland, College Park, MD
A-26
-------
J. NRC - Committee on Estimating the Health Risk Reduction Benefits of Proposed Air
Pollution Regulations (2002)
National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed
Air Pollution Regulations. Washington, DC: The National Academies Press.
Pg 109: "Linearity and Thresholds
"The shape of the concentration-response functions may influence the overall estimate of
benefits. The shape is particularly important for lower ambient air pollution concentrations to
which a large portion of the population is exposed. For this reason, the impact of the existence of
a threshold may be considerable.
"In epidemiological studies, air pollution concentrations are usually measured and modeled as
continuous variables. Thus, it may be feasible to test linearity and the existence of thresholds,
depending on the study design. In time-series studies with the large number of repeated
measurements, linearity and thresholds have been formally addressed with reasonable statistical
power. For pollutants such as PMi0 and PM2 5, there is no evidence for any departure of linearity
in the observed range of exposure, nor any indication of a threshold. For example, examination
of the mortality effects of short-term exposure to PMi0 in 88 cities indicates that the
concentration-response functions are not due to the high concentrations and that the slopes of
these functions do not appear to increase at higher concentrations (Samet et al. 2000). Many
other mortality studies have examined the shape of the concentration-response function and
indicated that a linear (nonthreshold) model fit the data well (Pope 2000). Furthermore, studies
conducted in cities with very low ambient pollution concentrations have similar effects per unit
change in concentration as those studies conducted in cities with higher concentrations. Again,
this finding suggests a fairly linear concentration-response function over the observed range of
exposures.
"Regarding the studies of long-term exposure, Krewski et al. (2000) found that the assumption of
a linear concentration-response function for mortality outcomes was not unreasonable. However,
the statistical power to assess the shape of these functions is weakest at the upper and lower end
of the observed exposure ranges. Most of the studies examining the effects of long-term
exposure on morbidity compare subjects living in a small number of communities (Dockery et al.
1996; Ackermmann-Liebrich 1997; Braun-Fahrlander et al. 1997). Because the number of long-
term effects studies are few and the number of communities studied is relatively small (8 to 24),
the ability to test formally the absence or existence of a no-effect threshold is not feasible.
However, even if thresholds exist, they may not be at the same concentration for all health
outcomes.
"A review of the time-series and cohort studies may lead to the conclusion that although a
threshold is not apparent at commonly observed concentrations, one may exist at lower levels.
An important point to acknowledge regarding thresholds is that for health benefits analysis a key
threshold is the population threshold (the lowest of the individual thresholds). However, the
population threshold would be very difficult to observe empirically through epidemiology,
A-27
-------
because epidemiology integrates information from very large groups of people (thousands). Air
pollution regulations affect even larger groups of people (millions). It is reasonable to assume
that among such large groups susceptibility to air pollution health effects varies considerably
across individuals and depends on a large set of underlying factors, including genetic makeup,
age, exposure measurement error, preexisting disease, and simultaneous exposures from smoking
and occupational hazards. This variation in individual susceptibilities and the resulting
distribution of individual thresholds underlies the concentration-response function observed in
epidemiology. Thus, until biologically based models of the distribution of individual thresholds
are developed, it may be productive to assume that the population concentration-response
function is continuous and to focus on finding evidence of changes in its slope as one approaches
lower concentrations.
EPA 's Use of Thresholds
"In EPA's benefits analyses, threshold issues were discussed and interpreted. For the PM and
ozone National Ambient Air Quality Standards (NAAQS), EPA investigated the effects of a
potential threshold or reference value below which health consequences were assumed to be zero
(EPA 1997). Specifically, the high-end benefits estimate assumed a 12-microgram per cubic
meter (|ig/m3) mean threshold for mortality associated with long-term exposure to PM2.5. The
low-end benefits estimate assumed a 15-|ig/m3 threshold for all PM-related health effects. The
studies, however, included concentrations as low as 7.5 |ig/m3. For the Tier 2 rule and the HD
engine and diesel-fuel rule, no threshold was assumed (EPA 1999, 2000). EPA in these analyses
acknowledged that there was no evidence for a threshold for PM.
"Several points should be noted regarding the threshold assumptions. If a threshold is assumed
where one was not apparent in the original study, then the data should be refit and a new curve
generated with the assumption of a zero slope over a segment of the concentration-response
function that was originally found to be positively sloped. The assumption of a zero slope over a
portion of the curve will force the slope in the remaining segment of the positively sloped
concentration-response function to be greater than was indicated in the original study. A new
concentration-response function was not generated for EPA's benefits analysis for the PM and
ozone NAAQS for which threshold assumptions were made. The generation of the steeper slope
in the remaining portion of the concentration-response function may fully offset the effect of
assuming a threshold. These aspects of assuming a threshold in a benefits analysis where one
was not indicated in the original study should be conveyed to the reader. The committee notes
that the treatment of thresholds should be evaluated in a consistent and transparent framework by
using different explicit assumptions in the formal uncertainty analyses (see Chapter 5)."
Pg 117: "Although the assumption of no thresholds in the most recent EPA benefits analyses was
appropriate, EPA should evaluate threshold assumptions in a consistent and transparent
framework using several alternative assumptions in the formal uncertainty analysis."
Pg 136: "Two additional illustrative examples are thresholds for adverse effects and lag
structures.- EPA considers implausible any threshold for mortality in the particulate matter (PM)
exposure ranges under consideration (EPA 1999a, p. 3-8). Although the agency conducts
sensitivity analyses incorporating thresholds, it provides no judgment as to their relative
A-28
-------
plausibility. In a probabilistic uncertainty analysis, EPA could assign appropriate weights to
various threshold models. For PM-related mortality in the Tier 2 analysis, the committee expects
that this approach would have resulted in only a slight widening of the probability distribution
for avoided mortality and a slight reduction in the mean of that distribution, thus reflecting
EPA's views about the implausibility of thresholds. The committee finds that such formal
incorporation of EPA's expert judgments about the plausibility of thresholds into its primary
analysis would have been an improvement.
"Uncertainty about thresholds is a special aspect of uncertainty about the shape of concentration-
response functions. Typically, EPA and authors of epidemiological studies assume that these
functions are linear on some scale. Often, the scale is a logarithmic transformation of the risk or
rate of the health outcome, but when a rate or risk is low, a linear function on the logarithmic
scale is approximately linear on the scale of the rate or risk itself. Increasingly, epidemiological
investigators are employing analytic methods that permit the estimation of nonlinear shapes for
concentration-response functions (Greenland et al. 1999). As a consequence, EPA will need to be
prepared to incorporate nonlinear concentration-response functions from epidemiological studies
into the agency's health benefits analyses. Any source of error or bias that can distort an
epidemiological association can also distort the shape of an estimated concentration -response
function, as can variation in individual susceptibility (Hattis and Burmaster 1994; Hattis et al.
2001)."
Pg 137: "In principle, many components of the health benefits model need realistic probabilistic
models (see Table 5-1 for a listing of such components), in addition to concentration-response
thresholds and time lags between exposure and response. For example, additional features of the
concentration-response function—such as projection of the results from the study population to
the target populations (which may have etiologically relevant characteristics outside the range
seen in the study population) and the projection of baseline frequencies of morbidity and
mortality into the future—must be characterized probabilistically. Other uncertainties that might
affect the probability distributions are the estimations of population exposure (or even
concentration) from emissions, estimates of emissions themselves, and the relative toxicity of
various classes of particles. Similarly, many aspects of the analysis of the impact of regulation on
ambient concentrations and on population exposure involve considerable uncertainty and,
therefore, may be beneficially modeled in this way. Depending on the analytic approach used,
joint probability distributions will have to be specified to incorporate correlations between model
components that are structurally dependent upon each other, or the analysis will have to be
conducted in a sequential fashion that follows the model for the data-generating process.
"EPA should explore alternative options for incorporating expert judgment into its probabilistic
uncertainty analyses. The agency possesses considerable internal expertise, which should be
employed as fully as possible. Outside experts should also be consulted as needed, individually
or in panels. In all cases, when expert judgment is used in the construction of a model
component, the experts should be identified and the rationales and empirical bases for their
judgments should be made available."
A-29
-------
NRC members
Dr. JOHN C. BAILAR, III (Chair), (emeritus) University of Chicago, Chicago, Illinois
Dr. HUGH ROSS ANDERSON, University of London, London, England
Dr. MAUREEN L. CROPPER, University of Maryland, College Park
Dr. JOHN S. EVANS, Harvard University, Boston, Massachusetts
Dr. DALE B. HATTIS, Clark University, Worcester, Massachusetts
Dr. ROGENE F. HENDERSON, Lovelace Respiratory Research Institute, Albuquerque, New
Mexico
Dr. PATRICK L. KINNEY, Columbia University, New York, New York
Dr. NINO KUNZLI, University of Basel, Basel, Switzerland; as of September 2002, University
of Southern California, Los Angeles
Dr. BART D. OSTRO, California Environmental Protection Agency, Oakland
Dr. CHARLES POOLE, University of North Carolina, Chapel Hill
Dr. KIRKR. SMITH, University of California, Berkeley
Dr. PETER A. VALBERG, Gradient Corporation, Cambridge, Massachusetts
Dr. SCOTT L. ZEGER, Johns Hopkins University, Baltimore, Maryland
A-30
-------
APPENDIX B
LOWEST MEASURED LEVEL (LML) ASSESSMENT FOR RULES WITHOUT
POLICY-SPECIFIC AIR QUALITY DATA AVAILABLE: TECHNICAL SUPPORT
DOCUMENT (TSD)
B-l
-------
Lowest Measured Level (LML) Assessment for Rules
without Policy-Specific Air Quality Data Available
Technical Support Document (TSD)
June 2010
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impact Division
Air Benefit-Cost Group
Research Triangle Park, North Carolina
B-2
-------
Inherent in any complex Regulatory Impact Analysis (RIA) are multiple sources of
uncertainty. Health benefits analysis 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.
There are a variety of methods to characterizing the uncertainty associated with the human health
benefits of air pollution, including quantitative and qualitative methods. When evaluated within
the context of these uncertainties, the health impact and monetized benefits estimates in an RIA
can provide useful information regarding the magnitude of the public health impacts attributable
to reducing air pollution.
Reductions in premature mortality typically dominate the size of the overall monetized
benefits. Therefore, most of the uncertainty characterization generally focuses on the mortality-
related benefits. Typically, EPA employs two primary techniques for quantifying this
uncertainty. First, because this characterization of random statistical error may omit important
sources of uncertainty, we employ the results of an expert elicitation on the relationship between
premature mortality and ambient PM2.5 concentration (Roman et al., 2008); this provides
additional insight into the likelihood of different outcomes and about the state of knowledge
regarding the benefits estimates. Second, when we have air quality modeling specific to the
policy we are evaluating and it can be used as an input to the health impact and economic
analysis, we use Monte Carlo methods for characterizing random sampling error associated with
the concentration response functions from epidemiological studies and economic valuation
functions.81 Both approaches have different strengths and weaknesses, which are fully described
in Chapter 5 of the PM NAAQS RIA (U.S. EPA, 2006).
In addition, some RIAs, including the PM NAAQS RIA (2006d) and Ozone NAAQS
RIA (2008a), also contain a suite of sensitivity analyses that evaluate the sensitivity of the
monetized benefits to the specification of alternate mortality cessation lags and income growth
adjustment factors. Cessation lags and income growth adjustments are simply multipliers applied
to the valuation function, which generally affect monetized benefits estimates in the same
manner. Thus, it is possible for readers to infer the sensitivity of these parameters by referring to
those previous analyses.82 Other RIAs contain unique sensitivity analyses that are specific to the
81 Currently, we are unable to characterize the random sampling error from the underlying studies when applying
national average benefit-per-ton estimates.
82 For example, in the PM NAAQS RIA, the use of an alternate lag structure would change the PM2 5-related
mortality benefits discounted at 3% discounted by between 10.4% and -27%; when discounted at 7%, these
B-3
-------
input parameters of that analysis, such as blood lead level (U.S. EPA, 2008b) or rollback method
(U.S. EPA, 2010a). Other sources of uncertainty, including the projection of atmospheric
conditions and source-level emissions, the projection of baseline morbidity rates, incomes and
technological development are typically unquantified in our RIAs. For these sources, we
typically provide a qualitative uncertainty characterization associated with these input
parameters.
One particular aspect of uncertainty has received extensive quantitative and qualitative
attention in recent RIAs: the existence of a threshold in the concentration-response function for
PM2.5-related mortality. A threshold is a specific type of discontinuity in the concentration-
response function where there are no benefits associated with reducing PM2.5 levels in areas
where the baseline air quality is less than the threshold. Previously, EPA had included a
sensitivity analysis with an arbitrary assumed threshold at 10 |ig/m3 in the PM-mortality health
impact function in the RIA to illustrate that the fraction of benefits that occur at lower air
pollution concentration levels are inherently more uncertain. A threshold of 10 |ig/m3 does not
necessarily have any stronger technical basis than any other threshold, and we could have instead
assumed a threshold at 4, 7.5, or 12 |ig/m3 for the sensitivity analysis. In addition to identifying
the most support for a non-threshold model, the underlying scientific evidence does not support
any specific "bright line".
Based on our review of the current body of scientific literature, EPA now estimates PM-
related mortality without applying an assumed concentration threshold. EPA's Integrated
Science Assessment for Particulate Matter (U.S. EPA, 2009b), which was recently 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
recognizing potential uncertainty about the exact shape of the concentration-response function.83
Since then, the Health Effects Subcommittee (U.S. EPA-SAB, 2010) of EPA's Council
concluded, "The HES fully supports EPA's decision to use a no-threshold model to estimate
mortality reductions. This decision is supported by the data, which are quite consistent in
benefits change by between 31% and -49%. When applying higher and lower income growth adjustments, the
monetary value of PM25 and ozone-related premature changes between 30% and -10%; the value of chronic
endpoints change between 5% and -2% and the value of acute endpoints change between 6% and -7%. (U.S. EPA,
2006)
83It is important to note that uncertainty regarding the shape of the concentration-response function is conceptually
distinct from an assumed threshold. An assumed threshold (below which there are no health effects) is a
discontinuity, which is a specific example of non-linearity.
B-4
-------
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. Therefore, there is no evidence to support a truncation of the CRF."
For a summary of these scientific review statements and the panel members please consult the
Technical Support Document (TSD) Summary of Expert Opinions on the Existence of a
Threshold (U.S. EPA, 2010c).
Consistent with this finding, we have conformed the previous threshold sensitivity
analysis to the current state of the PM science by incorporating a new "Lowest Measured Level"
(LML) assessment. While an LML assessment provides some insight into the level of uncertainty
in the estimated PM mortality benefits, EPA does not view the LML as a threshold and continues
to quantify PM-related mortality impacts using a full range of modeled air quality
concentrations. Unlike an assumed threshold, which is a modeling assumption that reduces the
magnitude of the estimated health impacts, the LML is a characterization of the fraction of
benefits that are more uncertain. It is important to emphasize that just because we have greater
confidence in the benefits above the LML, this does not mean that we have no confidence that
benefits occur below the LML.
While the LML of each study is important to consider when characterizing and
interpreting the overall level PM-related benefits, EPA believes that large cohort-based mortality
estimates are suitable for use in air pollution health impact analyses. When estimating PM
mortality impacts using risk coefficients drawn from the Harvard Six Cities and the American
Cancer Society cohorts there are innumerable other attributes that may affect the size of the
reported risk estimates—including differences in population demographics, the size of the
cohort, activity patterns and particle composition among others. The LML assessment provides a
limited representation of one key difference between the two studies. For the purpose of
estimating the benefits associated with reducing PM2.5 levels, we utilize the effect coefficients
from Pope et al. (2002) for the American Cancer Society cohort and from Laden et al. (2006) for
the Harvard Six Cities cohort.
Analyses of these cohorts using data from more recent years, during which time PM
concentrations have fallen, continue to report strong associations with mortality. For example,
the Krewski et al. (2009) follow-up study of the American Cancer Society cohort had an LML of
5.8 |ig/m3. As we model mortality impacts among populations exposed to levels of PM2.5 that are
successively lower than the LML of each study, our confidence in the results diminishes. As air
pollution emissions continue to decrease over time, there will be more people in areas where we
do not have published epidemiology studies. However, each successive cohort study has shown
B-5
-------
evidence of effects at successively lower levels of PM2.5. As more large cohort studies follow
populations over time, we will likely have more studies with lower LML as air quality levels
continue to improve. Even in the absence of a definable threshold, we have more confidence in
the benefits estimates above the LML of the large cohort studies. To account for the uncertainty
in each of the studies that we base our mortality estimates on, we provide the LML for each of
the cohort studies. However, the finding of effects at the lowest LML from the recent Krewski et
al (2009) study indicates that confidence in PM2.5-related mortality effects down to at least 5.8
|ig/m3 is high.
In the recently proposed Transport Rule RIA (U.S. EPA, 201 Ob), we included the new
LML assessment in which we binned the estimated number of avoided PM2.s-related premature
mortalities resulting from the implementation of the Transport Rule according to the projected
2014 baseline PM2.5 air quality levels. This presentation is consistent with our approach to
applying PM2.5 mortality risk coefficients that have not been adjusted to incorporate an assumed
threshold. A very large proportion of the avoided PM-related impacts occurred among
populations initially exposed at or above the LML of each study, which gave us a high level of
confidence in the PM mortality estimates. This assessment summarized the distribution of
avoided PM mortality impacts according to the baseline PM2.5 levels experienced by the
population receiving the PM2 5 mortality benefit. Approximately 80% of the avoided impacts
occurred at or above a baseline annual mean PM2.5 level of 10 |ig/m3 (the LML of the Laden et
al. 2006 study); about 97% occur at or above an annual mean PM2 5 level of 7.5 |ig/m3 (the LML
of the Pope et al. 2002 study). This assessment confirmed that the great majority of the impacts
associated with the Transport Rule occurred at or above each study's LML.
For the Transport Rule, policy-specific air quality modeling data for the year 2014 was
available as an input into the benefits analysis. For some rules, especially New Source
Performance Standards (NSPS) or National Emissions Standards for Hazardous Air Pollutant
(NESHAP) rules, policy-specific air quality data is not available due to time or resource
limitations. For these rules, we provide the following LML assessment as a characterization of
the baseline exposure to PM2 5 levels in the U.S. Many of the upcoming NSPS and NESHAP
rules have compliance dates between 2013 and 2016 and represent marginal improvements in air
quality levels. Although it the data is not a perfect match, we believe that the air quality data
B-6
-------
from the Transport Rule is a reasonable approximation of the baseline exposure in the U.S. for
upcoming NSPS and NESHAP rules.84
For rules without air quality modeling, we generally estimate the monetized benefits and
health impacts using benefit-per-ton estimates (Fann, Fulcher and Hubbell, 2009). Using this
method, we are unable to estimate the percentage of premature mortality associated with the
specific rules' emission reductions at each PM2.5 level. However, we believe that it is still
important to characterize the uncertainty associated with the distribution of the baseline air
quality. As a surrogate measure of mortality impacts, we provide the percentage of baseline
exposure at each PM2.5 level. If air quality levels in the baseline are above the LML, the marginal
changes anticipated from these rules would likely also lead to post-policy air quality levels above
the LML. Therefore, we have high confidence that the magnitude of the benefits estimated for
these rules, as the marginal changes would also be above the LML.
It is important to note that baseline exposure is only one parameter in the health impact
function, along with baseline incidence rates population, and change in air quality. In other
words, the percentage of the population exposed to air pollution below the LML is not the same
as the percentage of the population experiencing health impacts as a result of a specific emission
reduction policy. The most important aspect, which we are unable to quantify for rules without
air quality modeling, is the shift in exposure associated with the specific rule. Therefore, caution
is warranted when interpreting the following assessment.
A very large proportion of the population is exposed at or above the lowest LML of the
cohort studies (Figures 1 and 2), increasing our confidence in the PM mortality analysis. Figure 1
shows a bar chart of the percentage of the population exposed to various air quality levels in the
pre- and post-policy policy. Figure 2 shows a cumulative distribution function of the same data.
In addition, Figure 2 also demonstrates that policy had a greater impact on reducing exposure to
the portion of the population in areas with high PM2.5 levels relative to the portion of the
population at low PM2.5 levels. Both figures identify the LML for each of the major cohort
studies. As the policy shifts the distribution of air quality levels, fewer people are exposed to
PM2.5 levels above the LML. Under baseline conditions, about 96 percent of the population is
84 Because the Transport Rule is not yet promulgated, the baseline exposure obtained from this modeling data would
slightly overestimate the fraction of the population exposed to air quality levels below the LML. As additional
rules continue to reduce the ambient PM2 5 levels over time, a larger fraction of the population would be exposed
to air quality levels below the LML. However, the emission reductions anticipated from the rules without air
quality modeling available are comparatively small and represent marginal changes. We intend to update this
LML assessment as necessary to correspond with the successively lower baseline air quality levels anticipated as
the result of promulgating significant upcoming rules.
B-7
-------
exposed to annual mean PM2.5 levels of at least 5.8 iig/m3, which is the lowest air quality level
considered in the most recent study of the American Cancer Society cohort by Krewski et al.
(2009). Using the Pope et al. (2002) study, the 85% of the population is exposed at or above the
LML of 7.5 Lig/m3. Using the Laden et al. (2006) study, 40% of the population is exposed above
the LML of 10 Lig/m3. As we model mortality impacts among populations exposed to levels of
PM2.5 that are successively lower than the LML of the lowest cohort study, our confidence in the
results diminishes. However, the analysis above confirms that the great majority of the impacts
occur at or above the lowest cohort study's LML. It is important to emphasize that we have high
confidence in PM2.s-related effects down to the lowest LML of the major cohort studies, which is
5.8 Lig/m3. Just because we have greater confidence in the benefits above the LML, this does not
mean that we have no confidence that benefits occur below the LML.
Figure 1: Percentage of Adult Population by Annual Mean PMi.s Exposure (pre- and post-
policy)
Krewski etal. 2009 Pope etal. 2002
Laden et al. 2006
The control strategy lowers PMis
levels substantially, particularly
among highly exposed
populations. In the baseline, 96%
of the population lived in areas
where PM25 levels above the
lowest measured levels of the
Krewski study, increasing our
confidence in the estimated
mortality reductions for this rule.
10 '_ 13 14 15 16 17 IB 19 20
Post-contro • Ease ine
B-c
-------
Figure 2: Cumulative Distribution of Adult Population at Annual Mean PM2.s levels (pre-
and post-policy)
100%
90% —
Krewski et al. 2009 Pope et al. 2002 Laden et al. 2006
SU:-:.
The control strategy lowers PM1S levels
substantially, particularly among highly
exposed populations. In the baseline, 96% of
the population lived in areas where PM2s
levels above the lowest measured levels of
the Krewski study, increasing our
confidence in the estimated mortality
reductions for this rule.
10 11 12 13 14 15 16 17 IS 19 20
— Baseline
There are several important differences between the assessment conducted for the
Transport Rule and the assessment presented here. If you compare the graphics in the Transport
Rule to those provided here, you will notice that these graphs show a larger percentage of the
population below the LML. It is imperative to point out that the Transport Rule graphics
represented mortality impacts attributable to the Transport Rule, whereas these graphics
represent exposure. Mortality impacts are the result of the incremental change in exposure
between the baseline and control. However, the baseline population exposure at lower air quality
levels is so much larger than the impacts among these same populations. In other words, the
population exposed to lower PM2.5 levels are not receiving very much of the air quality benefit
between the base and the control case.
B-9
-------
References
Fann, N., C.M. Fulcher, BJ. Hubbell. 2009. The influence of location, source, and emission type
in estimates of the human health benefits of reducing a ton of air pollution. Air Qual Atmos
Health (2009) 2:169-176.
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. Estimating the Public Health Benefits of Proposed Air Pollution Regulations.
Washington, DC: The National Academies Press.
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.
Roman, H.A., K.D. Walker, T.L. Walsh, L. Conner, H.M. Richmond, J. Hubbell, and P.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.
U.S. Environmental Protection Agency (U.S. EPA). 2006. 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). 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. 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. March. Available at .
U.S. Environmental Protection Agency (U.S. EPA). 2010a. Final Regulatory Impact Analysis
(RIA) for the SO2 National Ambient Air Quality Standards (NAAQS). Office of Air Quality
Planning and Standards, Research Triangle Park, NC. June. Available at
.
U.S. Environmental Protection Agency (U.S. EPA). 2010b. Regulatory Impact Analysis for the
Transport Rule. Office of Air Quality Planning and Standards, Research Triangle Park, NC.
June. Available at .
B-10
-------
U.S. Environmental Protection Agency (U.S. EPA). 2010c. Summary of Expert Opinions on the
Existence of a Threshold in the Concentration-Response Function for PM2.s-related
Mortality: Technical Support Document. Compiled by Office of Air Quality Planning and
Standards, Research Triangle Park, NC. July. Available 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
.
B-ll
-------
United States Office of Air Quality Planning and Publication No. EPA
Environmental Protection Standards 452/R-10-010
Agency Health and Environmental Impacts August 2010
Division
Research Triangle Park, NC
------- |